diff options
| author | Larry Knox <lrknox@hdfgroup.org> | 2019-07-25 16:36:37 (GMT) |
|---|---|---|
| committer | Larry Knox <lrknox@hdfgroup.org> | 2019-07-25 16:47:12 (GMT) |
| commit | 8008294578b5a133907d7ab1dd20e34735c54535 (patch) | |
| tree | d1b9228d468afc05da9333567ea43a04bb0c4272 | |
| parent | d3fdcd8a680ad0f8b21304b35e8564b774a88ef0 (diff) | |
| download | hdf5-8008294578b5a133907d7ab1dd20e34735c54535.zip hdf5-8008294578b5a133907d7ab1dd20e34735c54535.tar.gz hdf5-8008294578b5a133907d7ab1dd20e34735c54535.tar.bz2 | |
Squashed commit of the following:
Merge changes from update_merged_S3_HDFS branch into develop.
commit d5034315aea88629929ac0c9c59ebfafd5f21a31
Merge: 9c48823 d3fdcd8
Author: Larry Knox <lrknox@hdfgroup.org>
Date: Thu Jul 25 08:24:53 2019 -0500
Merge branch 'develop' into update_merged_S3_HDFS
68 files changed, 19629 insertions, 225 deletions
diff --git a/CMakeLists.txt b/CMakeLists.txt index 95d24bb..dc6a94f 100644 --- a/CMakeLists.txt +++ b/CMakeLists.txt @@ -574,6 +574,27 @@ include (${HDF_RESOURCES_DIR}/HDFCompilerFlags.cmake) set (CMAKE_MODULE_PATH ${HDF_RESOURCES_DIR} ${HDF_RESOURCES_EXT_DIR} ${CMAKE_MODULE_PATH}) #----------------------------------------------------------------------------- +# Option to Enable HDFS +#----------------------------------------------------------------------------- +option (HDF5_ENABLE_HDFS "Enable HDFS" OFF) +if (HDF5_ENABLE_HDFS) + find_package(JNI REQUIRED) + if (JNI_FOUND) + set (H5_HAVE_LIBJVM 1) + endif () + find_package(HDFS REQUIRED) + if (HDFS_FOUND) + set (H5_HAVE_LIBHDFS 1) + set (H5_HAVE_HDFS_H 1) + if (NOT MSVC) + list (APPEND LINK_LIBS -pthread) + endif () + else () + message (FATAL_ERROR "Set to use libhdfs library, but could not find or use libhdfs. Please verify that the path to HADOOP_HOME is valid, and/or reconfigure without HDF5_ENABLE_HDFS") + endif () +endif () + +#----------------------------------------------------------------------------- # Option to Enable MPI Parallel #----------------------------------------------------------------------------- option (HDF5_ENABLE_PARALLEL "Enable parallel build (requires MPI)" OFF) @@ -636,6 +636,8 @@ ./src/H5FDdrvr_module.h ./src/H5FDfamily.c ./src/H5FDfamily.h +./src/H5FDhdfs.c +./src/H5FDhdfs.h ./src/H5FDint.c ./src/H5FDlog.c ./src/H5FDlog.h @@ -646,9 +648,13 @@ ./src/H5FDmpio.h ./src/H5FDmulti.c ./src/H5FDmulti.h +./src/H5FDros3.c +./src/H5FDros3.h ./src/H5FDpkg.h ./src/H5FDprivate.h ./src/H5FDpublic.h +./src/H5FDs3comms.h +./src/H5FDs3comms.c ./src/H5FDsec2.c ./src/H5FDsec2.h ./src/H5FDspace.c @@ -1069,6 +1075,7 @@ ./test/h5fc_ext_none.h5 ./test/h5test.c ./test/h5test.h +./test/hdfs.c ./test/hyperslab.c ./test/istore.c ./test/le_data.h5 @@ -1094,7 +1101,9 @@ ./test/paged_nopersist.h5 ./test/paged_persist.h5 ./test/reserved.c +./test/ros3.c ./test/pool.c +./test/s3comms.c ./test/set_extent.c # ====distribute this for now. See HDFFV-8236==== ./test/space_overflow.c @@ -1546,6 +1555,9 @@ ./tools/lib/io_timer.c ./tools/lib/io_timer.h +./tools/libtest/Makefile.am +./tools/libtest/h5tools_utils.c + ./tools/src/misc/Makefile.am ./tools/src/misc/h5clear.c ./tools/src/misc/h5debug.c @@ -3015,6 +3027,8 @@ ./java/src/hdf/hdf5lib/structs/H5AC_cache_config_t.java ./java/src/hdf/hdf5lib/structs/H5E_error2_t.java ./java/src/hdf/hdf5lib/structs/H5F_info2_t.java +./java/src/hdf/hdf5lib/structs/H5FD_hdfs_fapl_t.java +./java/src/hdf/hdf5lib/structs/H5FD_ros3_fapl_t.java ./java/src/hdf/hdf5lib/structs/H5G_info_t.java ./java/src/hdf/hdf5lib/structs/H5L_info_t.java ./java/src/hdf/hdf5lib/structs/H5O_hdr_info_t.java @@ -3178,6 +3192,8 @@ ./java/test/testfiles/JUnit-TestH5P.txt ./java/test/testfiles/JUnit-TestH5PData.txt ./java/test/testfiles/JUnit-TestH5Pfapl.txt +./java/test/testfiles/JUnit-TestH5Pfaplhdfs.txt +./java/test/testfiles/JUnit-TestH5Pfapls3.txt ./java/test/testfiles/JUnit-TestH5Plist.txt ./java/test/testfiles/JUnit-TestH5Pvirtual.txt ./java/test/testfiles/JUnit-TestH5PL.txt @@ -3216,6 +3232,8 @@ ./java/test/TestH5P.java ./java/test/TestH5PData.java ./java/test/TestH5Pfapl.java +./java/test/TestH5Pfaplhdfs.java +./java/test/TestH5Pfapls3.java ./java/test/TestH5Plist.java ./java/test/TestH5Pvirtual.java ./java/test/TestH5PL.java @@ -3249,6 +3267,7 @@ ./config/cmake/ConfigureChecks.cmake ./config/cmake/CPack.Info.plist.in ./config/cmake/CTestCustom.cmake +./config/cmake/FindHDFS.cmake ./config/cmake/H5cxx_config.h.in ./config/cmake/H5pubconf.h.in ./config/cmake/hdf5-config.cmake.in @@ -3358,6 +3377,8 @@ ./testpar/CMakeVFDTests.cmake ./tools/CMakeLists.txt ./tools/lib/CMakeLists.txt +./tools/libtest/CMakeLists.txt +./tools/libtest/CMakeTests.cmake ./tools/src/CMakeLists.txt ./tools/test/CMakeLists.txt ./tools/src/h5copy/CMakeLists.txt @@ -3478,6 +3499,7 @@ ./testpar/Makefile.in ./tools/Makefile.in ./tools/lib/Makefile.in +./tools/libtest/Makefile.in ./tools/src/Makefile.in ./tools/src/h5copy/Makefile.in ./tools/src/h5diff/Makefile.in @@ -139,6 +139,8 @@ $Source = ""; "H5FD_t" => "x", "H5FD_class_t" => "x", "H5FD_stream_fapl_t" => "x", + "H5FD_ros3_fapl_t" => "x", + "H5FD_hdfs_fapl_t" => "x", "H5FD_file_image_callbacks_t" => "x", "H5G_iterate_t" => "x", "H5G_info_t" => "x", diff --git a/config/cmake/ConfigureChecks.cmake b/config/cmake/ConfigureChecks.cmake index 3bd0553..f295a1c 100644 --- a/config/cmake/ConfigureChecks.cmake +++ b/config/cmake/ConfigureChecks.cmake @@ -155,6 +155,21 @@ if (NOT WINDOWS) endif () #----------------------------------------------------------------------------- +# Check if ROS3 driver works +#----------------------------------------------------------------------------- +option (HDF5_ENABLE_ROS3_VFD "Build the ROS3 Virtual File Driver" OFF) + if (HDF5_ENABLE_ROS3_VFD) + find_package(CURL REQUIRED) + find_package(OpenSSL REQUIRED) + if (${CURL_FOUND} AND ${OPENSSL_FOUND}) + set (${HDF_PREFIX}_HAVE_ROS3_VFD 1) + list (APPEND LINK_LIBS ${CURL_LIBRARIES} ${OPENSSL_LIBRARIES}) + else () + message (STATUS "The Read-Only S3 VFD was requested but cannot be built.\nPlease check that openssl and cURL are available on your\nsystem, and/or re-configure without option HDF5_ENABLE_ROS3_VFD.") + endif () +endif () + +#----------------------------------------------------------------------------- # Check if C has __float128 extension #----------------------------------------------------------------------------- diff --git a/config/cmake/FindHDFS.cmake b/config/cmake/FindHDFS.cmake new file mode 100644 index 0000000..e401a94 --- /dev/null +++ b/config/cmake/FindHDFS.cmake @@ -0,0 +1,70 @@ + +# DerivedFrom: https://github.com/cloudera/Impala/blob/cdh5-trunk/cmake_modules/FindHDFS.cmake +# - Find HDFS (hdfs.h and libhdfs.so) +# This module defines +# Hadoop_VERSION, version string of ant if found +# HDFS_INCLUDE_DIR, directory containing hdfs.h +# HDFS_LIBRARIES, location of libhdfs.so +# HDFS_FOUND, whether HDFS is found. + +exec_program($ENV{HADOOP_HOME}/bin/hadoop ARGS version OUTPUT_VARIABLE Hadoop_VERSION + RETURN_VALUE Hadoop_RETURN) + +# currently only looking in HADOOP_HOME +find_path(HDFS_INCLUDE_DIR hdfs.h PATHS + $ENV{HADOOP_HOME}/include/ + # make sure we don't accidentally pick up a different version + NO_DEFAULT_PATH +) + +if ("${CMAKE_SIZEOF_VOID_P}" STREQUAL "8") + set(arch_hint "x64") +elseif ("$ENV{LIB}" MATCHES "(amd64|ia64)") + set(arch_hint "x64") +else () + set(arch_hint "x86") +endif() + +message(STATUS "Architecture: ${arch_hint}") + +if ("${arch_hint}" STREQUAL "x64") + set(HDFS_LIB_PATHS $ENV{HADOOP_HOME}/lib/native) +else () + set(HDFS_LIB_PATHS $ENV{HADOOP_HOME}/lib/native) +endif () + +message(STATUS "HDFS_LIB_PATHS: ${HDFS_LIB_PATHS}") + +find_library(HDFS_LIB NAMES hdfs PATHS + ${HDFS_LIB_PATHS} + # make sure we don't accidentally pick up a different version + NO_DEFAULT_PATH +) + +if (HDFS_LIB) + set(HDFS_FOUND TRUE) + set(HDFS_LIBRARIES ${HDFS_LIB}) + set(HDFS_STATIC_LIB ${HDFS_LIB_PATHS}/${CMAKE_STATIC_LIBRARY_PREFIX}hdfs${CMAKE_STATIC_LIBRARY_SUFFIX}) + + add_library(hdfs_static STATIC IMPORTED) + set_target_properties(hdfs_static PROPERTIES IMPORTED_LOCATION ${HDFS_STATIC_LIB}) +else () + set(HDFS_FOUND FALSE) +endif () + +if (HDFS_FOUND) + if (NOT HDFS_FIND_QUIETLY) + message(STATUS "${Hadoop_VERSION}") + message(STATUS "HDFS_INCLUDE_DIR: ${HDFS_INCLUDE_DIR}") + message(STATUS "HDFS_LIBRARIES: ${HDFS_LIBRARIES}") + message(STATUS "hdfs_static: ${HDFS_STATIC_LIB}") + endif () +else () + message(FATAL_ERROR "HDFS includes and libraries NOT found." + "(${HDFS_INCLUDE_DIR}, ${HDFS_LIB})") +endif () + +mark_as_advanced( + HDFS_LIBRARIES + HDFS_INCLUDE_DIR +) diff --git a/config/cmake/H5pubconf.h.in b/config/cmake/H5pubconf.h.in index 9e7b8b7..cb05974 100644 --- a/config/cmake/H5pubconf.h.in +++ b/config/cmake/H5pubconf.h.in @@ -110,6 +110,9 @@ /* Define if the function stack tracing code is to be compiled in */ #cmakedefine H5_HAVE_CODESTACK @H5_HAVE_CODESTACK@ +/* Define to 1 if you have the <curl/curl.h> header file. */ +#cmakedefine H5_HAVE_CURL_H @H5_HAVE_CURL_H@ + /* Define if Darwin or Mac OS X */ #cmakedefine H5_HAVE_DARWIN @H5_HAVE_DARWIN@ @@ -185,6 +188,9 @@ /* Define to 1 if you have the `gettimeofday' function. */ #cmakedefine H5_HAVE_GETTIMEOFDAY @H5_HAVE_GETTIMEOFDAY@ +/* Define to 1 if you have the <hdfs.h> header file. */ +#cmakedefine H5_HAVE_HDFS_H @H5_HAVE_HDFS_H@ + /* Define if the compiler understands inline */ #cmakedefine H5_HAVE_INLINE @H5_HAVE_INLINE@ @@ -201,12 +207,24 @@ /* Define to 1 if you have the <io.h> header file. */ #cmakedefine H5_HAVE_IO_H @H5_HAVE_IO_H@ +/* Define to 1 if you have the `crypto' library (-lcrypto). */ +#cmakedefine H5_HAVE_LIBCRYPTO @H5_HAVE_LIBCRYPTO@ + +/* Define to 1 if you have the `curl' library (-lcurl). */ +#cmakedefine H5_HAVE_LIBCURL @H5_HAVE_LIBCURL@ + /* Define to 1 if you have the `dl' library (-ldl). */ #cmakedefine H5_HAVE_LIBDL @H5_HAVE_LIBDL@ /* Define to 1 if you have the `dmalloc' library (-ldmalloc). */ #cmakedefine H5_HAVE_LIBDMALLOC @H5_HAVE_LIBDMALLOC@ +/* Proceed to build with libhdfs */ +#cmakedefine H5_HAVE_LIBHDFS @H5_HAVE_LIBHDFS@ + +/* Define to 1 if you have the `jvm' library (-ljvm). */ +#cmakedefine H5_HAVE_LIBJVM @H5_HAVE_LIBJVM@ + /* Define to 1 if you have the `m' library (-lm). */ #cmakedefine H5_HAVE_LIBM @H5_HAVE_LIBM@ @@ -264,6 +282,15 @@ /* Define if MPI_Info_c2f and MPI_Info_f2c exists */ #cmakedefine H5_HAVE_MPI_MULTI_LANG_Info @H5_HAVE_MPI_MULTI_LANG_Info@ +/* Define to 1 if you have the <openssl/evp.h> header file. */ +#cmakedefine H5_HAVE_OPENSSL_EVP_H @H5_HAVE_OPENSSL_EVP_H@ + +/* Define to 1 if you have the <openssl/hmac.h> header file. */ +#cmakedefine H5_HAVE_OPENSSL_HMAC_H @H5_HAVE_OPENSSL_HMAC_H@ + +/* Define to 1 if you have the <openssl/sha.h> header file. */ +#cmakedefine H5_HAVE_OPENSSL_SHA_H @H5_HAVE_OPENSSL_SHA_H@ + /* Define if we have parallel support */ #cmakedefine H5_HAVE_PARALLEL @H5_HAVE_PARALLEL@ @@ -282,6 +309,10 @@ /* Define to 1 if you have the `rand_r' function. */ #cmakedefine H5_HAVE_RAND_R @H5_HAVE_RAND_R@ +/* Define whether the Read-Only S3 virtual file driver (VFD) should be + compiled */ +#cmakedefine H5_HAVE_ROS3_VFD @H5_HAVE_ROS3_VFD@ + /* Define to 1 if you have the `round' function. */ #cmakedefine H5_HAVE_ROUND @H5_HAVE_ROUND@ diff --git a/config/cmake/libhdf5.settings.cmake.in b/config/cmake/libhdf5.settings.cmake.in index 2117f3b..3451545 100644 --- a/config/cmake/libhdf5.settings.cmake.in +++ b/config/cmake/libhdf5.settings.cmake.in @@ -76,6 +76,8 @@ Parallel Filtered Dataset Writes: @PARALLEL_FILTERED_WRITES@ I/O filters (external): @EXTERNAL_FILTERS@ MPE: @H5_HAVE_LIBLMPE@ Direct VFD: @H5_HAVE_DIRECT@ + (Read-Only) S3 VFD: @H5_HAVE_ROS3_VFD@ + (Read-Only) HDFS VFD: @H5_HAVE_LIBHDFS@ dmalloc: @H5_HAVE_LIBDMALLOC@ Packages w/ extra debug output: @INTERNAL_DEBUG_OUTPUT@ API Tracing: @HDF5_ENABLE_TRACE@ diff --git a/configure.ac b/configure.ac index d1d209f..3d6ab10 100644 --- a/configure.ac +++ b/configure.ac @@ -2809,6 +2809,130 @@ fi AM_CONDITIONAL([DIRECT_VFD_CONDITIONAL], [test "X$DIRECT_VFD" = "Xyes"]) ## ---------------------------------------------------------------------- +## Check if Read-Only S3 virtual file driver is enabled by --enable-ros3-vfd +## +AC_SUBST([ROS3_VFD]) + +## Default is no Read-Only S3 VFD +ROS3_VFD=no + +AC_ARG_ENABLE([ros3-vfd], + [AS_HELP_STRING([--enable-ros3-vfd], + [Build the Read-Only S3 virtual file driver (VFD). + [default=no]])], + [ROS3_VFD=$enableval], [ROS3_VFD=no]) + +if test "X$ROS3_VFD" = "Xyes"; then + AC_CHECK_HEADERS([curl/curl.h],, [unset ROS3_VFD]) + AC_CHECK_HEADERS([openssl/evp.h],, [unset ROS3_VFD]) + AC_CHECK_HEADERS([openssl/hmac.h],, [unset ROS3_VFD]) + AC_CHECK_HEADERS([openssl/sha.h],, [unset ROS3_VFD]) + if test "X$ROS3_VFD" = "Xyes"; then + AC_CHECK_LIB([curl], [curl_global_init],, [unset ROS3_VFD]) + AC_CHECK_LIB([crypto], [EVP_sha256],, [unset ROS3_VFD]) + fi + + AC_MSG_CHECKING([if the Read-Only S3 virtual file driver (VFD) is enabled]) + if test "X$ROS3_VFD" = "Xyes"; then + AC_DEFINE([HAVE_ROS3_VFD], [1], + [Define whether the Read-Only S3 virtual file driver (VFD) should be compiled]) + AC_MSG_RESULT([yes]) + else + AC_MSG_RESULT([no]) + ROS3_VFD=no + AC_MSG_ERROR([The Read-Only S3 VFD was requested but cannot be built. + Please check that openssl and cURL are available on your + system, and/or re-configure without option + --enable-ros3-vfd.]) + fi +else + AC_MSG_CHECKING([if the Read-Only S3 virtual file driver (VFD) is enabled]) + AC_MSG_RESULT([no]) + ROS3_VFD=no + +fi +## ---------------------------------------------------------------------- +## Is libhdfs (Hadoop Distributed File System) present? +## It might be specified with the `--with-libhdfs' command-line switch. +## If found, enables the HDFS VFD. +## +AC_SUBST([HAVE_LIBHDFS]) +AC_ARG_WITH([libhdfs], + [AS_HELP_STRING([--with-libhdfs=DIR], + [Provide libhdfs library to enable HDFS virtual file driver (VFD) [default=no]])],, + [withval=no]) + +case $withval in + no) + HAVE_LIBHDFS="no" + AC_MSG_CHECKING([for libhdfs]) + AC_MSG_RESULT([suppressed]) + ;; + *) + HAVE_LIBHDFS="yes" + case "$withval" in + *,*) + libhdfs_inc="`echo $withval |cut -f1 -d,`" + libhdfs_lib="`echo $withval |cut -f2 -d, -s`" + ;; + yes) + libhdfs_inc="$HADOOP_HOME/include" + libhdfs_lib="$HADOOP_HOME/lib" + ;; + *) + if test -n "$withval"; then + libhdfs_inc="$withval/include" + libhdfs_lib="$withval/lib" + fi + ;; + esac + + if test -n "$libhdfs_inc"; then + CPPFLAGS="$CPPFLAGS -I$libhdfs_inc" + AM_CPPFLAGS="$AM_CPPFLAGS -I$libhdfs_inc" + fi + AC_CHECK_HEADERS([hdfs.h],, + [unset HAVE_LIBHDFS]) + + if test "x$HAVE_LIBHDFS" = "xyes"; then + dnl Check for '-ljvm' needed by libhdfs + JNI_LDFLAGS="" + if test $JAVA_HOME != "" + then + JNI_LDFLAGS="-L$JAVA_HOME/jre/lib/$OS_ARCH -L$JAVA_HOME/jre/lib/$OS_ARCH/server" + fi + ldflags_bak=$LDFLAGS + LDFLAGS="$LDFLAGS $JNI_LDFLAGS" + AC_CHECK_LIB([jvm], [JNI_GetCreatedJavaVMs]) + LDFLAGS=$ldflags_bak + AC_SUBST([JNI_LDFLAGS]) + if test -n "$libhdfs_lib"; then + ## Hadoop distribution hides libraries down one level in 'lib/native' + libhdfs_lib="$libhdfs_lib/native" + LDFLAGS="$LDFLAGS -L$libhdfs_lib $JNI_LDFLAGS" + AM_LDFLAGS="$AM_LDFLAGS -L$libhdfs_lib $JNI_LDFLAGS" + fi + AC_CHECK_LIB([hdfs], [hdfsConnect],, + [unset HAVE_LIBHDFS]) + fi + + if test -z "$HAVE_LIBHDFS"; then + AC_MSG_ERROR([Set to use libhdfs library, but could not find or use + libhdfs. Please verify that the path to HADOOP_HOME is + valid, and/or reconfigure without --with-libhdfs.]) + fi + ;; +esac + +if test "x$HAVE_LIBHDFS" = "xyes"; then + AC_DEFINE([HAVE_LIBHDFS], [1], + [Proceed to build with libhdfs]) +fi + +## Checkpoint the cache +AC_CACHE_SAVE + +## ---------------------------------------------------------------------- ## Enable custom plugin default path for library. It requires SHARED support. ## AC_MSG_CHECKING([for custom plugin default path definition]) @@ -3537,6 +3661,7 @@ AC_CONFIG_FILES([src/libhdf5.settings testpar/testpflush.sh tools/Makefile tools/lib/Makefile + tools/libtest/Makefile tools/src/Makefile tools/src/h5dump/Makefile tools/src/h5import/Makefile diff --git a/java/examples/groups/JavaGroupExample.sh.in b/java/examples/groups/JavaGroupExample.sh.in index 9d4673f..3ba512a 100644 --- a/java/examples/groups/JavaGroupExample.sh.in +++ b/java/examples/groups/JavaGroupExample.sh.in @@ -64,6 +64,8 @@ $HDFTEST_HOME/h5ex_g_iterate.h5 $HDFTEST_HOME/h5ex_g_visit.h5 " LIST_DATA_FILES=" +$HDFTEST_HOME/h5ex_g_iterate.h5 +$HDFTEST_HOME/h5ex_g_visit.h5 $HDFTEST_HOME/../testfiles/examples.groups.H5Ex_G_Create.txt $HDFTEST_HOME/../testfiles/examples.groups.H5Ex_G_Iterate.txt $HDFTEST_HOME/../testfiles/examples.groups.H5Ex_G_Compact.txt diff --git a/java/src/Makefile.am b/java/src/Makefile.am index bd55c39..fcdeae9 100644 --- a/java/src/Makefile.am +++ b/java/src/Makefile.am @@ -98,6 +98,8 @@ hdf5_java_JAVA = \ ${pkgpath}/structs/H5A_info_t.java \ ${pkgpath}/structs/H5E_error2_t.java \ ${pkgpath}/structs/H5F_info2_t.java \ + ${pkgpath}/structs/H5FD_hdfs_fapl_t.java \ + ${pkgpath}/structs/H5FD_ros3_fapl_t.java \ ${pkgpath}/structs/H5G_info_t.java \ ${pkgpath}/structs/H5L_info_t.java \ ${pkgpath}/structs/H5O_info_t.java \ diff --git a/java/src/hdf/hdf5lib/CMakeLists.txt b/java/src/hdf/hdf5lib/CMakeLists.txt index c171ea8..be8f60a 100644 --- a/java/src/hdf/hdf5lib/CMakeLists.txt +++ b/java/src/hdf/hdf5lib/CMakeLists.txt @@ -73,6 +73,8 @@ set (HDF5_JAVA_HDF_HDF5_STRUCTS_SOURCES structs/H5AC_cache_config_t.java structs/H5E_error2_t.java structs/H5F_info2_t.java + structs/H5FD_ros3_fapl_t.java + structs/H5FD_hdfs_fapl_t.java structs/H5G_info_t.java structs/H5L_info_t.java structs/H5O_hdr_info_t.java diff --git a/java/src/hdf/hdf5lib/H5.java b/java/src/hdf/hdf5lib/H5.java index a1a8ede..481ca8c 100644 --- a/java/src/hdf/hdf5lib/H5.java +++ b/java/src/hdf/hdf5lib/H5.java @@ -50,6 +50,8 @@ import hdf.hdf5lib.structs.H5AC_cache_config_t; import hdf.hdf5lib.structs.H5A_info_t; import hdf.hdf5lib.structs.H5E_error2_t; import hdf.hdf5lib.structs.H5F_info2_t; +import hdf.hdf5lib.structs.H5FD_hdfs_fapl_t; +import hdf.hdf5lib.structs.H5FD_ros3_fapl_t; import hdf.hdf5lib.structs.H5G_info_t; import hdf.hdf5lib.structs.H5L_info_t; import hdf.hdf5lib.structs.H5O_info_t; @@ -7796,6 +7798,10 @@ public class H5 implements java.io.Serializable { public synchronized static native int H5Pset_fapl_family(long fapl_id, long memb_size, long memb_fapl_id) throws HDF5LibraryException, NullPointerException; + public synchronized static native int H5Pset_fapl_hdfs(long fapl_id, H5FD_hdfs_fapl_t fapl_conf) throws HDF5LibraryException, NullPointerException; + + public synchronized static native H5FD_hdfs_fapl_t H5Pget_fapl_hdfs(long fapl_id) throws HDF5LibraryException, NullPointerException; + /** * H5Pget_fapl_multi Sets up use of the multi I/O driver. * @@ -7880,6 +7886,10 @@ public class H5 implements java.io.Serializable { public synchronized static native int H5Pset_fapl_windows(long fapl_id) throws HDF5LibraryException, NullPointerException; + public synchronized static native int H5Pset_fapl_ros3(long fapl_id, H5FD_ros3_fapl_t fapl_conf) throws HDF5LibraryException, NullPointerException; + + public synchronized static native H5FD_ros3_fapl_t H5Pget_fapl_ros3(long fapl_id) throws HDF5LibraryException, NullPointerException; + // /////// unimplemented //////// // Generic property list routines // diff --git a/java/src/hdf/hdf5lib/HDF5Constants.java b/java/src/hdf/hdf5lib/HDF5Constants.java index cb5ed22..2e80f2e 100644 --- a/java/src/hdf/hdf5lib/HDF5Constants.java +++ b/java/src/hdf/hdf5lib/HDF5Constants.java @@ -246,6 +246,8 @@ public class HDF5Constants { public static final long H5FD_SEC2 = H5FD_SEC2(); public static final long H5FD_STDIO = H5FD_STDIO(); public static final long H5FD_WINDOWS = H5FD_WINDOWS(); + public static final long H5FD_ROS3 = H5FD_ROS3(); + public static final long H5FD_HDFS = H5FD_HDFS(); public static final int H5FD_LOG_LOC_READ = H5FD_LOG_LOC_READ(); public static final int H5FD_LOG_LOC_WRITE = H5FD_LOG_LOC_WRITE(); public static final int H5FD_LOG_LOC_SEEK = H5FD_LOG_LOC_SEEK(); @@ -1111,6 +1113,10 @@ public class HDF5Constants { private static native final long H5FD_WINDOWS(); + private static native final long H5FD_ROS3(); + + private static native final long H5FD_HDFS(); + private static native final int H5FD_LOG_LOC_READ(); private static native final int H5FD_LOG_LOC_WRITE(); diff --git a/java/src/hdf/hdf5lib/structs/H5FD_hdfs_fapl_t.java b/java/src/hdf/hdf5lib/structs/H5FD_hdfs_fapl_t.java new file mode 100644 index 0000000..f56a038 --- /dev/null +++ b/java/src/hdf/hdf5lib/structs/H5FD_hdfs_fapl_t.java @@ -0,0 +1,102 @@ +/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * Read-Only HDFS Virtual File Driver (VFD) * + * Copyright (c) 2018, The HDF Group. * + * * + * All rights reserved. * + * * + * NOTICE: * + * All information contained herein is, and remains, the property of The HDF * + * Group. The intellectual and technical concepts contained herein are * + * proprietary to The HDF Group. Dissemination of this information or * + * reproduction of this material is strictly forbidden unless prior written * + * permission is obtained from The HDF Group. * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ + +package hdf.hdf5lib.structs; + +import java.io.Serializable; + +/* + * Java representation of the HDFS VFD file access property list (fapl) + * structure. + * + * Used for the access of files hosted on the Hadoop Distributed File System. + */ + +@SuppressWarnings("serial") // mute default serialUID warnings until someone knowledgeable comes along or something breaks horribly +public class H5FD_hdfs_fapl_t implements Serializable { + + private long version; + private String namenode_name; + private String user_name; + private String kerberos_ticket_cache; + private int namenode_port; + private int stream_buffer_size; + + /** + * Create a fapl_t structure with the specified components. + */ + public H5FD_hdfs_fapl_t( + String namenode_name, + int namenode_port, + String user_name, + String kerberos_ticket_cache, + int stream_buffer_size) + { + this.version = 1; + this.namenode_name = namenode_name; + this.namenode_port = namenode_port; + this.user_name = user_name; + this.kerberos_ticket_cache = kerberos_ticket_cache; + this.stream_buffer_size = stream_buffer_size; + } + + @Override + public boolean equals(Object o) { + if (o == null) + return false; + if (!(o instanceof H5FD_hdfs_fapl_t)) + return false; + + H5FD_hdfs_fapl_t other = (H5FD_hdfs_fapl_t)o; + if (this.version != other.version) + return false; + if (!this.namenode_name.equals(other.namenode_name)) + return false; + if (this.namenode_port != other.namenode_port) + return false; + if (!this.user_name.equals(other.user_name)) + return false; + if (!this.kerberos_ticket_cache.equals(other.kerberos_ticket_cache)) + return false; + if (this.stream_buffer_size != other.stream_buffer_size) + return false; + return true; + } + + @Override + public int hashCode() { + /* this is a _very bad_ hash algorithm for purposes of hashing! */ + /* implemented to satisfy the "contract" regarding equality */ + int k = (int)this.version; + k += this.namenode_name.length(); + k += this.user_name.length(); + k += this.kerberos_ticket_cache.length(); + k += namenode_port; + k += stream_buffer_size; + return k; + } + + @Override + public String toString() { + return "H5FD_hdfs_fapl_t (Version: " + this.version + ") {" + + "\n namenode_name: '" + this.namenode_name + + "'\n namenode_port: " + this.namenode_port + + "\n user_name: '" + this.user_name + + "'\n kerberos_ticket_cache: '" + this.kerberos_ticket_cache + + "'\n stream_buffer_size: " + this.stream_buffer_size + + "\n}\n"; + } +} + + diff --git a/java/src/hdf/hdf5lib/structs/H5FD_ros3_fapl_t.java b/java/src/hdf/hdf5lib/structs/H5FD_ros3_fapl_t.java new file mode 100644 index 0000000..6b086c3 --- /dev/null +++ b/java/src/hdf/hdf5lib/structs/H5FD_ros3_fapl_t.java @@ -0,0 +1,121 @@ +/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * Read-Only S3 Virtual File Driver (VFD) * + * Copyright (c) 2017-2018, The HDF Group. * + * * + * All rights reserved. * + * * + * NOTICE: * + * All information contained herein is, and remains, the property of The HDF * + * Group. The intellectual and technical concepts contained herein are * + * proprietary to The HDF Group. Dissemination of this information or * + * reproduction of this material is strictly forbidden unless prior written * + * permission is obtained from The HDF Group. * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ + +package hdf.hdf5lib.structs; + +import java.io.Serializable; + +/* + * Java representation of the ROS3 VFD file access property list (fapl) + * structure. + * + * Used for the access of files hosted remotely on S3 by Amazon. + * + * For simplicity, implemented assuming that all ROS3 fapls have components: + * - version + * - aws_region + * - secret_id + * - secret_key + * + * Future implementations may be created to enable different fapl "shapes" + * depending on provided version. + * + * proposed: + * + * H5FD_ros3_fapl_t (super class, has only version field) + * H5FD_ros3_fapl_v1_t (extends super with Version 1 components) + * H5FD_ros3_fapl_v2_t (extends super with Version 2 components) + * and so on, for each version + * + * "super" is passed around, and is version-checked and re-cast as + * appropriate + */ + +@SuppressWarnings("serial") // mute default serialUID warnings until someone knowledgeable comes along or something breaks horribly +public class H5FD_ros3_fapl_t implements Serializable { + + private long version; + private String aws_region; + private String secret_id; + private String secret_key; + + /** + * Create a "default" fapl_t structure, for anonymous access. + */ + public H5FD_ros3_fapl_t () { + /* H5FD_ros3_fapl_t("", "", ""); */ /* defer */ + this.version = 1; + this.aws_region = ""; + this.secret_id = ""; + this.secret_key = ""; + } + + /** + * Create a fapl_t structure with the specified components. + * If all are the empty string, is anonymous (non-authenticating). + * Region and ID must both be supplied for authentication. + * + * @param region "aws region" for authenticating request + * @param id "secret id" or "access id" for authenticating request + * @param key "secret key" or "access key" for authenticating request + */ + public H5FD_ros3_fapl_t (String region, String id, String key) { + this.version = 1; /* must equal H5FD__CURR_ROS3_FAPL_T_VERSION */ + /* as found in H5FDros3.h */ + this.aws_region = region; + this.secret_id = id; + this.secret_key = key; + } + + @Override + public boolean equals(Object o) { + if (o == null) + return false; + if (!(o instanceof H5FD_ros3_fapl_t)) + return false; + + H5FD_ros3_fapl_t other = (H5FD_ros3_fapl_t)o; + if (this.version != other.version) + return false; + if (!this.aws_region.equals(other.aws_region)) + return false; + if (!this.secret_key.equals(other.secret_key)) + return false; + if (!this.secret_id.equals(other.secret_id)) + return false; + return true; + } + + @Override + public int hashCode() { + /* this is a _very bad_ hash algorithm for purposes of hashing! */ + /* implemented to satisfy the "contract" regarding equality */ + int k = (int)this.version; + k += this.aws_region.length(); + k += this.secret_id.length(); + k += this.secret_key.length(); + return k; + } + + @Override + public String toString() { + return "H5FD_ros3_fapl_t (Version:" + this.version + ") {" + + "\n aws_region : " + this.aws_region + + "\n secret_id : " + this.secret_id + + "\n secret_key : " + this.secret_key + + "\n}\n"; + } +} + + diff --git a/java/src/jni/h5Constants.c b/java/src/jni/h5Constants.c index b9e320f..ac55a13 100644 --- a/java/src/jni/h5Constants.c +++ b/java/src/jni/h5Constants.c @@ -453,6 +453,8 @@ Java_hdf_hdf5lib_HDF5Constants_H5FD_1DIRECT(JNIEnv *env, jclass cls) { JNIEXPORT jlong JNICALL Java_hdf_hdf5lib_HDF5Constants_H5FD_1FAMILY(JNIEnv *env, jclass cls) { return H5FD_FAMILY; } JNIEXPORT jlong JNICALL +Java_hdf_hdf5lib_HDF5Constants_H5FD_1HDFS(JNIEnv *env, jclass cls) { return H5FD_HDFS; } +JNIEXPORT jlong JNICALL Java_hdf_hdf5lib_HDF5Constants_H5FD_1LOG(JNIEnv *env, jclass cls) { return H5FD_LOG; } JNIEXPORT jlong JNICALL Java_hdf_hdf5lib_HDF5Constants_H5FD_1MPIO(JNIEnv *env, jclass cls) { return H5FD_MPIO; } @@ -461,6 +463,8 @@ Java_hdf_hdf5lib_HDF5Constants_H5FD_1MULTI(JNIEnv *env, jclass cls) { return H5F JNIEXPORT jlong JNICALL Java_hdf_hdf5lib_HDF5Constants_H5FD_1SEC2(JNIEnv *env, jclass cls) { return H5FD_SEC2; } JNIEXPORT jlong JNICALL +Java_hdf_hdf5lib_HDF5Constants_H5FD_1ROS3(JNIEnv *env, jclass cls) { return H5FD_ROS3; } +JNIEXPORT jlong JNICALL Java_hdf_hdf5lib_HDF5Constants_H5FD_1STDIO(JNIEnv *env, jclass cls) { return H5FD_STDIO; } JNIEXPORT jlong JNICALL Java_hdf_hdf5lib_HDF5Constants_H5FD_1WINDOWS(JNIEnv *env, jclass cls) { diff --git a/java/test/CMakeLists.txt b/java/test/CMakeLists.txt index 3298a47..dfb6e72 100644 --- a/java/test/CMakeLists.txt +++ b/java/test/CMakeLists.txt @@ -51,6 +51,20 @@ if (NOT HDF5_ENABLE_DEBUG_APIS) ) endif () +if (HDF5_ENABLE_ROS3_VFD) + set (HDF5_JAVA_TEST_SOURCES + ${HDF5_JAVA_TEST_SOURCES} + TestH5Pfapls3 + ) +endif () + +if (HDF5_ENABLE_HDFS) + set (HDF5_JAVA_TEST_SOURCES + ${HDF5_JAVA_TEST_SOURCES} + TestH5Pfaplhdfs + ) +endif () + set (CMAKE_JAVA_INCLUDE_PATH "${HDF5_JAVA_LIB_DIR}/junit.jar;${HDF5_JAVA_LIB_DIR}/hamcrest-core.jar;${HDF5_JAVA_JARS};${HDF5_JAVA_LOGGING_JAR};${HDF5_JAVA_LOGGING_SIMPLE_JAR}") foreach (test_file ${HDF5_JAVA_TEST_SOURCES}) diff --git a/java/test/Makefile.am b/java/test/Makefile.am index af99d92..c375938 100644 --- a/java/test/Makefile.am +++ b/java/test/Makefile.am @@ -61,6 +61,8 @@ noinst_JAVA = \ TestH5P.java \ TestH5PData.java \ TestH5Pfapl.java \ + TestH5Pfaplhdfs.java \ + TestH5Pfapls3.java \ TestH5Pvirtual.java \ TestH5Plist.java \ TestH5A.java \ diff --git a/java/test/TestAll.java b/java/test/TestAll.java index 13cb597..c7c206c 100644 --- a/java/test/TestAll.java +++ b/java/test/TestAll.java @@ -27,6 +27,7 @@ import org.junit.runners.Suite; TestH5Lparams.class, TestH5Lbasic.class, TestH5Lcreate.class, TestH5R.class, TestH5P.class, TestH5PData.class, TestH5Pfapl.class, TestH5Pvirtual.class, TestH5Plist.class, + TestH5Pfapls3.class, TestH5Pfaplhdfs.class, TestH5A.class, TestH5Oparams.class, TestH5Obasic.class, TestH5Ocopy.class, TestH5Ocreate.class, TestH5PL.class, TestH5Z.class diff --git a/java/test/TestH5Pfapl.java b/java/test/TestH5Pfapl.java index 10a79dd..81a7ecb 100644 --- a/java/test/TestH5Pfapl.java +++ b/java/test/TestH5Pfapl.java @@ -15,6 +15,7 @@ package test; import static org.junit.Assert.assertEquals; import static org.junit.Assert.assertFalse; +import static org.junit.Assert.assertNull; import static org.junit.Assert.assertTrue; import static org.junit.Assert.fail; @@ -28,6 +29,8 @@ import hdf.hdf5lib.exceptions.HDF5Exception; import hdf.hdf5lib.exceptions.HDF5LibraryException; import hdf.hdf5lib.exceptions.HDF5PropertyListInterfaceException; import hdf.hdf5lib.structs.H5AC_cache_config_t; +import hdf.hdf5lib.structs.H5FD_hdfs_fapl_t; +import hdf.hdf5lib.structs.H5FD_ros3_fapl_t; import org.junit.After; import org.junit.Before; @@ -1398,4 +1401,5 @@ public class TestH5Pfapl { fail("H5P_evict_on_close: " + err); } } + } diff --git a/java/test/TestH5Pfaplhdfs.java b/java/test/TestH5Pfaplhdfs.java new file mode 100644 index 0000000..2b0a808 --- /dev/null +++ b/java/test/TestH5Pfaplhdfs.java @@ -0,0 +1,393 @@ +/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * Copyright by The HDF Group. * + * Copyright by the Board of Trustees of the University of Illinois. * + * All rights reserved. * + * * + * This file is part of HDF5. The full HDF5 copyright notice, including * + * terms governing use, modification, and redistribution, is contained in * + * the COPYING file, which can be found at the root of the source code * + * distribution tree, or in https://support.hdfgroup.org/ftp/HDF5/releases. * + * If you do not have access to either file, you may request a copy from * + * help@hdfgroup.org. * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ + +package test; + +import static org.junit.Assert.assertEquals; +import static org.junit.Assert.assertFalse; +import static org.junit.Assert.assertNull; +import static org.junit.Assert.assertTrue; +import static org.junit.Assert.fail; + +import java.io.File; +import java.text.DecimalFormat; +import java.text.NumberFormat; + +import hdf.hdf5lib.H5; +import hdf.hdf5lib.HDF5Constants; +import hdf.hdf5lib.exceptions.HDF5Exception; +import hdf.hdf5lib.exceptions.HDF5LibraryException; +import hdf.hdf5lib.exceptions.HDF5PropertyListInterfaceException; +import hdf.hdf5lib.structs.H5AC_cache_config_t; +import hdf.hdf5lib.structs.H5FD_hdfs_fapl_t; +import hdf.hdf5lib.structs.H5FD_ros3_fapl_t; + +import org.junit.After; +import org.junit.Before; +import org.junit.Rule; +import org.junit.Test; +import org.junit.rules.TestName; + +public class TestH5Pfaplhdfs { + @Rule public TestName testname = new TestName(); + + private static final String H5_FILE = "testPf.h5"; + private static final String H5_LOG_FILE = "testPf.log"; + private static final String H5_FAMILY_FILE = "testPf%05d"; + private static final String H5_MULTI_FILE = "testPfmulti"; + private static char MULTI_LETTERS[] = {'X','s','b','r','g','l','o'}; + private static final int DIM_X = 4; + private static final int DIM_Y = 6; + private static final int DIMF_X = 12; + private static final int DIMF_Y = 18; + long H5fid = -1; + long H5dsid = -1; + long H5did = -1; + long H5Fdsid = -1; + long H5Fdid = -1; + long[] H5dims = { DIM_X, DIM_Y }; + long fapl_id = -1; + long plapl_id = -1; + long dapl_id = -1; + long plist_id = -1; + long btplist_id = -1; + long[] H5Fdims = { DIMF_X, DIMF_Y }; + double windchillF[][] = + {{36.0, 31.0, 25.0, 19.0, 13.0, 7.0, 1.0, -5.0, -11.0, -16.0, -22.0, -28.0, -34.0, -40.0, -46.0, -52.0, -57.0, -63.0}, + {34.0, 27.0, 21.0, 15.0, 9.0, 3.0, -4.0, -10.0, -16.0, -22.0, -28.0, -35.0, -41.0, -47.0, -53.0, -59.0, -66.0, -72.0}, + {32.0, 25.0, 19.0, 13.0, 6.0, 0.0, -7.0, -13.0, -19.0, -26.0, -32.0, -39.0, -45.0, -51.0, -58.0, -64.0, -71.0, -77.0}, + {30.0, 24.0, 17.0, 11.0, 4.0, -2.0, -9.0, -15.0, -22.0, -29.0, -35.0, -42.0, -48.0, -55.0, -61.0, -68.0, -74.0, -81.0}, + {29.0, 23.0, 16.0, 9.0, 3.0, -4.0, -11.0, -17.0, -24.0, -31.0, -37.0, -44.0, -51.0, -58.0, -64.0, -71.0, -78.0, -84.0}, + {28.0, 22.0, 15.0, 8.0, 1.0, -5.0, -12.0, -19.0, -26.0, -33.0, -39.0, -46.0, -53.0, -60.0, -67.0, -73.0, -80.0, -87.0}, + {28.0, 21.0, 14.0, 7.0, 0.0, -7.0, -14.0, -21.0, -27.0, -34.0, -41.0, -48.0, -55.0, -62.0, -69.0, -76.0, -82.0, -89.0}, + {27.0, 20.0, 13.0, 6.0, -1.0, -8.0, -15.0, -22.0, -29.0, -36.0, -43.0, -50.0, -57.0, -64.0, -71.0, -78.0, -84.0, -91.0}, + {26.0, 19.0, 12.0, 5.0, -2.0, -9.0, -16.0, -23.0, -30.0, -37.0, -44.0, -51.0, -58.0, -65.0, -72.0, -79.0, -86.0, -93.0}, + {26.0, 19.0, 12.0, 4.0, -3.0, -10.0, -17.0, -24.0, -31.0, -38.0, -45.0, -52.0, -60.0, -67.0, -74.0, -81.0, -88.0, -95.0}, + {25.0, 18.0, 11.0, 4.0, -3.0, -11.0, -18.0, -25.0, -32.0, -39.0, -46.0, -54.0, -61.0, -68.0, -75.0, -82.0, -89.0, -97.0}, + {25.0, 17.0, 10.0, 3.0, -4.0, -11.0, -19.0, -26.0, -33.0, -40.0, -48.0, -55.0, -62.0, -69.0, -76.0, -84.0, -91.0, -98.0} + }; + + private final void _deleteFile(String filename) { + File file = null; + try { + file = new File(filename); + } + catch (Throwable err) {} + + if (file.exists()) { + try {file.delete();} catch (SecurityException e) {} + } + } + + private final void _deleteLogFile() { + File file = null; + try { + file = new File(H5_LOG_FILE); + } + catch (Throwable err) {} + + if (file.exists()) { + try {file.delete();} catch (SecurityException e) {} + } + } + + private final void _deleteFamilyFile() { + File file = null; + for(int indx = 0; ;indx++) { + java.text.DecimalFormat myFormat = new java.text.DecimalFormat("00000"); + try { + file = new File("test"+myFormat.format(new Integer(indx))+".h5"); + } + catch (Throwable err) {} + + if (file.exists()) { + try {file.delete();} catch (SecurityException e) {} + } + else + return; + } + } + + private final void _deleteMultiFile() { + File file = null; + for(int indx = 1;indx<7;indx++) { + try { + file = new File(H5_MULTI_FILE+"-"+MULTI_LETTERS[indx]+".h5"); + } + catch (Throwable err) {} + + if (file.exists()) { + try {file.delete();} catch (SecurityException e) {} + } + } + } + + private final long _createDataset(long fid, long dsid, String name, long dapl) { + long did = -1; + try { + did = H5.H5Dcreate(fid, name, HDF5Constants.H5T_STD_I32BE, dsid, + HDF5Constants.H5P_DEFAULT, HDF5Constants.H5P_DEFAULT, dapl); + } + catch (Throwable err) { + err.printStackTrace(); + fail("H5.H5Dcreate: " + err); + } + assertTrue("TestH5Pfapl._createDataset: ", did > 0); + + return did; + } + + private final void _createFloatDataset() { + try { + H5Fdsid = H5.H5Screate_simple(2, H5Fdims, null); + H5Fdid = H5.H5Dcreate(H5fid, "dsfloat", HDF5Constants.H5T_NATIVE_FLOAT, H5Fdsid, + HDF5Constants.H5P_DEFAULT, HDF5Constants.H5P_DEFAULT, HDF5Constants.H5P_DEFAULT); + } + catch (Throwable err) { + err.printStackTrace(); + fail("H5.H5Dcreate: " + err); + } + assertTrue("TestH5Pfapl._createFloatDataset: ", H5Fdid > 0); + + try { + H5.H5Fflush(H5fid, HDF5Constants.H5F_SCOPE_LOCAL); + } + catch (Throwable err) { + err.printStackTrace(); + } + } + + private final void _createH5multiFileDS() { + try { + H5did = _createDataset(H5fid, H5dsid, "dset", HDF5Constants.H5P_DEFAULT); + } + catch (Throwable err) { + err.printStackTrace(); + fail("TestH5Pfapl.createH5file: " + err); + } + assertTrue("TestH5Pfapl.createH5file: _createDataset: ", H5did > 0); + + try { + H5.H5Fflush(H5fid, HDF5Constants.H5F_SCOPE_LOCAL); + } + catch (Throwable err) { + err.printStackTrace(); + } + } + + private final void _createH5File(long fapl) { + try { + H5fid = H5.H5Fcreate(H5_FILE, HDF5Constants.H5F_ACC_TRUNC, + HDF5Constants.H5P_DEFAULT, fapl); + H5dsid = H5.H5Screate_simple(2, H5dims, null); + H5did = _createDataset(H5fid, H5dsid, "dset", HDF5Constants.H5P_DEFAULT); + } + catch (Throwable err) { + err.printStackTrace(); + fail("TestH5Pfapl.createH5file: " + err); + } + assertTrue("TestH5Pfapl.createH5file: H5.H5Fcreate: ", H5fid > 0); + assertTrue("TestH5Pfapl.createH5file: H5.H5Screate_simple: ", H5dsid > 0); + assertTrue("TestH5Pfapl.createH5file: _createDataset: ", H5did > 0); + + try { + H5.H5Fflush(H5fid, HDF5Constants.H5F_SCOPE_LOCAL); + } + catch (Throwable err) { + err.printStackTrace(); + } + } + + private final void _createH5familyFile(long fapl) { + try { + H5fid = H5.H5Fcreate(H5_FAMILY_FILE+".h5", HDF5Constants.H5F_ACC_TRUNC, + HDF5Constants.H5P_DEFAULT, fapl); + H5dsid = H5.H5Screate_simple(2, H5dims, null); + H5did = _createDataset(H5fid, H5dsid, "dset", HDF5Constants.H5P_DEFAULT); + } + catch (Throwable err) { + err.printStackTrace(); + fail("TestH5Pfapl.createH5file: " + err); + } + assertTrue("TestH5Pfapl.createH5file: H5.H5Fcreate: ", H5fid > 0); + assertTrue("TestH5Pfapl.createH5file: H5.H5Screate_simple: ", H5dsid > 0); + assertTrue("TestH5Pfapl.createH5file: _createDataset: ", H5did > 0); + + try { + H5.H5Fflush(H5fid, HDF5Constants.H5F_SCOPE_LOCAL); + } + catch (Throwable err) { + err.printStackTrace(); + } + } + + private final void _createH5multiFile(long fapl) { + try { + H5fid = H5.H5Fcreate(H5_MULTI_FILE, HDF5Constants.H5F_ACC_TRUNC, + HDF5Constants.H5P_DEFAULT, fapl); + H5dsid = H5.H5Screate_simple(2, H5dims, null); + } + catch (Throwable err) { + err.printStackTrace(); + fail("TestH5Pfapl.createH5file: " + err); + } + assertTrue("TestH5Pfapl.createH5file: H5.H5Fcreate: ", H5fid > 0); + assertTrue("TestH5Pfapl.createH5file: H5.H5Screate_simple: ", H5dsid > 0); + + try { + H5.H5Fflush(H5fid, HDF5Constants.H5F_SCOPE_LOCAL); + } + catch (Throwable err) { + err.printStackTrace(); + } + } + + public void deleteH5file() { + _deleteFile(H5_FILE); + } + + public void deleteH5familyfile() { + _deleteFamilyFile(); + } + + public void deleteH5multifile() { + _deleteMultiFile(); + } + + @Before + public void createFileAccess() + throws NullPointerException, HDF5Exception { + assertTrue("H5 open ids is 0",H5.getOpenIDCount()==0); + System.out.print(testname.getMethodName()); + + try { + fapl_id = H5.H5Pcreate(HDF5Constants.H5P_FILE_ACCESS); + } + catch (Throwable err) { + err.printStackTrace(); + fail("TestH5Pfapl.createFileAccess: " + err); + } + assertTrue(fapl_id > 0); + try { + plapl_id = H5.H5Pcreate(HDF5Constants.H5P_LINK_ACCESS); + } + catch (Throwable err) { + err.printStackTrace(); + fail("TestH5Pfapl.createFileAccess: " + err); + } + assertTrue(plapl_id > 0); + try { + plist_id = H5.H5Pcreate(HDF5Constants.H5P_DATASET_XFER); + btplist_id = H5.H5Pcreate(HDF5Constants.H5P_DATASET_XFER); + dapl_id = H5.H5Pcreate(HDF5Constants.H5P_DATASET_ACCESS); + } + catch (Throwable err) { + err.printStackTrace(); + fail("TestH5Pfapl.createFileAccess: " + err); + } + assertTrue(plist_id > 0); + assertTrue(btplist_id > 0); + assertTrue(dapl_id > 0); + } + + @After + public void deleteFileAccess() throws HDF5LibraryException { + if (fapl_id > 0) + try {H5.H5Pclose(fapl_id);} catch (Exception ex) {} + if (plapl_id > 0) + try {H5.H5Pclose(plapl_id);} catch (Exception ex) {} + if (dapl_id > 0) + try {H5.H5Pclose(dapl_id);} catch (Exception ex) {} + if (plist_id > 0) + try {H5.H5Pclose(plist_id);} catch (Exception ex) {} + if (btplist_id > 0) + try {H5.H5Pclose(btplist_id);} catch (Exception ex) {} + + if (H5Fdsid > 0) + try {H5.H5Sclose(H5Fdsid);} catch (Exception ex) {} + if (H5Fdid > 0) + try {H5.H5Dclose(H5Fdid);} catch (Exception ex) {} + if (H5dsid > 0) + try {H5.H5Sclose(H5dsid);} catch (Exception ex) {} + if (H5did > 0) + try {H5.H5Dclose(H5did);} catch (Exception ex) {} + if (H5fid > 0) + try {H5.H5Fclose(H5fid);} catch (Exception ex) {} + System.out.println(); + } + + @Test + public void testHDFS_fapl() + throws Exception + { + if (HDF5Constants.H5FD_HDFS < 0) + throw new HDF5LibraryException("skip"); + + String nodename = "blues"; + int nodeport = 12345; + String username = "sparticus"; + String kerbcache = "/dev/null"; + int streamsize = 1024; + + final H5FD_hdfs_fapl_t config = new H5FD_hdfs_fapl_t( + nodename, + nodeport, + username, + kerbcache, + streamsize + ); + assertTrue("setting fapl should succeed", + -1 < H5.H5Pset_fapl_hdfs(fapl_id, config)); + + assertEquals("driver types should match", + HDF5Constants.H5FD_HDFS, + H5.H5Pget_driver(fapl_id)); + + H5FD_hdfs_fapl_t copy = H5.H5Pget_fapl_hdfs(fapl_id); + assertEquals("fapl contents should match", + new H5FD_hdfs_fapl_t( + nodename, + nodeport, + username, + kerbcache, + streamsize), + copy); + } + + @Test(expected = HDF5LibraryException.class) + public void testH5Pget_fapl_hdfs_invalid_fapl_id() + throws Exception + { + if (HDF5Constants.H5FD_HDFS < 0) + throw new HDF5LibraryException("skip"); + H5FD_hdfs_fapl_t fails = H5.H5Pget_fapl_hdfs(-1); + } + + @Test(expected = HDF5LibraryException.class) + public void testH5Pget_fapl_hdfs_fapl_id_of_wrong_driver_type() + throws Exception + { + if (HDF5Constants.H5FD_HDFS < 0) + throw new HDF5LibraryException("skip"); + if (HDF5Constants.H5FD_SEC2 < 0 ) + throw new HDF5LibraryException("skip"); + /* TODO: for now, test against a sec2 fapl only */ + + H5.H5Pset_fapl_sec2(fapl_id); + assertEquals("fapl_id was not set properly", + HDF5Constants.H5FD_SEC2, + H5.H5Pget_driver(fapl_id)); + H5FD_hdfs_fapl_t fails = H5.H5Pget_fapl_hdfs(fapl_id); + } + +} diff --git a/java/test/TestH5Pfapls3.java b/java/test/TestH5Pfapls3.java new file mode 100644 index 0000000..3107bc8 --- /dev/null +++ b/java/test/TestH5Pfapls3.java @@ -0,0 +1,406 @@ +/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * Copyright by The HDF Group. * + * Copyright by the Board of Trustees of the University of Illinois. * + * All rights reserved. * + * * + * This file is part of HDF5. The full HDF5 copyright notice, including * + * terms governing use, modification, and redistribution, is contained in * + * the COPYING file, which can be found at the root of the source code * + * distribution tree, or in https://support.hdfgroup.org/ftp/HDF5/releases. * + * If you do not have access to either file, you may request a copy from * + * help@hdfgroup.org. * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ + +package test; + +import static org.junit.Assert.assertEquals; +import static org.junit.Assert.assertFalse; +import static org.junit.Assert.assertNull; +import static org.junit.Assert.assertTrue; +import static org.junit.Assert.fail; + +import java.io.File; +import java.text.DecimalFormat; +import java.text.NumberFormat; + +import hdf.hdf5lib.H5; +import hdf.hdf5lib.HDF5Constants; +import hdf.hdf5lib.exceptions.HDF5Exception; +import hdf.hdf5lib.exceptions.HDF5LibraryException; +import hdf.hdf5lib.exceptions.HDF5PropertyListInterfaceException; +import hdf.hdf5lib.structs.H5AC_cache_config_t; +import hdf.hdf5lib.structs.H5FD_hdfs_fapl_t; +import hdf.hdf5lib.structs.H5FD_ros3_fapl_t; + +import org.junit.After; +import org.junit.Before; +import org.junit.Rule; +import org.junit.Test; +import org.junit.rules.TestName; + +public class TestH5Pfapls3 { + @Rule public TestName testname = new TestName(); + + private static final String H5_FILE = "testPf.h5"; + private static final String H5_LOG_FILE = "testPf.log"; + private static final String H5_FAMILY_FILE = "testPf%05d"; + private static final String H5_MULTI_FILE = "testPfmulti"; + private static char MULTI_LETTERS[] = {'X','s','b','r','g','l','o'}; + private static final int DIM_X = 4; + private static final int DIM_Y = 6; + private static final int DIMF_X = 12; + private static final int DIMF_Y = 18; + long H5fid = -1; + long H5dsid = -1; + long H5did = -1; + long H5Fdsid = -1; + long H5Fdid = -1; + long[] H5dims = { DIM_X, DIM_Y }; + long fapl_id = -1; + long plapl_id = -1; + long dapl_id = -1; + long plist_id = -1; + long btplist_id = -1; + long[] H5Fdims = { DIMF_X, DIMF_Y }; + double windchillF[][] = + {{36.0, 31.0, 25.0, 19.0, 13.0, 7.0, 1.0, -5.0, -11.0, -16.0, -22.0, -28.0, -34.0, -40.0, -46.0, -52.0, -57.0, -63.0}, + {34.0, 27.0, 21.0, 15.0, 9.0, 3.0, -4.0, -10.0, -16.0, -22.0, -28.0, -35.0, -41.0, -47.0, -53.0, -59.0, -66.0, -72.0}, + {32.0, 25.0, 19.0, 13.0, 6.0, 0.0, -7.0, -13.0, -19.0, -26.0, -32.0, -39.0, -45.0, -51.0, -58.0, -64.0, -71.0, -77.0}, + {30.0, 24.0, 17.0, 11.0, 4.0, -2.0, -9.0, -15.0, -22.0, -29.0, -35.0, -42.0, -48.0, -55.0, -61.0, -68.0, -74.0, -81.0}, + {29.0, 23.0, 16.0, 9.0, 3.0, -4.0, -11.0, -17.0, -24.0, -31.0, -37.0, -44.0, -51.0, -58.0, -64.0, -71.0, -78.0, -84.0}, + {28.0, 22.0, 15.0, 8.0, 1.0, -5.0, -12.0, -19.0, -26.0, -33.0, -39.0, -46.0, -53.0, -60.0, -67.0, -73.0, -80.0, -87.0}, + {28.0, 21.0, 14.0, 7.0, 0.0, -7.0, -14.0, -21.0, -27.0, -34.0, -41.0, -48.0, -55.0, -62.0, -69.0, -76.0, -82.0, -89.0}, + {27.0, 20.0, 13.0, 6.0, -1.0, -8.0, -15.0, -22.0, -29.0, -36.0, -43.0, -50.0, -57.0, -64.0, -71.0, -78.0, -84.0, -91.0}, + {26.0, 19.0, 12.0, 5.0, -2.0, -9.0, -16.0, -23.0, -30.0, -37.0, -44.0, -51.0, -58.0, -65.0, -72.0, -79.0, -86.0, -93.0}, + {26.0, 19.0, 12.0, 4.0, -3.0, -10.0, -17.0, -24.0, -31.0, -38.0, -45.0, -52.0, -60.0, -67.0, -74.0, -81.0, -88.0, -95.0}, + {25.0, 18.0, 11.0, 4.0, -3.0, -11.0, -18.0, -25.0, -32.0, -39.0, -46.0, -54.0, -61.0, -68.0, -75.0, -82.0, -89.0, -97.0}, + {25.0, 17.0, 10.0, 3.0, -4.0, -11.0, -19.0, -26.0, -33.0, -40.0, -48.0, -55.0, -62.0, -69.0, -76.0, -84.0, -91.0, -98.0} + }; + + private final void _deleteFile(String filename) { + File file = null; + try { + file = new File(filename); + } + catch (Throwable err) {} + + if (file.exists()) { + try {file.delete();} catch (SecurityException e) {} + } + } + + private final void _deleteLogFile() { + File file = null; + try { + file = new File(H5_LOG_FILE); + } + catch (Throwable err) {} + + if (file.exists()) { + try {file.delete();} catch (SecurityException e) {} + } + } + + private final void _deleteFamilyFile() { + File file = null; + for(int indx = 0; ;indx++) { + java.text.DecimalFormat myFormat = new java.text.DecimalFormat("00000"); + try { + file = new File("test"+myFormat.format(new Integer(indx))+".h5"); + } + catch (Throwable err) {} + + if (file.exists()) { + try {file.delete();} catch (SecurityException e) {} + } + else + return; + } + } + + private final void _deleteMultiFile() { + File file = null; + for(int indx = 1;indx<7;indx++) { + try { + file = new File(H5_MULTI_FILE+"-"+MULTI_LETTERS[indx]+".h5"); + } + catch (Throwable err) {} + + if (file.exists()) { + try {file.delete();} catch (SecurityException e) {} + } + } + } + + private final long _createDataset(long fid, long dsid, String name, long dapl) { + long did = -1; + try { + did = H5.H5Dcreate(fid, name, HDF5Constants.H5T_STD_I32BE, dsid, + HDF5Constants.H5P_DEFAULT, HDF5Constants.H5P_DEFAULT, dapl); + } + catch (Throwable err) { + err.printStackTrace(); + fail("H5.H5Dcreate: " + err); + } + assertTrue("TestH5Pfapl._createDataset: ", did > 0); + + return did; + } + + private final void _createFloatDataset() { + try { + H5Fdsid = H5.H5Screate_simple(2, H5Fdims, null); + H5Fdid = H5.H5Dcreate(H5fid, "dsfloat", HDF5Constants.H5T_NATIVE_FLOAT, H5Fdsid, + HDF5Constants.H5P_DEFAULT, HDF5Constants.H5P_DEFAULT, HDF5Constants.H5P_DEFAULT); + } + catch (Throwable err) { + err.printStackTrace(); + fail("H5.H5Dcreate: " + err); + } + assertTrue("TestH5Pfapl._createFloatDataset: ", H5Fdid > 0); + + try { + H5.H5Fflush(H5fid, HDF5Constants.H5F_SCOPE_LOCAL); + } + catch (Throwable err) { + err.printStackTrace(); + } + } + + private final void _createH5multiFileDS() { + try { + H5did = _createDataset(H5fid, H5dsid, "dset", HDF5Constants.H5P_DEFAULT); + } + catch (Throwable err) { + err.printStackTrace(); + fail("TestH5Pfapl.createH5file: " + err); + } + assertTrue("TestH5Pfapl.createH5file: _createDataset: ", H5did > 0); + + try { + H5.H5Fflush(H5fid, HDF5Constants.H5F_SCOPE_LOCAL); + } + catch (Throwable err) { + err.printStackTrace(); + } + } + + private final void _createH5File(long fapl) { + try { + H5fid = H5.H5Fcreate(H5_FILE, HDF5Constants.H5F_ACC_TRUNC, + HDF5Constants.H5P_DEFAULT, fapl); + H5dsid = H5.H5Screate_simple(2, H5dims, null); + H5did = _createDataset(H5fid, H5dsid, "dset", HDF5Constants.H5P_DEFAULT); + } + catch (Throwable err) { + err.printStackTrace(); + fail("TestH5Pfapl.createH5file: " + err); + } + assertTrue("TestH5Pfapl.createH5file: H5.H5Fcreate: ", H5fid > 0); + assertTrue("TestH5Pfapl.createH5file: H5.H5Screate_simple: ", H5dsid > 0); + assertTrue("TestH5Pfapl.createH5file: _createDataset: ", H5did > 0); + + try { + H5.H5Fflush(H5fid, HDF5Constants.H5F_SCOPE_LOCAL); + } + catch (Throwable err) { + err.printStackTrace(); + } + } + + private final void _createH5familyFile(long fapl) { + try { + H5fid = H5.H5Fcreate(H5_FAMILY_FILE+".h5", HDF5Constants.H5F_ACC_TRUNC, + HDF5Constants.H5P_DEFAULT, fapl); + H5dsid = H5.H5Screate_simple(2, H5dims, null); + H5did = _createDataset(H5fid, H5dsid, "dset", HDF5Constants.H5P_DEFAULT); + } + catch (Throwable err) { + err.printStackTrace(); + fail("TestH5Pfapl.createH5file: " + err); + } + assertTrue("TestH5Pfapl.createH5file: H5.H5Fcreate: ", H5fid > 0); + assertTrue("TestH5Pfapl.createH5file: H5.H5Screate_simple: ", H5dsid > 0); + assertTrue("TestH5Pfapl.createH5file: _createDataset: ", H5did > 0); + + try { + H5.H5Fflush(H5fid, HDF5Constants.H5F_SCOPE_LOCAL); + } + catch (Throwable err) { + err.printStackTrace(); + } + } + + private final void _createH5multiFile(long fapl) { + try { + H5fid = H5.H5Fcreate(H5_MULTI_FILE, HDF5Constants.H5F_ACC_TRUNC, + HDF5Constants.H5P_DEFAULT, fapl); + H5dsid = H5.H5Screate_simple(2, H5dims, null); + } + catch (Throwable err) { + err.printStackTrace(); + fail("TestH5Pfapl.createH5file: " + err); + } + assertTrue("TestH5Pfapl.createH5file: H5.H5Fcreate: ", H5fid > 0); + assertTrue("TestH5Pfapl.createH5file: H5.H5Screate_simple: ", H5dsid > 0); + + try { + H5.H5Fflush(H5fid, HDF5Constants.H5F_SCOPE_LOCAL); + } + catch (Throwable err) { + err.printStackTrace(); + } + } + + public void deleteH5file() { + _deleteFile(H5_FILE); + } + + public void deleteH5familyfile() { + _deleteFamilyFile(); + } + + public void deleteH5multifile() { + _deleteMultiFile(); + } + + @Before + public void createFileAccess() + throws NullPointerException, HDF5Exception { + assertTrue("H5 open ids is 0",H5.getOpenIDCount()==0); + System.out.print(testname.getMethodName()); + + try { + fapl_id = H5.H5Pcreate(HDF5Constants.H5P_FILE_ACCESS); + } + catch (Throwable err) { + err.printStackTrace(); + fail("TestH5Pfapl.createFileAccess: " + err); + } + assertTrue(fapl_id > 0); + try { + plapl_id = H5.H5Pcreate(HDF5Constants.H5P_LINK_ACCESS); + } + catch (Throwable err) { + err.printStackTrace(); + fail("TestH5Pfapl.createFileAccess: " + err); + } + assertTrue(plapl_id > 0); + try { + plist_id = H5.H5Pcreate(HDF5Constants.H5P_DATASET_XFER); + btplist_id = H5.H5Pcreate(HDF5Constants.H5P_DATASET_XFER); + dapl_id = H5.H5Pcreate(HDF5Constants.H5P_DATASET_ACCESS); + } + catch (Throwable err) { + err.printStackTrace(); + fail("TestH5Pfapl.createFileAccess: " + err); + } + assertTrue(plist_id > 0); + assertTrue(btplist_id > 0); + assertTrue(dapl_id > 0); + } + + @After + public void deleteFileAccess() throws HDF5LibraryException { + if (fapl_id > 0) + try {H5.H5Pclose(fapl_id);} catch (Exception ex) {} + if (plapl_id > 0) + try {H5.H5Pclose(plapl_id);} catch (Exception ex) {} + if (dapl_id > 0) + try {H5.H5Pclose(dapl_id);} catch (Exception ex) {} + if (plist_id > 0) + try {H5.H5Pclose(plist_id);} catch (Exception ex) {} + if (btplist_id > 0) + try {H5.H5Pclose(btplist_id);} catch (Exception ex) {} + + if (H5Fdsid > 0) + try {H5.H5Sclose(H5Fdsid);} catch (Exception ex) {} + if (H5Fdid > 0) + try {H5.H5Dclose(H5Fdid);} catch (Exception ex) {} + if (H5dsid > 0) + try {H5.H5Sclose(H5dsid);} catch (Exception ex) {} + if (H5did > 0) + try {H5.H5Dclose(H5did);} catch (Exception ex) {} + if (H5fid > 0) + try {H5.H5Fclose(H5fid);} catch (Exception ex) {} + System.out.println(); + } + + @Test + public void testH5Pset_fapl_ros3() + throws Exception + { + if (HDF5Constants.H5FD_ROS3 < 0) + return; + + final H5FD_ros3_fapl_t config = new H5FD_ros3_fapl_t(); + assertEquals("Default fapl has unexpected contents", + new H5FD_ros3_fapl_t("", "", ""), + config); + + H5.H5Pset_fapl_ros3(fapl_id, config); + + assertEquals("driver types don't match", + HDF5Constants.H5FD_ROS3, + H5.H5Pget_driver(fapl_id)); + + /* get_fapl_ros3 can throw exception in error cases */ + H5FD_ros3_fapl_t copy = H5.H5Pget_fapl_ros3(fapl_id); + assertEquals("contents of fapl set and get don't match", + new H5FD_ros3_fapl_t("", "", ""), + copy); + } + + @Test(expected = HDF5LibraryException.class) + public void testH5Pget_fapl_ros3_invalid_fapl_id() + throws Exception + { + if (HDF5Constants.H5FD_ROS3 < 0) + throw new HDF5LibraryException("skip"); + H5FD_ros3_fapl_t fails = H5.H5Pget_fapl_ros3(-1); + } + + @Test(expected = HDF5LibraryException.class) + public void testH5Pget_fapl_ros3_fapl_id_of_wrong_driver_type() + throws Exception + { + if (HDF5Constants.H5FD_ROS3 < 0) + throw new HDF5LibraryException("skip"); + if (HDF5Constants.H5FD_SEC2 < 0 ) + throw new HDF5LibraryException("skip"); + /* TODO: for now, test against a sec2 fapl only */ + + H5.H5Pset_fapl_sec2(fapl_id); + assertEquals("fapl_id was not set properly", + HDF5Constants.H5FD_SEC2, + H5.H5Pget_driver(fapl_id)); + H5FD_ros3_fapl_t fails = H5.H5Pget_fapl_ros3(fapl_id); + } + + @Test + public void testH5Pset_fapl_ros3_specified() + throws Exception + { + if (HDF5Constants.H5FD_ROS3 < 0) + return; + + String region = "us-east-1"; + String acc_id = "my_access_id"; + String acc_key = "my_access_key"; + + final H5FD_ros3_fapl_t config = new H5FD_ros3_fapl_t( + region, + acc_id, + acc_key); + H5.H5Pset_fapl_ros3(fapl_id, config); + assertEquals("driver types don't match", + HDF5Constants.H5FD_ROS3, + H5.H5Pget_driver(fapl_id)); + + H5FD_ros3_fapl_t copy = H5.H5Pget_fapl_ros3(fapl_id); + assertEquals("contents of fapl set and get don't match", + new H5FD_ros3_fapl_t(region, acc_id, acc_key), + copy); + } + +} diff --git a/java/test/junit.sh.in b/java/test/junit.sh.in index 7fb1bae..008c89c 100644 --- a/java/test/junit.sh.in +++ b/java/test/junit.sh.in @@ -18,6 +18,8 @@ srcdir=@srcdir@ USE_FILTER_SZIP="@USE_FILTER_SZIP@" USE_FILTER_DEFLATE="@USE_FILTER_DEFLATE@" +USE_ROS3_VFD="@HAVE_ROS3_VFD@" +USE_HDFS_VFD="@HAVE_LIBHDFS@" TESTNAME=JUnitInterface EXIT_SUCCESS=0 @@ -93,6 +95,8 @@ $HDFTEST_HOME/testfiles/JUnit-TestH5R.txt $HDFTEST_HOME/testfiles/JUnit-TestH5P.txt $HDFTEST_HOME/testfiles/JUnit-TestH5PData.txt $HDFTEST_HOME/testfiles/JUnit-TestH5Pfapl.txt +$HDFTEST_HOME/testfiles/JUnit-TestH5Pfapls3.txt +$HDFTEST_HOME/testfiles/JUnit-TestH5Pfaplhdfs.txt $HDFTEST_HOME/testfiles/JUnit-TestH5Pvirtual.txt $HDFTEST_HOME/testfiles/JUnit-TestH5Plist.txt $HDFTEST_HOME/testfiles/JUnit-TestH5A.txt @@ -1096,6 +1100,50 @@ if test $USE_FILTER_SZIP = "yes"; then test yes = "$verbose" && $DIFF JUnit-TestH5Giterate.txt JUnit-TestH5Giterate.out |sed 's/^/ /' fi fi +if test $ROS3_VFD = "yes"; then + echo "$JAVAEXE $JAVAEXEFLAGS -Xmx1024M -Dorg.slf4j.simpleLogger.defaultLog=trace -Djava.library.path=$BLDLIBDIR -cp $CLASSPATH -ea org.junit.runner.JUnitCore test.TestH5Pfapls3" + TESTING JUnit-TestH5Pfapls3 + ($RUNSERIAL $JAVAEXE $JAVAEXEFLAGS -Xmx1024M -Dorg.slf4j.simpleLogger.defaultLog=trace -Djava.library.path=$BLDLIBDIR -cp $CLASSPATH -ea org.junit.runner.JUnitCore test.TestH5Pfapls3 > JUnit-TestH5Pfapls3.ext) + + # Extract file name, line number, version and thread IDs because they may be different + sed -e 's/thread [0-9]*/thread (IDs)/' -e 's/: .*\.c /: (file name) /' \ + -e 's/line [0-9]*/line (number)/' \ + -e 's/Time: [0-9]*\.[0-9]*/Time: XXXX/' \ + -e 's/v[1-9]*\.[0-9]*\./version (number)\./' \ + -e 's/[1-9]*\.[0-9]*\.[0-9]*[^)]*/version (number)/' \ + JUnit-TestH5Pfapls3.ext > JUnit-TestH5Pfapls3.out + + if diff JUnit-TestH5Pfapls3.out JUnit-TestH5Pfapls3.txt > /dev/null; then + echo " PASSED JUnit-TestH5Pfapls3" + else + echo "**FAILED** JUnit-TestH5Pfapls3" + echo " Expected result differs from actual result" + nerrors="`expr $nerrors + 1`" + test yes = "$verbose" && $DIFF JUnit-TestH5Pfapls3.txt JUnit-TestH5Pfapls3.out |sed 's/^/ /' + fi +fi +if test $HAVE_LIBHDFS = "yes"; then + echo "$JAVAEXE $JAVAEXEFLAGS -Xmx1024M -Dorg.slf4j.simpleLogger.defaultLog=trace -Djava.library.path=$BLDLIBDIR -cp $CLASSPATH -ea org.junit.runner.JUnitCore test.TestH5Pfaplhdfs" + TESTING JUnit-TestH5Pfaplhdfs + ($RUNSERIAL $JAVAEXE $JAVAEXEFLAGS -Xmx1024M -Dorg.slf4j.simpleLogger.defaultLog=trace -Djava.library.path=$BLDLIBDIR -cp $CLASSPATH -ea org.junit.runner.JUnitCore test.TestH5Pfaplhdfs > JUnit-TestH5Pfaplhdfs.ext) + + # Extract file name, line number, version and thread IDs because they may be different + sed -e 's/thread [0-9]*/thread (IDs)/' -e 's/: .*\.c /: (file name) /' \ + -e 's/line [0-9]*/line (number)/' \ + -e 's/Time: [0-9]*\.[0-9]*/Time: XXXX/' \ + -e 's/v[1-9]*\.[0-9]*\./version (number)\./' \ + -e 's/[1-9]*\.[0-9]*\.[0-9]*[^)]*/version (number)/' \ + JUnit-TestH5Pfaplhdfs.ext > JUnit-TestH5Pfaplhdfs.out + + if diff JUnit-TestH5Pfaplhdfs.out JUnit-TestH5Pfaplhdfs.txt > /dev/null; then + echo " PASSED JUnit-TestH5Pfaplhdfs" + else + echo "**FAILED** JUnit-TestH5Pfaplhdfs" + echo " Expected result differs from actual result" + nerrors="`expr $nerrors + 1`" + test yes = "$verbose" && $DIFF JUnit-TestH5Pfaplhdfs.txt JUnit-TestH5Pfaplhdfs.out |sed 's/^/ /' + fi +fi # Clean up temporary files/directories diff --git a/java/test/testfiles/JUnit-TestH5Pfaplhdfs.txt b/java/test/testfiles/JUnit-TestH5Pfaplhdfs.txt new file mode 100644 index 0000000..47a00a4 --- /dev/null +++ b/java/test/testfiles/JUnit-TestH5Pfaplhdfs.txt @@ -0,0 +1,9 @@ +JUnit version 4.11 +.testH5Pget_fapl_hdfs_invalid_fapl_id +.testH5Pget_fapl_hdfs_fapl_id_of_wrong_driver_type +.testHDFS_fapl + +Time: XXXX + +OK (3 tests) + diff --git a/java/test/testfiles/JUnit-TestH5Pfapls3.txt b/java/test/testfiles/JUnit-TestH5Pfapls3.txt new file mode 100644 index 0000000..3f46342 --- /dev/null +++ b/java/test/testfiles/JUnit-TestH5Pfapls3.txt @@ -0,0 +1,10 @@ +JUnit version 4.11 +.testH5Pset_fapl_ros3_specified +.testH5Pset_fapl_ros3 +.testH5Pget_fapl_ros3_invalid_fapl_id +.testH5Pget_fapl_ros3_fapl_id_of_wrong_driver_type + +Time: XXXX + +OK (4 tests) + diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt index 4106515..2b693bd 100644 --- a/src/CMakeLists.txt +++ b/src/CMakeLists.txt @@ -227,11 +227,14 @@ set (H5FD_SOURCES ${HDF5_SRC_DIR}/H5FDcore.c ${HDF5_SRC_DIR}/H5FDdirect.c ${HDF5_SRC_DIR}/H5FDfamily.c + ${HDF5_SRC_DIR}/H5FDhdfs.c ${HDF5_SRC_DIR}/H5FDint.c ${HDF5_SRC_DIR}/H5FDlog.c ${HDF5_SRC_DIR}/H5FDmpi.c ${HDF5_SRC_DIR}/H5FDmpio.c ${HDF5_SRC_DIR}/H5FDmulti.c + ${HDF5_SRC_DIR}/H5FDros3.c + ${HDF5_SRC_DIR}/H5FDs3comms.c ${HDF5_SRC_DIR}/H5FDsec2.c ${HDF5_SRC_DIR}/H5FDspace.c ${HDF5_SRC_DIR}/H5FDstdio.c @@ -243,11 +246,14 @@ set (H5FD_HDRS ${HDF5_SRC_DIR}/H5FDcore.h ${HDF5_SRC_DIR}/H5FDdirect.h ${HDF5_SRC_DIR}/H5FDfamily.h + ${HDF5_SRC_DIR}/H5FDhdfs.h ${HDF5_SRC_DIR}/H5FDlog.h ${HDF5_SRC_DIR}/H5FDmpi.h ${HDF5_SRC_DIR}/H5FDmpio.h ${HDF5_SRC_DIR}/H5FDmulti.h ${HDF5_SRC_DIR}/H5FDpublic.h + ${HDF5_SRC_DIR}/H5FDros3.h + ${HDF5_SRC_DIR}/H5FDs3comms.c ${HDF5_SRC_DIR}/H5FDsec2.h ${HDF5_SRC_DIR}/H5FDstdio.h ${HDF5_SRC_DIR}/H5FDwindows.h @@ -1142,6 +1148,7 @@ if (BUILD_SHARED_LIBS) add_library (${HDF5_LIBSH_TARGET} SHARED ${common_SRCS} ${shared_gen_SRCS} ${H5_PUBLIC_HEADERS} ${H5_PRIVATE_HEADERS} ${H5_GENERATED_HEADERS}) target_include_directories (${HDF5_LIBSH_TARGET} PRIVATE "${HDF5_SRC_DIR};${HDF5_BINARY_DIR};$<$<BOOL:${HDF5_ENABLE_PARALLEL}>:${MPI_C_INCLUDE_DIRS}>" + PUBLIC "$<$<BOOL:${HDF5_ENABLE_HDFS}>:${HDFS_INCLUDE_DIR}>" INTERFACE "$<INSTALL_INTERFACE:$<INSTALL_PREFIX>/include>" ) target_compile_definitions(${HDF5_LIBSH_TARGET} diff --git a/src/H5FDhdfs.c b/src/H5FDhdfs.c new file mode 100644 index 0000000..e3e11b2 --- /dev/null +++ b/src/H5FDhdfs.c @@ -0,0 +1,2070 @@ +/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * Read-Only HDFS Virtual File Driver (VFD) * + * Copyright (c) 2018, The HDF Group. * + * * + * All rights reserved. * + * * + * NOTICE: * + * All information contained herein is, and remains, the property of The HDF * + * Group. The intellectual and technical concepts contained herein are * + * proprietary to The HDF Group. Dissemination of this information or * + * reproduction of this material is strictly forbidden unless prior written * + * permission is obtained from The HDF Group. * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ + +/* + * Programmer: Jacob Smith + * 2018-04-23 + * + * Purpose: Provide read-only access to files on the Hadoop Distributed + * File System (HDFS). + */ + +/* This source code file is part of the H5FD driver module */ +#include "H5FDdrvr_module.h" + +#include "H5private.h" /* Generic Functions */ +#include "H5Eprivate.h" /* Error handling */ +#include "H5FDprivate.h" /* File drivers */ +#include "H5FDhdfs.h" /* hdfs file driver */ +#include "H5FLprivate.h" /* Free Lists */ +#include "H5Iprivate.h" /* IDs */ +#include "H5MMprivate.h" /* Memory management */ + +#ifdef H5_HAVE_LIBHDFS +#include "hdfs.h" +#endif + +/* toggle function call prints: 1 turns on */ +#define HDFS_DEBUG 0 + +/* toggle stats collection and reporting */ +#define HDFS_STATS 0 + +/* The driver identification number, initialized at runtime */ +static hid_t H5FD_HDFS_g = 0; + +#if HDFS_STATS + +/* arbitrarily large value, such that any reasonable size read will be "less" + * than this value and set a true minimum + * not 0 because that may be a valid recorded minimum in degenerate cases + */ +#define HDFS_STATS_STARTING_MIN 0xfffffffful + +/* Configuration definitions for stats collection and breakdown + * + * 2^10 = 1024 + * Reads up to 1024 bytes (1 kB) fall in bin 0 + * 2^(10+(1*16)) = 2^26 = 64MB + * Reads of 64MB or greater fall in "overflow" bin[BIN_COUNT] + */ +#define HDFS_STATS_BASE 2 +#define HDFS_STATS_INTERVAL 1 +#define HDFS_STATS_START_POWER 10 +#define HDFS_STATS_BIN_COUNT 16 /* MUST BE GREATER THAN 0 */ + + +/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * Calculate `BASE ^ (START_POWER + (INTERVAL * bin_i))` + * Stores result at `(unsigned long long *) out_ptr`. + * Used in computing boundaries between stats bins. + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + */ +#define HDFS_STATS_POW(bin_i, out_ptr) { \ + unsigned long long donotshadowresult = 1; \ + unsigned donotshadowindex = 0; \ + for (donotshadowindex = 0; \ + donotshadowindex < (((bin_i) * HDFS_STATS_INTERVAL) + \ + HDFS_STATS_START_POWER); \ + donotshadowindex++) \ + { \ + donotshadowresult *= HDFS_STATS_BASE; \ + } \ + *(out_ptr) = donotshadowresult; \ +} + +/* array to hold pre-computed boundaries for stats bins */ +static unsigned long long hdfs_stats_boundaries[HDFS_STATS_BIN_COUNT]; + + +/*************************************************************************** + * + * Structure: hdfs_statsbin + * + * Purpose: + * + * Structure for storing per-file hdfs VFD usage statistics. + * + * + * + * `count` (unsigned long long) + * + * Number of reads with size in this bin's range. + * + * `bytes` (unsigned long long) + * + * Total number of bytes read through this bin. + * + * `min` (unsigned long long) + * + * Smallest read size in this bin. + * + * `max` (unsigned long long) + * + * Largest read size in this bin. + * + * + * + * Programmer: Jacob Smith + * + * Changes: None + * + ***************************************************************************/ +typedef struct { + unsigned long long count; + unsigned long long bytes; + unsigned long long min; + unsigned long long max; +} hdfs_statsbin; + +#endif /* HDFS_STATS */ + +/* "unique" identifier for `hdfs_t` structures. + * Randomly generated by unweighted dice rolls. + */ +#define HDFS_HDFST_MAGIC 0x1AD5DE84 + + +/*************************************************************************** + * + * Structure: hdfs_t + * + * Purpose: + * + * Contain/retain information associated with a file hosted on Hadoop + * Distributed File System (HDFS). Instantiated and populated via + * `H5FD_hdfs_handle_open()` and cleaned up via `H5FD_hdfs_handle_close()`. + * + * + * + * `magic` (unisgned long) + * + * Number to indicate that this structure is of the promised + * type and should still be valid; should be HDFS_HDFST_MAGIC throughout + * the lifespan of the structure. Upon deletion of the structure, the + * programmer should set magic to anything but HDFS_HDFST_MAGIC, to + * indicate that the structure is to no longer be trusted. + * + * `filesystem` (hdfsFS) + * + * A libhdfs file system handle. + * + * `fileinfo` (hdfsFileInfo*) + * + * A pointer to a libhdfs file info structure. + * + * `file` (hdfsFile) + * + * A libhdfs file handle. + * + * + * + * Programmer: Jacob Smith + * May 2018 + * + * Changes: None + * + *************************************************************************** + */ +typedef struct { + unsigned long magic; +#ifdef H5_HAVE_LIBHDFS + hdfsFS filesystem; + hdfsFileInfo *fileinfo; + hdfsFile file; +#endif +} hdfs_t; + +#ifdef H5_HAVE_LIBHDFS + +/*-------------------------------------------------------------------------- + * Function: H5FD_hdfs_handle_open + * + * Purpose: Create a HDFS file handle, 'opening' the target file. + * + * Return: Success: Pointer to HDFS container/handle of opened file. + * Failure: NULL + * + * Programmer: Gerd Herber + * May 2018 + * + * Changes: None. + *-------------------------------------------------------------------------- + */ +static hdfs_t * +H5FD_hdfs_handle_open( + const char *path, + const char *namenode_name, + const int32_t namenode_port, + const char *user_name, + const char *kerberos_ticket_cache, + const int32_t stream_buffer_size) +{ + struct hdfsBuilder *builder = NULL; + hdfs_t *handle = NULL; + hdfs_t *ret_value = NULL; + + FUNC_ENTER_NOAPI_NOINIT + +#if HDFS_DEBUG + HDfprintf(stdout, "called H5FD_hdfs_handle_open.\n"); +#endif + + if (path == NULL || path[0] == '\0') { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "path cannot be null.\n") + } + if (namenode_name == NULL /* || namenode_name[0] == '\0' */ ) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "namenode name cannot be null.\n") + } + if (namenode_port < 0 || namenode_port > 65535) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "namenode port must be non-negative and <= 65535.\n") + } + if (stream_buffer_size < 0) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "buffer size must non-negative.\n") + } + + handle = (hdfs_t *)H5MM_malloc(sizeof(hdfs_t)); + if (handle == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, NULL, + "could not malloc space for handle.\n") + } + + handle->magic = (unsigned long)HDFS_HDFST_MAGIC; + handle->filesystem = NULL; /* TODO: not a pointer; NULL may cause bug */ + handle->fileinfo = NULL; + handle->file = NULL; /* TODO: not a pointer; NULL may cause bug */ + + builder = hdfsNewBuilder(); + if (!builder) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "(hdfs) failed to create builder") + } + hdfsBuilderSetNameNode(builder, namenode_name); + hdfsBuilderSetNameNodePort(builder, (tPort)namenode_port); + if (user_name != NULL && user_name[0] != '\0') { + hdfsBuilderSetUserName(builder, user_name); + } + if (kerberos_ticket_cache != NULL && kerberos_ticket_cache[0] != '\0') { + hdfsBuilderSetKerbTicketCachePath(builder, kerberos_ticket_cache); + } + /* Call to `hdfsBuilderConnect` releases builder, regardless of success. */ + handle->filesystem = hdfsBuilderConnect(builder); + if (!handle->filesystem) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "(hdfs) could not connect to default namenode") + } + handle->fileinfo = hdfsGetPathInfo(handle->filesystem, path); + if (!handle->fileinfo) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "hdfsGetPathInfo failed") + } + handle->file = hdfsOpenFile( + handle->filesystem, + path, + O_RDONLY, + stream_buffer_size, + 0, + 0); + if (!handle->file) { + HGOTO_ERROR(H5E_VFL, H5E_CANTOPENFILE, NULL, + "(hdfs) could not open") + } + + ret_value = handle; + +done: + if (ret_value == NULL && handle != NULL) { + /* error; clean up */ + HDassert(handle->magic == HDFS_HDFST_MAGIC); + handle->magic++; + if (handle->file != NULL) { + if (FAIL == (hdfsCloseFile(handle->filesystem, handle->file))) { + HDONE_ERROR(H5E_VFL, H5E_CANTCLOSEFILE, NULL, + "unable to close hdfs file handle") + } + } + if (handle->fileinfo != NULL) { + hdfsFreeFileInfo(handle->fileinfo, 1); + } + if (handle->filesystem != NULL) { + if (FAIL == (hdfsDisconnect(handle->filesystem))) { + HDONE_ERROR(H5E_VFL, H5E_CANTCLOSEFILE, NULL, + "unable to disconnect from hdfs") + } + } + H5MM_xfree(handle); + } + + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_hdfs_handle_open() */ + + +/*-------------------------------------------------------------------------- + * Function: H5FD_hdfs_handle_close + * + * Purpose: 'Close' an HDFS file container/handle, releasing underlying + * resources. + * + * Return: Success: `SUCCEED` (0) + * Failure: `FAIL` (-1) + * + * Programmer: Gerd Herber + * May 2018 + * + * Changes: None. + *-------------------------------------------------------------------------- + */ +static herr_t +H5FD_hdfs_handle_close(hdfs_t *handle) +{ + herr_t ret_value = SUCCEED; + + FUNC_ENTER_NOAPI_NOINIT + +#if HDFS_DEBUG + HDfprintf(stdout, "called H5FD_hdfs_close.\n"); +#endif + + if (handle == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle cannot be null.\n") + } + if (handle->magic != HDFS_HDFST_MAGIC) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle has invalid magic.\n") + } + + handle->magic++; + if (handle->file != NULL) { + if (FAIL == (hdfsCloseFile(handle->filesystem, handle->file))) { + HDONE_ERROR(H5E_VFL, H5E_CANTCLOSEFILE, FAIL, + "unable to close hdfs file handle") + } + } + if (handle->fileinfo != NULL) { + hdfsFreeFileInfo(handle->fileinfo, 1); + } + if (handle->filesystem != NULL) { + if (FAIL == (hdfsDisconnect(handle->filesystem))) { + HDONE_ERROR(H5E_VFL, H5E_CANTCLOSEFILE, FAIL, + "unable to disconnect hdfs file system") + } + } + + H5MM_xfree(handle); + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_hdfs_close() */ +#endif /* H5_HAVE_LIBHDFS */ + + +/*************************************************************************** + * + * Structure: H5FD_hdfs_t + * + * Purpose: + * + * H5FD_hdfs_t is a structure used to store all information needed to + * maintain R/O access to a single HDF5 file in an HDFS file system. + * This structure is created when such a file is "opened" and + * discarded when it is "closed". + * + * + * `pub` (H5FD_t) + * + * Instance of H5FD_t which contains all fields common to all VFDs. + * It must be the first item in this structure, since at higher levels, + * this structure will be treated as an instance of H5FD_t. + * + * `fa` (H5FD_hdfs_fapl_t) + * + * Instance of `H5FD_hdfs_fapl_t` containing the HDFS configuration data + * needed to "open" the HDF5 file. + * + * `eoa` (haddr_t) + * + * End of addressed space in file. After open, it should always + * equal the file size. + * + * `hdfs_handle` (hdfs_t *) + * + * Instance of HDFS Request handle associated with the target resource. + * Responsible for communicating with remote host and presenting file + * contents as indistinguishable from a file on the local filesystem. + * + * *** present only if HDFS_SATS is flagged to enable stats collection *** + * + * `meta` (hdfs_statsbin[]) + * `raw` (hdfs_statsbin[]) + * + * Only present if hdfs stats collection is enabled. + * + * Arrays of `hdfs_statsbin` structures to record raw- and metadata reads. + * + * Records count and size of reads performed by the VFD, and is used to + * print formatted usage statistics to stdout upon VFD shutdown. + * + * Reads of each raw- and metadata type are recorded in an individual bin + * determined by the size of the read. The last bin of each type is + * reserved for "big" reads, with no defined upper bound. + * + * *** end HDFS_STATS *** + * + * + * + * Programmer: Jacob Smith + * + * Changes: None. + * + *************************************************************************** + */ +typedef struct H5FD_hdfs_t { + H5FD_t pub; + H5FD_hdfs_fapl_t fa; + haddr_t eoa; +#ifdef H5_HAVE_LIBHDFS + hdfs_t *hdfs_handle; +#endif +#if HDFS_STATS + hdfs_statsbin meta[HDFS_STATS_BIN_COUNT + 1]; + hdfs_statsbin raw[HDFS_STATS_BIN_COUNT + 1]; +#endif +} H5FD_hdfs_t; + +/* + * These macros check for overflow of various quantities. These macros + * assume that HDoff_t is signed and haddr_t and size_t are unsigned. + * + * ADDR_OVERFLOW: Checks whether a file address of type `haddr_t' + * is too large to be represented by the second argument + * of the file seek function. + * + */ +#define MAXADDR (((haddr_t)1<<(8*sizeof(HDoff_t)-1))-1) +#define ADDR_OVERFLOW(A) (HADDR_UNDEF==(A) || ((A) & ~(haddr_t)MAXADDR)) + +/* Prototypes */ +static herr_t H5FD_hdfs_term(void); +static void *H5FD_hdfs_fapl_get(H5FD_t *_file); +static void *H5FD_hdfs_fapl_copy(const void *_old_fa); +static herr_t H5FD_hdfs_fapl_free(void *_fa); +static H5FD_t *H5FD_hdfs_open(const char *name, unsigned flags, hid_t fapl_id, + haddr_t maxaddr); +static herr_t H5FD_hdfs_close(H5FD_t *_file); +static int H5FD_hdfs_cmp(const H5FD_t *_f1, const H5FD_t *_f2); +static herr_t H5FD_hdfs_query(const H5FD_t *_f1, unsigned long *flags); +static haddr_t H5FD_hdfs_get_eoa(const H5FD_t *_file, H5FD_mem_t type); +static herr_t H5FD_hdfs_set_eoa(H5FD_t *_file, H5FD_mem_t type, haddr_t addr); +static haddr_t H5FD_hdfs_get_eof(const H5FD_t *_file, H5FD_mem_t type); +static herr_t H5FD_hdfs_get_handle(H5FD_t *_file, hid_t fapl, + void** file_handle); +static herr_t H5FD_hdfs_read(H5FD_t *_file, H5FD_mem_t type, hid_t fapl_id, + haddr_t addr, size_t size, void *buf); +static herr_t H5FD_hdfs_write(H5FD_t *_file, H5FD_mem_t type, hid_t fapl_id, + haddr_t addr, size_t size, const void *buf); +static herr_t H5FD_hdfs_truncate(H5FD_t *_file, hid_t dxpl_id, + hbool_t closing); +static herr_t H5FD_hdfs_lock(H5FD_t *_file, hbool_t rw); +static herr_t H5FD_hdfs_unlock(H5FD_t *_file); +static herr_t H5FD_hdfs_validate_config(const H5FD_hdfs_fapl_t * fa); + +static const H5FD_class_t H5FD_hdfs_g = { + "hdfs", /* name */ + MAXADDR, /* maxaddr */ + H5F_CLOSE_WEAK, /* fc_degree */ + H5FD_hdfs_term, /* terminate */ + NULL, /* sb_size */ + NULL, /* sb_encode */ + NULL, /* sb_decode */ + sizeof(H5FD_hdfs_fapl_t), /* fapl_size */ + H5FD_hdfs_fapl_get, /* fapl_get */ + H5FD_hdfs_fapl_copy, /* fapl_copy */ + H5FD_hdfs_fapl_free, /* fapl_free */ + 0, /* dxpl_size */ + NULL, /* dxpl_copy */ + NULL, /* dxpl_free */ + H5FD_hdfs_open, /* open */ + H5FD_hdfs_close, /* close */ + H5FD_hdfs_cmp, /* cmp */ + H5FD_hdfs_query, /* query */ + NULL, /* get_type_map */ + NULL, /* alloc */ + NULL, /* free */ + H5FD_hdfs_get_eoa, /* get_eoa */ + H5FD_hdfs_set_eoa, /* set_eoa */ + H5FD_hdfs_get_eof, /* get_eof */ + H5FD_hdfs_get_handle, /* get_handle */ + H5FD_hdfs_read, /* read */ + H5FD_hdfs_write, /* write */ + NULL, /* flush */ + H5FD_hdfs_truncate, /* truncate */ + H5FD_hdfs_lock, /* lock */ + H5FD_hdfs_unlock, /* unlock */ + H5FD_FLMAP_DICHOTOMY /* fl_map */ +}; + +/* Declare a free list to manage the H5FD_hdfs_t struct */ +H5FL_DEFINE_STATIC(H5FD_hdfs_t); + + +/*------------------------------------------------------------------------- + * Function: H5FD__init_package + * + * Purpose: Initializes any interface-specific data or routines. + * + * Return: Non-negative on success/Negative on failure + * + * Changes: Rename as appropriate for hdfs vfd. + * Jacob Smith 2018 + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD__init_package(void) +{ + herr_t ret_value = SUCCEED; + + FUNC_ENTER_STATIC + + if (H5FD_hdfs_init() < 0) { + HGOTO_ERROR(H5E_VFL, H5E_CANTINIT, FAIL, + "unable to initialize hdfs VFD") + } + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD__init_package() */ + + +/*------------------------------------------------------------------------- + * Function: H5FD_hdfs_init + * + * Purpose: Initialize this driver by registering the driver with the + * library. + * + * Return: Success: The driver ID for the hdfs driver. + * Failure: Negative + * + * Programmer: Robb Matzke + * Thursday, July 29, 1999 + * + * Changes: Rename as appropriate for hdfs vfd. + * Jacob Smith 2018 + * + *------------------------------------------------------------------------- + */ +hid_t +H5FD_hdfs_init(void) +{ + hid_t ret_value = H5I_INVALID_HID; /* Return value */ + + FUNC_ENTER_NOAPI(FAIL) + +#if HDFS_DEBUG + HDfprintf(stdout, "H5FD_hdfs_init() called.\n"); +#endif + + if (H5I_VFL != H5I_get_type(H5FD_HDFS_g)) { + H5FD_HDFS_g = H5FD_register( + &H5FD_hdfs_g, + sizeof(H5FD_class_t), + FALSE); + } + +#if HDFS_STATS + /* pre-compute statsbin boundaries + */ + for (unsigned bin_i = 0; bin_i < HDFS_STATS_BIN_COUNT; bin_i++) { + unsigned long long value = 0; + HDFS_STATS_POW(bin_i, &value) + hdfs_stats_boundaries[bin_i] = value; + } +#endif + + ret_value = H5FD_HDFS_g; + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* end H5FD_hdfs_init() */ + + +/*--------------------------------------------------------------------------- + * Function: H5FD_hdfs_term + * + * Purpose: Shut down the VFD + * + * Returns: SUCCEED (Can't fail) + * + * Programmer: Quincey Koziol + * Friday, Jan 30, 2004 + * + * Changes: Rename as appropriate for hdfs vfd. + * Jacob Smith 2018 + * + *--------------------------------------------------------------------------- + */ +static herr_t +H5FD_hdfs_term(void) +{ + FUNC_ENTER_NOAPI_NOINIT_NOERR + +#if HDFS_DEBUG + HDfprintf(stdout, "H5FD_hdfs_term() called.\n"); +#endif + + /* Reset VFL ID */ + H5FD_HDFS_g = 0; + + FUNC_LEAVE_NOAPI(SUCCEED) +} /* end H5FD_hdfs_term() */ + + +/*------------------------------------------------------------------------- + * Function: H5Pset_fapl_hdfs + * + * Purpose: Modify the file access property list to use the H5FD_HDFS + * driver defined in this source file. All driver specfic + * properties are passed in as a pointer to a suitably + * initialized instance of H5FD_hdfs_fapl_t + * + * Return: SUCCEED/FAIL + * + * Programmer: John Mainzer + * 9/10/17 + * + * Changes: Rename as appropriate for hdfs vfd. + * Jacob Smith 2018 + * + *------------------------------------------------------------------------- + */ +herr_t +H5Pset_fapl_hdfs(hid_t fapl_id, + H5FD_hdfs_fapl_t *fa) +{ + H5P_genplist_t *plist = NULL; /* Property list pointer */ + herr_t ret_value = FAIL; + + FUNC_ENTER_API(FAIL) + H5TRACE2("e", "i*x", fapl_id, fa); + + HDassert(fa != NULL); + +#if HDFS_DEBUG + HDfprintf(stdout, "H5Pset_fapl_hdfs() called.\n"); +#endif + + plist = H5P_object_verify(fapl_id, H5P_FILE_ACCESS); + if (plist == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADTYPE, FAIL, \ + "not a file access property list") + } + + if (FAIL == H5FD_hdfs_validate_config(fa)) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "invalid hdfs config") + } + + ret_value = H5P_set_driver(plist, H5FD_HDFS, (void *)fa); + +done: + FUNC_LEAVE_API(ret_value) + +} /* H5Pset_fapl_hdfs() */ + + +/*------------------------------------------------------------------------- + * Function: H5FD_hdfs_validate_config() + * + * Purpose: Test to see if the supplied instance of H5FD_hdfs_fapl_t + * contains internally consistant data. Return SUCCEED if so, + * and FAIL otherwise. + * + * Note the difference between internally consistant and + * correct. As we will have to try to access the target + * object to determine whether the supplied data is correct, + * we will settle for internal consistancy at this point + * + * Return: SUCCEED if instance of H5FD_hdfs_fapl_t contains internally + * consistant data, FAIL otherwise. + * + * Programmer: Jacob Smith + * 9/10/17 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_hdfs_validate_config(const H5FD_hdfs_fapl_t * fa) +{ + herr_t ret_value = SUCCEED; + + FUNC_ENTER_NOAPI_NOINIT + + HDassert(fa != NULL); + + if ( fa->version != H5FD__CURR_HDFS_FAPL_T_VERSION ) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "Unknown H5FD_hdfs_fapl_t version"); + } + + if ( fa->namenode_port > 65535 ) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "Invalid namenode port number"); + } + if ( fa->namenode_port < 0 ) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "Invalid namenode port number"); + } + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_hdfs_validate_config() */ + + +/*------------------------------------------------------------------------- + * Function: H5Pget_fapl_hdfs + * + * Purpose: Returns information about the hdfs file access property + * list though the function arguments. + * + * Return: Success: Non-negative + * + * Failure: Negative + * + * Programmer: John Mainzer + * 9/10/17 + * + * Modifications: + * + *------------------------------------------------------------------------- + */ +herr_t +H5Pget_fapl_hdfs(hid_t fapl_id, + H5FD_hdfs_fapl_t *fa_out) +{ + const H5FD_hdfs_fapl_t *fa = NULL; + H5P_genplist_t *plist = NULL; + herr_t ret_value = SUCCEED; + + FUNC_ENTER_API(FAIL) + H5TRACE2("e", "i*x", fapl_id, fa_out); + +#if HDFS_DEBUG + HDfprintf(stdout, "H5Pget_fapl_hdfs() called.\n"); +#endif + + if (fa_out == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "fa_out is NULL") + } + plist = H5P_object_verify(fapl_id, H5P_FILE_ACCESS); + if (plist == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADTYPE, FAIL, + "not a file access list") + } + if (H5FD_HDFS != H5P_peek_driver(plist)) { + HGOTO_ERROR(H5E_PLIST, H5E_BADVALUE, FAIL, + "incorrect VFL driver") + } + + fa = (const H5FD_hdfs_fapl_t *)H5P_peek_driver_info(plist); + if (fa == NULL) { + HGOTO_ERROR(H5E_PLIST, H5E_BADVALUE, FAIL, + "bad VFL driver info") + } + + /* Copy the hdfs fapl data out */ + HDmemcpy(fa_out, fa, sizeof(H5FD_hdfs_fapl_t)); + +done: + FUNC_LEAVE_API(ret_value) + +} /* H5Pget_fapl_hdfs() */ + + +/*------------------------------------------------------------------------- + * Function: H5FD_hdfs_fapl_get + * + * Purpose: Gets a file access property list which could be used to + * create an identical file. + * + * Return: Success: Ptr to new file access property list value. + * + * Failure: NULL + * + * Programmer: John Mainzer + * 9/8/17 + * + * Modifications: + * + *------------------------------------------------------------------------- + */ +static void * +H5FD_hdfs_fapl_get(H5FD_t *_file) +{ + H5FD_hdfs_t *file = (H5FD_hdfs_t*)_file; + H5FD_hdfs_fapl_t *fa = NULL; + void *ret_value = NULL; + + FUNC_ENTER_NOAPI_NOINIT + + fa = (H5FD_hdfs_fapl_t *)H5MM_calloc(sizeof(H5FD_hdfs_fapl_t)); + if (fa == NULL) { + HGOTO_ERROR(H5E_RESOURCE, H5E_NOSPACE, NULL, + "memory allocation failed") + } + + /* Copy the fields of the structure */ + HDmemcpy(fa, &(file->fa), sizeof(H5FD_hdfs_fapl_t)); + + ret_value = fa; + +done: + if (ret_value == NULL && fa != NULL) { + H5MM_xfree(fa); /* clean up on error */ + } + + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_hdfs_fapl_get() */ + + +/*------------------------------------------------------------------------- + * Function: H5FD_hdfs_fapl_copy + * + * Purpose: Copies the hdfs-specific file access properties. + * + * Return: Success: Ptr to a new property list + * + * Failure: NULL + * + * Programmer: John Mainzer + * 9/8/17 + * + * Modifications: + * + *------------------------------------------------------------------------- + */ +static void * +H5FD_hdfs_fapl_copy(const void *_old_fa) +{ + const H5FD_hdfs_fapl_t *old_fa = (const H5FD_hdfs_fapl_t*)_old_fa; + H5FD_hdfs_fapl_t *new_fa = NULL; + void *ret_value = NULL; + + FUNC_ENTER_NOAPI_NOINIT + + new_fa = (H5FD_hdfs_fapl_t *)H5MM_malloc(sizeof(H5FD_hdfs_fapl_t)); + if (new_fa == NULL) { + HGOTO_ERROR(H5E_RESOURCE, H5E_NOSPACE, NULL, + "memory allocation failed") + } + + HDmemcpy(new_fa, old_fa, sizeof(H5FD_hdfs_fapl_t)); + ret_value = new_fa; + +done: + if (ret_value == NULL && new_fa != NULL) { + H5MM_xfree(new_fa); /* clean up on error */ + } + + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_hdfs_fapl_copy() */ + + +/*------------------------------------------------------------------------- + * Function: H5FD_hdfs_fapl_free + * + * Purpose: Frees the hdfs-specific file access properties. + * + * Return: SUCCEED (cannot fail) + * + * Programmer: John Mainzer + * 9/8/17 + * + * Modifications: + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_hdfs_fapl_free(void *_fa) +{ + H5FD_hdfs_fapl_t *fa = (H5FD_hdfs_fapl_t*)_fa; + + FUNC_ENTER_NOAPI_NOINIT_NOERR + + HDassert(fa != NULL); /* sanity check */ + + H5MM_xfree(fa); + + FUNC_LEAVE_NOAPI(SUCCEED) + +} /* H5FD_hdfs_fapl_free() */ + +#if HDFS_STATS + +/*---------------------------------------------------------------------------- + * + * Function: hdfs_reset_stats() + * + * Purpose: + * + * Reset the stats collection elements in this virtual file structure. + * + * Clears any set data in stats bins; initializes/zeroes values. + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - FAILURE: `FAIL` + * - Occurs if the file is invalid somehow + * + * Programmer: Jacob Smith + * 2017-12-08 + * + * Changes: None. + * + *---------------------------------------------------------------------------- + */ +static herr_t +hdfs_reset_stats(H5FD_hdfs_t *file) +{ + unsigned i = 0; + herr_t ret_value = SUCCEED; + + FUNC_ENTER_NOAPI_NOINIT + +#if HDFS_DEBUG + HDprintf("hdfs_reset_stats() called\n"); +#endif + + if (file == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "file was null") + } + + for (i = 0; i <= HDFS_STATS_BIN_COUNT; i++) { + file->raw[i].bytes = 0; + file->raw[i].count = 0; + file->raw[i].min = (unsigned long long)HDFS_STATS_STARTING_MIN; + file->raw[i].max = 0; + + file->meta[i].bytes = 0; + file->meta[i].count = 0; + file->meta[i].min = (unsigned long long)HDFS_STATS_STARTING_MIN; + file->meta[i].max = 0; + } + +done: + FUNC_LEAVE_NOAPI(ret_value); + +} /* hdfs_reset_stats */ +#endif /* HDFS_STATS */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_hdfs_open() + * + * Purpose: + * + * Create and/or opens a file as an HDF5 file. + * + * Any flag except H5F_ACC_RDONLY will cause an error. + * + * Return: + * + * Success: A pointer to a new file data structure. + * The public fields will be initialized by the caller, which is + * always H5FD_open(). + * + * Failure: NULL + * + * Programmer: Jacob Smith + * 2017-11-02 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static H5FD_t * +H5FD_hdfs_open( + const char *path, + unsigned flags, + hid_t fapl_id, + haddr_t maxaddr) +{ + H5FD_t *ret_value = NULL; +#ifdef H5_HAVE_LIBHDFS + H5FD_hdfs_t *file = NULL; + hdfs_t *handle = NULL; + H5FD_hdfs_fapl_t fa; +#endif + + FUNC_ENTER_NOAPI_NOINIT + +#ifndef H5_HAVE_LIBHDFS + HGOTO_ERROR(H5E_VFL, H5E_UNSUPPORTED, NULL, + "Illegal open of unsupported virtual file (hdfs)"); +#else +#if HDFS_DEBUG + HDfprintf(stdout, "H5FD_hdfs_open() called.\n"); +#endif /* HDFS_DEBUG */ + + /* Sanity check on file offsets */ + HDcompile_assert(sizeof(HDoff_t) >= sizeof(size_t)); + + /* Check arguments */ + if (!path || !*path) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "invalid file name") + } + if (0 == maxaddr || HADDR_UNDEF == maxaddr) { + HGOTO_ERROR(H5E_ARGS, H5E_BADRANGE, NULL, + "bogus maxaddr") + } + if (ADDR_OVERFLOW(maxaddr)) { + HGOTO_ERROR(H5E_ARGS, H5E_OVERFLOW, NULL, + "bogus maxaddr") + } + if (flags != H5F_ACC_RDONLY) { + HGOTO_ERROR(H5E_ARGS, H5E_UNSUPPORTED, NULL, + "only Read-Only access allowed") + } + if (fapl_id == H5P_DEFAULT || fapl_id == H5P_FILE_ACCESS_DEFAULT) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "fapl cannot be H5P_DEFAULT") + } + if (FAIL == H5Pget_fapl_hdfs(fapl_id, &fa)) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "can't get property list") + } + + handle = H5FD_hdfs_handle_open( + path, + fa.namenode_name, + fa.namenode_port, + fa.user_name, + fa.kerberos_ticket_cache, + fa.stream_buffer_size); + + if (handle == NULL) { + HGOTO_ERROR(H5E_VFL, H5E_CANTOPENFILE, NULL, + "could not open") + } + + HDassert(handle->magic == HDFS_HDFST_MAGIC); + + /* create new file struct + */ + file = H5FL_CALLOC(H5FD_hdfs_t); + if (file == NULL) { + HGOTO_ERROR(H5E_RESOURCE, H5E_NOSPACE, NULL, + "unable to allocate file struct") + } + file->hdfs_handle = handle; + HDmemcpy(&(file->fa), &fa, sizeof(H5FD_hdfs_fapl_t)); + +#if HDFS_STATS + if (FAIL == hdfs_reset_stats(file)) { + HGOTO_ERROR(H5E_INTERNAL, H5E_UNINITIALIZED, NULL, + "unable to reset file statistics") + } +#endif /* HDFS_STATS */ + + ret_value = (H5FD_t*)file; +#endif /* H5_HAVE_LIBHDFS */ + +done: +#ifdef H5_HAVE_LIBHDFS + if (ret_value == NULL) { + if (handle != NULL) { + if (FAIL == H5FD_hdfs_handle_close(handle)) { + HDONE_ERROR(H5E_VFL, H5E_CANTCLOSEFILE, NULL, + "unable to close HDFS file handle") + } + } + if (file != NULL) { + file = H5FL_FREE(H5FD_hdfs_t, file); + } + } /* if null return value (error) */ +#endif /* H5_HAVE_LIBHDFS */ + + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_hdfs_open() */ + +#if HDFS_STATS + +/*---------------------------------------------------------------------------- + * + * Function: hdfs_fprint_stats() + * + * Purpose: + * + * Tabulate and pretty-print statistics for this virtual file. + * + * Should be called upon file close. + * + * Shows number of reads and bytes read, broken down by + * "raw" (H5FD_MEM_DRAW) + * or "meta" (any other flag) + * + * Prints filename and listing of total number of reads and bytes read, + * both as a grand total and separate meta- and rawdata reads. + * + * If any reads were done, prints out two tables: + * + * 1. overview of raw- and metadata reads + * - min (smallest size read) + * - average of size read + * - k,M,G suffixes by powers of 1024 (2^10) + * - max (largest size read) + * 2. tabulation of "bins", sepraring reads into exponentially-larger + * ranges of size. + * - columns for number of reads, total bytes, and average size, with + * separate sub-colums for raw- and metadata reads. + * - each row represents one bin, identified by the top of its range + * + * Bin ranges can be modified with pound-defines at the top of this file. + * + * Bins without any reads in their bounds are not printed. + * + * An "overflow" bin is also present, to catch "big" reads. + * + * Output for all bins (and range ceiling and average size report) + * is divied by powers of 1024. By corollary, four digits before the decimal + * is valid. + * + * - 41080 bytes is represented by 40.177k, not 41.080k + * - 1004.831M represents approx. 1052642000 bytes + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - FAILURE: `FAIL` + * - occurs if the file passed in is invalid + * - TODO: if stream is invalid? how can we check this? + * + * Programmer: Jacob Smith + * + * Changes: None. + * + *---------------------------------------------------------------------------- + */ +static herr_t +hdfs_fprint_stats( + FILE *stream, + const H5FD_hdfs_t *file) +{ + herr_t ret_value = SUCCEED; + parsed_url_t *purl = NULL; + unsigned i = 0; + unsigned long count_meta = 0; + unsigned long count_raw = 0; + double average_meta = 0.0; + double average_raw = 0.0; + unsigned long long min_meta = (unsigned long long)HDFS_STATS_STARTING_MIN; + unsigned long long min_raw = (unsigned long long)HDFS_STATS_STARTING_MIN; + unsigned long long max_meta = 0; + unsigned long long max_raw = 0; + unsigned long long bytes_raw = 0; + unsigned long long bytes_meta = 0; + double re_dub = 0.0; /* re-usable double variable */ + unsigned suffix_i = 0; + const char suffixes[] = { ' ', 'K', 'M', 'G', 'T', 'P' }; + + FUNC_ENTER_NOAPI_NOINIT + + if (stream == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "file stream cannot be null" ) + } + if (file == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "file cannot be null") + } + if (file->hdfs_handle == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "hdfs handle cannot be null") + } + if (file->hdfs_handle->magic != HDFS_HDFST_MAGIC) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "hdfs handle has invalid magic") + } + + /* TODO: See what libhdfs exposes to us. */ + +#if 0 + if (file->s3r_handle->purl == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "parsed url structure cannot be null") + } + purl = file->s3r_handle->purl; + + /****************** + * PRINT FILENAME * + ******************/ + + HDfprintf(stream, "stats for %s://%s", purl->scheme, purl->host); + if (purl->port != NULL && purl->port[0] != '\0') { + HDfprintf(stream, ":%s", purl->port); + } + if (purl->query != NULL && purl->query[0] != '\0') { + if (purl->path != NULL && purl->path[0] != '\0') { + HDfprintf(stream, "/%s", purl->path); + } else { + HDfprintf(stream, "/"); + } + HDfprintf(stream, "?%s", purl->query); + } else if (purl->path != NULL && purl->path[0] != '\0') { + HDfprintf(stream, "/%s", purl->path); + } + HDfprintf(stream, "\n"); +#endif + + /******************* + * AGGREGATE STATS * + *******************/ + + for (i = 0; i <= HDFS_STATS_BIN_COUNT; i++) { + const hdfs_statsbin *r = &file->raw[i]; + const hdfs_statsbin *m = &file->meta[i]; + + if (m->min < min_meta) min_meta = m->min; + if (r->min < min_raw) min_raw = r->min; + if (m->max > max_meta) max_meta = m->max; + if (r->max > max_raw) max_raw = r->max; + + count_raw += r->count; + count_meta += m->count; + bytes_raw += r->bytes; + bytes_meta += m->bytes; + } + if (count_raw > 0) average_raw = (double)bytes_raw / (double)count_raw; + if (count_meta > 0) average_meta = (double)bytes_meta / (double)count_meta; + + /****************** + * PRINT OVERVIEW * + ******************/ + + HDfprintf(stream, "TOTAL READS: %llu (%llu meta, %llu raw)\n", + count_raw + count_meta, count_meta, count_raw); + HDfprintf(stream, "TOTAL BYTES: %llu (%llu meta, %llu raw)\n", + bytes_raw + bytes_meta, bytes_meta, bytes_raw); + + if (count_raw + count_meta == 0) { + goto done; + } + + /************************* + * PRINT AGGREGATE STATS * + *************************/ + + HDfprintf(stream, "SIZES meta raw\n"); + HDfprintf(stream, " min "); + if (count_meta == 0) { + HDfprintf(stream, " 0.000 "); + } else { + re_dub = (double)min_meta; + for (suffix_i = 0; re_dub >= 1024.0; suffix_i++) { + re_dub /= 1024.0; + } + HDassert(suffix_i < sizeof(suffixes)); + HDfprintf(stream, "%8.3lf%c ", re_dub, suffixes[suffix_i]); + } + + if (count_raw == 0) { + HDfprintf(stream, " 0.000 \n"); + } else { + re_dub = (double)min_raw; + for (suffix_i = 0; re_dub >= 1024.0; suffix_i++) { + re_dub /= 1024.0; + } + HDassert(suffix_i < sizeof(suffixes)); + HDfprintf(stream, "%8.3lf%c\n", re_dub, suffixes[suffix_i]); + } + + HDfprintf(stream, " avg "); + re_dub = (double)average_meta; + for (suffix_i = 0; re_dub >= 1024.0; suffix_i++) { + re_dub /= 1024.0; + } + HDassert(suffix_i < sizeof(suffixes)); + HDfprintf(stream, "%8.3lf%c ", re_dub, suffixes[suffix_i]); + + re_dub = (double)average_raw; + for (suffix_i = 0; re_dub >= 1024.0; suffix_i++) { + re_dub /= 1024.0; + } + HDassert(suffix_i < sizeof(suffixes)); + HDfprintf(stream, "%8.3lf%c\n", re_dub, suffixes[suffix_i]); + + HDfprintf(stream, " max "); + re_dub = (double)max_meta; + for (suffix_i = 0; re_dub >= 1024.0; suffix_i++) { + re_dub /= 1024.0; + } + HDassert(suffix_i < sizeof(suffixes)); + HDfprintf(stream, "%8.3lf%c ", re_dub, suffixes[suffix_i]); + + re_dub = (double)max_raw; + for (suffix_i = 0; re_dub >= 1024.0; suffix_i++) { + re_dub /= 1024.0; + } + HDassert(suffix_i < sizeof(suffixes)); + HDfprintf(stream, "%8.3lf%c\n", re_dub, suffixes[suffix_i]); + + /****************************** + * PRINT INDIVIDUAL BIN STATS * + ******************************/ + + HDfprintf(stream, + "BINS # of reads total bytes average size\n"); + HDfprintf(stream, + " up-to meta raw meta raw meta raw\n"); + + for (i = 0; i <= HDFS_STATS_BIN_COUNT; i++) { + const hdfs_statsbin *m; + const hdfs_statsbin *r; + unsigned long long range_end = 0; + char bm_suffix = ' '; /* bytes-meta */ + double bm_val = 0.0; + char br_suffix = ' '; /* bytes-raw */ + double br_val = 0.0; + char am_suffix = ' '; /* average-meta */ + double am_val = 0.0; + char ar_suffix = ' '; /* average-raw */ + double ar_val = 0.0; + + m = &file->meta[i]; + r = &file->raw[i]; + if (r->count == 0 && m->count == 0) { + continue; + } + + range_end = hdfs_stats_boundaries[i]; + + if (i == HDFS_STATS_BIN_COUNT) { + range_end = hdfs_stats_boundaries[i-1]; + HDfprintf(stream, ">"); + } else { + HDfprintf(stream, " "); + } + + bm_val = (double)m->bytes; + for (suffix_i = 0; bm_val >= 1024.0; suffix_i++) { + bm_val /= 1024.0; + } + HDassert(suffix_i < sizeof(suffixes)); + bm_suffix = suffixes[suffix_i]; + + br_val = (double)r->bytes; + for (suffix_i = 0; br_val >= 1024.0; suffix_i++) { + br_val /= 1024.0; + } + HDassert(suffix_i < sizeof(suffixes)); + br_suffix = suffixes[suffix_i]; + + if (m->count > 0) { + am_val = (double)(m->bytes) / (double)(m->count); + } + for (suffix_i = 0; am_val >= 1024.0; suffix_i++) { + am_val /= 1024.0; + } + HDassert(suffix_i < sizeof(suffixes)); + am_suffix = suffixes[suffix_i]; + + if (r->count > 0) { + ar_val = (double)(r->bytes) / (double)(r->count); + } + for (suffix_i = 0; ar_val >= 1024.0; suffix_i++) { + ar_val /= 1024.0; + } + HDassert(suffix_i < sizeof(suffixes)); + ar_suffix = suffixes[suffix_i]; + + re_dub = (double)range_end; + for (suffix_i = 0; re_dub >= 1024.0; suffix_i++) { + re_dub /= 1024.0; + } + HDassert(suffix_i < sizeof(suffixes)); + + HDfprintf( + stream, + " %8.3f%c %7d %7d %8.3f%c %8.3f%c %8.3f%c %8.3f%c\n", + re_dub, suffixes[suffix_i], /* bin ceiling */ + m->count, /* metadata reads */ + r->count, /* rawdata reads */ + bm_val, bm_suffix, /* metadata bytes */ + br_val, br_suffix, /* rawdata bytes */ + am_val, am_suffix, /* metadata average */ + ar_val, ar_suffix); /* rawdata average */ + fflush(stream); + } + +done: + FUNC_LEAVE_NOAPI(ret_value); + +} /* hdfs_fprint_stats */ +#endif /* HDFS_STATS */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_hdfs_close() + * + * Purpose: + * + * Close an HDF5 file. + * + * Return: + * + * SUCCEED/FAIL + * + * Programmer: Jacob Smith + * 2017-11-02 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_hdfs_close(H5FD_t *_file) +{ + herr_t ret_value = SUCCEED; +#ifdef H5_HAVE_LIBHDFS + H5FD_hdfs_t *file = (H5FD_hdfs_t *)_file; +#endif + + FUNC_ENTER_NOAPI_NOINIT + +#ifndef H5_HAVE_LIBHDFS + HGOTO_ERROR(H5E_VFL, H5E_CANTCLOSEFILE, FAIL, + "Illegal close of unsupported Virtual File (hdfs)") +#else +#if HDFS_DEBUG + HDfprintf(stdout, "H5FD_hdfs_close() called.\n"); +#endif + + /* Sanity checks + */ + HDassert(file != NULL); + HDassert(file->hdfs_handle != NULL); + HDassert(file->hdfs_handle->magic == HDFS_HDFST_MAGIC); + + /* Close the underlying request handle + */ + if (file->hdfs_handle != NULL) { + if (FAIL == H5FD_hdfs_handle_close(file->hdfs_handle)) { + HGOTO_ERROR(H5E_VFL, H5E_CANTCLOSEFILE, FAIL, + "unable to close HDFS file handle") + } + } + +#if HDFS_STATS + /* TODO: mechanism to re-target stats printout */ + if (FAIL == hdfs_fprint_stats(stdout, file)) { + HGOTO_ERROR(H5E_INTERNAL, H5E_ERROR, FAIL, + "problem while writing file statistics") + } +#endif /* HDFS_STATS */ + + /* Release the file info + */ + file = H5FL_FREE(H5FD_hdfs_t, file); +#endif /* H5_HAVE_LIBHDFS */ + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* end H5FD_hdfs_close() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_hdfs_cmp() + * + * Purpose: + * + * Compares two files using this driver by their HDFS-provided file info, + * field-by-field. + * + * Return: + * + Equivalent: 0 + * + Not Equivalent: -1 + * + * Programmer: Gerd Herber + * May 2018 + * + * Changes: + * + * + Replace `if (ret_value == 0)` chain with `HGOTO_DONE` jumps. + * Jacob Smith 17 May 2018 + * + *------------------------------------------------------------------------- + */ +static int +H5FD_hdfs_cmp(const H5FD_t *_f1, + const H5FD_t *_f2) +{ + int ret_value = 0; +#ifdef H5_HAVE_LIBHDFS + const H5FD_hdfs_t *f1 = (const H5FD_hdfs_t *)_f1; + const H5FD_hdfs_t *f2 = (const H5FD_hdfs_t *)_f2; + hdfsFileInfo *finfo1 = NULL; + hdfsFileInfo *finfo2 = NULL; +#endif /* H5_HAVE_LIBHDFS */ + + FUNC_ENTER_NOAPI_NOINIT_NOERR + +#ifdef H5_HAVE_LIBHDFS +#if HDFS_DEBUG + HDfprintf(stdout, "H5FD_hdfs_cmp() called.\n"); +#endif /* HDFS_DEBUG */ + + HDassert(f1->hdfs_handle != NULL); + HDassert(f2->hdfs_handle != NULL); + HDassert(f1->hdfs_handle->magic == HDFS_HDFST_MAGIC); + HDassert(f2->hdfs_handle->magic == HDFS_HDFST_MAGIC); + + finfo1 = f1->hdfs_handle->fileinfo; + finfo2 = f2->hdfs_handle->fileinfo; + HDassert(finfo1 != NULL); + HDassert(finfo2 != NULL); + + if (finfo1->mKind != finfo2->mKind) HGOTO_DONE(-1); + if (finfo1->mName != finfo2->mName) HGOTO_DONE(-1); + if (finfo1->mLastMod != finfo2->mLastMod) HGOTO_DONE(-1); + if (finfo1->mSize != finfo2->mSize) HGOTO_DONE(-1); + if (finfo1->mReplication != finfo2->mReplication) HGOTO_DONE(-1); + if (finfo1->mBlockSize != finfo2->mBlockSize) HGOTO_DONE(-1); + if (strcmp(finfo1->mOwner, finfo2->mOwner)) HGOTO_DONE(-1); + if (strcmp(finfo1->mGroup, finfo2->mGroup)) HGOTO_DONE(-1); + if (finfo1->mPermissions != finfo2->mPermissions) HGOTO_DONE(-1); + if (finfo1->mLastAccess != finfo2->mLastAccess) HGOTO_DONE(-1); +#endif /* H5_HAVE_LIBHDFS */ + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_hdfs_cmp() */ + + +/*------------------------------------------------------------------------- + * Function: H5FD_hdfs_query + * + * Purpose: Set the flags that this VFL driver is capable of supporting. + * (listed in H5FDpublic.h) + * + * Note that since the HDFS VFD is read only, most flags + * are irrelevant. + * + * The term "set" is highly misleading... + * stores/copies the supported flags in the out-pointer `flags`. + * + * Return: SUCCEED (Can't fail) + * + * Programmer: John Mainzer + * 9/11/17 + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_hdfs_query( + const H5FD_t H5_ATTR_UNUSED *_file, + unsigned long *flags) /* out variable */ +{ + FUNC_ENTER_NOAPI_NOINIT_NOERR + +#if HDFS_DEBUG + HDfprintf(stdout, "H5FD_hdfs_query() called.\n"); +#endif + + if (flags) { + *flags = 0; + *flags |= H5FD_FEAT_DATA_SIEVE; + } + + FUNC_LEAVE_NOAPI(SUCCEED) + +} /* H5FD_hdfs_query() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_hdfs_get_eoa() + * + * Purpose: + * + * Gets the end-of-address marker for the file. The EOA marker + * is the first address past the last byte allocated in the + * format address space. + * + * Return: + * + * The end-of-address marker. + * + * Programmer: Jacob Smith + * 2017-11-02 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static haddr_t +H5FD_hdfs_get_eoa( + const H5FD_t *_file, + H5FD_mem_t H5_ATTR_UNUSED type) +{ +#ifdef H5_HAVE_LIBHDFS + const H5FD_hdfs_t *file = (const H5FD_hdfs_t *)_file; +#endif /* H5_HAVE_LIBHDFS */ + + FUNC_ENTER_NOAPI_NOINIT_NOERR + +#if HDFS_DEBUG + HDfprintf(stdout, "H5FD_hdfs_get_eoa() called.\n"); +#endif + +#ifdef H5_HAVE_LIBHDFS + FUNC_LEAVE_NOAPI(file->eoa) +#else + FUNC_LEAVE_NOAPI(0) +#endif /* H5_HAVE_LIBHDFS */ + +} /* end H5FD_hdfs_get_eoa() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_hdfs_set_eoa() + * + * Purpose: + * + * Set the end-of-address marker for the file. + * + * Return: + * + * SUCCEED (can't fail) + * + * Programmer: Jacob Smith + * 2017-11-03 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_hdfs_set_eoa( + H5FD_t *_file, + H5FD_mem_t H5_ATTR_UNUSED type, + haddr_t addr) +{ +#ifdef H5_HAVE_LIBHDFS + H5FD_hdfs_t *file = (H5FD_hdfs_t *)_file; +#endif /* H5_HAVE_LIBHDFS */ + + FUNC_ENTER_NOAPI_NOINIT_NOERR + +#if HDFS_DEBUG + HDfprintf(stdout, "H5FD_hdfs_set_eoa() called.\n"); +#endif + +#ifdef H5_HAVE_LIBHDFS + file->eoa = addr; + + FUNC_LEAVE_NOAPI(SUCCEED) +#else + FUNC_LEAVE_NOAPI(FAIL) +#endif /* H5_HAVE_LIBHDFS */ + +} /* H5FD_hdfs_set_eoa() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_hdfs_get_eof() + * + * Purpose: + * + * Returns the end-of-file marker. + * + * Return: + * + * EOF: the first address past the end of the "file", either the + * filesystem file or the HDF5 file. + * + * Programmer: Jacob Smith + * 2017-11-02 + * + *------------------------------------------------------------------------- + */ +static haddr_t +H5FD_hdfs_get_eof( + const H5FD_t *_file, + H5FD_mem_t H5_ATTR_UNUSED type) +{ +#ifdef H5_HAVE_LIBHDFS + const H5FD_hdfs_t *file = (const H5FD_hdfs_t *)_file; +#endif /* H5_HAVE_LIBHDFS */ + + FUNC_ENTER_NOAPI_NOINIT_NOERR + +#if HDFS_DEBUG + HDfprintf(stdout, "H5FD_hdfs_get_eof() called.\n"); +#endif + +#ifdef H5_HAVE_LIBHDFS + HDassert(file->hdfs_handle != NULL); + HDassert(file->hdfs_handle->magic == HDFS_HDFST_MAGIC); + + FUNC_LEAVE_NOAPI((size_t) file->hdfs_handle->fileinfo->mSize) +#else + FUNC_LEAVE_NOAPI((size_t)0) +#endif /* H5_HAVE_LIBHDFS */ + +} /* end H5FD_hdfs_get_eof() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_hdfs_get_handle() + * + * Purpose: + * + * Returns the HDFS handle (hdfs_t) of hdfs file driver. + * + * Returns: + * + * SUCCEED/FAIL + * + * Programmer: Jacob Smith + * 2017-11-02 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_hdfs_get_handle( + H5FD_t *_file, + hid_t H5_ATTR_UNUSED fapl, + void **file_handle) +{ + herr_t ret_value = SUCCEED; +#ifdef H5_HAVE_LIBHDFS + H5FD_hdfs_t *file = (H5FD_hdfs_t *)_file; +#endif + + FUNC_ENTER_NOAPI_NOINIT + +#if HDFS_DEBUG + HDfprintf(stdout, "H5FD_hdfs_get_handle() called.\n"); +#endif /* HDFS_DEBUG */ + +#ifdef H5_HAVE_LIBHDFS + if (!file_handle) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "file handle not valid") + } + + *file_handle = file->hdfs_handle; +#else + HGOTO_ERROR(H5E_VFL, H5E_UNSUPPORTED, FAIL, + "Illegal get-handle of unsupported virtual file (hdfs)"); +#endif /* H5_HAVE_LIBHDFS */ + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* end H5FD_hdfs_get_handle() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_hdfs_read() + * + * Purpose: + * + * Reads SIZE bytes of data from FILE beginning at address ADDR + * into buffer BUF according to data transfer properties in DXPL_ID. + * + * Return: + * + * Success: `SUCCEED` + * - Result is stored in caller-supplied buffer BUF. + * Failure: `FAIL` + * - Unable to complete read. + * - Contents of buffer `buf` are undefined. + * + * Programmer: Jacob Smith + * 2017-11-?? + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_hdfs_read( + H5FD_t *_file, + H5FD_mem_t H5_ATTR_UNUSED type, + hid_t H5_ATTR_UNUSED dxpl_id, + haddr_t addr, /* start offset */ + size_t size, /* length of read */ + void *buf) /* out */ +{ + herr_t ret_value = SUCCEED; +#if H5_HAVE_LIBHDFS + H5FD_hdfs_t *file = (H5FD_hdfs_t *)_file; + size_t filesize = 0; +#endif /* H5_HAVE_LIBHDFS */ +#if HDFS_STATS + /* working variables for storing stats */ + hdfs_statsbin *bin = NULL; + unsigned bin_i = 0; +#endif /* HDFS_STATS */ + + FUNC_ENTER_NOAPI_NOINIT + +#if HDFS_DEBUG + HDfprintf(stdout, "H5FD_hdfs_read() called.\n"); +#endif /* HDFS_DEBUG */ + +#ifndef H5_HAVE_LIBHDFS + HGOTO_ERROR(H5E_VFL, H5E_UNSUPPORTED, FAIL, + "Illegal get-handle of unsupported virtual file (hdfs)"); +#else + HDassert(file != NULL); + HDassert(file->hdfs_handle != NULL); + HDassert(file->hdfs_handle->magic == HDFS_HDFST_MAGIC); + HDassert(buf != NULL); + + filesize = (size_t) file->hdfs_handle->fileinfo->mSize; + + if ((addr > filesize) || ((addr + size) > filesize)) { + HGOTO_ERROR(H5E_ARGS, H5E_OVERFLOW, FAIL, + "range exceeds file address") + } + + if (FAIL == hdfsPread( + file->hdfs_handle->filesystem, + file->hdfs_handle->file, + (tOffset)addr, + buf, + (tSize)size)) + { + HGOTO_ERROR(H5E_VFL, H5E_READERROR, FAIL, + "unable to execute read") + } + +#if HDFS_STATS + + /* Find which "bin" this read fits in. Can be "overflow" bin. + */ + for (bin_i = 0; bin_i < HDFS_STATS_BIN_COUNT; bin_i++) { + if ((unsigned long long)size < hdfs_stats_boundaries[bin_i]) { + break; + } + } + bin = (type == H5FD_MEM_DRAW) + ? &file->raw[bin_i] + : &file->meta[bin_i]; + + /* Store collected stats in appropriate bin + */ + if (bin->count == 0) { + bin->min = size; + bin->max = size; + } else { + if (size < bin->min) bin->min = size; + if (size > bin->max) bin->max = size; + } + bin->count++; + bin->bytes += (unsigned long long)size; + +#endif /* HDFS_STATS */ +#endif /* H5_HAVE_LIBHDFS */ + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* end H5FD_hdfs_read() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_hdfs_write() + * + * Purpose: + * + * Write bytes to file. + * UNSUPPORTED IN READ-ONLY HDFS VFD. + * + * Return: + * + * FAIL (Not possible with Read-Only S3 file.) + * + * Programmer: Jacob Smith + * 2017-10-23 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_hdfs_write( + H5FD_t H5_ATTR_UNUSED *_file, + H5FD_mem_t H5_ATTR_UNUSED type, + hid_t H5_ATTR_UNUSED dxpl_id, + haddr_t H5_ATTR_UNUSED addr, + size_t H5_ATTR_UNUSED size, + const void H5_ATTR_UNUSED *buf) +{ + herr_t ret_value = FAIL; + + FUNC_ENTER_NOAPI_NOINIT + +#if HDFS_DEBUG + HDfprintf(stdout, "H5FD_hdfs_write() called.\n"); +#endif + + HGOTO_ERROR(H5E_VFL, H5E_UNSUPPORTED, FAIL, + "cannot write to read-only file.") + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_hdfs_write() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_hdfs_truncate() + * + * Purpose: + * + * Makes sure that the true file size is the same (or larger) + * than the end-of-address. + * + * NOT POSSIBLE ON READ-ONLY S3 FILES. + * + * Return: + * + * FAIL (Not possible on Read-Only S3 files.) + * + * Programmer: Jacob Smith + * 2017-10-23 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_hdfs_truncate( + H5FD_t H5_ATTR_UNUSED *_file, + hid_t H5_ATTR_UNUSED dxpl_id, + hbool_t H5_ATTR_UNUSED closing) +{ + herr_t ret_value = SUCCEED; + + FUNC_ENTER_NOAPI_NOINIT + +#if HDFS_DEBUG + HDfprintf(stdout, "H5FD_hdfs_truncate() called.\n"); +#endif + + HGOTO_ERROR(H5E_VFL, H5E_UNSUPPORTED, FAIL, + "cannot truncate read-only file.") + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* end H5FD_hdfs_truncate() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_hdfs_lock() + * + * Purpose: + * + * Place an advisory lock on a file. + * No effect on Read-Only S3 file. + * + * Suggestion: remove lock/unlock from class + * > would result in error at H5FD_[un]lock() (H5FD.c) + * + * Return: + * + * SUCCEED (No-op always succeeds) + * + * Programmer: Jacob Smith + * 2017-11-03 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_hdfs_lock( + H5FD_t H5_ATTR_UNUSED *_file, + hbool_t H5_ATTR_UNUSED rw) +{ + FUNC_ENTER_NOAPI_NOINIT_NOERR + FUNC_LEAVE_NOAPI(SUCCEED) + +} /* end H5FD_hdfs_lock() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_hdfs_unlock() + * + * Purpose: + * + * Remove the existing lock on the file. + * No effect on Read-Only S3 file. + * + * Return: + * + * SUCCEED (No-op always succeeds) + * + * Programmer: Jacob Smith + * 2017-11-03 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_hdfs_unlock(H5FD_t H5_ATTR_UNUSED *_file) +{ + FUNC_ENTER_NOAPI_NOINIT_NOERR + FUNC_LEAVE_NOAPI(SUCCEED) + +} /* end H5FD_hdfs_unlock() */ + diff --git a/src/H5FDhdfs.h b/src/H5FDhdfs.h new file mode 100644 index 0000000..3d4128d --- /dev/null +++ b/src/H5FDhdfs.h @@ -0,0 +1,122 @@ +/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * Read-Only HDFS Virtual File Driver (VFD) * + * Copyright (c) 2018, The HDF Group. * + * * + * All rights reserved. * + * * + * NOTICE: * + * All information contained herein is, and remains, the property of The HDF * + * Group. The intellectual and technical concepts contained herein are * + * proprietary to The HDF Group. Dissemination of this information or * + * reproduction of this material is strictly forbidden unless prior written * + * permission is obtained from The HDF Group. * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ + +/* + * Programmer: Jacob Smith + * 2018-04-23 + * + * Purpose: The public header file for the hdfs driver. + */ + +#ifndef H5FDhdfs_H +#define H5FDhdfs_H + +#define H5FD_HDFS (H5FD_hdfs_init()) + +#ifdef __cplusplus +extern "C" { +#endif + +/**************************************************************************** + * + * Structure: H5FD_hdfs_fapl_t + * + * Purpose: + * + * H5FD_hdfs_fapl_t is a public structure that is used to pass + * configuration information to the appropriate HDFS VFD via the FAPL. + * A pointer to an instance of this structure is a parameter to + * H5Pset_fapl_hdfs() and H5Pget_fapl_hdfs(). + * + * + * + * `version` (int32_t) + * + * Version number of the `H5FD_hdfs_fapl_t` structure. Any instance passed + * to the above calls must have a recognized version number, or an error + * will be flagged. + * + * This field should be set to `H5FD__CURR_HDFS_FAPL_T_VERSION`. + * + * `namenode_name` (const char[]) + * + * Name of "Name Node" to access as the HDFS server. + * + * Must not be longer than `H5FD__HDFS_NODE_NAME_SPACE`. + * + * TBD: Can be NULL. + * + * `namenode_port` (int32_t) TBD + * + * Port number to use to connect with Name Node. + * + * TBD: If 0, uses a default port. + * + * `kerberos_ticket_cache` (const char[]) + * + * Path to the location of the Kerberos authentication cache. + * + * Must not be longer than `H5FD__HDFS_KERB_CACHE_PATH_SPACE`. + * + * TBD: Can be NULL. + * + * `user_name` (const char[]) + * + * Username to use when accessing file. + * + * Must not be longer than `H5FD__HDFS_USER_NAME_SPACE`. + * + * TBD: Can be NULL. + * + * `stream_buffer_size` (int32_t) + * + * Size (in bytes) of the file read stream buffer. + * + * TBD: If -1, relies on a default value. + * + * + * + * Programmer: Jacob Smith + * 2018-04-23 + * + * Changes: None + * + ****************************************************************************/ + +#define H5FD__CURR_HDFS_FAPL_T_VERSION 1 + +#define H5FD__HDFS_NODE_NAME_SPACE 128 +#define H5FD__HDFS_USER_NAME_SPACE 128 +#define H5FD__HDFS_KERB_CACHE_PATH_SPACE 128 + +typedef struct H5FD_hdfs_fapl_t { + int32_t version; + char namenode_name[H5FD__HDFS_NODE_NAME_SPACE + 1]; + int32_t namenode_port; + char user_name[H5FD__HDFS_USER_NAME_SPACE + 1]; + char kerberos_ticket_cache[H5FD__HDFS_KERB_CACHE_PATH_SPACE + 1]; + int32_t stream_buffer_size; +} H5FD_hdfs_fapl_t; + +H5_DLL hid_t H5FD_hdfs_init(void); +H5_DLL herr_t H5Pget_fapl_hdfs(hid_t fapl_id, H5FD_hdfs_fapl_t *fa_out); +H5_DLL herr_t H5Pset_fapl_hdfs(hid_t fapl_id, H5FD_hdfs_fapl_t *fa); + +#ifdef __cplusplus +} +#endif + +#endif /* ifndef H5FDhdfs_H */ + + diff --git a/src/H5FDros3.c b/src/H5FDros3.c new file mode 100644 index 0000000..8bf0420 --- /dev/null +++ b/src/H5FDros3.c @@ -0,0 +1,1847 @@ +/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * Read-Only S3 Virtual File Driver (VFD) * + * Copyright (c) 2017-2018, The HDF Group. * + * * + * All rights reserved. * + * * + * NOTICE: * + * All information contained herein is, and remains, the property of The HDF * + * Group. The intellectual and technical concepts contained herein are * + * proprietary to The HDF Group. Dissemination of this information or * + * reproduction of this material is strictly forbidden unless prior written * + * permission is obtained from The HDF Group. * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ + +/* + * Programmer: Jacob Smith <jake.smith@hdfgroup.org> + * 2017-10-13 + * + * Purpose: + * + * Provide read-only access to files hosted on Amazon's S3 service. + * Relies on "s3comms" utility layer to implement the AWS REST API. + */ + +/* This source code file is part of the H5FD driver module */ +#include "H5FDdrvr_module.h" + +#include "H5private.h" /* Generic Functions */ +#include "H5Eprivate.h" /* Error handling */ +#include "H5FDprivate.h" /* File drivers */ +#include "H5FDros3.h" /* ros3 file driver */ +#include "H5FLprivate.h" /* Free Lists */ +#include "H5Iprivate.h" /* IDs */ +#include "H5MMprivate.h" /* Memory management */ +#include "H5FDs3comms.h" /* S3 Communications */ + +/* toggle function call prints: 1 turns on + */ +#define ROS3_DEBUG 0 + +/* toggle stats collection and reporting + */ +#define ROS3_STATS 0 + +/* The driver identification number, initialized at runtime + */ +static hid_t H5FD_ROS3_g = 0; + +#if ROS3_STATS + +/* arbitrarily large value, such that any reasonable size read will be "less" + * than this value and set a true minimum + * not 0 because that may be a valid recorded minimum in degenerate cases + */ +#define ROS3_STATS_STARTING_MIN 0xfffffffful + +/* Configuration definitions for stats collection and breakdown + * + * 2^10 = 1024 + * Reads up to 1024 bytes (1 kB) fall in bin 0 + * 2^(10+(1*16)) = 2^26 = 64MB + * Reads of 64MB or greater fall in "overflow" bin[BIN_COUNT] + */ +#define ROS3_STATS_BASE 2 +#define ROS3_STATS_INTERVAL 1 +#define ROS3_STATS_START_POWER 10 +#define ROS3_STATS_BIN_COUNT 16 /* MUST BE GREATER THAN 0 */ + + +/* + * Calculate `BASE ^ (START_POWER + (INTERVAL * bin_i))` + * Stores result at `(unsigned long long *) out_ptr`. + * Used in computing boundaries between stats bins. + */ +#define ROS3_STATS_POW(bin_i, out_ptr) { \ + unsigned long long donotshadowresult = 1; \ + unsigned donotshadowindex = 0; \ + for (donotshadowindex = 0; \ + donotshadowindex < (((bin_i) * ROS3_STATS_INTERVAL) + \ + ROS3_STATS_START_POWER); \ + donotshadowindex++) \ + { \ + donotshadowresult *= ROS3_STATS_BASE; \ + } \ + *(out_ptr) = donotshadowresult; \ +} + +/* array to hold pre-computed boundaries for stats bins + */ +static unsigned long long ros3_stats_boundaries[ROS3_STATS_BIN_COUNT]; + +/*************************************************************************** + * + * Structure: ros3_statsbin + * + * Purpose: + * + * Structure for storing per-file ros3 VFD usage statistics. + * + * + * + * `count` (unsigned long long) + * + * Number of reads with size in this bin's range. + * + * `bytes` (unsigned long long) + * + * Total number of bytes read through this bin. + * + * `min` (unsigned long long) + * + * Smallest read size in this bin. + * + * `max` (unsigned long long) + * + * Largest read size in this bin. + * + * + * + * Programmer: Jacob Smith + * + * Changes: None + * + ***************************************************************************/ +typedef struct { + unsigned long long count; + unsigned long long bytes; + unsigned long long min; + unsigned long long max; +} ros3_statsbin; + +#endif /* ROS3_STATS */ + +/*************************************************************************** + * + * Structure: H5FD_ros3_t + * + * Purpose: + * + * H5FD_ros3_t is a structure used to store all information needed to + * maintain R/O access to a single HDF5 file that has been stored as a + * S3 object. This structure is created when such a file is "opened" and + * discarded when it is "closed". + * + * Presents an S3 object as a file to the HDF5 library. + * + * + * + * `pub` (H5FD_t) + * + * Instance of H5FD_t which contains all fields common to all VFDs. + * It must be the first item in this structure, since at higher levels, + * this structure will be treated as an instance of H5FD_t. + * + * `fa` (H5FD_ros3_fapl_t) + * + * Instance of `H5FD_ros3_fapl_t` containing the S3 configuration data + * needed to "open" the HDF5 file. + * + * `eoa` (haddr_t) + * + * End of addressed space in file. After open, it should always + * equal the file size. + * + * `s3r_handle` (s3r_t *) + * + * Instance of S3 Request handle associated with the target resource. + * Responsible for communicating with remote host and presenting file + * contents as indistinguishable from a file on the local filesystem. + * + * *** present only if ROS3_SATS is flagged to enable stats collection *** + * + * `meta` (ros3_statsbin[]) + * `raw` (ros3_statsbin[]) + * + * Only present if ros3 stats collection is enabled. + * + * Arrays of `ros3_statsbin` structures to record raw- and metadata reads. + * + * Records count and size of reads performed by the VFD, and is used to + * print formatted usage statistics to stdout upon VFD shutdown. + * + * Reads of each raw- and metadata type are recorded in an individual bin + * determined by the size of the read. The last bin of each type is + * reserved for "big" reads, with no defined upper bound. + * + * *** end ROS3_STATS *** + * + * + * + * Programmer: Jacob Smith + * + * Changes: None. + * + ***************************************************************************/ +typedef struct H5FD_ros3_t { + H5FD_t pub; + H5FD_ros3_fapl_t fa; + haddr_t eoa; + s3r_t *s3r_handle; +#if ROS3_STATS + ros3_statsbin meta[ROS3_STATS_BIN_COUNT + 1]; + ros3_statsbin raw[ROS3_STATS_BIN_COUNT + 1]; +#endif +} H5FD_ros3_t; + +/* + * These macros check for overflow of various quantities. These macros + * assume that HDoff_t is signed and haddr_t and size_t are unsigned. + * + * ADDR_OVERFLOW: Checks whether a file address of type `haddr_t' + * is too large to be represented by the second argument + * of the file seek function. + * + */ +#define MAXADDR (((haddr_t)1<<(8*sizeof(HDoff_t)-1))-1) +#define ADDR_OVERFLOW(A) (HADDR_UNDEF==(A) || ((A) & ~(haddr_t)MAXADDR)) + +/* Prototypes */ +static herr_t H5FD_ros3_term(void); +static void *H5FD_ros3_fapl_get(H5FD_t *_file); +static void *H5FD_ros3_fapl_copy(const void *_old_fa); +static herr_t H5FD_ros3_fapl_free(void *_fa); +static H5FD_t *H5FD_ros3_open(const char *name, unsigned flags, hid_t fapl_id, + haddr_t maxaddr); +static herr_t H5FD_ros3_close(H5FD_t *_file); +static int H5FD_ros3_cmp(const H5FD_t *_f1, const H5FD_t *_f2); +static herr_t H5FD_ros3_query(const H5FD_t *_f1, unsigned long *flags); +static haddr_t H5FD_ros3_get_eoa(const H5FD_t *_file, H5FD_mem_t type); +static herr_t H5FD_ros3_set_eoa(H5FD_t *_file, H5FD_mem_t type, haddr_t addr); +static haddr_t H5FD_ros3_get_eof(const H5FD_t *_file, H5FD_mem_t type); +static herr_t H5FD_ros3_get_handle(H5FD_t *_file, hid_t fapl, + void** file_handle); +static herr_t H5FD_ros3_read(H5FD_t *_file, H5FD_mem_t type, hid_t fapl_id, + haddr_t addr, size_t size, void *buf); +static herr_t H5FD_ros3_write(H5FD_t *_file, H5FD_mem_t type, hid_t fapl_id, + haddr_t addr, size_t size, const void *buf); +static herr_t H5FD_ros3_truncate(H5FD_t *_file, hid_t dxpl_id, + hbool_t closing); +static herr_t H5FD_ros3_lock(H5FD_t *_file, hbool_t rw); +static herr_t H5FD_ros3_unlock(H5FD_t *_file); +static herr_t H5FD_ros3_validate_config(const H5FD_ros3_fapl_t * fa); + +static const H5FD_class_t H5FD_ros3_g = { + "ros3", /* name */ + MAXADDR, /* maxaddr */ + H5F_CLOSE_WEAK, /* fc_degree */ + H5FD_ros3_term, /* terminate */ + NULL, /* sb_size */ + NULL, /* sb_encode */ + NULL, /* sb_decode */ + sizeof(H5FD_ros3_fapl_t), /* fapl_size */ + H5FD_ros3_fapl_get, /* fapl_get */ + H5FD_ros3_fapl_copy, /* fapl_copy */ + H5FD_ros3_fapl_free, /* fapl_free */ + 0, /* dxpl_size */ + NULL, /* dxpl_copy */ + NULL, /* dxpl_free */ + H5FD_ros3_open, /* open */ + H5FD_ros3_close, /* close */ + H5FD_ros3_cmp, /* cmp */ + H5FD_ros3_query, /* query */ + NULL, /* get_type_map */ + NULL, /* alloc */ + NULL, /* free */ + H5FD_ros3_get_eoa, /* get_eoa */ + H5FD_ros3_set_eoa, /* set_eoa */ + H5FD_ros3_get_eof, /* get_eof */ + H5FD_ros3_get_handle, /* get_handle */ + H5FD_ros3_read, /* read */ + H5FD_ros3_write, /* write */ + NULL, /* flush */ + H5FD_ros3_truncate, /* truncate */ + H5FD_ros3_lock, /* lock */ + H5FD_ros3_unlock, /* unlock */ + H5FD_FLMAP_DICHOTOMY /* fl_map */ +}; + +/* Declare a free list to manage the H5FD_ros3_t struct */ +H5FL_DEFINE_STATIC(H5FD_ros3_t); + + +/*------------------------------------------------------------------------- + * Function: H5FD__init_package + * + * Purpose: Initializes any interface-specific data or routines. + * + * Return: Non-negative on success/Negative on failure + * + * Changes: Rename as appropriate for ros3 vfd. + * Jacob Smith 2017 + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD__init_package(void) +{ + herr_t ret_value = SUCCEED; + + FUNC_ENTER_STATIC + + if (H5FD_ros3_init() < 0) { + HGOTO_ERROR(H5E_VFL, H5E_CANTINIT, FAIL, + "unable to initialize ros3 VFD") + } + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD__init_package() */ + + +/*------------------------------------------------------------------------- + * Function: H5FD_ros3_init + * + * Purpose: Initialize this driver by registering the driver with the + * library. + * + * Return: Success: The driver ID for the ros3 driver. + * Failure: Negative + * + * Programmer: Robb Matzke + * Thursday, July 29, 1999 + * + * Changes: Rename as appropriate for ros3 vfd. + * Jacob Smith 2017 + * + *------------------------------------------------------------------------- + */ +hid_t +H5FD_ros3_init(void) +{ + hid_t ret_value = H5I_INVALID_HID; /* Return value */ + + FUNC_ENTER_NOAPI(FAIL) + +#if ROS3_DEBUG + HDfprintf(stdout, "H5FD_ros3_init() called.\n"); +#endif + + if (H5I_VFL != H5I_get_type(H5FD_ROS3_g)) + H5FD_ROS3_g = H5FD_register(&H5FD_ros3_g, sizeof(H5FD_class_t), FALSE); + +#if ROS3_STATS + /* pre-compute statsbin boundaries + */ + for (unsigned bin_i = 0; bin_i < ROS3_STATS_BIN_COUNT; bin_i++) { + unsigned long long value = 0; + ROS3_STATS_POW(bin_i, &value) + ros3_stats_boundaries[bin_i] = value; + } +#endif + + /* Set return value */ + ret_value = H5FD_ROS3_g; + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* end H5FD_ros3_init() */ + + +/*--------------------------------------------------------------------------- + * Function: H5FD_ros3_term + * + * Purpose: Shut down the VFD + * + * Returns: SUCCEED (Can't fail) + * + * Programmer: Quincey Koziol + * Friday, Jan 30, 2004 + * + * Changes: Rename as appropriate for ros3 vfd. + * Jacob Smith 2017 + * + *--------------------------------------------------------------------------- + */ +static herr_t +H5FD_ros3_term(void) +{ + FUNC_ENTER_NOAPI_NOINIT_NOERR + +#if ROS3_DEBUG + HDfprintf(stdout, "H5FD_ros3_term() called.\n"); +#endif + + /* Reset VFL ID */ + H5FD_ROS3_g = 0; + + FUNC_LEAVE_NOAPI(SUCCEED) +} /* end H5FD_ros3_term() */ + + +/*------------------------------------------------------------------------- + * Function: H5Pset_fapl_ros3 + * + * Purpose: Modify the file access property list to use the H5FD_ROS3 + * driver defined in this source file. All driver specfic + * properties are passed in as a pointer to a suitably + * initialized instance of H5FD_ros3_fapl_t + * + * Return: SUCCEED/FAIL + * + * Programmer: John Mainzer + * 9/10/17 + * + *------------------------------------------------------------------------- + */ +herr_t +H5Pset_fapl_ros3(hid_t fapl_id, + H5FD_ros3_fapl_t *fa) +{ + H5P_genplist_t *plist = NULL; /* Property list pointer */ + herr_t ret_value = FAIL; + + + + FUNC_ENTER_API(FAIL) + H5TRACE2("e", "i*x", fapl_id, fa); + + HDassert(fa != NULL); + +#if ROS3_DEBUG + HDfprintf(stdout, "H5Pset_fapl_ros3() called.\n"); +#endif + + plist = H5P_object_verify(fapl_id, H5P_FILE_ACCESS); + if (plist == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADTYPE, FAIL, \ + "not a file access property list") + } + + if (FAIL == H5FD_ros3_validate_config(fa)) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, "invalid ros3 config") + + ret_value = H5P_set_driver(plist, H5FD_ROS3, (void *)fa); + +done: + + FUNC_LEAVE_API(ret_value) + +} /* H5Pset_fapl_ros3() */ + + +/*------------------------------------------------------------------------- + * Function: H5FD_ros3_validate_config() + * + * Purpose: Test to see if the supplied instance of H5FD_ros3_fapl_t + * contains internally consistant data. Return SUCCEED if so, + * and FAIL otherwise. + * + * Note the difference between internally consistant and + * correct. As we will have to try to access the target + * object to determine whether the supplied data is correct, + * we will settle for internal consistancy at this point + * + * Return: SUCCEED if instance of H5FD_ros3_fapl_t contains internally + * consistant data, FAIL otherwise. + * + * Programmer: Jacob Smith + * 9/10/17 + * + * Changes: Add checks for authenticate flag requring populated + * `aws_region` and `secret_id` strings. + * -- Jacob Smith 2017-11-01 + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_ros3_validate_config(const H5FD_ros3_fapl_t * fa) +{ + herr_t ret_value = SUCCEED; + + FUNC_ENTER_NOAPI_NOINIT + + HDassert(fa != NULL); + + if ( fa->version != H5FD__CURR_ROS3_FAPL_T_VERSION ) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "Unknown H5FD_ros3_fapl_t version"); + } + + /* if set to authenticate, region and id cannot be empty strings + */ + if (fa->authenticate == TRUE) { + if ((fa->aws_region[0] == '\0') || + (fa->secret_id[0] == '\0')) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "Inconsistent authentication information"); + } + } + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_ros3_validate_config() */ + + +/*------------------------------------------------------------------------- + * Function: H5Pget_fapl_ros3 + * + * Purpose: Returns information about the ros3 file access property + * list though the function arguments. + * + * Return: Success: Non-negative + * + * Failure: Negative + * + * Programmer: John Mainzer + * 9/10/17 + * + * Modifications: + * + *------------------------------------------------------------------------- + */ +herr_t +H5Pget_fapl_ros3(hid_t fapl_id, + H5FD_ros3_fapl_t *fa_out) +{ + const H5FD_ros3_fapl_t *fa; + H5P_genplist_t *plist = NULL; /* Property list pointer */ + herr_t ret_value = SUCCEED; /* Return value */ + + FUNC_ENTER_API(FAIL) + H5TRACE2("e", "i*x", fapl_id, fa_out); + +#if ROS3_DEBUG + HDfprintf(stdout, "H5Pget_fapl_ros3() called.\n"); +#endif + + if (fa_out == NULL) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, "fa_out is NULL") + + plist = H5P_object_verify(fapl_id, H5P_FILE_ACCESS); + if (plist == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADTYPE, FAIL, "not a file access list") + } + + if (H5FD_ROS3 != H5P_peek_driver(plist)) { + HGOTO_ERROR(H5E_PLIST, H5E_BADVALUE, FAIL, "incorrect VFL driver") + } + + fa = (const H5FD_ros3_fapl_t *)H5P_peek_driver_info(plist); + if (fa == NULL) { + HGOTO_ERROR(H5E_PLIST, H5E_BADVALUE, FAIL, "bad VFL driver info") + } + + /* Copy the ros3 fapl data out */ + HDmemcpy(fa_out, fa, sizeof(H5FD_ros3_fapl_t)); + +done: + FUNC_LEAVE_API(ret_value) + +} /* H5Pget_fapl_ros3() */ + + +/*------------------------------------------------------------------------- + * Function: H5FD_ros3_fapl_get + * + * Purpose: Gets a file access property list which could be used to + * create an identical file. + * + * Return: Success: Ptr to new file access property list value. + * + * Failure: NULL + * + * Programmer: John Mainzer + * 9/8/17 + * + * Modifications: + * + *------------------------------------------------------------------------- + */ +static void * +H5FD_ros3_fapl_get(H5FD_t *_file) +{ + H5FD_ros3_t *file = (H5FD_ros3_t*)_file; + H5FD_ros3_fapl_t *fa = NULL; + void *ret_value = NULL; + + FUNC_ENTER_NOAPI_NOINIT + + fa = (H5FD_ros3_fapl_t *)H5MM_calloc(sizeof(H5FD_ros3_fapl_t)); + if (fa == NULL) { + HGOTO_ERROR(H5E_RESOURCE, H5E_NOSPACE, NULL, + "memory allocation failed") + } + + /* Copy the fields of the structure */ + HDmemcpy(fa, &(file->fa), sizeof(H5FD_ros3_fapl_t)); + + /* Set return value */ + ret_value = fa; + +done: + if (ret_value == NULL) { + if (fa != NULL) + H5MM_xfree(fa); + } + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_ros3_fapl_get() */ + + +/*------------------------------------------------------------------------- + * Function: H5FD_ros3_fapl_copy + * + * Purpose: Copies the ros3-specific file access properties. + * + * Return: Success: Ptr to a new property list + * + * Failure: NULL + * + * Programmer: John Mainzer + * 9/8/17 + * + * Modifications: + * + *------------------------------------------------------------------------- + */ +static void * +H5FD_ros3_fapl_copy(const void *_old_fa) +{ + const H5FD_ros3_fapl_t *old_fa = (const H5FD_ros3_fapl_t*)_old_fa; + H5FD_ros3_fapl_t *new_fa = NULL; + void *ret_value = NULL; /* Return value */ + + FUNC_ENTER_NOAPI_NOINIT + + new_fa = (H5FD_ros3_fapl_t *)H5MM_malloc(sizeof(H5FD_ros3_fapl_t)); + if (new_fa == NULL) { + HGOTO_ERROR(H5E_RESOURCE, H5E_NOSPACE, NULL, + "memory allocation failed"); + } + + HDmemcpy(new_fa, old_fa, sizeof(H5FD_ros3_fapl_t)); + ret_value = new_fa; + +done: + if (ret_value == NULL) { + if (new_fa != NULL) + H5MM_xfree(new_fa); + } + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_ros3_fapl_copy() */ + + +/*------------------------------------------------------------------------- + * Function: H5FD_ros3_fapl_free + * + * Purpose: Frees the ros3-specific file access properties. + * + * Return: SUCCEED (cannot fail) + * + * Programmer: John Mainzer + * 9/8/17 + * + * Modifications: + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_ros3_fapl_free(void *_fa) +{ + H5FD_ros3_fapl_t *fa = (H5FD_ros3_fapl_t*)_fa; + + FUNC_ENTER_NOAPI_NOINIT_NOERR + + HDassert(fa != NULL); /* sanity check */ + + H5MM_xfree(fa); + + FUNC_LEAVE_NOAPI(SUCCEED) + +} /* H5FD_ros3_fapl_free() */ + +#if ROS3_STATS + +/*---------------------------------------------------------------------------- + * + * Function: ros3_reset_stats() + * + * Purpose: + * + * Reset the stats collection elements in this virtual file structure. + * + * Clears any set data in stats bins; initializes/zeroes values. + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - FAILURE: `FAIL` + * - Occurs if the file is invalid somehow + * + * Programmer: Jacob Smith + * 2017-12-08 + * + * Changes: None. + * + *---------------------------------------------------------------------------- + */ +static herr_t +ros3_reset_stats(H5FD_ros3_t *file) +{ + unsigned i = 0; + herr_t ret_value = SUCCEED; + + + FUNC_ENTER_NOAPI_NOINIT + +#if ROS3_DEBUG + HDprintf("ros3_reset_stats() called\n"); +#endif + + if (file == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "file was null"); + } + + for (i = 0; i <= ROS3_STATS_BIN_COUNT; i++) { + file->raw[i].bytes = 0; + file->raw[i].count = 0; + file->raw[i].min = (unsigned long long)ROS3_STATS_STARTING_MIN; + file->raw[i].max = 0; + + file->meta[i].bytes = 0; + file->meta[i].count = 0; + file->meta[i].min = (unsigned long long)ROS3_STATS_STARTING_MIN; + file->meta[i].max = 0; + } + +done: + FUNC_LEAVE_NOAPI(ret_value); + +} /* ros3_reset_stats */ +#endif /* ROS3_STATS */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_ros3_open() + * + * Purpose: + * + * Create and/or opens a file as an HDF5 file. + * + * Any flag except H5F_ACC_RDONLY will cause an error. + * + * Name (as received from `H5FD_open()`) must conform to web url: + * NAME :: HTTP "://" DOMAIN [PORT] ["/" [URI] [QUERY] ] + * HTTP :: "http" [ "s" ] + * DOMAIN :: e.g., "mybucket.host.org" + * PORT :: ":" <number> (e.g., ":9000" ) + * URI :: <string> (e.g., "path/to/resource.hd5" ) + * QUERY :: "?" <string> (e.g., "arg1=param1&arg2=param2") + * + * Return: + * + * Success: A pointer to a new file data structure. + * The public fields will be initialized by the caller, which is + * always H5FD_open(). + * + * Failure: NULL + * + * Programmer: Jacob Smith + * 2017-11-02 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static H5FD_t * +H5FD_ros3_open(const char *url, + unsigned flags, + hid_t fapl_id, + haddr_t maxaddr) +{ +#ifdef H5_HAVE_ROS3_VFD + H5FD_ros3_t *file = NULL; + struct tm *now = NULL; + char iso8601now[ISO8601_SIZE]; + unsigned char signing_key[SHA256_DIGEST_LENGTH]; + s3r_t *handle = NULL; + H5FD_ros3_fapl_t fa; +#endif + H5FD_t *ret_value = NULL; + + + + FUNC_ENTER_NOAPI_NOINIT +#ifdef H5_HAVE_ROS3_VFD + +#if ROS3_DEBUG + HDfprintf(stdout, "H5FD_ros3_open() called.\n"); +#endif + + + /* Sanity check on file offsets */ + HDcompile_assert(sizeof(HDoff_t) >= sizeof(size_t)); + + /* Check arguments */ + if (!url || !*url) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, "invalid file name") + if (0 == maxaddr || HADDR_UNDEF == maxaddr) + HGOTO_ERROR(H5E_ARGS, H5E_BADRANGE, NULL, "bogus maxaddr") + if (ADDR_OVERFLOW(maxaddr)) + HGOTO_ERROR(H5E_ARGS, H5E_OVERFLOW, NULL, "bogus maxaddr") + if (flags != H5F_ACC_RDONLY) + HGOTO_ERROR(H5E_ARGS, H5E_UNSUPPORTED, NULL, + "only Read-Only access allowed") + + if (FAIL == H5Pget_fapl_ros3(fapl_id, &fa)) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, "can't get property list") + } + + if (CURLE_OK != curl_global_init(CURL_GLOBAL_DEFAULT)) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "unable to initialize curl global (placeholder flags)") + } + + /* open file; procedure depends on whether or not the fapl instructs to + * authenticate requests or not. + */ + if (fa.authenticate == TRUE) { + /* compute signing key (part of AWS/S3 REST API) + * can be re-used by user/key for 7 days after creation. + * find way to re-use/share + */ + now = gmnow(); + HDassert( now != NULL ); + if (ISO8601NOW(iso8601now, now) != (ISO8601_SIZE - 1)) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "problem while writing iso8601 timestamp") + } + if (FAIL == H5FD_s3comms_signing_key(signing_key, + (const char *)fa.secret_key, + (const char *)fa.aws_region, + (const char *)iso8601now) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "problem while computing signing key") + } + + handle = H5FD_s3comms_s3r_open( + url, + (const char *)fa.aws_region, + (const char *)fa.secret_id, + (const unsigned char *)signing_key); + } else { + handle = H5FD_s3comms_s3r_open(url, NULL, NULL, NULL); + } /* if/else should authenticate */ + + if (handle == NULL) { + /* If we want to check CURL's say on the matter in a controlled + * fashion, this is the place to do it, but would need to make a + * few minor changes to s3comms `s3r_t` and `s3r_read()`. + */ + HGOTO_ERROR(H5E_VFL, H5E_CANTOPENFILE, NULL, "could not open"); + } + + /* create new file struct + */ + file = H5FL_CALLOC(H5FD_ros3_t); + if (file == NULL) { + HGOTO_ERROR(H5E_RESOURCE, H5E_NOSPACE, NULL, + "unable to allocate file struct") + } + + file->s3r_handle = handle; + HDmemcpy(&(file->fa), &fa, sizeof(H5FD_ros3_fapl_t)); + +#if ROS3_STATS + if (FAIL == ros3_reset_stats(file)) { + HGOTO_ERROR(H5E_INTERNAL, H5E_UNINITIALIZED, NULL, + "unable to reset file statistics") + } +#endif /* ROS3_STATS */ + + ret_value = (H5FD_t*)file; + +done: + if (ret_value == NULL) { + if (handle != NULL) { + if (FAIL == H5FD_s3comms_s3r_close(handle)) { + HDONE_ERROR(H5E_VFL, H5E_CANTCLOSEFILE, NULL, + "unable to close s3 file handle") + } + } + if (file != NULL) { + file = H5FL_FREE(H5FD_ros3_t, file); + } + curl_global_cleanup(); /* early cleanup because open failed */ + } /* if null return value (error) */ +#endif /* H5_HAVE_ROS3_VFD */ + + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_ros3_open() */ + +#if ROS3_STATS + +/*---------------------------------------------------------------------------- + * + * Function: ros3_fprint_stats() + * + * Purpose: + * + * Tabulate and pretty-print statistics for this virtual file. + * + * Should be called upon file close. + * + * Shows number of reads and bytes read, broken down by + * "raw" (H5FD_MEM_DRAW) + * or "meta" (any other flag) + * + * Prints filename and listing of total number of reads and bytes read, + * both as a grand total and separate meta- and rawdata reads. + * + * If any reads were done, prints out two tables: + * + * 1. overview of raw- and metadata reads + * - min (smallest size read) + * - average of size read + * - k,M,G suffixes by powers of 1024 (2^10) + * - max (largest size read) + * 2. tabulation of "bins", sepraring reads into exponentially-larger + * ranges of size. + * - columns for number of reads, total bytes, and average size, with + * separate sub-colums for raw- and metadata reads. + * - each row represents one bin, identified by the top of its range + * + * Bin ranges can be modified with pound-defines at the top of this file. + * + * Bins without any reads in their bounds are not printed. + * + * An "overflow" bin is also present, to catch "big" reads. + * + * Output for all bins (and range ceiling and average size report) + * is divied by powers of 1024. By corollary, four digits before the decimal + * is valid. + * + * - 41080 bytes is represented by 40.177k, not 41.080k + * - 1004.831M represents approx. 1052642000 bytes + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - FAILURE: `FAIL` + * - occurs if the file passed in is invalid + * - TODO: if stream is invalid? how can we check this? + * + * Programmer: Jacob Smith + * + * Changes: None. + * + *---------------------------------------------------------------------------- + */ +static herr_t +ros3_fprint_stats(FILE *stream, + const H5FD_ros3_t *file) +{ + herr_t ret_value = SUCCEED; + parsed_url_t *purl = NULL; + unsigned i = 0; + unsigned long count_meta = 0; + unsigned long count_raw = 0; + double average_meta = 0.0; + double average_raw = 0.0; + unsigned long long min_meta = (unsigned long long)ROS3_STATS_STARTING_MIN; + unsigned long long min_raw = (unsigned long long)ROS3_STATS_STARTING_MIN; + unsigned long long max_meta = 0; + unsigned long long max_raw = 0; + unsigned long long bytes_raw = 0; + unsigned long long bytes_meta = 0; + double re_dub = 0.0; /* re-usable double variable */ + unsigned suffix_i = 0; + const char suffixes[] = { ' ', 'K', 'M', 'G', 'T', 'P' }; + + + + FUNC_ENTER_NOAPI_NOINIT + + if (stream == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "file stream cannot be null" ); + } + if (file == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "file cannot be null"); + } + if (file->s3r_handle == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "s3 request handle cannot be null"); + } + if (file->s3r_handle->purl == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "parsed url structure cannot be null"); + } + purl = file->s3r_handle->purl; + + /****************** + * PRINT FILENAME * + ******************/ + + HDfprintf(stream, "stats for %s://%s", purl->scheme, purl->host); + if (purl->port != NULL && purl->port[0] != '\0') + HDfprintf(stream, ":%s", purl->port); + if (purl->query != NULL && purl->query[0] != '\0') { + if (purl->path != NULL && purl->path[0] != '\0') + HDfprintf(stream, "/%s", purl->path); + else + HDfprintf(stream, "/"); + HDfprintf(stream, "?%s", purl->query); + } else if (purl->path != NULL && purl->path[0] != '\0') { + HDfprintf(stream, "/%s", purl->path); + } + HDfprintf(stream, "\n"); + + /******************* + * AGGREGATE STATS * + *******************/ + + for (i = 0; i <= ROS3_STATS_BIN_COUNT; i++) { + const ros3_statsbin *r = &file->raw[i]; + const ros3_statsbin *m = &file->meta[i]; + + if (m->min < min_meta) min_meta = m->min; + if (r->min < min_raw) min_raw = r->min; + if (m->max > max_meta) max_meta = m->max; + if (r->max > max_raw) max_raw = r->max; + + count_raw += r->count; + count_meta += m->count; + bytes_raw += r->bytes; + bytes_meta += m->bytes; + } + if (count_raw > 0) + average_raw = (double)bytes_raw / (double)count_raw; + if (count_meta > 0) + average_meta = (double)bytes_meta / (double)count_meta; + + /****************** + * PRINT OVERVIEW * + ******************/ + + HDfprintf(stream, "TOTAL READS: %llu (%llu meta, %llu raw)\n", + count_raw + count_meta, count_meta, count_raw); + HDfprintf(stream, "TOTAL BYTES: %llu (%llu meta, %llu raw)\n", + bytes_raw + bytes_meta, bytes_meta, bytes_raw); + + if (count_raw + count_meta == 0) + goto done; + + /************************* + * PRINT AGGREGATE STATS * + *************************/ + + HDfprintf(stream, "SIZES meta raw\n"); + HDfprintf(stream, " min "); + if (count_meta == 0) { + HDfprintf(stream, " 0.000 "); + } else { + re_dub = (double)min_meta; + for (suffix_i = 0; re_dub >= 1024.0; suffix_i++) + re_dub /= 1024.0; + HDassert(suffix_i < sizeof(suffixes)); + HDfprintf(stream, "%8.3lf%c ", re_dub, suffixes[suffix_i]); + } + + if (count_raw == 0) { + HDfprintf(stream, " 0.000 \n"); + } else { + re_dub = (double)min_raw; + for (suffix_i = 0; re_dub >= 1024.0; suffix_i++) + re_dub /= 1024.0; + HDassert(suffix_i < sizeof(suffixes)); + HDfprintf(stream, "%8.3lf%c\n", re_dub, suffixes[suffix_i]); + } + + HDfprintf(stream, " avg "); + re_dub = (double)average_meta; + for (suffix_i = 0; re_dub >= 1024.0; suffix_i++) + re_dub /= 1024.0; + HDassert(suffix_i < sizeof(suffixes)); + HDfprintf(stream, "%8.3lf%c ", re_dub, suffixes[suffix_i]); + + re_dub = (double)average_raw; + for (suffix_i = 0; re_dub >= 1024.0; suffix_i++) + re_dub /= 1024.0; + HDassert(suffix_i < sizeof(suffixes)); + HDfprintf(stream, "%8.3lf%c\n", re_dub, suffixes[suffix_i]); + + HDfprintf(stream, " max "); + re_dub = (double)max_meta; + for (suffix_i = 0; re_dub >= 1024.0; suffix_i++) + re_dub /= 1024.0; + HDassert(suffix_i < sizeof(suffixes)); + HDfprintf(stream, "%8.3lf%c ", re_dub, suffixes[suffix_i]); + + re_dub = (double)max_raw; + for (suffix_i = 0; re_dub >= 1024.0; suffix_i++) + re_dub /= 1024.0; + HDassert(suffix_i < sizeof(suffixes)); + HDfprintf(stream, "%8.3lf%c\n", re_dub, suffixes[suffix_i]); + + /****************************** + * PRINT INDIVIDUAL BIN STATS * + ******************************/ + + HDfprintf(stream, + "BINS # of reads total bytes average size\n"); + HDfprintf(stream, + " up-to meta raw meta raw meta raw\n"); + + for (i = 0; i <= ROS3_STATS_BIN_COUNT; i++) { + const ros3_statsbin *m; + const ros3_statsbin *r; + unsigned long long range_end = 0; + char bm_suffix = ' '; /* bytes-meta */ + double bm_val = 0.0; + char br_suffix = ' '; /* bytes-raw */ + double br_val = 0.0; + char am_suffix = ' '; /* average-meta */ + double am_val = 0.0; + char ar_suffix = ' '; /* average-raw */ + double ar_val = 0.0; + + m = &file->meta[i]; + r = &file->raw[i]; + if (r->count == 0 && m->count == 0) + continue; + + range_end = ros3_stats_boundaries[i]; + + if (i == ROS3_STATS_BIN_COUNT) { + range_end = ros3_stats_boundaries[i-1]; + HDfprintf(stream, ">"); + } else { + HDfprintf(stream, " "); + } + + bm_val = (double)m->bytes; + for (suffix_i = 0; bm_val >= 1024.0; suffix_i++) + bm_val /= 1024.0; + HDassert(suffix_i < sizeof(suffixes)); + bm_suffix = suffixes[suffix_i]; + + br_val = (double)r->bytes; + for (suffix_i = 0; br_val >= 1024.0; suffix_i++) + br_val /= 1024.0; + HDassert(suffix_i < sizeof(suffixes)); + br_suffix = suffixes[suffix_i]; + + if (m->count > 0) + am_val = (double)(m->bytes) / (double)(m->count); + for (suffix_i = 0; am_val >= 1024.0; suffix_i++) + am_val /= 1024.0; + HDassert(suffix_i < sizeof(suffixes)); + am_suffix = suffixes[suffix_i]; + + if (r->count > 0) + ar_val = (double)(r->bytes) / (double)(r->count); + for (suffix_i = 0; ar_val >= 1024.0; suffix_i++) + ar_val /= 1024.0; + HDassert(suffix_i < sizeof(suffixes)); + ar_suffix = suffixes[suffix_i]; + + re_dub = (double)range_end; + for (suffix_i = 0; re_dub >= 1024.0; suffix_i++) + re_dub /= 1024.0; + HDassert(suffix_i < sizeof(suffixes)); + + HDfprintf(stream, + " %8.3f%c %7d %7d %8.3f%c %8.3f%c %8.3f%c %8.3f%c\n", + re_dub, suffixes[suffix_i], /* bin ceiling */ + m->count, /* metadata reads */ + r->count, /* rawdata reads */ + bm_val, bm_suffix, /* metadata bytes */ + br_val, br_suffix, /* rawdata bytes */ + am_val, am_suffix, /* metadata average */ + ar_val, ar_suffix); /* rawdata average */ + + fflush(stream); + } + +done: + FUNC_LEAVE_NOAPI(ret_value); + +} /* ros3_fprint_stats */ +#endif /* ROS3_STATS */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_ros3_close() + * + * Purpose: + * + * Close an HDF5 file. + * + * Return: + * + * SUCCEED/FAIL + * + * Programmer: Jacob Smith + * 2017-11-02 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_ros3_close(H5FD_t *_file) +{ +#ifdef H5_HAVE_ROS3_VFD + H5FD_ros3_t *file = (H5FD_ros3_t *)_file; + herr_t ret_value = SUCCEED; +#else + herr_t ret_value = FAIL; +#endif + + + + FUNC_ENTER_NOAPI_NOINIT +#ifdef H5_HAVE_ROS3_VFD + +#if ROS3_DEBUG + HDfprintf(stdout, "H5FD_ros3_close() called.\n"); +#endif + + /* Sanity checks + */ + HDassert(file != NULL); + HDassert(file->s3r_handle != NULL); + + /* Close the underlying request handle + */ + if (FAIL == H5FD_s3comms_s3r_close(file->s3r_handle)) { + HGOTO_ERROR(H5E_VFL, H5E_CANTCLOSEFILE, FAIL, + "unable to close S3 request handle") + } + +#if ROS3_STATS + /* TODO: mechanism to re-target stats printout */ + if (FAIL == ros3_fprint_stats(stdout, file)) { + HGOTO_ERROR(H5E_INTERNAL, H5E_ERROR, FAIL, + "problem while writing file statistics") + } +#endif /* ROS3_STATS */ + + /* Release the file info + */ + file = H5FL_FREE(H5FD_ros3_t, file); + +done: + curl_global_cleanup(); /* cleanup to answer init on open */ +#endif /* H5_HAVE_ROS3_VFD */ + + FUNC_LEAVE_NOAPI(ret_value) + +} /* end H5FD_ros3_close() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_ros3_cmp() + * + * Purpose: + * + * Compares two files belonging to this driver using an arbitrary + * (but consistent) ordering: + * + * + url scheme + * + url host + * + url port + * + url path + * + url query + * + fapl aws_region + * + fapl secret_id + * + fapl secret_key + * + * tl;dr -> check URL, check crentials + * + * Return: + * + * - Equivalent: 0 + * - Not Equivalent: -1 + * + * Programmer: Jacob Smith + * 2017-11-06 + * + * Changes: + * + * + Change from strcmp-like return values (-1, 0, 1) to instead return + * binary equivalence (0) or inequality (-1). + * + Replace "if still equal then check this" waterfall with GOTO jumps. + * Jacob Smith 2018-05-17 + * + *------------------------------------------------------------------------- + */ +static int +H5FD_ros3_cmp(const H5FD_t *_f1, + const H5FD_t *_f2) +{ +#ifdef H5_HAVE_ROS3_VFD + const H5FD_ros3_t *f1 = (const H5FD_ros3_t *)_f1; + const H5FD_ros3_t *f2 = (const H5FD_ros3_t *)_f2; + const parsed_url_t *purl1 = NULL; + const parsed_url_t *purl2 = NULL; +#endif + int ret_value = 0; + + + + FUNC_ENTER_NOAPI_NOINIT_NOERR +#ifdef H5_HAVE_ROS3_VFD + +#if ROS3_DEBUG + HDfprintf(stdout, "H5FD_ros3_cmp() called.\n"); +#endif + + HDassert(f1->s3r_handle != NULL); + HDassert(f2->s3r_handle != NULL); + + purl1 = (const parsed_url_t *)f1->s3r_handle->purl; + purl2 = (const parsed_url_t *)f2->s3r_handle->purl; + HDassert(purl1 != NULL); + HDassert(purl2 != NULL); + HDassert(purl1->scheme != NULL); + HDassert(purl2->scheme != NULL); + HDassert(purl1->host != NULL); + HDassert(purl2->host != NULL); + + /* URL: SCHEME */ + if (strcmp(purl1->scheme, purl2->scheme)) HGOTO_DONE(-1); + + /* URL: HOST */ + if (strcmp(purl1->host, purl2->host)) HGOTO_DONE(-1); + + /* URL: PORT */ + if (purl1->port && purl2->port) { + if (strcmp(purl1->port, purl2->port)) HGOTO_DONE(-1); + } else if (purl1->port) { + HGOTO_DONE(-1); + } else if (purl2->port) { + HGOTO_DONE(-1); + } + + /* URL: PATH */ + if (purl1->path && purl2->path) { + if (strcmp(purl1->path, purl2->path)) HGOTO_DONE(-1); + } else if (purl1->path && !purl2->path) { + HGOTO_DONE(-1); + } else if (purl2->path && !purl1->path) { + HGOTO_DONE(-1); + } + + /* URL: QUERY */ + if (purl1->query && purl2->query) { + if (strcmp(purl1->query, purl2->query)) HGOTO_DONE(-1); + } else if (purl1->query && !purl2->query) { + HGOTO_DONE(-1); + } else if (purl2->query && !purl1->query) { + HGOTO_DONE(-1); + } + + /* FAPL: AWS_REGION */ + if (f1->fa.aws_region[0] != '\0' && f1->fa.aws_region[0] != '\0') { + if (strcmp(f1->fa.aws_region, f2->fa.aws_region)) HGOTO_DONE(-1); + } else if (f1->fa.aws_region[0] != '\0') { + HGOTO_DONE(-1); + } else if (f2->fa.aws_region[0] != '\0') { + HGOTO_DONE(-1); + } + + /* FAPL: SECRET_ID */ + if (f1->fa.secret_id[0] != '\0' && f1->fa.secret_id[0] != '\0') { + if (strcmp(f1->fa.secret_id, f2->fa.secret_id)) HGOTO_DONE(-1); + } else if (f1->fa.secret_id[0] != '\0') { + HGOTO_DONE(-1); + } else if (f2->fa.secret_id[0] != '\0') { + HGOTO_DONE(-1); + } + + /* FAPL: SECRET_KEY */ + if (f1->fa.secret_key[0] != '\0' && f1->fa.secret_key[0] != '\0') { + if (strcmp(f1->fa.secret_key, f2->fa.secret_key)) HGOTO_DONE(-1); + } else if (f1->fa.secret_key[0] != '\0') { + HGOTO_DONE(-1); + } else if (f2->fa.secret_key[0] != '\0') { + HGOTO_DONE(-1); + } +#endif /* H5_HAVE_ROS3_VFD */ + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_ros3_cmp() */ + + +/*------------------------------------------------------------------------- + * Function: H5FD_ros3_query + * + * Purpose: Set the flags that this VFL driver is capable of supporting. + * (listed in H5FDpublic.h) + * + * Note that since the ROS3 VFD is read only, most flags + * are irrelevant. + * + * The term "set" is highly misleading... + * stores/copies the supported flags in the out-pointer `flags`. + * + * Return: SUCCEED (Can't fail) + * + * Programmer: John Mainzer + * 9/11/17 + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_ros3_query(const H5FD_t H5_ATTR_UNUSED *_file, + unsigned long *flags /* out */) +{ + FUNC_ENTER_NOAPI_NOINIT_NOERR + +#if ROS3_DEBUG + HDfprintf(stdout, "H5FD_ros3_query() called.\n"); +#endif + + /* Set the VFL feature flags that this driver supports */ + if (flags) { + *flags = 0; + /* OK to perform data sieving for faster raw data reads & writes */ + *flags |= H5FD_FEAT_DATA_SIEVE; + } /* end if */ + + FUNC_LEAVE_NOAPI(SUCCEED) + +} /* H5FD_ros3_query() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_ros3_get_eoa() + * + * Purpose: + * + * Gets the end-of-address marker for the file. The EOA marker + * is the first address past the last byte allocated in the + * format address space. + * + * Return: + * + * The end-of-address marker. + * + * Programmer: Jacob Smith + * 2017-11-02 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static haddr_t +H5FD_ros3_get_eoa(const H5FD_t *_file, + H5FD_mem_t H5_ATTR_UNUSED type) +{ + const H5FD_ros3_t *file = (const H5FD_ros3_t *)_file; + + FUNC_ENTER_NOAPI_NOINIT_NOERR + +#if ROS3_DEBUG + HDfprintf(stdout, "H5FD_ros3_get_eoa() called.\n"); +#endif + + FUNC_LEAVE_NOAPI(file->eoa) + +} /* end H5FD_ros3_get_eoa() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_ros3_set_eoa() + * + * Purpose: + * + * Set the end-of-address marker for the file. + * + * Return: + * + * SUCCEED (can't fail) + * + * Programmer: Jacob Smith + * 2017-11-03 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_ros3_set_eoa(H5FD_t *_file, + H5FD_mem_t H5_ATTR_UNUSED type, + haddr_t addr) +{ + H5FD_ros3_t *file = (H5FD_ros3_t *)_file; + + FUNC_ENTER_NOAPI_NOINIT_NOERR + +#if ROS3_DEBUG + HDfprintf(stdout, "H5FD_ros3_set_eoa() called.\n"); +#endif + + file->eoa = addr; + + FUNC_LEAVE_NOAPI(SUCCEED) + +} /* H5FD_ros3_set_eoa() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_ros3_get_eof() + * + * Purpose: + * + * Returns the end-of-file marker. + * + * Return: + * + * EOF: the first address past the end of the "file", either the + * filesystem file or the HDF5 file. + * + * Programmer: Jacob Smith + * 2017-11-02 + * + *------------------------------------------------------------------------- + */ +static haddr_t +H5FD_ros3_get_eof(const H5FD_t *_file, + H5FD_mem_t H5_ATTR_UNUSED type) +{ + const H5FD_ros3_t *file = (const H5FD_ros3_t *)_file; + + FUNC_ENTER_NOAPI_NOINIT_NOERR + +#if ROS3_DEBUG + HDfprintf(stdout, "H5FD_ros3_get_eof() called.\n"); +#endif + + FUNC_LEAVE_NOAPI(H5FD_s3comms_s3r_get_filesize(file->s3r_handle)) + +} /* end H5FD_ros3_get_eof() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_ros3_get_handle() + * + * Purpose: + * + * Returns the S3 Request handle (s3r_t) of ros3 file driver. + * + * Returns: + * + * SUCCEED/FAIL + * + * Programmer: Jacob Smith + * 2017-11-02 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_ros3_get_handle(H5FD_t *_file, + hid_t H5_ATTR_UNUSED fapl, + void **file_handle) +{ + H5FD_ros3_t *file = (H5FD_ros3_t *)_file; + herr_t ret_value = SUCCEED; + + FUNC_ENTER_NOAPI_NOINIT + +#if ROS3_DEBUG + HDfprintf(stdout, "H5FD_ros3_get_handle() called.\n"); +#endif + + if(!file_handle) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, "file handle not valid") + + *file_handle = file->s3r_handle; + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* end H5FD_ros3_get_handle() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_ros3_read() + * + * Purpose: + * + * Reads SIZE bytes of data from FILE beginning at address ADDR + * into buffer BUF according to data transfer properties in DXPL_ID. + * + * Return: + * + * Success: `SUCCEED` + * - Result is stored in caller-supplied buffer BUF. + * Failure: `FAIL` + * - Unable to complete read. + * - Contents of buffer `buf` are undefined. + * + * Programmer: Jacob Smith + * 2017-11-?? + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_ros3_read(H5FD_t *_file, + H5FD_mem_t H5_ATTR_UNUSED type, + hid_t H5_ATTR_UNUSED dxpl_id, + haddr_t addr, /* start offset */ + size_t size, /* length of read */ + void *buf) /* out */ +{ + H5FD_ros3_t *file = (H5FD_ros3_t *)_file; + size_t filesize = 0; + herr_t ret_value = SUCCEED; /* Return value */ +#if ROS3_STATS + /* working variables for storing stats */ + ros3_statsbin *bin = NULL; + unsigned bin_i = 0; +#endif /* ROS3_STATS */ + + + FUNC_ENTER_NOAPI_NOINIT + +#if ROS3_DEBUG + HDfprintf(stdout, "H5FD_ros3_read() called.\n"); +#endif + + HDassert(file != NULL); + HDassert(file->s3r_handle != NULL); + HDassert(buf != NULL); + + filesize = H5FD_s3comms_s3r_get_filesize(file->s3r_handle); + + if ((addr > filesize) || ((addr + size) > filesize)) { + HGOTO_ERROR(H5E_ARGS, H5E_OVERFLOW, FAIL, "range exceeds file address") + } + + if (FAIL == H5FD_s3comms_s3r_read(file->s3r_handle, addr, size, buf) ) { + HGOTO_ERROR(H5E_VFL, H5E_READERROR, FAIL, "unable to execute read") + } + +#if ROS3_STATS + + /* Find which "bin" this read fits in. Can be "overflow" bin. + */ + for (bin_i = 0; bin_i < ROS3_STATS_BIN_COUNT; bin_i++) { + if ((unsigned long long)size < ros3_stats_boundaries[bin_i]) + break; + } + bin = (type == H5FD_MEM_DRAW) + ? &file->raw[bin_i] + : &file->meta[bin_i]; + + /* Store collected stats in appropriate bin + */ + if (bin->count == 0) { + bin->min = size; + bin->max = size; + } else { + if (size < bin->min) + bin->min = size; + if (size > bin->max) + bin->max = size; + } + bin->count++; + bin->bytes += (unsigned long long)size; + +#endif /* ROS3_STATS */ + + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* end H5FD_ros3_read() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_ros3_write() + * + * Purpose: + * + * Write bytes to file. + * UNSUPPORTED IN READ-ONLY ROS3 VFD. + * + * Return: + * + * FAIL (Not possible with Read-Only S3 file.) + * + * Programmer: Jacob Smith + * 2017-10-23 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_ros3_write(H5FD_t H5_ATTR_UNUSED *_file, + H5FD_mem_t H5_ATTR_UNUSED type, + hid_t H5_ATTR_UNUSED dxpl_id, + haddr_t H5_ATTR_UNUSED addr, + size_t H5_ATTR_UNUSED size, + const void H5_ATTR_UNUSED *buf) +{ + herr_t ret_value = FAIL; + + FUNC_ENTER_NOAPI_NOINIT + +#if ROS3_DEBUG + HDfprintf(stdout, "H5FD_ros3_write() called.\n"); +#endif + + HGOTO_ERROR(H5E_VFL, H5E_UNSUPPORTED, FAIL, + "cannot write to read-only file.") + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_ros3_write() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_ros3_truncate() + * + * Purpose: + * + * Makes sure that the true file size is the same (or larger) + * than the end-of-address. + * + * NOT POSSIBLE ON READ-ONLY S3 FILES. + * + * Return: + * + * FAIL (Not possible on Read-Only S3 files.) + * + * Programmer: Jacob Smith + * 2017-10-23 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_ros3_truncate(H5FD_t H5_ATTR_UNUSED *_file, + hid_t H5_ATTR_UNUSED dxpl_id, + hbool_t H5_ATTR_UNUSED closing) +{ + herr_t ret_value = SUCCEED; + + FUNC_ENTER_NOAPI_NOINIT + +#if ROS3_DEBUG + HDfprintf(stdout, "H5FD_ros3_truncate() called.\n"); +#endif + + HGOTO_ERROR(H5E_VFL, H5E_UNSUPPORTED, FAIL, + "cannot truncate read-only file.") + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* end H5FD_ros3_truncate() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_ros3_lock() + * + * Purpose: + * + * Place an advisory lock on a file. + * No effect on Read-Only S3 file. + * + * Suggestion: remove lock/unlock from class + * > would result in error at H5FD_[un]lock() (H5FD.c) + * + * Return: + * + * SUCCEED (No-op always succeeds) + * + * Programmer: Jacob Smith + * 2017-11-03 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_ros3_lock(H5FD_t H5_ATTR_UNUSED *_file, + hbool_t H5_ATTR_UNUSED rw) +{ + FUNC_ENTER_NOAPI_NOINIT_NOERR + FUNC_LEAVE_NOAPI(SUCCEED) + +} /* end H5FD_ros3_lock() */ + + +/*------------------------------------------------------------------------- + * + * Function: H5FD_ros3_unlock() + * + * Purpose: + * + * Remove the existing lock on the file. + * No effect on Read-Only S3 file. + * + * Return: + * + * SUCCEED (No-op always succeeds) + * + * Programmer: Jacob Smith + * 2017-11-03 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static herr_t +H5FD_ros3_unlock(H5FD_t H5_ATTR_UNUSED *_file) +{ + FUNC_ENTER_NOAPI_NOINIT_NOERR + FUNC_LEAVE_NOAPI(SUCCEED) + +} /* end H5FD_ros3_unlock() */ + + diff --git a/src/H5FDros3.h b/src/H5FDros3.h new file mode 100644 index 0000000..49e757c --- /dev/null +++ b/src/H5FDros3.h @@ -0,0 +1,105 @@ +/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * Read-Only S3 Virtual File Driver (VFD) * + * Copyright (c) 2017-2018, The HDF Group. * + * * + * All rights reserved. * + * * + * NOTICE: * + * All information contained herein is, and remains, the property of The HDF * + * Group. The intellectual and technical concepts contained herein are * + * proprietary to The HDF Group. Dissemination of this information or * + * reproduction of this material is strictly forbidden unless prior written * + * permission is obtained from The HDF Group. * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ + +/* + * Programmer: John Mainzer + * 2017-10-10 + * + * Purpose: The public header file for the ros3 driver. + */ +#ifndef H5FDros3_H +#define H5FDros3_H + +#define H5FD_ROS3 (H5FD_ros3_init()) + +#ifdef __cplusplus +extern "C" { +#endif + +/**************************************************************************** + * + * Structure: H5FD_ros3_fapl_t + * + * Purpose: + * + * H5FD_ros3_fapl_t is a public structure that is used to pass S3 + * authentication data to the appropriate S3 VFD via the FAPL. A pointer + * to an instance of this structure is a parameter to H5Pset_fapl_ros3() + * and H5Pget_fapl_ros3(). + * + * + * + * `version` (int32_t) + * + * Version number of the H5FD_ros3_fapl_t structure. Any instance passed + * to the above calls must have a recognized version number, or an error + * will be flagged. + * + * This field should be set to H5FD__CURR_ROS3_FAPL_T_VERSION. + * + * `authenticate` (hbool_t) + * + * Flag TRUE or FALSE whether or not requests are to be authenticated + * with the AWS4 algorithm. + * If TRUE, `aws_region`, `secret_id`, and `secret_key` must be populated. + * If FALSE, those three components are unused. + * + * `aws_region` (char[]) + * + * String: name of the AWS "region" of the host, e.g. "us-east-1". + * + * `secret_id` (char[]) + * + * String: "Access ID" for the resource. + * + * `secret_key` (char[]) + * + * String: "Secret Access Key" associated with the ID and resource. + * + * + * + * Programmer: John Mainzer + * + * Changes: + * + * - Add documentation of fields (except `version`) + * --- Jacob Smith 2017-12-04 + * + ****************************************************************************/ + +#define H5FD__CURR_ROS3_FAPL_T_VERSION 1 + +#define H5FD__ROS3_MAX_REGION_LEN 32 +#define H5FD__ROS3_MAX_SECRET_ID_LEN 128 +#define H5FD__ROS3_MAX_SECRET_KEY_LEN 128 + +typedef struct H5FD_ros3_fapl_t { + int32_t version; + hbool_t authenticate; + char aws_region[H5FD__ROS3_MAX_REGION_LEN + 1]; + char secret_id[H5FD__ROS3_MAX_SECRET_ID_LEN + 1]; + char secret_key[H5FD__ROS3_MAX_SECRET_KEY_LEN + 1]; +} H5FD_ros3_fapl_t; + +H5_DLL hid_t H5FD_ros3_init(void); +H5_DLL herr_t H5Pget_fapl_ros3(hid_t fapl_id, H5FD_ros3_fapl_t * fa_out); +H5_DLL herr_t H5Pset_fapl_ros3(hid_t fapl_id, H5FD_ros3_fapl_t * fa); + +#ifdef __cplusplus +} +#endif + +#endif /* ifndef H5FDros3_H */ + + diff --git a/src/H5FDs3comms.c b/src/H5FDs3comms.c new file mode 100644 index 0000000..7caeacb --- /dev/null +++ b/src/H5FDs3comms.c @@ -0,0 +1,3770 @@ +/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * Read-Only S3 Virtual File Driver (VFD) * + * Copyright (c) 2017-2018, The HDF Group. * + * * + * All rights reserved. * + * * + * NOTICE: * + * All information contained herein is, and remains, the property of The HDF * + * Group. The intellectual and technical concepts contained herein are * + * proprietary to The HDF Group. Dissemination of this information or * + * reproduction of this material is strictly forbidden unless prior written * + * permission is obtained from The HDF Group. * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ + +/***************************************************************************** + * Source for S3 Communications module + * + * ***NOT A FILE DRIVER*** + * + * Provide functions and structures required for interfacing with Amazon + * Simple Storage Service (S3). + * + * Provide S3 object access as if it were a local file. + * + * Connect to remote host, send and receive HTTP requests and responses + * as part of the AWS REST API, authenticating requests as appropriate. + * + * Programmer: Jacob Smith + * 2017-11-30 + * + *****************************************************************************/ + +/****************/ +/* Module Setup */ +/****************/ + +/***********/ +/* Headers */ +/***********/ + +#include "H5private.h" /* generic functions */ +#include "H5Eprivate.h" /* error handling */ +#include "H5MMprivate.h" /* memory management */ +#include "H5FDs3comms.h" /* S3 Communications */ + +/****************/ +/* Local Macros */ +/****************/ + +/* toggle debugging (enable with 1) + */ +#define S3COMMS_DEBUG 0 + +/* manipulate verbosity of CURL output + * operates separately from S3COMMS_DEBUG + * + * 0 -> no explicit curl output + * 1 -> on error, print failure info to stderr + * 2 -> in addition to above, print information for all performs; sets all + * curl handles with CURLOPT_VERBOSE + */ +#define S3COMMS_CURL_VERBOSITY 0 + +/* size to allocate for "bytes=<first_byte>[-<last_byte>]" HTTP Range value + */ +#define S3COMMS_MAX_RANGE_STRING_SIZE 128 + +/******************/ +/* Local Typedefs */ +/******************/ + +/********************/ +/* Local Structures */ +/********************/ + +/* struct s3r_datastruct + * Structure passed to curl write callback + * pointer to data region and record of bytes written (offset) + */ +struct s3r_datastruct { + unsigned long magic; + char *data; + size_t size; +}; +#define S3COMMS_CALLBACK_DATASTRUCT_MAGIC 0x28c2b2ul + +/********************/ +/* Local Prototypes */ +/********************/ + +size_t curlwritecallback(char *ptr, + size_t size, + size_t nmemb, + void *userdata); + +herr_t H5FD_s3comms_s3r_getsize(s3r_t *handle); + +/*********************/ +/* Package Variables */ +/*********************/ + +/*****************************/ +/* Library Private Variables */ +/*****************************/ + +/*******************/ +/* Local Variables */ +/*******************/ + +/*************/ +/* Functions */ +/*************/ + + +/*---------------------------------------------------------------------------- + * + * Function: curlwritecallback() + * + * Purpose: + * + * Function called by CURL to write received data. + * + * Writes bytes to `userdata`. + * + * Internally manages number of bytes processed. + * + * Return: + * + * - Number of bytes processed. + * - Should equal number of bytes passed to callback. + * - Failure will result in curl error: CURLE_WRITE_ERROR. + * + * Programmer: Jacob Smith + * 2017-08-17 + * + * Changes: None. + * + *---------------------------------------------------------------------------- + */ +size_t +curlwritecallback(char *ptr, + size_t size, + size_t nmemb, + void *userdata) +{ + struct s3r_datastruct *sds = (struct s3r_datastruct *)userdata; + size_t product = (size * nmemb); + size_t written = 0; + + if (sds->magic != S3COMMS_CALLBACK_DATASTRUCT_MAGIC) + return written; + + if (size > 0) { + HDmemcpy(&(sds->data[sds->size]), ptr, product); + sds->size += product; + written = product; + } + + return written; + +} /* curlwritecallback */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_hrb_node_set() + * + * Purpose: + * + * Create, insert, modify, and remove elements in a field node list. + * + * `name` cannot be null; will return FAIL and list will be unaltered. + * + * Entries are accessed via the lowercase representation of their name: + * "Host", "host", and "hOSt" would all access the same node, + * but name's case is relevant in HTTP request output. + * + * List pointer `L` must always point to either of : + * - header node with lowest alphabetical order (by lowername) + * - NULL, if list is empty + * + * Types of operations: + * + * - CREATE + * - If `L` is NULL and `name` and `value` are not NULL, + * a new node is created at `L`, starting a list. + * - MODIFY + * - If a node is found with a matching lowercase name and `value` + * is not NULL, the existing name, value, and cat values are released + * and replaced with the new data. + * - No modifications are made to the list pointers. + * - REMOVE + * - If `value` is NULL, will attempt to remove node with matching + * lowercase name. + * - If no match found, returns FAIL and list is not modified. + * - When removing a node, all its resources is released. + * - If removing the last node in the list, list pointer is set to NULL. + * - INSERT + * - If no nodes exists with matching lowercase name and `value` + * is not NULL, a new node is created, inserted into list + * alphabetically by lowercase name. + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - List was successfully modified + * - FAILURE: `FAIL` + * - Unable to perform operation + * - Forbidden (attempting to remove absent node, e.g.) + * - Internal error + * + * Programmer: Jacob Smith + * 2017-09-22 + * + * Changes: + * + * - Change return value to herr_t + * - Change list pointer to pointer-to-pointer-to-node + * - Change to use singly-linked list (from twin doubly-linked lists) + * with modification to hrb_node_t + * --- Jake Smith 2017-01-17 + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_hrb_node_set(hrb_node_t **L, + const char *name, + const char *value) +{ + size_t i = 0; + char *valuecpy = NULL; + char *namecpy = NULL; + size_t namelen = 0; + char *lowername = NULL; + char *nvcat = NULL; + hrb_node_t *node_ptr = NULL; + hrb_node_t *new_node = NULL; + hbool_t is_looking = TRUE; + herr_t ret_value = SUCCEED; + + + + FUNC_ENTER_NOAPI_NOINIT + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_hrb_node_set.\n"); + HDprintf("NAME: %s\n", name); + HDprintf("VALUE: %s\n", value); + HDprintf("LIST:\n->"); + for (node_ptr = (*L); node_ptr != NULL; node_ptr = node_ptr->next) + HDfprintf(stdout, "{%s}\n->", node_ptr->cat); + HDprintf("(null)\n"); + fflush(stdout); + node_ptr = NULL; +#endif + + if (name == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to operate on null name.\n"); + } + namelen = HDstrlen(name); + + /*********************** + * PREPARE ALL STRINGS * + **********************/ + + /* copy and lowercase name + */ + lowername = (char *)H5MM_malloc(sizeof(char) * (namelen + 1)); + if (lowername == NULL) { + HGOTO_ERROR(H5E_RESOURCE, H5E_NOSPACE, FAIL, + "cannot make space for lowercase name copy.\n"); + } + for (i = 0; i < namelen; i++) { + lowername[i] = (char)tolower((int)name[i]); + } + lowername[namelen] = 0; + + /* If value supplied, copy name, value, and concatenated "name: value". + * If NULL, we will be removing a node or doing nothing, so no need for + * copies + */ + if (value != NULL) { + size_t valuelen = HDstrlen(value); + size_t catlen = namelen + valuelen + 2; /* HDstrlen(": ") -> +2 */ + int sprint_ret = 0; + + namecpy = (char *)H5MM_malloc(sizeof(char) * (namelen + 1)); + if (namecpy == NULL) { + HGOTO_ERROR(H5E_RESOURCE, H5E_NOSPACE, FAIL, + "cannot make space for name copy.\n"); + } + HDmemcpy(namecpy, name, namelen + 1); + + valuecpy = (char *)H5MM_malloc(sizeof(char) * (valuelen + 1)); + if (valuecpy == NULL) { + HGOTO_ERROR(H5E_RESOURCE, H5E_NOSPACE, FAIL, + "cannot make space for value copy.\n"); + } + HDmemcpy(valuecpy, value, valuelen + 1); + + nvcat = (char *)H5MM_malloc(sizeof(char) * (catlen + 1)); + if (nvcat == NULL) { + HGOTO_ERROR(H5E_RESOURCE, H5E_NOSPACE, FAIL, + "cannot make space for concatenated string.\n"); + } + sprint_ret = HDsnprintf(nvcat, (catlen + 1), "%s: %s", name, value); + if (sprint_ret <= 0) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "error while concatenating `%s: %s", name, value); + HDassert( catlen == (size_t)sprint_ret ); + + /* create new_node, should we need it + */ + new_node = (hrb_node_t *)H5MM_malloc(sizeof(hrb_node_t)); + if (new_node == NULL) { + HGOTO_ERROR(H5E_RESOURCE, H5E_NOSPACE, FAIL, + "cannot make space for new set.\n"); + } + + new_node->magic = S3COMMS_HRB_NODE_MAGIC; + new_node->name = NULL; + new_node->value = NULL; + new_node->cat = NULL; + new_node->lowername = NULL; + new_node->next = NULL; + } + + /*************** + * ACT ON LIST * + ***************/ + + if (*L == NULL) { + if (value == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "trying to remove node from empty list"); + } else { +#if S3COMMS_DEBUG +HDprintf("CREATE NEW\n"); fflush(stdout); +#endif + /******************* + * CREATE NEW LIST * + *******************/ + + new_node->cat = nvcat; + new_node->name = namecpy; + new_node->lowername = lowername; + new_node->value = valuecpy; + + *L = new_node; + goto done; /* bypass further seeking */ + } + } + + /* sanity-check pointer passed in + */ + HDassert( (*L) != NULL ); + HDassert( (*L)->magic == S3COMMS_HRB_NODE_MAGIC ); + node_ptr = (*L); + + /* Check whether to modify/remove first node in list + */ + if (strcmp(lowername, node_ptr->lowername) == 0) { + + is_looking = FALSE; + + if (value == NULL) { +#if S3COMMS_DEBUG +HDprintf("REMOVE HEAD\n"); fflush(stdout); +#endif + /*************** + * REMOVE HEAD * + ***************/ + + *L = node_ptr->next; + +#if S3COMMS_DEBUG +HDprintf("FREEING CAT (node)\n"); fflush(stdout); +#endif + H5MM_xfree(node_ptr->cat); +#if S3COMMS_DEBUG +HDprintf("FREEING LOWERNAME (node)\n"); fflush(stdout); +#endif + H5MM_xfree(node_ptr->lowername); +#if S3COMMS_DEBUG +HDprintf("FREEING NAME (node)\n"); fflush(stdout); +#endif + H5MM_xfree(node_ptr->name); +#if S3COMMS_DEBUG +HDprintf("FREEING VALUE (node)\n"); fflush(stdout); +#endif + H5MM_xfree(node_ptr->value); +#if S3COMMS_DEBUG +HDprintf("MAGIC OK? %s\n", + (node_ptr->magic == S3COMMS_HRB_NODE_MAGIC) ? "YES" : "NO"); +fflush(stdout); +#endif + HDassert( node_ptr->magic == S3COMMS_HRB_NODE_MAGIC ); + node_ptr->magic += 1ul; +#if S3COMMS_DEBUG +HDprintf("FREEING POINTER\n"); fflush(stdout); +#endif + H5MM_xfree(node_ptr); + +#if S3COMMS_DEBUG +HDprintf("FREEING WORKING LOWERNAME\n"); fflush(stdout); +#endif + H5MM_xfree(lowername); lowername = NULL; + } else { +#if S3COMMS_DEBUG +HDprintf("MODIFY HEAD\n"); fflush(stdout); +#endif + /*************** + * MODIFY HEAD * + ***************/ + + H5MM_xfree(node_ptr->cat); + H5MM_xfree(node_ptr->name); + H5MM_xfree(node_ptr->value); + + node_ptr->name = namecpy; + node_ptr->value = valuecpy; + node_ptr->cat = nvcat; + + H5MM_xfree(lowername); + lowername = NULL; + new_node->magic += 1ul; + H5MM_xfree(new_node); + new_node = NULL; + } + } else if (strcmp(lowername, node_ptr->lowername) < 0) { + + is_looking = FALSE; + + if (value == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "trying to remove a node 'before' head"); + } else { +#if S3COMMS_DEBUG +HDprintf("PREPEND NEW HEAD\n"); fflush(stdout); +#endif + /******************* + * INSERT NEW HEAD * + *******************/ + + new_node->name = namecpy; + new_node->value = valuecpy; + new_node->lowername = lowername; + new_node->cat = nvcat; + new_node->next = node_ptr; + *L = new_node; + } + } + + /*************** + * SEARCH LIST * + ***************/ + + while (is_looking) { + if (node_ptr->next == NULL) { + + is_looking = FALSE; + + if (value == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "trying to remove absent node"); + } else { +#if S3COMMS_DEBUG +HDprintf("APPEND A NODE\n"); fflush(stdout); +#endif + /******************* + * APPEND NEW NODE * + *******************/ + + HDassert( strcmp(lowername, node_ptr->lowername) > 0 ); + new_node->name = namecpy; + new_node->value = valuecpy; + new_node->lowername = lowername; + new_node->cat = nvcat; + node_ptr->next = new_node; + } + } else if (strcmp(lowername, node_ptr->next->lowername) < 0) { + + is_looking = FALSE; + + if (value == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "trying to remove absent node"); + } else { +#if S3COMMS_DEBUG +HDprintf("INSERT A NODE\n"); fflush(stdout); +#endif + /******************* + * INSERT NEW NODE * + *******************/ + + HDassert( strcmp(lowername, node_ptr->lowername) > 0 ); + new_node->name = namecpy; + new_node->value = valuecpy; + new_node->lowername = lowername; + new_node->cat = nvcat; + new_node->next = node_ptr->next; + node_ptr->next = new_node; + } + } else if (strcmp(lowername, node_ptr->next->lowername) == 0) { + + is_looking = FALSE; + + if (value == NULL) { + /***************** + * REMOVE A NODE * + *****************/ + + hrb_node_t *tmp = node_ptr->next; + node_ptr->next = tmp->next; + +#if S3COMMS_DEBUG +HDprintf("REMOVE A NODE\n"); fflush(stdout); +#endif + H5MM_xfree(tmp->cat); + H5MM_xfree(tmp->lowername); + H5MM_xfree(tmp->name); + H5MM_xfree(tmp->value); + + HDassert( tmp->magic == S3COMMS_HRB_NODE_MAGIC ); + tmp->magic += 1ul; + H5MM_xfree(tmp); + + H5MM_xfree(lowername); + lowername = NULL; + } else { +#if S3COMMS_DEBUG +HDprintf("MODIFY A NODE\n"); fflush(stdout); +#endif + /***************** + * MODIFY A NODE * + *****************/ + + node_ptr = node_ptr->next; + H5MM_xfree(node_ptr->name); + H5MM_xfree(node_ptr->value); + H5MM_xfree(node_ptr->cat); + + HDassert( new_node->magic == S3COMMS_HRB_NODE_MAGIC ); + new_node->magic += 1ul; + H5MM_xfree(new_node); + H5MM_xfree(lowername); + new_node = NULL; + lowername = NULL; + + node_ptr->name = namecpy; + node_ptr->value = valuecpy; + node_ptr->cat = nvcat; + } + } else { + /**************** + * KEEP LOOKING * + ****************/ + + node_ptr = node_ptr->next; + } + } + +done: + if (ret_value == FAIL) { + /* clean up + */ + if (nvcat != NULL) H5MM_xfree(nvcat); + if (namecpy != NULL) H5MM_xfree(namecpy); + if (lowername != NULL) H5MM_xfree(lowername); + if (valuecpy != NULL) H5MM_xfree(valuecpy); + if (new_node != NULL) { + HDassert( new_node->magic == S3COMMS_HRB_NODE_MAGIC ); + new_node->magic += 1ul; + H5MM_xfree(new_node); + } + } + + FUNC_LEAVE_NOAPI(ret_value); + +} /* H5FD_s3comms_hrb_node_set */ + + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_hrb_destroy() + * + * Purpose: + * + * Destroy and free resources _directly_ associated with an HTTP Buffer. + * + * Takes a pointer to pointer to the buffer structure. + * This allows for the pointer itself to be NULLed from within the call. + * + * If buffer or buffer pointer is NULL, there is no effect. + * + * Headers list at `first_header` is not touched. + * + * - Programmer should re-use or destroy `first_header` pointer + * (hrb_node_t *) as suits their purposes. + * - Recommend fetching prior to destroy() + * e.g., `reuse_node = hrb_to_die->first_header; destroy(hrb_to_die);` + * or maintaining an external reference. + * - Destroy node/list separately as appropriate + * - Failure to account for this will result in a memory leak. + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - successfully released buffer resources + * - if `buf` is NULL or `*buf` is NULL, no effect + * - FAILURE: `FAIL` + * - `buf->magic != S3COMMS_HRB_MAGIC` + * + * Programmer: Jacob Smith + * 2017-07-21 + * + * Changes: + * + * - Conditional free() of `hrb_node_t` pointer properties based on + * `which_free` property. + * --- Jacob Smith 2017-08-08 + * + * - Integrate with HDF5. + * - Returns herr_t instead of nothing. + * --- Jacob Smith 2017-09-21 + * + * - Change argument to from *buf to **buf, to null pointer within call + * --- Jacob Smith 2017-20-05 + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_hrb_destroy(hrb_t **_buf) +{ + hrb_t *buf = NULL; + herr_t ret_value = SUCCEED; + + + + FUNC_ENTER_NOAPI_NOINIT + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_hrb_destroy.\n"); +#endif + + if (_buf != NULL && *_buf != NULL) { + buf = *_buf; + if (buf->magic != S3COMMS_HRB_MAGIC) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "pointer's magic does not match.\n"); + } + + H5MM_xfree(buf->verb); + H5MM_xfree(buf->version); + H5MM_xfree(buf->resource); + buf->magic += 1ul; + H5MM_xfree(buf); + *_buf = NULL; + } + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_s3comms_hrb_destroy */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_hrb_init_request() + * + * Purpose: + * + * Create a new HTTP Request Buffer + * + * All non-null arguments should be null-terminated strings. + * + * If `verb` is NULL, defaults to "GET". + * If `http_version` is NULL, defaults to "HTTP/1.1". + * + * `resource` cannot be NULL; should be string beginning with slash + * character ('/'). + * + * All strings are copied into the structure, making them safe from + * modification in source strings. + * + * Return: + * + * - SUCCESS: pointer to new `hrb_t` + * - FAILURE: `NULL` + * + * Programmer: Jacob Smith + * 2017-07-21 + * + * Changes: + * + * - Update struct membership for newer 'generic' `hrb_t` format. + * --- Jacob Smith, 2017-07-24 + * + * - Rename from `hrb_new()` to `hrb_request()` + * --- Jacob Smith, 2017-07-25 + * + * - Integrate with HDF5. + * - Rename from 'hrb_request()` to `H5FD_s3comms_hrb_init_request()`. + * - Remove `host` from input parameters. + * - Host, as with all other fields, must now be added through the + * add-field functions. + * - Add `version` (HTTP version string, e.g. "HTTP/1.1") to parameters. + * --- Jacob Smith 2017-09-20 + * + * - Update to use linked-list `hrb_node_t` headers in structure. + * --- Jacob Smith 2017-10-05 + * + *---------------------------------------------------------------------------- + */ +hrb_t * +H5FD_s3comms_hrb_init_request(const char *_verb, + const char *_resource, + const char *_http_version) +{ + hrb_t *request = NULL; + char *res = NULL; + size_t reslen = 0; + hrb_t *ret_value = NULL; + char *verb = NULL; + size_t verblen = 0; + char *vrsn = NULL; + size_t vrsnlen = 0; + + + + FUNC_ENTER_NOAPI_NOINIT + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_hrb_init_request.\n"); +#endif + + if (_resource == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "resource string cannot be null.\n"); + } + + /* populate valid NULLs with defaults + */ + if (_verb == NULL) + _verb = "GET"; + + if (_http_version == NULL) + _http_version = "HTTP/1.1"; + + /* malloc space for and prepare structure + */ + request = (hrb_t *)H5MM_malloc(sizeof(hrb_t)); + if (request == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, NULL, + "no space for request structure"); + } + request->magic = S3COMMS_HRB_MAGIC; + request->body = NULL; + request->body_len = 0; + request->first_header = NULL; + + + + /* malloc and copy strings for the structure + */ + if (_resource[0] == '/') { + reslen = HDstrlen(_resource) + 1; + res = (char *)H5MM_malloc(sizeof(char) * reslen); + if (res == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, NULL, + "no space for resource string"); + } + HDstrncpy(res, _resource, reslen); + } else { + int sprint_ret = 0; + reslen = HDstrlen(_resource) + 2; + res = (char *)H5MM_malloc(sizeof(char) * reslen); + if (res == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, NULL, + "no space for resource string"); + } + sprint_ret = HDsnprintf(res, reslen, "/%s", _resource); + if (sprint_ret <= 0) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "errro while appending resource string %s", _resource); + HDassert( (reslen - 1) == (size_t)sprint_ret ); + } /* start resource string with '/' */ + + verblen = HDstrlen(_verb) + 1; + verb = (char *)H5MM_malloc(sizeof(char) * verblen); + if (verb == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "no space for verb string"); + } + HDstrncpy(verb, _verb, verblen); + + vrsnlen = HDstrlen(_http_version) + 1; + vrsn = (char *)H5MM_malloc(sizeof(char) * vrsnlen); + if (vrsn == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "no space for http-version string"); + } + HDstrncpy(vrsn, _http_version, vrsnlen); + + + + /* place new copies into structure + */ + request->resource = res; + request->verb = verb; + request->version = vrsn; + + ret_value = request; + +done: + + /* if there is an error, clean up after ourselves + */ + if (ret_value == NULL) { + if (request != NULL) H5MM_xfree(request); + if (vrsn != NULL) H5MM_xfree(vrsn); + if (verb != NULL) H5MM_xfree(verb); + if (res != NULL) H5MM_xfree(res); + } + + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_s3comms_hrb_init_request */ + + + +/**************************************************************************** + * S3R FUNCTIONS + ****************************************************************************/ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_s3r_close() + * + * Purpose: + * + * Close communications through given S3 Request Handle (`s3r_t`) + * and clean up associated resources. + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - FAILURE: `FAIL` + * - fails if handle is null or has invalid magic number + * + * + * Programmer: Jacob Smith + * 2017-08-31 + * + * Changes: + * + * - Remove all messiness related to the now-gone "setopt" utility + * as it no longer exists in the handle. + * - Return type to `void`. + * --- Jacob Smith 2017-09-01 + * + * - Incorporate into HDF environment. + * - Rename from `s3r_close()` to `H5FD_s3comms_s3r_close()`. + * --- Jacob Smith 2017-10-06 + * + * - Change separate host, resource, port info to `parsed_url_t` struct ptr. + * --- Jacob Smith 2017-11-01 + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_s3r_close(s3r_t *handle) +{ + herr_t ret_value = SUCCEED; + + + + FUNC_ENTER_NOAPI_NOINIT + +#ifdef H5_HAVE_ROS3_VFD +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_s3r_close.\n"); +#endif + + if (handle == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle cannot be null.\n"); + } + if (handle->magic != S3COMMS_S3R_MAGIC) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle has invalid magic.\n"); + } + + curl_easy_cleanup(handle->curlhandle); + + H5MM_xfree(handle->secret_id); + H5MM_xfree(handle->region); + H5MM_xfree(handle->signing_key); + + HDassert( handle->httpverb != NULL ); + H5MM_xfree(handle->httpverb); + + if (FAIL == H5FD_s3comms_free_purl(handle->purl)) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to release parsed url structure") + } + + H5MM_xfree(handle); + +#endif /* H5_HAVE_ROS3_VFD */ + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_s3comms_s3r_close */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_s3r_get_filesize() + * + * Purpose: + * + * Retrieve the filesize of an open request handle. + * + * Wrapper "getter" to hide implementation details. + * + * + * Return: + * + * - SUCCESS: size of file, in bytes, if handle is valid. + * - FAILURE: 0, if handle is NULL or undefined. + * + * Programmer: Jacob Smith 2017-01-14 + * + * Changes: None + * + *---------------------------------------------------------------------------- + */ +size_t +H5FD_s3comms_s3r_get_filesize(s3r_t *handle) { + + size_t ret_value = 0; + + FUNC_ENTER_NOAPI_NOINIT_NOERR + +#ifdef H5_HAVE_ROS3_VFD + if (handle != NULL) + ret_value = handle->filesize; +#endif /* H5_HAVE_ROS3_VFD */ + + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_s3comms_s3r_get_filesize */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_s3r_getsize() + * + * Purpose: + * + * Get the number of bytes of handle's target resource. + * + * Sets handle and curlhandle with to enact an HTTP HEAD request on file, + * and parses received headers to extract "Content-Length" from response + * headers, storing file size at `handle->filesize`. + * + * Critical step in opening (initiating) an `s3r_t` handle. + * + * Wraps `s3r_read()`. + * Sets curlhandle to write headers to a temporary buffer (using extant + * write callback) and provides no buffer for body. + * + * Upon exit, unsets HTTP HEAD settings from curl handle, returning to + * initial state. In event of error, curl handle state is undefined and is + * not to be trusted. + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - FAILURE: `FAIL` + * + * Programmer: Jacob Smith + * 2017-08-23 + * + * Changes: + * + * - Update to revised `s3r_t` format and life cycle. + * --- Jacob Smith 2017-09-01 + * + * - Conditional change to static header buffer and structure. + * --- Jacob Smith 2017-09-05 + * + * - Incorporate into HDF environment. + * - Rename from `s3r_getsize()` to `H5FD_s3comms_s3r_getsize()`. + * --- Jacob Smith 2017-10-06 + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_s3r_getsize(s3r_t *handle) +{ +#ifdef H5_HAVE_ROS3_VFD + unsigned long int content_length = 0; + CURL *curlh = NULL; + char *end = NULL; + char *headerresponse = NULL; + herr_t ret_value = SUCCEED; + struct s3r_datastruct sds = { + S3COMMS_CALLBACK_DATASTRUCT_MAGIC, + NULL, + 0 }; + char *start = NULL; +#else + herr_t ret_value = FAIL; +#endif /* H5_HAVE_ROS3_VFD */ + + + + FUNC_ENTER_NOAPI_NOINIT + +#ifdef H5_HAVE_ROS3_VFD + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_s3r_getsize.\n"); +#endif + + if (handle == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle cannot be null.\n"); + } + if (handle->magic != S3COMMS_S3R_MAGIC) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle has invalid magic.\n"); + } + if (handle->curlhandle == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle has bad (null) curlhandle.\n") + } + + /******************** + * PREPARE FOR HEAD * + ********************/ + + curlh = handle->curlhandle; + + if ( CURLE_OK != + curl_easy_setopt(curlh, + CURLOPT_NOBODY, + 1L) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "error while setting CURL option (CURLOPT_NOBODY). " + "(placeholder flags)"); + } + + if ( CURLE_OK != + curl_easy_setopt(curlh, + CURLOPT_HEADERDATA, + &sds) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "error while setting CURL option (CURLOPT_HEADERDATA). " + "(placeholder flags)"); + } + + HDassert( handle->httpverb == NULL ); + handle->httpverb = (char *)H5MM_malloc(sizeof(char) * 16); + if (handle->httpverb == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, FAIL, + "unable to allocate space for S3 request HTTP verb"); + } + HDmemcpy(handle->httpverb, "HEAD", 5); + + headerresponse = (char *)H5MM_malloc(sizeof(char) * CURL_MAX_HTTP_HEADER); + if (headerresponse == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, FAIL, + "unable to allocate space for curl header response"); + } + sds.data = headerresponse; + + /******************* + * PERFORM REQUEST * + *******************/ + + /* these parameters fetch the entire file, + * but, with a NULL destination and NOBODY and HEADERDATA supplied above, + * only http metadata will be sent by server and recorded by s3comms + */ + if (FAIL == + H5FD_s3comms_s3r_read(handle, 0, 0, NULL) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem in reading during getsize.\n"); + } + + if (sds.size > CURL_MAX_HTTP_HEADER) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "HTTP metadata buffer overrun\n"); + } else if (sds.size == 0) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "No HTTP metadata\n"); +#if S3COMMS_DEBUG + } else { + HDfprintf(stderr, "GETSIZE: OK\n"); +#endif + } + + + /****************** + * PARSE RESPONSE * + ******************/ + + start = strstr(headerresponse, + "\r\nContent-Length: "); + if (start == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "could not find \"Content-Length\" in response.\n"); + } + + /* move "start" to beginning of value in line; find end of line + */ + start = start + HDstrlen("\r\nContent-Length: "); + end = strstr(start, "\r\n"); + if (end == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "could not find end of content length line"); + } + + /* place null terminator at end of numbers + */ + *end = '\0'; + + content_length = strtoul((const char *)start, + NULL, + 0); + if (content_length == 0 || + content_length == ULONG_MAX || + errno == ERANGE) /* errno set by strtoul */ + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "could not convert found \"Content-Length\" response (\"%s\")", + start); /* range is null-terminated, remember */ + } + + handle->filesize = (size_t)content_length; + + /********************** + * UNDO HEAD SETTINGS * + **********************/ + + if ( CURLE_OK != + curl_easy_setopt(curlh, + CURLOPT_NOBODY, + 0) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "error while setting CURL option (CURLOPT_NOBODY). " + "(placeholder flags)"); + } + + if ( CURLE_OK != + curl_easy_setopt(curlh, + CURLOPT_HEADERDATA, + 0) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "error while setting CURL option (CURLOPT_HEADERDATA). " + "(placeholder flags)"); + } + +done: + H5MM_xfree(headerresponse); + sds.magic += 1; /* set to bad magic */ + +#endif /* H5_HAVE_ROS3_VFD */ + + FUNC_LEAVE_NOAPI(ret_value); + +} /* H5FD_s3comms_s3r_getsize */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_s3r_open() + * + * Purpose: + * + * Logically 'open' a file hosted on S3. + * + * - create new Request Handle + * - copy supplied url + * - copy authentication info if supplied + * - create CURL handle + * - fetch size of file + * - connect with server and execute HEAD request + * - return request handle ready for reads + * + * To use 'default' port to connect, `port` should be 0. + * + * To prevent AWS4 authentication, pass null pointer to `region`, `id`, + * and `signing_key`. + * + * Uses `H5FD_s3comms_parse_url()` to validate and parse url input. + * + * Return: + * + * - SUCCESS: Pointer to new request handle. + * - FAILURE: NULL + * - occurs if: + * - authentication strings are inconsistent + * - must _all_ be null, or have at least `region` and `id` + * - url is NULL (no filename) + * - unable to parse url (malformed?) + * - error while performing `getsize()` + * + * Programmer: Jacob Smith + * 2017-09-01 + * + * Changes: + * + * - Incorporate into HDF environment. + * - Rename from `s3r_open()` to `H5FD_s3comms_s3r_open()`. + * --- Jacob Smith 2017-10-06 + * + * - Remove port number from signature. + * - Name (`url`) must be complete url with http scheme and optional port + * number in string. + * - e.g., "http://bucket.aws.com:9000/myfile.dat?query=param" + * - Internal storage of host, resource, and port information moved into + * `parsed_url_t` struct pointer. + * --- Jacob Smith 2017-11-01 + * + *---------------------------------------------------------------------------- + */ +s3r_t * +H5FD_s3comms_s3r_open(const char *url, + const char *region, + const char *id, + const unsigned char *signing_key) +{ +#ifdef H5_HAVE_ROS3_VFD + size_t tmplen = 0; + CURL *curlh = NULL; + s3r_t *handle = NULL; + parsed_url_t *purl = NULL; +#endif + s3r_t *ret_value = NULL; + + + + FUNC_ENTER_NOAPI_NOINIT + +#ifdef H5_HAVE_ROS3_VFD + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_s3r_open.\n"); +#endif + + + + if (url == NULL || url[0] == '\0') { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "url cannot be null.\n"); + } + + if (FAIL == H5FD_s3comms_parse_url(url, &purl)) { + /* probably a malformed url, but could be internal error */ + HGOTO_ERROR(H5E_ARGS, H5E_CANTCREATE, NULL, + "unable to create parsed url structure"); + } + HDassert( purl != NULL ); /* if above passes, this must be true */ + HDassert( purl->magic == S3COMMS_PARSED_URL_MAGIC ); + + handle = (s3r_t *)H5MM_malloc(sizeof(s3r_t)); + if (handle == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, NULL, + "could not malloc space for handle.\n"); + } + + handle->magic = S3COMMS_S3R_MAGIC; + handle->purl = purl; + handle->filesize = 0; + handle->region = NULL; + handle->secret_id = NULL; + handle->signing_key = NULL; + handle->httpverb = NULL; + + /************************************* + * RECORD AUTHENTICATION INFORMATION * + *************************************/ + + if ((region != NULL && *region != '\0') || + (id != NULL && *id != '\0') || + (signing_key != NULL && *signing_key != '\0')) + { + /* if one exists, all three must exist + */ + if (region == NULL || region[0] == '\0') { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "region cannot be null.\n"); + } + if (id == NULL || id[0] == '\0') { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "secret id cannot be null.\n"); + } + if (signing_key == NULL || signing_key[0] == '\0') { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "signing key cannot be null.\n"); + } + + /* copy strings + */ + tmplen = HDstrlen(region) + 1; + handle->region = (char *)H5MM_malloc(sizeof(char) * tmplen); + if (handle->region == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "could not malloc space for handle region copy.\n"); + } + HDmemcpy(handle->region, region, tmplen); + + tmplen = HDstrlen(id) + 1; + handle->secret_id = (char *)H5MM_malloc(sizeof(char) * tmplen); + if (handle->secret_id == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "could not malloc space for handle ID copy.\n"); + } + HDmemcpy(handle->secret_id, id, tmplen); + + tmplen = SHA256_DIGEST_LENGTH; + handle->signing_key = + (unsigned char *)H5MM_malloc(sizeof(unsigned char) * tmplen); + if (handle->signing_key == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "could not malloc space for handle key copy.\n"); + } + HDmemcpy(handle->signing_key, signing_key, tmplen); + } /* if authentication information provided */ + + /************************ + * INITIATE CURL HANDLE * + ************************/ + + curlh = curl_easy_init(); + + if (curlh == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "problem creating curl easy handle!\n"); + } + + if ( CURLE_OK != + curl_easy_setopt(curlh, + CURLOPT_HTTPGET, + 1L) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "error while setting CURL option (CURLOPT_HTTPGET). " + "(placeholder flags)"); + } + + if ( CURLE_OK != + curl_easy_setopt(curlh, + CURLOPT_HTTP_VERSION, + CURL_HTTP_VERSION_1_1) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "error while setting CURL option (CURLOPT_HTTP_VERSION). " + "(placeholder flags)"); + } + + if ( CURLE_OK != + curl_easy_setopt(curlh, + CURLOPT_FAILONERROR, + 1L) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "error while setting CURL option (CURLOPT_FAILONERROR). " + "(placeholder flags)"); + } + + if ( CURLE_OK != + curl_easy_setopt(curlh, + CURLOPT_WRITEFUNCTION, + curlwritecallback) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "error while setting CURL option (CURLOPT_WRITEFUNCTION). " + "(placeholder flags)"); + } + + if ( CURLE_OK != + curl_easy_setopt(curlh, + CURLOPT_URL, + url) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "error while setting CURL option (CURLOPT_URL). " + "(placeholder flags)"); + } + +#if S3COMMS_CURL_VERBOSITY > 1 + /* CURL will print (to stdout) information for each operation + */ + curl_easy_setopt(curlh, CURLOPT_VERBOSE, 1L); +#endif + + handle->curlhandle = curlh; + + /******************* + * OPEN CONNECTION * + * * * * * * * * * * + * GET FILE SIZE * + *******************/ + + if (FAIL == + H5FD_s3comms_s3r_getsize(handle) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "problem in H5FD_s3comms_s3r_getsize.\n"); + } + + /********************* + * FINAL PREPARATION * + *********************/ + + HDassert( handle->httpverb != NULL ); + HDmemcpy(handle->httpverb, "GET", 4); + + ret_value = handle; +#endif /* H5_HAVE_ROS3_VFD */ + +done: + if (ret_value == NULL) { +#ifdef H5_HAVE_ROS3_VFD + if (curlh != NULL) { + curl_easy_cleanup(curlh); + } + if (FAIL == H5FD_s3comms_free_purl(purl)) { + HDONE_ERROR(H5E_ARGS, H5E_BADVALUE, NULL, + "unable to free parsed url structure") + } + if (handle != NULL) { + H5MM_xfree(handle->region); + H5MM_xfree(handle->secret_id); + H5MM_xfree(handle->signing_key); + if (handle->httpverb != NULL) { + H5MM_xfree(handle->httpverb); + } + H5MM_xfree(handle); + } +#endif /* H5_HAVE_ROS3_VFD */ + } + + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_s3comms_s3r_open */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_s3r_read() + * + * Purpose: + * + * Read file pointed to by request handle, writing specified + * `offset` .. `offset + len` bytes to buffer `dest`. + * + * If `len` is 0, reads entirety of file starting at `offset`. + * If `offset` and `len` are both 0, reads entire file. + * + * If `offset` or `offset+len` is greater than the file size, read is + * aborted and returns `FAIL`. + * + * Uses configured "curl easy handle" to perform request. + * + * In event of error, buffer should remain unaltered. + * + * If handle is set to authorize a request, creates a new (temporary) + * HTTP Request object (hrb_t) for generating requisite headers, + * which is then translated to a `curl slist` and set in the curl handle + * for the request. + * + * `dest` _may_ be NULL, but no body data will be recorded. + * + * - In general practice, NULL should never be passed in as `dest`. + * - NULL `dest` passed in by internal function `s3r_getsize()`, in + * conjunction with CURLOPT_NOBODY to preempt transmission of file data + * from server. + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - FAILURE: `FAIL` + * + * Programmer: Jacob Smith + * 2017-08-22 + * + * Changes: + * + * - Revise structure to prevent unnecessary hrb_t element creation. + * - Rename tmprstr -> rangebytesstr to reflect purpose. + * - Insert needed `free()`s, particularly for `sds`. + * --- Jacob Smith 2017-08-23 + * + * - Revise heavily to accept buffer, range as parameters. + * - Utilize modified s3r_t format. + * --- Jacob Smith 2017-08-31 + * + * - Incorporate into HDF library. + * - Rename from `s3r_read()` to `H5FD_s3comms_s3r_read()`. + * - Return `herr_t` succeed/fail instead of S3code. + * - Update to use revised `hrb_t` and `hrb_node_t` structures. + * --- Jacob Smith 2017-10-06 + * + * - Update to use `parsed_url_t *purl` in handle. + * --- Jacob Smith 2017-11-01 + * + * - Better define behavior upon read past EOF + * --- Jacob Smith 2017-01-19 + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_s3r_read(s3r_t *handle, + haddr_t offset, + size_t len, + void *dest) +{ +#ifdef H5_HAVE_ROS3_VFD + CURL *curlh = NULL; + CURLcode p_status = CURLE_OK; + struct curl_slist *curlheaders = NULL; + hrb_node_t *headers = NULL; + hrb_node_t *node = NULL; + struct tm *now = NULL; + char *rangebytesstr = NULL; + hrb_t *request = NULL; + int ret = 0; /* working variable to check */ + /* return value of HDsnprintf */ + struct s3r_datastruct *sds = NULL; + herr_t ret_value = SUCCEED; +#else + herr_t ret_value = FAIL; +#endif /* H5_HAVE_ROS3_VFD */ + + + + FUNC_ENTER_NOAPI_NOINIT + +#ifdef H5_HAVE_ROS3_VFD + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_s3r_read.\n"); +#endif + + /************************************** + * ABSOLUTELY NECESSARY SANITY-CHECKS * + **************************************/ + + if (handle == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle cannot be null.\n"); + } + if (handle->magic != S3COMMS_S3R_MAGIC) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle has invalid magic.\n"); + } + if (handle->curlhandle == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle has bad (null) curlhandle.\n") + } + if (handle->purl == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle has bad (null) url.\n") + } + HDassert( handle->purl->magic == S3COMMS_PARSED_URL_MAGIC ); + if (offset > handle->filesize || (len + offset) > handle->filesize) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to read past EoF") + } + + curlh = handle->curlhandle; + + /********************* + * PREPARE WRITEDATA * + *********************/ + + if (dest != NULL) { + sds = (struct s3r_datastruct *)H5MM_malloc( + sizeof(struct s3r_datastruct)); + if (sds == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, FAIL, + "could not malloc destination datastructure.\n"); + } + + sds->magic = S3COMMS_CALLBACK_DATASTRUCT_MAGIC; + sds->data = (char *)dest; + sds->size = 0; + if (CURLE_OK != + curl_easy_setopt(curlh, + CURLOPT_WRITEDATA, + sds) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_UNINITIALIZED, FAIL, + "error while setting CURL option (CURLOPT_WRITEDATA). " + "(placeholder flags)"); + } + } + + /********************* + * FORMAT HTTP RANGE * + *********************/ + + if (len > 0) { + rangebytesstr = (char *)H5MM_malloc(sizeof(char) * \ + S3COMMS_MAX_RANGE_STRING_SIZE ); + if (rangebytesstr == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, FAIL, + "could not malloc range format string.\n"); + } + ret = HDsnprintf(rangebytesstr, + (S3COMMS_MAX_RANGE_STRING_SIZE), + "bytes="H5_PRINTF_HADDR_FMT"-"H5_PRINTF_HADDR_FMT, + offset, + offset + len - 1); + if (ret == 0 || ret >= S3COMMS_MAX_RANGE_STRING_SIZE) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to format HTTP Range value"); + } else if (offset > 0) { + rangebytesstr = (char *)H5MM_malloc(sizeof(char) * \ + S3COMMS_MAX_RANGE_STRING_SIZE); + if (rangebytesstr == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, FAIL, + "could not malloc range format string.\n"); + } + ret = HDsnprintf(rangebytesstr, + (S3COMMS_MAX_RANGE_STRING_SIZE), + "bytes="H5_PRINTF_HADDR_FMT"-", + offset); + if (ret == 0 || ret >= S3COMMS_MAX_RANGE_STRING_SIZE) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to format HTTP Range value"); + } + + /******************* + * COMPILE REQUEST * + *******************/ + + if (handle->signing_key == NULL) { + /* Do not authenticate. + */ + if (rangebytesstr != NULL) { + /* Pass in range directly + */ + char *bytesrange_ptr = NULL; /* pointer past "bytes=" portion */ + + bytesrange_ptr = strchr(rangebytesstr, '='); + HDassert( bytesrange_ptr != NULL ); + bytesrange_ptr++; /* move to first char past '=' */ + HDassert( *bytesrange_ptr != '\0' ); + + if (CURLE_OK != + curl_easy_setopt(curlh, + CURLOPT_RANGE, + bytesrange_ptr) ) + { + HGOTO_ERROR(H5E_VFL, H5E_UNINITIALIZED, FAIL, + "error while setting CURL option (CURLOPT_RANGE). "); + } + } + } else { + /* authenticate request + */ + char authorization[512]; + /* 512 := approximate max length... + * 67 <len("AWS4-HMAC-SHA256 Credential=///s3/aws4_request," + * "SignedHeaders=,Signature=")> + * + 8 <yyyyMMDD> + * + 64 <hex(sha256())> + * + 128 <max? len(secret_id)> + * + 20 <max? len(region)> + * + 128 <max? len(signed_headers)> + */ + char buffer1[512]; /* -> Canonical Request -> Signature */ + char buffer2[256]; /* -> String To Sign -> Credential */ + char iso8601now[ISO8601_SIZE]; + char signed_headers[48]; + /* should be large enough for nominal listing: + * "host;range;x-amz-content-sha256;x-amz-date" + * + '\0', with "range;" possibly absent + */ + + /* zero start of strings */ + authorization[0] = 0; + buffer1[0] = 0; + buffer2[0] = 0; + iso8601now[0] = 0; + signed_headers[0] = 0; + + /**** VERIFY INFORMATION EXISTS ****/ + + if (handle->region == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle must have non-null region.\n"); + } + if (handle->secret_id == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle must have non-null secret_id.\n"); + } + if (handle->signing_key == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle must have non-null signing_key.\n"); + } + if (handle->httpverb == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle must have non-null httpverb.\n"); + } + if (handle->purl->host == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle must have non-null host.\n"); + } + if (handle->purl->path == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "handle must have non-null resource.\n"); + } + + /**** CREATE HTTP REQUEST STRUCTURE (hrb_t) ****/ + + request = H5FD_s3comms_hrb_init_request( + (const char *)handle->httpverb, + (const char *)handle->purl->path, + "HTTP/1.1"); + if (request == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "could not allocate hrb_t request.\n"); + } + HDassert( request->magic == S3COMMS_HRB_MAGIC ); + + now = gmnow(); + if (ISO8601NOW(iso8601now, now) != (ISO8601_SIZE - 1)) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "could not format ISO8601 time.\n"); + } + + if (FAIL == + H5FD_s3comms_hrb_node_set( + &headers, + "x-amz-date", + (const char *)iso8601now) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to set x-amz-date header") + } + if (headers == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem building headers list. " + "(placeholder flags)\n"); + } + HDassert( headers->magic == S3COMMS_HRB_NODE_MAGIC ); + + if (FAIL == + H5FD_s3comms_hrb_node_set( + &headers, + "x-amz-content-sha256", + (const char *)EMPTY_SHA256) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to set x-amz-content-sha256 header") + } + if (headers == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem building headers list. " + "(placeholder flags)\n"); + } + HDassert( headers->magic == S3COMMS_HRB_NODE_MAGIC ); + + if (rangebytesstr != NULL) { + if (FAIL == + H5FD_s3comms_hrb_node_set( + &headers, + "Range", + (const char *)rangebytesstr) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to set range header") + } + if (headers == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem building headers list. " + "(placeholder flags)\n"); + } + HDassert( headers->magic == S3COMMS_HRB_NODE_MAGIC ); + } + + if (FAIL == + H5FD_s3comms_hrb_node_set( + &headers, + "Host", + (const char *)handle->purl->host) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to set host header") + } + if (headers == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem building headers list. " + "(placeholder flags)\n"); + } + HDassert( headers->magic == S3COMMS_HRB_NODE_MAGIC ); + + request->first_header = headers; + + /**** COMPUTE AUTHORIZATION ****/ + + if (FAIL == /* buffer1 -> canonical request */ + H5FD_s3comms_aws_canonical_request(buffer1, + signed_headers, + request) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "(placeholder flags)\n"); + } + if ( FAIL == /* buffer2->string-to-sign */ + H5FD_s3comms_tostringtosign(buffer2, + buffer1, + iso8601now, + handle->region) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "(placeholder flags)\n"); + } + if (FAIL == /* buffer1 -> signature */ + H5FD_s3comms_HMAC_SHA256(handle->signing_key, + SHA256_DIGEST_LENGTH, + buffer2, + HDstrlen(buffer2), + buffer1) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "(placeholder flags)\n"); + } + + iso8601now[8] = 0; /* trim to yyyyMMDD */ + ret = S3COMMS_FORMAT_CREDENTIAL(buffer2, + handle->secret_id, + iso8601now, + handle->region, + "s3"); + if (ret == 0 || ret >= S3COMMS_MAX_CREDENTIAL_SIZE) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to format aws4 credential string"); + + ret = HDsnprintf(authorization, + 512, + "AWS4-HMAC-SHA256 Credential=%s,SignedHeaders=%s,Signature=%s", + buffer2, + signed_headers, + buffer1); + if (ret == 0 || ret >= 512) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to format aws4 authorization string"); + + /* append authorization header to http request buffer + */ + if (FAIL == + H5FD_s3comms_hrb_node_set( + &headers, + "Authorization", + (const char *)authorization) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to set Authorization header") + } + if (headers == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem building headers list. " + "(placeholder flags)\n"); + } + + /* update hrb's "first header" pointer + */ + request->first_header = headers; + + /**** SET CURLHANDLE HTTP HEADERS FROM GENERATED DATA ****/ + + node = request->first_header; + while (node != NULL) { + HDassert( node->magic == S3COMMS_HRB_NODE_MAGIC ); + curlheaders = curl_slist_append(curlheaders, + (const char *)node->cat); + if (curlheaders == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "could not append header to curl slist. " + "(placeholder flags)\n"); + } + node = node->next; + } + + /* sanity-check + */ + if (curlheaders == NULL) { + /* above loop was probably never run */ + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "curlheaders was never populated.\n"); + } + + /* finally, set http headers in curl handle + */ + if (CURLE_OK != + curl_easy_setopt(curlh, + CURLOPT_HTTPHEADER, + curlheaders) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "error while setting CURL option " + "(CURLOPT_HTTPHEADER). (placeholder flags)"); + } + + } /* if should authenticate (info provided) */ + + /******************* + * PERFORM REQUEST * + *******************/ + +#if S3COMMS_CURL_VERBOSITY > 0 + /* In event of error, print detailed information to stderr + * This is not the default behavior. + */ + { + long int httpcode = 0; + char curlerrbuf[CURL_ERROR_SIZE]; + curlerrbuf[0] = '\0'; + + if (CURLE_OK != + curl_easy_setopt(curlh, CURLOPT_ERRORBUFFER, curlerrbuf) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem setting error buffer") + } + + p_status = curl_easy_perform(curlh); + + if (p_status != CURLE_OK) { + if (CURLE_OK != + curl_easy_getinfo(curlh, CURLINFO_RESPONSE_CODE, &httpcode) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem getting response code") + } + HDfprintf(stderr, "CURL ERROR CODE: %d\nHTTP CODE: %d\n", + p_status, httpcode); + HDfprintf(stderr, "%s\n", curl_easy_strerror(p_status)); + HGOTO_ERROR(H5E_VFL, H5E_CANTOPENFILE, FAIL, + "problem while performing request.\n"); + } + if (CURLE_OK != + curl_easy_setopt(curlh, CURLOPT_ERRORBUFFER, NULL) ) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem unsetting error buffer") + } + } /* verbose error reporting */ +#else + p_status = curl_easy_perform(curlh); + + if (p_status != CURLE_OK) { + HGOTO_ERROR(H5E_VFL, H5E_CANTOPENFILE, FAIL, + "curl cannot perform request\n") + } +#endif + +#if S3COMMS_DEBUG + if (dest != NULL) { + HDfprintf(stderr, "len: %d\n", (int)len); + HDfprintf(stderr, "CHECKING FOR BUFFER OVERFLOW\n"); + if (sds == NULL) { + HDfprintf(stderr, "sds is NULL!\n"); + } else { + HDfprintf(stderr, "sds: 0x%lx\n", (long long)sds); + HDfprintf(stderr, "sds->size: %d\n", (int)sds->size); + if (len > sds->size) { + HDfprintf(stderr, "buffer overwrite\n"); + } + } + } else { + HDfprintf(stderr, "performed on entire file\n"); + } +#endif + +done: + /* clean any malloc'd resources + */ + if (curlheaders != NULL) { + curl_slist_free_all(curlheaders); + curlheaders = NULL; + } + if (rangebytesstr != NULL) { + H5MM_xfree(rangebytesstr); + rangebytesstr = NULL; + } + if (sds != NULL) { + H5MM_xfree(sds); + sds = NULL; + } + if (request != NULL) { + while (headers != NULL) + if (FAIL == + H5FD_s3comms_hrb_node_set(&headers, headers->name, NULL)) + { + HDONE_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "cannot release header node") + } + HDassert( NULL == headers ); + if (FAIL == H5FD_s3comms_hrb_destroy(&request)) { + HDONE_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "cannot release header request structure") + } + HDassert( NULL == request ); + } + + if (curlh != NULL) { + /* clear any Range */ + if (CURLE_OK != curl_easy_setopt(curlh, CURLOPT_RANGE, NULL) ) + HDONE_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "cannot unset CURLOPT_RANGE") + + /* clear headers */ + if (CURLE_OK != curl_easy_setopt(curlh, CURLOPT_HTTPHEADER, NULL) ) + HDONE_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "cannot unset CURLOPT_HTTPHEADER") + } + +#endif /* H5_HAVE_ROS3_VFD */ + + FUNC_LEAVE_NOAPI(ret_value); + +} /* H5FD_s3comms_s3r_read */ + + + +/**************************************************************************** + * MISCELLANEOUS FUNCTIONS + ****************************************************************************/ + + +/*---------------------------------------------------------------------------- + * + * Function: gmnow() + * + * Purpose: + * + * Get the output of `time.h`'s `gmtime()` call while minimizing setup + * clutter where important. + * + * Return: + * + * Pointer to resulting `struct tm`,as created by gmtime(time_t * T). + * + * Programmer: Jacob Smith + * 2017-07-12 + * + * Changes: None. + * + *---------------------------------------------------------------------------- + */ +struct tm * +gmnow(void) +{ + time_t now; + time_t *now_ptr = &now; + struct tm *ret_value = NULL; + + /* Doctor assert, checks against error in time() */ + if ( (time_t)(-1) != time(now_ptr) ) + ret_value = gmtime(now_ptr); + + HDassert( ret_value != NULL ); + + return ret_value; + +} /* gmnow */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_aws_canonical_request() + * + * Purpose: + * + * Compose AWS "Canonical Request" (and signed headers string) + * as defined in the REST API documentation. + * + * Both destination strings are null-terminated. + * + * Destination string arguments must be provided with adequate space. + * + * Canonical Request format: + * + * <HTTP VERB>"\n" + * <resource path>"\n" + * <query string>"\n" + * <header1>"\n" (`lowercase(name)`":"`trim(value)`) + * <header2>"\n" + * ... (headers sorted by name) + * <header_n>"\n" + * "\n" + * <signed headers>"\n" (`lowercase(header 1 name)`";"`header 2 name`;...) + * <hex-string of sha256sum of body> ("e3b0c4429...", e.g.) + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - writes canonical request to respective `...dest` strings + * - FAILURE: `FAIL` + * - one or more input argument was NULL + * - internal error + * + * Programmer: Jacob Smith + * 2017-10-04 + * + * Changes: None. + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_aws_canonical_request(char *canonical_request_dest, + char *signed_headers_dest, + hrb_t *http_request) +{ + hrb_node_t *node = NULL; + const char *query_params = ""; /* unused at present */ + herr_t ret_value = SUCCEED; + int ret = 0; /* return value of HDsnprintf */ + size_t len = 0; /* working string length variable */ + char tmpstr[256]; + + /* "query params" refers to the optional element in the URL, e.g. + * http://bucket.aws.com/myfile.txt?max-keys=2&prefix=J + * ^-----------------^ + * + * Not handled/implemented as of 2017-10-xx. + * Element introduced as empty placeholder and reminder. + * Further research to be done if this is ever relevant for the + * VFD use-cases. + */ + + + + FUNC_ENTER_NOAPI_NOINIT + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_aws_canonical_request.\n"); +#endif + + if (http_request == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "hrb object cannot be null.\n"); + } + HDassert( http_request->magic == S3COMMS_HRB_MAGIC ); + + if (canonical_request_dest == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "canonical request destination cannot be null.\n"); + } + + if (signed_headers_dest == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "signed headers destination cannot be null.\n"); + } + + /* HTTP verb, resource path, and query string lines + */ + len = (HDstrlen(http_request->verb) + + HDstrlen(http_request->resource) + + HDstrlen(query_params) + + 3 ); + ret = HDsnprintf(canonical_request_dest, + len + 1, + "%s\n%s\n%s\n", + http_request->verb, + http_request->resource, + query_params); + if (ret == 0 || (size_t)ret > len) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to compose canonical request first line"); + + /* write in canonical headers, building signed headers concurrently + */ + node = http_request->first_header; /* assumed at first sorted */ + while (node != NULL) { + size_t join_len = 0; /* string len of joined header-value */ + + HDassert( node->magic == S3COMMS_HRB_NODE_MAGIC ); + + len = HDstrlen(node->lowername); + join_len = HDstrlen(node->value) + len + 2; /* +2 <- ":\n" */ + ret = HDsnprintf(tmpstr, + join_len + 1, /* +1 for null terminator */ + "%s:%s\n", + node->lowername, + node->value); + if (ret == 0 || (size_t)ret > join_len) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to concatenate HTTP header %s:%s", + node->lowername, + node->value); + strcat(canonical_request_dest, tmpstr); + + len += 1; /* semicolon */ + ret = HDsnprintf(tmpstr, + len + 1, + "%s;", + node->lowername); + if (ret == 0 || (size_t)ret > len) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to append signed header %s", + node->lowername); + strcat(signed_headers_dest, tmpstr); + + node = node->next; + } + + /* remove tailing ';' from signed headers sequence + */ + signed_headers_dest[HDstrlen(signed_headers_dest) - 1] = '\0'; + + /* append signed headers and payload hash + * NOTE: at present, no HTTP body is handled, per the nature of + * requests/range-gets + */ + strcat(canonical_request_dest, "\n"); + strcat(canonical_request_dest, signed_headers_dest); + strcat(canonical_request_dest, "\n"); + strcat(canonical_request_dest, EMPTY_SHA256); + +done: + FUNC_LEAVE_NOAPI(ret_value); + +} /* H5FD_s3comms_aws_canonical_request */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_bytes_to_hex() + * + * Purpose: + * + * Produce human-readable hex string [0-9A-F] from sequence of bytes. + * + * For each byte (char), writes two-character hexadecimal representation. + * + * No null-terminator appended. + * + * Assumes `dest` is allocated to enough size (msg_len * 2). + * + * Fails if either `dest` or `msg` are null. + * + * `msg_len` message length of 0 has no effect. + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - hex string written to `dest` (not null-terminated) + * - FAILURE: `FAIL` + * - `dest == NULL` + * - `msg == NULL` + * + * Programmer: Jacob Smith + * 2017-07-12 + * + * Changes: + * + * - Integrate into HDF. + * - Rename from hex() to H5FD_s3comms_bytes_to_hex. + * - Change return type from `void` to `herr_t`. + * --- Jacob Smtih 2017-09-14 + * + * - Add bool parameter `lowercase` to configure upper/lowercase output + * of a-f hex characters. + * --- Jacob Smith 2017-09-19 + * + * - Change bool type to `hbool_t` + * --- Jacob Smtih 2017-10-11 + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_bytes_to_hex(char *dest, + const unsigned char *msg, + size_t msg_len, + hbool_t lowercase) +{ + size_t i = 0; + herr_t ret_value = SUCCEED; + + + + FUNC_ENTER_NOAPI_NOINIT + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_bytes_to_hex.\n"); +#endif + + if (dest == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "hex destination cannot be null.\n") + } + if (msg == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "bytes sequence cannot be null.\n") + } + + for (i = 0; i < msg_len; i++) { + int chars_written = + HDsnprintf(&(dest[i * 2]), + 3, /* 'X', 'X', '\n' */ + (lowercase == TRUE) ? "%02x" + : "%02X", + msg[i]); + if (chars_written != 2) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem while writing hex chars for %c", + msg[i]); + } + +done: + FUNC_LEAVE_NOAPI(ret_value); + +} /* H5FD_s3comms_bytes_to_hex */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_free_purl() + * + * Purpose: + * + * Release resources from a parsed_url_t pointer. + * + * If pointer is null, nothing happens. + * + * Return: + * + * `SUCCEED` (never fails) + * + * Programmer: Jacob Smith + * 2017-11-01 + * + * Changes: None. + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_free_purl(parsed_url_t *purl) +{ + FUNC_ENTER_NOAPI_NOINIT_NOERR + +#if S3COMMS_DEBUG + HDprintf("called H5FD_s3comms_free_purl.\n"); +#endif + + if (purl != NULL) { + HDassert( purl->magic == S3COMMS_PARSED_URL_MAGIC ); + if (purl->scheme != NULL) H5MM_xfree(purl->scheme); + if (purl->host != NULL) H5MM_xfree(purl->host); + if (purl->port != NULL) H5MM_xfree(purl->port); + if (purl->path != NULL) H5MM_xfree(purl->path); + if (purl->query != NULL) H5MM_xfree(purl->query); + purl->magic += 1ul; + H5MM_xfree(purl); + } + + FUNC_LEAVE_NOAPI(SUCCEED) +} /* H5FD_s3comms_free_purl */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_HMAC_SHA256() + * + * Purpose: + * + * Generate Hash-based Message Authentication Checksum using the SHA-256 + * hashing algorithm. + * + * Given a key, message, and respective lengths (to accommodate null + * characters in either), generate _hex string_ of authentication checksum + * and write to `dest`. + * + * `dest` must be at least `SHA256_DIGEST_LENGTH * 2` characters in size. + * Not enforceable by this function. + * `dest` will _not_ be null-terminated by this function. + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - hex string written to `dest` (not null-terminated) + * - FAILURE: `FAIL` + * - `dest == NULL` + * - error while generating hex string output + * + * Programmer: Jacob Smith + * 2017-07-?? + * + * Changes: + * + * - Integrate with HDF5. + * - Rename from `HMAC_SHA256` to `H5FD_s3comms_HMAC_SHA256`. + * - Rename output parameter from `md` to `dest`. + * - Return `herr_t` type instead of `void`. + * - Call `H5FD_s3comms_bytes_to_hex` to generate hex cleartext for output. + * --- Jacob Smith 2017-09-19 + * + * - Use static char array instead of malloc'ing `md` + * --- Jacob Smith 2017-10-10 + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_HMAC_SHA256(const unsigned char *key, + size_t key_len, + const char *msg, + size_t msg_len, + char *dest) +{ +#ifdef H5_HAVE_ROS3_VFD + unsigned char md[SHA256_DIGEST_LENGTH]; + unsigned int md_len = SHA256_DIGEST_LENGTH; + herr_t ret_value = SUCCEED; +#else + herr_t ret_value = FAIL; +#endif /* H5_HAVE_ROS3_VFD */ + + + + FUNC_ENTER_NOAPI_NOINIT + +#ifdef H5_HAVE_ROS3_VFD + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_HMAC_SHA256.\n"); +#endif + + if (dest == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "destination cannot be null."); + } + + HMAC(EVP_sha256(), + key, + (int)key_len, + (const unsigned char *)msg, + msg_len, + md, + &md_len); + + if (FAIL == + H5FD_s3comms_bytes_to_hex(dest, + (const unsigned char *)md, + (size_t)md_len, + true)) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "could not convert to hex string."); + } + +#endif /* H5_HAVE_ROS3_VFD */ + +done: + FUNC_LEAVE_NOAPI(ret_value); + +} /* H5FD_s3comms_HMAC_SHA256 */ + + +/*----------------------------------------------------------------------------- + * + * Function: H5FD__s3comms_load_aws_creds_from_file() + * + * Purpose: + * + * Extract AWS configuration information from a target file. + * + * Given a file and a profile name, e.g. "ros3_vfd_test", attempt to locate + * that region in the file. If not found, returns in error and output + * pointers are not modified. + * + * If the profile label is found, attempts to locate and parse configuration + * data, stopping at the first line where: + * + reached end of file + * + line does not start with a recognized setting name + * + * Following AWS documentation, looks for any of: + * + aws_access_key_id + * + aws_secret_access_key + * + region + * + * To be valid, the setting must begin the line with one of the keywords, + * followed immediately by an equals sign '=', and have some data before + * newline at end of line. + * + `spam=eggs` would be INVALID because name is unrecognized + * + `region = us-east-2` would be INVALID because of spaces + * + `region=` would be INVALID because no data. + * + * Upon successful parsing of a setting line, will store the result in the + * corresponding output pointer. If the output pointer is NULL, will skip + * any matching setting line while parsing -- useful to prevent overwrite + * when reading from multiple files. + * + * Return: + * + * + SUCCESS: `SUCCEED` + * + no error. settings may or may not have been loaded. + * + FAILURE: `FAIL` + * + internal error occurred. + * + -1 :: unable to format profile label + * + -2 :: profile name/label not found in file + * + -3 :: some other error + * + * Programmer: Jacob Smith + * 2018-02-27 + * + * Changes: None + * + *----------------------------------------------------------------------------- + */ +static herr_t +H5FD__s3comms_load_aws_creds_from_file( + FILE *file, + const char *profile_name, + char *key_id, + char *access_key, + char *aws_region) +{ + char profile_line[32]; + char buffer[128]; + const char *setting_names[] = { + "region", + "aws_access_key_id", + "aws_secret_access_key", + }; + char * const setting_pointers[] = { + aws_region, + key_id, + access_key, + }; + unsigned setting_count = 3; + herr_t ret_value = SUCCEED; + unsigned buffer_i = 0; + unsigned setting_i = 0; + int found_setting = 0; + char *line_buffer = &(buffer[0]); + + FUNC_ENTER_NOAPI_NOINIT + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called load_aws_creds_from_file.\n"); +#endif + + /* format target line for start of profile */ + if (32 < HDsnprintf(profile_line, 32, "[%s]", profile_name)) + HGOTO_ERROR(H5E_ARGS, H5E_CANTCOPY, FAIL, + "unable to format profile label") + + /* look for start of profile */ + do { + /* clear buffer */ + for (buffer_i=0; buffer_i < 128; buffer_i++) buffer[buffer_i] = 0; + + line_buffer = fgets(line_buffer, 128, file); + if (line_buffer == NULL) /* reached end of file */ + goto done; + } while (strncmp(line_buffer, profile_line, HDstrlen(profile_line))); + + /* extract credentials from lines */ + do { + size_t setting_name_len = 0; + const char *setting_name = NULL; + char line_prefix[128]; + + /* clear buffer */ + for (buffer_i=0; buffer_i < 128; buffer_i++) buffer[buffer_i] = 0; + + /* collect a line from file */ + line_buffer = fgets(line_buffer, 128, file); + if (line_buffer == NULL) + goto done; /* end of file */ + + /* loop over names to see if line looks like assignment */ + for (setting_i = 0; setting_i < setting_count; setting_i++) { + setting_name = setting_names[setting_i]; + setting_name_len = HDstrlen(setting_name); + if (128 < HDsnprintf( + line_prefix, + setting_name_len+2, + "%s=", + setting_name)) + HGOTO_ERROR(H5E_ARGS, H5E_CANTCOPY, FAIL, + "unable to format line prefix") + + /* found a matching name? */ + if (!strncmp(line_buffer, line_prefix, setting_name_len + 1)) { + found_setting = 1; + + /* skip NULL destination buffer */ + if (setting_pointers[setting_i] == NULL) + break; + + /* advance to end fo name in string */ + do { + line_buffer++; + } while (*line_buffer != 0 && *line_buffer != '='); + + if (*line_buffer == 0 || *(line_buffer+1) == 0) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "incomplete assignment in file") + line_buffer++; /* was pointing at '='; advance */ + + /* copy line buffer into out pointer */ + strcpy(setting_pointers[setting_i], (const char *)line_buffer); + + /* "trim" tailing whitespace by replacing with null terminator*/ + buffer_i = 0; + while (!isspace(setting_pointers[setting_i][buffer_i])) + buffer_i++; + setting_pointers[setting_i][buffer_i] = '\0'; + + break; /* have read setting; don't compare with others */ + } + } + } while (found_setting); + +done: + FUNC_LEAVE_NOAPI(ret_value); + +} /* H5FD__s3comms_load_aws_creds_from_file */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_load_aws_profile() + * + * Purpose : + * + * Read aws profile elements from standard location on system and store + * settings in memory. + * + * Looks for both `~/.aws/config` and `~/.aws/credentials`, the standard + * files for AWS tools. If a file exists (can be opened), looks for the + * given profile name and reads the settings into the relevant buffer. + * + * Any setting duplicated in both files will be set to that from + * `credentials`. + * + * Settings are stored in the supplied buffers as null-terminated strings. + * + * Return: + * + * + SUCCESS: `SUCCEED` (0) + * + no error occurred and all settings were populated + * + FAILURE: `FAIL` (-1) + * + internal error occurred + * + unable to locate profile + * + region, key id, and secret key were not all found and set + * + * Programmer: Jacob Smith + * 2018-02-27 + * + * Changes: None + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_load_aws_profile(const char *profile_name, + char *key_id_out, + char *secret_access_key_out, + char *aws_region_out) +{ + herr_t ret_value = SUCCEED; + FILE *credfile = NULL; + char awspath[117]; + char filepath[128]; + + FUNC_ENTER_NOAPI_NOINIT + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_load_aws_profile.\n"); +#endif + + /* TODO: Windows and other path gotchas */ + if (117 < HDsnprintf(awspath, 117, "%s/.aws/", getenv("HOME"))) + HGOTO_ERROR(H5E_ARGS, H5E_CANTCOPY, FAIL, + "unable to format home-aws path") + if (128 < HDsnprintf(filepath, 128, "%s%s", awspath, "credentials")) + HGOTO_ERROR(H5E_ARGS, H5E_CANTCOPY, FAIL, + "unable to format credentials path") + + credfile = fopen(filepath, "r"); + if (credfile != NULL) { + if (FAIL == H5FD__s3comms_load_aws_creds_from_file( + credfile, + profile_name, + key_id_out, + secret_access_key_out, + aws_region_out)) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to load from aws credentials") + if (EOF == fclose(credfile)) + HGOTO_ERROR(H5E_FILE, H5E_CANTCLOSEFILE, FAIL, + "unable to close credentials file") + credfile = NULL; + } + + if (128 < HDsnprintf(filepath, 128, "%s%s", awspath, "config")) + HGOTO_ERROR(H5E_ARGS, H5E_CANTCOPY, FAIL, + "unable to format config path") + credfile = fopen(filepath, "r"); + if (credfile != NULL) { + if (FAIL == H5FD__s3comms_load_aws_creds_from_file( + credfile, + profile_name, + (*key_id_out == 0) ? key_id_out : NULL, + (*secret_access_key_out == 0) ? secret_access_key_out : NULL, + (*aws_region_out == 0) ? aws_region_out : NULL)) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to load from aws config") + if (EOF == fclose(credfile)) + HGOTO_ERROR(H5E_FILE, H5E_CANTCLOSEFILE, FAIL, + "unable to close config file") + credfile = NULL; + } + + /* fail if not all three settings were loaded */ + if (*key_id_out == 0 || + *secret_access_key_out == 0 || + *aws_region_out == 0) + { + ret_value = FAIL; + } + +done: + if (credfile != NULL) { + if (EOF == fclose(credfile)) + HDONE_ERROR(H5E_ARGS, H5E_ARGS, FAIL, + "problem error-closing aws configuration file") + } + + FUNC_LEAVE_NOAPI(ret_value); + +} /* H5FD_s3comms_load_aws_profile */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_nlowercase() + * + * Purpose: + * + * From string starting at `s`, write `len` characters to `dest`, + * converting all to lowercase. + * + * Behavior is undefined if `s` is NULL or `len` overruns the allocated + * space of either `s` or `dest`. + * + * Provided as convenience. + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - upon completion, `dest` is populated + * - FAILURE: `FAIL` + * - `dest == NULL` + * + * Programmer: Jacob Smith + * 2017-09-18 + * + * Changes: None. + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_nlowercase(char *dest, + const char *s, + size_t len) +{ + herr_t ret_value = SUCCEED; + + FUNC_ENTER_NOAPI_NOINIT + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_nlowercase.\n"); +#endif + + if (dest == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "destination cannot be null.\n"); + } + + if (len > 0) { + HDmemcpy(dest, s, len); + do { + len--; + dest[len] = (char)tolower( (int)dest[len] ); + } while (len > 0); + } + +done: + FUNC_LEAVE_NOAPI(ret_value); + +} /* H5FD_s3comms_nlowercase */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_parse_url() + * + * Purpose: + * + * Parse URL-like string and stuff URL components into + * `parsed_url` structure, if possible. + * + * Expects null-terminated string of format: + * SCHEME "://" HOST [":" PORT ] ["/" [ PATH ] ] ["?" QUERY] + * where SCHEME :: "[a-zA-Z/.-]+" + * PORT :: "[0-9]" + * + * Stores resulting structure in argument pointer `purl`, if successful, + * creating and populating new `parsed_url_t` structure pointer. + * Empty or absent elements are NULL in new purl structure. + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - `purl` pointer is populated + * - FAILURE: `FAIL` + * - unable to parse + * - `purl` is unaltered (probably NULL) + * + * Programmer: Jacob Smith + * 2017-10-30 + * + * Changes: None. + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_parse_url(const char *str, + parsed_url_t **_purl) +{ + parsed_url_t *purl = NULL; /* pointer to new structure */ + const char *tmpstr = NULL; /* working pointer in string */ + const char *curstr = str; /* "start" pointer in string */ + long int len = 0; /* substring length */ + long int urllen = 0; /* length of passed-in url string */ + unsigned int i = 0; + herr_t ret_value = FAIL; + + + + FUNC_ENTER_NOAPI_NOINIT; + +#if S3COMMS_DEBUG + HDprintf("called H5FD_s3comms_parse_url.\n"); +#endif + + if (str == NULL || *str == '\0') { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "invalid url string"); + } + + urllen = (long int)HDstrlen(str); + + purl = (parsed_url_t *)H5MM_malloc(sizeof(parsed_url_t)); + if (purl == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, FAIL, + "can't allocate space for parsed_url_t"); + } + purl->magic = S3COMMS_PARSED_URL_MAGIC; + purl->scheme = NULL; + purl->host = NULL; + purl->port = NULL; + purl->path = NULL; + purl->query = NULL; + + /*************** + * READ SCHEME * + ***************/ + + tmpstr = strchr(curstr, ':'); + if (tmpstr == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "invalid SCHEME construction: probably not URL"); + } + len = tmpstr - curstr; + HDassert( (0 <= len) && (len < urllen) ); + + /* check for restrictions + */ + for (i = 0; i < len; i++) { + /* scheme = [a-zA-Z+-.]+ (terminated by ":") */ + if (!isalpha(curstr[i]) && + '+' != curstr[i] && + '-' != curstr[i] && + '.' != curstr[i]) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "invalid SCHEME construction"); + } + } + /* copy lowercased scheme to structure + */ + purl->scheme = (char *)H5MM_malloc(sizeof(char) * (size_t)(len + 1)); + if (purl->scheme == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, FAIL, + "can't allocate space for SCHEME"); + } + (void)HDstrncpy(purl->scheme, curstr, (size_t)len); + purl->scheme[len] = '\0'; + for ( i = 0; i < len; i++ ) { + purl->scheme[i] = (char)tolower(purl->scheme[i]); + } + + /* Skip "://" */ + tmpstr += 3; + curstr = tmpstr; + + /************* + * READ HOST * + *************/ + + if (*curstr == '[') { + /* IPv6 */ + while (']' != *tmpstr) { + if (tmpstr == 0) { /* end of string reached! */ + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "reached end of URL: incomplete IPv6 HOST"); + } + tmpstr++; + } + tmpstr++; + } else { + while (0 != *tmpstr) { + if (':' == *tmpstr || + '/' == *tmpstr || + '?' == *tmpstr) + { + break; + } + tmpstr++; + } + } /* if IPv4 or IPv6 */ + len = tmpstr - curstr; + if (len == 0) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "HOST substring cannot be empty"); + } else if (len > urllen) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem with length of HOST substring"); + } + + /* copy host + */ + purl->host = (char *)H5MM_malloc(sizeof(char) * (size_t)(len + 1)); + if (purl->host == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, FAIL, + "can't allocate space for HOST"); + } + (void)HDstrncpy(purl->host, curstr, (size_t)len); + purl->host[len] = 0; + + /************* + * READ PORT * + *************/ + + if (':' == *tmpstr) { + tmpstr += 1; /* advance past ':' */ + curstr = tmpstr; + while ((0 != *tmpstr) && ('/' != *tmpstr) && ('?' != *tmpstr)) { + tmpstr++; + } + len = tmpstr - curstr; + if (len == 0) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "PORT element cannot be empty"); + } else if (len > urllen) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem with length of PORT substring"); + } + for (i = 0; i < len; i ++) { + if (!isdigit(curstr[i])) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "PORT is not a decimal string"); + } + } + + /* copy port + */ + purl->port = (char *)H5MM_malloc(sizeof(char) * (size_t)(len + 1)); + if (purl->port == NULL) { /* cannot malloc */ + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, FAIL, + "can't allocate space for PORT"); + } + (void)HDstrncpy(purl->port, curstr, (size_t)len); + purl->port[len] = 0; + } /* if PORT element */ + + /************* + * READ PATH * + *************/ + + if ('/' == *tmpstr) { + /* advance past '/' */ + tmpstr += 1; + curstr = tmpstr; + + /* seek end of PATH + */ + while ((0 != *tmpstr) && ('?' != *tmpstr)) { + tmpstr++; + } + len = tmpstr - curstr; + if (len > urllen) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem with length of PATH substring"); + } + if (len > 0) { + purl->path = (char *)H5MM_malloc(sizeof(char) * (size_t)(len + 1)); + if (purl->path == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, FAIL, + "can't allocate space for PATH"); + } /* cannot malloc path pointer */ + (void)HDstrncpy(purl->path, curstr, (size_t)len); + purl->path[len] = 0; + } + } /* if PATH element */ + + /************** + * READ QUERY * + **************/ + + if ('?' == *tmpstr) { + tmpstr += 1; + curstr = tmpstr; + while (0 != *tmpstr) { + tmpstr++; + } + len = tmpstr - curstr; + if (len == 0) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "QUERY cannot be empty"); + } else if (len > urllen) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem with length of QUERY substring"); + } + purl->query = (char *)H5MM_malloc(sizeof(char) * (size_t)(len + 1)); + if (purl->query == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_CANTALLOC, FAIL, + "can't allocate space for QUERY"); + } /* cannot malloc path pointer */ + (void)HDstrncpy(purl->query, curstr, (size_t)len); + purl->query[len] = 0; + } /* if QUERY exists */ + + + + *_purl = purl; + ret_value = SUCCEED; + +done: + if (ret_value == FAIL) { + H5FD_s3comms_free_purl(purl); + } + FUNC_LEAVE_NOAPI(ret_value); + +} /* H5FD_s3comms_parse_url */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_percent_encode_char() + * + * Purpose: + * + * "Percent-encode" utf-8 character `c`, e.g., + * '$' -> "%24" + * '¢' -> "%C2%A2" + * + * `c` cannot be null. + * + * Does not (currently) accept multi-byte characters... + * limit to (?) u+00ff, well below upper bound for two-byte utf-8 encoding + * (u+0080..u+07ff). + * + * Writes output to `repr`. + * `repr` cannot be null. + * Assumes adequate space i `repr`... + * >>> char[4] or [7] for most characters, + * >>> [13] as theoretical maximum. + * + * Representation `repr` is null-terminated. + * + * Stores length of representation (without null terminator) at pointer + * `repr_len`. + * + * Return : SUCCEED/FAIL + * + * - SUCCESS: `SUCCEED` + * - percent-encoded representation written to `repr` + * - 'repr' is null-terminated + * - FAILURE: `FAIL` + * - `c` or `repr` was NULL + * + * Programmer: Jacob Smith + * + * Changes: + * + * - Integrate into HDF. + * - Rename from `hexutf8` to `H5FD_s3comms_percent_encode_char`. + * --- Jacob Smith 2017-09-15 + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_percent_encode_char(char *repr, + const unsigned char c, + size_t *repr_len) +{ + unsigned int acc = 0; + unsigned int i = 0; + unsigned int k = 0; + unsigned int stack[4] = {0, 0, 0, 0}; + unsigned int stack_size = 0; + int chars_written = 0; + herr_t ret_value = SUCCEED; +#if S3COMMS_DEBUG + unsigned char s[2] = {c, 0}; + unsigned char hex[3] = {0, 0, 0}; +#endif + + + + FUNC_ENTER_NOAPI_NOINIT + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_percent_encode_char.\n"); +#endif + + if (repr == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, "no destination `repr`.\n") + } + +#if S3COMMS_DEBUG + H5FD_s3comms_bytes_to_hex((char *)hex, s, 1, FALSE); + HDfprintf(stdout, " CHAR: \'%s\'\n", s); + HDfprintf(stdout, " CHAR-HEX: \"%s\"\n", hex); +#endif + + if (c <= (unsigned char)0x7f) { + /* character represented in a single "byte" + * and single percent-code + */ +#if S3COMMS_DEBUG + HDfprintf(stdout, " SINGLE-BYTE\n"); +#endif + *repr_len = 3; + chars_written = HDsnprintf(repr, 4, "%%%02X", c); + if (chars_written != 3) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "cannot write char %c", + c); + } else { + /* multi-byte, multi-percent representation + */ +#if S3COMMS_DEBUG + HDfprintf(stdout, " MULTI-BYTE\n"); +#endif + stack_size = 0; + k = (unsigned int)c; + *repr_len = 0; + do { + /* push number onto stack in six-bit slices + */ + acc = k; + acc >>= 6; /* cull least */ + acc <<= 6; /* six bits */ + stack[stack_size++] = k - acc; /* max six-bit number */ + k = acc >> 6; + } while (k > 0); + + /* now have "stack" of two to four six-bit numbers + * to be put into UTF-8 byte fields + */ + +#if S3COMMS_DEBUG + HDfprintf(stdout, " STACK:\n {\n"); + for (i = 0; i < stack_size; i++) { + H5FD_s3comms_bytes_to_hex((char *)hex, + (unsigned char *)(&stack[i]), + 1, + FALSE); + hex[2] = 0; + HDfprintf(stdout, " %s,\n", hex); + } + HDfprintf(stdout, " }\n"); +#endif + + /**************** + * leading byte * + ****************/ + + /* prepend 11[1[1]]0 to first byte */ + /* 110xxxxx, 1110xxxx, or 11110xxx */ + acc = 0xC0; /* 2^7 + 2^6 -> 11000000 */ + acc += (stack_size > 2) ? 0x20 : 0; + acc += (stack_size > 3) ? 0x10 : 0; + stack_size -= 1; + chars_written = HDsnprintf(repr, 4, "%%%02X", acc + stack[stack_size]); + if (chars_written != 3) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "cannot write char %c", + c); + *repr_len += 3; + + /************************ + * continuation byte(s) * + ************************/ + + /* 10xxxxxx */ + for (i = 0; i < stack_size; i++) { + chars_written = HDsnprintf(&repr[i*3 + 3], + 4, + "%%%02X", + 128 + stack[stack_size - 1 - i]); + if (chars_written != 3) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "cannot write char %c", + c); + *repr_len += 3; + } + } + *(repr + *repr_len) = '\0'; + +done: + FUNC_LEAVE_NOAPI(ret_value); + +} /* H5FD_s3comms_percent_encode_char */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_signing_key() + * + * Purpose: + * + * Create AWS4 "Signing Key" from secret key, AWS region, and timestamp. + * + * Sequentially runs HMAC_SHA256 on strings in specified order, + * generating re-usable checksum (according to documentation, valid for + * 7 days from time given). + * + * `secret` is `access key id` for targeted service/bucket/resource. + * + * `iso8601now` must conform to format, yyyyMMDD'T'hhmmss'Z' + * e.g. "19690720T201740Z". + * + * `region` should be one of AWS service region names, e.g. "us-east-1". + * + * Hard-coded "service" algorithm requirement to "s3". + * + * Inputs must be null-terminated strings. + * + * Writes to `md` the raw byte data, length of `SHA256_DIGEST_LENGTH`. + * Programmer must ensure that `md` is appropriately allocated. + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - raw byte data of signing key written to `md` + * - FAILURE: `FAIL` + * - if any input arguments was NULL + * + * Programmer: Jacob Smith + * 2017-07-13 + * + * Changes: + * + * - Integrate into HDF5. + * - Return herr_t type. + * --- Jacob Smith 2017-09-18 + * + * - NULL check and fail of input parameters. + * --- Jacob Smith 2017-10-10 + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_signing_key(unsigned char *md, + const char *secret, + const char *region, + const char *iso8601now) +{ +#ifdef H5_HAVE_ROS3_VFD + char *AWS4_secret = NULL; + size_t AWS4_secret_len = 0; + unsigned char datekey[SHA256_DIGEST_LENGTH]; + unsigned char dateregionkey[SHA256_DIGEST_LENGTH]; + unsigned char dateregionservicekey[SHA256_DIGEST_LENGTH]; + int ret = 0; /* return value of HDsnprintf */ + herr_t ret_value = SUCCEED; +#else + herr_t ret_value = SUCCEED; +#endif /* H5_HAVE_ROS3_VFD */ + + + + FUNC_ENTER_NOAPI_NOINIT + +#ifdef H5_HAVE_ROS3_VFD + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_signing_key.\n"); +#endif + + if (md == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "Destination `md` cannot be NULL.\n") + } + if (secret == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "`secret` cannot be NULL.\n") + } + if (region == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "`region` cannot be NULL.\n") + } + if (iso8601now == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "`iso8601now` cannot be NULL.\n") + } + + AWS4_secret_len = 4 + HDstrlen(secret) + 1; + AWS4_secret = (char*)H5MM_malloc(sizeof(char *) * AWS4_secret_len); + if (AWS4_secret == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "Could not allocate space.\n") + } + + /* prepend "AWS4" to start of the secret key + */ + ret = HDsnprintf(AWS4_secret, AWS4_secret_len,"%s%s", "AWS4", secret); + if ((size_t)ret != (AWS4_secret_len - 1)) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem writing AWS4+secret `%s`", + secret); + + /* hash_func, key, len(key), msg, len(msg), digest_dest, digest_len_dest + * we know digest length, so ignore via NULL + */ + HMAC(EVP_sha256(), + (const unsigned char *)AWS4_secret, + (int)HDstrlen(AWS4_secret), + (const unsigned char*)iso8601now, + 8, /* 8 --> length of 8 --> "yyyyMMDD" */ + datekey, + NULL); + HMAC(EVP_sha256(), + (const unsigned char *)datekey, + SHA256_DIGEST_LENGTH, + (const unsigned char *)region, + HDstrlen(region), + dateregionkey, + NULL); + HMAC(EVP_sha256(), + (const unsigned char *)dateregionkey, + SHA256_DIGEST_LENGTH, + (const unsigned char *)"s3", + 2, + dateregionservicekey, + NULL); + HMAC(EVP_sha256(), + (const unsigned char *)dateregionservicekey, + SHA256_DIGEST_LENGTH, + (const unsigned char *)"aws4_request", + 12, + md, + NULL); + +done: + H5MM_xfree(AWS4_secret); + +#endif /* H5_HAVE_ROS3_VFD */ + + FUNC_LEAVE_NOAPI(ret_value); + +} /* H5FD_s3comms_signing_key */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_tostringtosign() + * + * Purpose: + * + * Get AWS "String to Sign" from Canonical Request, timestamp, + * and AWS "region". + * + * Common between single request and "chunked upload", + * conforms to: + * "AWS4-HMAC-SHA256\n" + + * <ISO8601 date format> + "\n" + // yyyyMMDD'T'hhmmss'Z' + * <yyyyMMDD> + "/" + <AWS Region> + "/s3/aws4-request\n" + + * hex(SHA256(<CANONICAL-REQUEST>)) + * + * Inputs `creq` (canonical request string), `now` (ISO8601 format), + * and `region` (s3 region designator string) must all be + * null-terminated strings. + * + * Result is written to `dest` with null-terminator. + * It is left to programmer to ensure `dest` has adequate space. + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - "string to sign" written to `dest` and null-terminated + * - FAILURE: `FAIL` + * - if any of the inputs are NULL + * - if an error is encountered while computing checksum + * + * Programmer: Jacob Smith + * 2017-07-?? + * + * Changes: + * + * - Integrate with HDF5. + * - Rename from `tostringtosign` to `H5FD_s3comms_tostringtosign`. + * - Return `herr_t` instead of characters written. + * - Use HDF-friendly bytes-to-hex function (`H5FD_s3comms_bytes_to_hex`) + * instead of general-purpose, deprecated `hex()`. + * - Adjust casts to openssl's `SHA256`. + * - Input strings are now `const`. + * --- Jacob Smith 2017-09-19 + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_tostringtosign(char *dest, + const char *req, + const char *now, + const char *region) +{ +#ifdef H5_HAVE_ROS3_VFD + unsigned char checksum[SHA256_DIGEST_LENGTH * 2 + 1]; + size_t d = 0; + char day[9]; + char hexsum[SHA256_DIGEST_LENGTH * 2 + 1]; + size_t i = 0; + int ret = 0; /* HDsnprintf return value */ + herr_t ret_value = SUCCEED; + char tmp[128]; +#else + herr_t ret_value = FAIL; +#endif /* H5_HAVE_ROS3_VFD */ + + + FUNC_ENTER_NOAPI_NOINIT + +#ifdef H5_HAVE_ROS3_VFD + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_tostringtosign.\n"); +#endif + + if (dest == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "destination buffer cannot be null.\n") + } + if (req == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "canonical request cannot be null.\n") + } + if (now == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "Timestring cannot be NULL.\n") + } + if (region == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "Region cannot be NULL.\n") + } + + + + for (i = 0; i < 128; i++) { + tmp[i] = '\0'; + } + for (i = 0; i < SHA256_DIGEST_LENGTH * 2 + 1; i++) { + checksum[i] = '\0'; + hexsum[i] = '\0'; + } + HDstrncpy(day, now, 8); + day[8] = '\0'; + ret = HDsnprintf(tmp, 127, "%s/%s/s3/aws4_request", day, region); + if (ret <= 0 || ret >= 127) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "problem adding day and region to string") + + + + HDmemcpy((dest + d), "AWS4-HMAC-SHA256\n", 17); + d = 17; + + HDmemcpy((dest+d), now, HDstrlen(now)); + d += HDstrlen(now); + dest[d++] = '\n'; + + HDmemcpy((dest + d), tmp, HDstrlen(tmp)); + d += HDstrlen(tmp); + dest[d++] = '\n'; + + SHA256((const unsigned char *)req, + HDstrlen(req), + checksum); + + if (FAIL == + H5FD_s3comms_bytes_to_hex(hexsum, + (const unsigned char *)checksum, + SHA256_DIGEST_LENGTH, + true)) + { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "could not create hex string"); + } + + for (i = 0; i < SHA256_DIGEST_LENGTH * 2; i++) { + dest[d++] = hexsum[i]; + } + + dest[d] = '\0'; + +#endif /* H5_HAVE_ROS3_VFD */ + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5ros3_tostringtosign */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_trim() + * + * Purpose: + * + * Remove all whitespace characters from start and end of a string `s` + * of length `s_len`, writing trimmed string copy to `dest`. + * Stores number of characters remaining at `n_written`. + * + * Destination for trimmed copy `dest` cannot be null. + * `dest` must have adequate space allocated for trimmed copy. + * If inadequate space, behavior is undefined, possibly resulting + * in segfault or overwrite of other data. + * + * If `s` is NULL or all whitespace, `dest` is untouched and `n_written` + * is set to 0. + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - FAILURE: `FAIL` + * - `dest == NULL` + * + * Programmer: Jacob Smith + * 2017-09-18 + * + * Changes: + * + * - Rename from `trim()` to `H5FD_s3comms_trim()`. + * - Incorporate into HDF5. + * - Returns `herr_t` type. + * --- Jacob Smith 2017-??-?? + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_trim(char *dest, + char *s, + size_t s_len, + size_t *n_written) +{ + herr_t ret_value = SUCCEED; + + + + FUNC_ENTER_NOAPI_NOINIT + +#if S3COMMS_DEBUG + HDfprintf(stdout, "called H5FD_s3comms_trim.\n"); +#endif + + if (dest == NULL) { + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "destination cannot be null.") + } + if (s == NULL) { + s_len = 0; + } + + + + if (s_len > 0) { + /* Find first non-whitespace character from start; + * reduce total length per character. + */ + while ((s_len > 0) && + isspace((unsigned char)s[0]) && s_len > 0) + { + s++; + s_len--; + } + + /* Find first non-whitespace character from tail; + * reduce length per-character. + * If length is 0 already, there is no non-whitespace character. + */ + if (s_len > 0) { + do { + s_len--; + } while( isspace((unsigned char)s[s_len]) ); + s_len++; + + /* write output into dest + */ + HDmemcpy(dest, s, s_len); + } + } + + *n_written = s_len; + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_s3comms_trim */ + + +/*---------------------------------------------------------------------------- + * + * Function: H5FD_s3comms_uriencode() + * + * Purpose: + * + * URIencode (percent-encode) every byte except "[a-zA-Z0-9]-._~". + * + * For each character in source string `_s` from `s[0]` to `s[s_len-1]`, + * writes to `dest` either the raw character or its percent-encoded + * equivalent. + * + * See `H5FD_s3comms_bytes_to_hex` for information on percent-encoding. + * + * Space (' ') character encoded as "%20" (not "+") + * + * Forward-slash ('/') encoded as "%2F" only when `encode_slash == true`. + * + * Records number of characters written at `n_written`. + * + * Assumes that `dest` has been allocated with enough space. + * + * Neither `dest` nor `s` can be NULL. + * + * `s_len == 0` will have no effect. + * + * Return: + * + * - SUCCESS: `SUCCEED` + * - FAILURE: `FAIL` + * - source strings `s` or destination `dest` are NULL + * - error while attempting to percent-encode a character + * + * Programmer: Jacob Smith + * 2017-07-?? + * + * Changes: + * + * - Integrate to HDF environment. + * - Rename from `uriencode` to `H5FD_s3comms_uriencode`. + * - Change return from characters written to herr_t; + * move to i/o parameter `n_written`. + * - No longer append null-terminator to string; + * programmer may append or not as appropriate upon return. + * --- Jacob Smith 2017-09-15 + * + *---------------------------------------------------------------------------- + */ +herr_t +H5FD_s3comms_uriencode(char *dest, + const char *s, + size_t s_len, + hbool_t encode_slash, + size_t *n_written) +{ + char c = 0; + size_t dest_off = 0; + char hex_buffer[13]; + size_t hex_off = 0; + size_t hex_len = 0; + herr_t ret_value = SUCCEED; + size_t s_off = 0; + + + + FUNC_ENTER_NOAPI_NOINIT + +#if S3COMMS_DEBUG + HDfprintf(stdout, "H5FD_s3comms_uriencode called.\n"); +#endif + + if (s == NULL) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "source string cannot be NULL"); + if (dest == NULL) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "destination cannot be NULL"); + + /* Write characters to destination, converting to percent-encoded + * "hex-utf-8" strings if necessary. + * e.g., '$' -> "%24" + */ + for (s_off = 0; s_off < s_len; s_off++) { + c = s[s_off]; + if (isalnum(c) || + c == '.' || + c == '-' || + c == '_' || + c == '~' || + (c == '/' && encode_slash == FALSE)) + { + dest[dest_off++] = c; + } else { + hex_off = 0; + if (FAIL == + H5FD_s3comms_percent_encode_char(hex_buffer, + (const unsigned char)c, + &hex_len)) + { + hex_buffer[0] = c; + hex_buffer[1] = 0; + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "unable to percent-encode character \'%s\' " + "at %d in \"%s\"", hex_buffer, (int)s_off, s); + } + + for (hex_off = 0; hex_off < hex_len; hex_off++) { + dest[dest_off++] = hex_buffer[hex_off]; + } + } + } + + if (dest_off < s_len) + HGOTO_ERROR(H5E_ARGS, H5E_BADVALUE, FAIL, + "buffer overflow"); + + *n_written = dest_off; + +done: + FUNC_LEAVE_NOAPI(ret_value) + +} /* H5FD_s3comms_uriencode */ + + diff --git a/src/H5FDs3comms.h b/src/H5FDs3comms.h new file mode 100644 index 0000000..0524c46 --- /dev/null +++ b/src/H5FDs3comms.h @@ -0,0 +1,634 @@ +/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * Read-Only S3 Virtual File Driver (VFD) * + * Copyright (c) 2017-2018, The HDF Group. * + * * + * All rights reserved. * + * * + * NOTICE: * + * All information contained herein is, and remains, the property of The HDF * + * Group. The intellectual and technical concepts contained herein are * + * proprietary to The HDF Group. Dissemination of this information or * + * reproduction of this material is strictly forbidden unless prior written * + * permission is obtained from The HDF Group. * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ + +/***************************************************************************** + * + * This is the header for the S3 Communications module + * + * ***NOT A FILE DRIVER*** + * + * Purpose: + * + * - Provide structures and functions related to communicating with + * Amazon S3 (Simple Storage Service). + * - Abstract away the REST API (HTTP, + * networked communications) behind a series of uniform function calls. + * - Handle AWS4 authentication, if appropriate. + * - Fail predictably in event of errors. + * - Eventually, support more S3 operations, such as creating, writing to, + * and removing Objects remotely. + * + * translates: + * `read(some_file, bytes_offset, bytes_length, &dest_buffer);` + * to: + * ``` + * GET myfile HTTP/1.1 + * Host: somewhere.me + * Range: bytes=4096-5115 + * ``` + * and places received bytes from HTTP response... + * ``` + * HTTP/1.1 206 Partial-Content + * Content-Range: 4096-5115/63239 + * + * <bytes> + * ``` + * ...in destination buffer. + * + * TODO: put documentation in a consistent place and point to it from here. + * + * Programmer: Jacob Smith + * 2017-11-30 + * + *****************************************************************************/ + +#include <ctype.h> +#include <stddef.h> +#include <stdlib.h> +#include <string.h> +#include <time.h> + +#ifdef H5_HAVE_ROS3_VFD +#include <curl/curl.h> +#include <openssl/evp.h> +#include <openssl/hmac.h> +#include <openssl/sha.h> +#endif /* ifdef H5_HAVE_ROS3_VFD */ + +/***************** + * PUBLIC MACROS * + *****************/ + +/* hexadecimal string of pre-computed sha256 checksum of the empty string + * hex(sha256sum("")) + */ +#define EMPTY_SHA256 \ +"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855" + +/* string length (plus null terminator) + * example ISO8601-format string: "20170713T145903Z" (YYYYmmdd'T'HHMMSS'_') + */ +#define ISO8601_SIZE 17 + +/* string length (plus null terminator) + * example RFC7231-format string: "Fri, 30 Jun 2017 20:41:55 GMT" + */ +#define RFC7231_SIZE 30 + +/*--------------------------------------------------------------------------- + * + * Macro: ISO8601NOW() + * + * Purpose: + * + * write "YYYYmmdd'T'HHMMSS'Z'" (less single-quotes) to dest + * e.g., "20170630T204155Z" + * + * wrapper for strftime() + * + * It is left to the programmer to check return value of + * ISO8601NOW (should equal ISO8601_SIZE - 1). + * + * Programmer: Jacob Smith + * 2017-07-?? + * + *--------------------------------------------------------------------------- + */ +#define ISO8601NOW(dest, now_gm) \ +strftime((dest), ISO8601_SIZE, "%Y%m%dT%H%M%SZ", (now_gm)) + +/*--------------------------------------------------------------------------- + * + * Macro: RFC7231NOW() + * + * Purpose: + * + * write "Day, dd Mmm YYYY HH:MM:SS GMT" to dest + * e.g., "Fri, 30 Jun 2017 20:41:55 GMT" + * + * wrapper for strftime() + * + * It is left to the programmer to check return value of + * RFC7231NOW (should equal RFC7231_SIZE - 1). + * + * Programmer: Jacob Smith + * 2017-07-?? + * + *--------------------------------------------------------------------------- + */ +#define RFC7231NOW(dest, now_gm) \ +strftime((dest), RFC7231_SIZE, "%a, %d %b %Y %H:%M:%S GMT", (now_gm)) + + +/* Reasonable maximum length of a credential string. + * Provided for error-checking S3COMMS_FORMAT_CREDENTIAL (below). + * 17 <- "////aws4_request\0" + * 2 < "s3" (service) + * 8 <- "YYYYmmdd" (date) + * 128 <- (access_id) + * 155 :: sum + */ +#define S3COMMS_MAX_CREDENTIAL_SIZE 155 + + +/*--------------------------------------------------------------------------- + * + * Macro: H5FD_S3COMMS_FORMAT_CREDENTIAL() + * + * Purpose: + * + * Format "S3 Credential" string from inputs, for AWS4. + * + * Wrapper for HDsnprintf(). + * + * _HAS NO ERROR-CHECKING FACILITIES_ + * It is left to programmer to ensure that return value confers success. + * e.g., + * ``` + * assert( S3COMMS_MAX_CREDENTIAL_SIZE >= + * S3COMMS_FORMAT_CREDENTIAL(...) ); + * ``` + * + * "<access-id>/<date>/<aws-region>/<aws-service>/aws4_request" + * assuming that `dest` has adequate space. + * + * ALL inputs must be null-terminated strings. + * + * `access` should be the user's access key ID. + * `date` must be of format "YYYYmmdd". + * `region` should be relevant AWS region, i.e. "us-east-1". + * `service` should be "s3". + * + * Programmer: Jacob Smith + * 2017-09-19 + * + * Changes: None. + * + *--------------------------------------------------------------------------- + */ +#define S3COMMS_FORMAT_CREDENTIAL(dest, access, iso8601_date, region, service) \ +HDsnprintf((dest), S3COMMS_MAX_CREDENTIAL_SIZE, \ + "%s/%s/%s/%s/aws4_request", \ + (access), (iso8601_date), (region), (service)) + +/********************* + * PUBLIC STRUCTURES * + *********************/ + + +/*---------------------------------------------------------------------------- + * + * Structure: hrb_node_t + * + * HTTP Header Field Node + * + * + * + * Maintain a ordered (linked) list of HTTP Header fields. + * + * Provides efficient access and manipulation of a logical sequence of + * HTTP header fields, of particular use when composing an + * "S3 Canonical Request" for authentication. + * + * - The creation of a Canoncial Request involves: + * - convert field names to lower case + * - sort by this lower-case name + * - convert ": " name-value separator in HTTP string to ":" + * - get sorted lowercase names without field or separator + * + * As HTTP headers allow headers in any order (excepting the case of multiple + * headers with the same name), the list ordering can be optimized for Canonical + * Request creation, suggesting alphabtical order. For more expedient insertion + * and removal of elements in the list, linked list seems preferable to a + * dynamically-expanding array. The usually-smaller number of entries (5 or + * fewer) makes performance overhead of traversing the list trivial. + * + * The above requirements of creating at Canonical Request suggests a reasonable + * trade-off of speed for space with the option to compute elements as needed + * or to have the various elements prepared and stored in the structure + * (e.g. name, value, lowername, concatenated name:value) + * The structure currently is implemented to pre-compute. + * + * At all times, the "first" node of the list should be the least, + * alphabetically. For all nodes, the `next` node should be either NULL or + * of greater alphabetical value. + * + * Each node contains its own header field information, plus a pointer to the + * next node. + * + * It is not allowed to have multiple nodes with the same _lowercase_ `name`s + * in the same list + * (i.e., name is case-insensitive for access and modification.) + * + * All data (`name`, `value`, `lowername`, and `cat`) are null-terminated + * strings allocated specifically for their node. + * + * + * + * `magic` (unsigned long) + * + * "unique" idenfier number for the structure type + * + * `name` (char *) + * + * Case-meaningful name of the HTTP field. + * Given case is how it is supplied to networking code. + * e.g., "Range" + * + * `lowername` (char *) + * + * Lowercase copy of name. + * e.g., "range" + * + * `value` (char *) + * + * Case-meaningful value of HTTP field. + * e.g., "bytes=0-9" + * + * `cat` (char *) + * + * Concatenated, null-terminated string of HTTP header line, + * as the field would appear in an HTTP request. + * e.g., "Range: bytes=0-9" + * + * `next` (hrb_node_t *) + * + * Pointers to next node in the list, or NULL sentinel as end of list. + * Next node must have a greater `lowername` as determined by strcmp(). + * + * + * + * Programmer: Jacob Smith + * 2017-09-22 + * + * Changes: + * + * - Change from twin doubly-linked lists to singly-linked list. + * --- Jake Smith 2017-01-17 + * + *---------------------------------------------------------------------------- + */ +typedef struct hrb_node_t { + unsigned long magic; + char *name; + char *value; + char *cat; + char *lowername; + struct hrb_node_t *next; +} hrb_node_t; +#define S3COMMS_HRB_NODE_MAGIC 0x7F5757UL + + +/*---------------------------------------------------------------------------- + * + * Structure: hrb_t + * + * HTTP Request Buffer structure + * + * + * + * Logically represent an HTTP request + * + * GET /myplace/myfile.h5 HTTP/1.1 + * Host: over.rainbow.oz + * Date: Fri, 01 Dec 2017 12:35:04 CST + * + * <body> + * + * ...with fast, efficient access to and modification of primary and field + * elements. + * + * Structure for building HTTP requests while hiding much of the string + * processing required "under the hood." + * + * Information about the request target -- the first line -- and the body text, + * if any, are managed directly with this structure. All header fields, e.g., + * "Host" and "Date" above, are created with a linked list of `hrb_node_t` and + * included in the request by a pointer to the head of the list. + * + * + * + * `magic` (unsigned long) + * + * "Magic" number confirming that this is an hrb_t structure and + * what operations are valid for it. + * + * Must be S3COMMS_HRB_MAGIC to be valid. + * + * `body` (char *) : + * + * Pointer to start of HTTP body. + * + * Can be NULL, in which case it is treated as the empty string, "". + * + * `body_len` (size_t) : + * + * Number of bytes (characters) in `body`. 0 if empty or NULL `body`. + * + * `first_header` (hrb_node_t *) : + * + * Pointer to first SORTED header node, if any. + * It is left to the programmer to ensure that this node and associated + * list is destroyed when done. + * + * `resource` (char *) : + * + * Pointer to resource URL string, e.g., "/folder/page.xhtml". + * + * `verb` (char *) : + * + * Pointer to HTTP verb string, e.g., "GET". + * + * `version` (char *) : + * + * Pointer to HTTP version string, e.g., "HTTP/1.1". + * + * + * + * Programmer: Jacob Smith + * + *---------------------------------------------------------------------------- + */ +typedef struct { + unsigned long magic; + char *body; + size_t body_len; + hrb_node_t *first_header; + char *resource; + char *verb; + char *version; +} hrb_t; +#define S3COMMS_HRB_MAGIC 0x6DCC84UL + + +/*---------------------------------------------------------------------------- + * + * Structure: parsed_url_t + * + * + * Represent a URL with easily-accessed pointers to logical elements within. + * These elements (components) are stored as null-terminated strings (or just + * NULLs). These components should be allocated for the structure, making the + * data as safe as possible from modification. If a component is NULL, it is + * either implicit in or absent from the URL. + * + * "http://mybucket.s3.amazonaws.com:8080/somefile.h5?param=value&arg=value" + * ^--^ ^-----------------------^ ^--^ ^---------^ ^-------------------^ + * Scheme Host Port Resource Query/-ies + * + * + * + * `magic` (unsigned long) + * + * Structure identification and validation identifier. + * Identifies as `parsed_url_t` type. + * + * `scheme` (char *) + * + * String representing which protocol is to be expected. + * _Must_ be present. + * "http", "https", "ftp", e.g. + * + * `host` (char *) + * + * String of host, either domain name, IPv4, or IPv6 format. + * _Must_ be present. + * "over.rainbow.oz", "192.168.0.1", "[0000:0000:0000:0001]" + * + * `port` (char *) + * + * String representation of specified port. Must resolve to a valid unsigned + * integer. + * "9000", "80" + * + * `path` (char *) + * + * Path to resource on host. If not specified, assumes root "/". + * "lollipop_guild.wav", "characters/witches/white.dat" + * + * `query` (char *) + * + * Single string of all query parameters in url (if any). + * "arg1=value1&arg2=value2" + * + * + * + * Programmer: Jacob Smith + * + *---------------------------------------------------------------------------- + */ +typedef struct { + unsigned long magic; + char *scheme; /* required */ + char *host; /* required */ + char *port; + char *path; + char *query; +} parsed_url_t; +#define S3COMMS_PARSED_URL_MAGIC 0x21D0DFUL + + +/*---------------------------------------------------------------------------- + * + * Structure: s3r_t + * + * + * + * S3 request structure "handle". + * + * Holds persistent information for Amazon S3 requests. + * + * Instantiated through `H5FD_s3comms_s3r_open()`, copies data into self. + * + * Intended to be re-used for operations on a remote object. + * + * Cleaned up through `H5FD_s3comms_s3r_close()`. + * + * _DO NOT_ share handle between threads: curl easy handle `curlhandle` has + * undefined behavior if called to perform in multiple threads. + * + * + * + * `magic` (unsigned long) + * + * "magic" number identifying this structure as unique type. + * MUST equal `S3R_MAGIC` to be valid. + * + * `curlhandle` (CURL) + * + * Pointer to the curl_easy handle generated for the request. + * + * `httpverb` (char *) + * + * Pointer to NULL-terminated string. HTTP verb, + * e.g. "GET", "HEAD", "PUT", etc. + * + * Default is NULL, resulting in a "GET" request. + * + * `purl` (parsed_url_t *) + * + * Pointer to structure holding the elements of URL for file open. + * + * e.g., "http://bucket.aws.com:8080/myfile.dat?q1=v1&q2=v2" + * parsed into... + * { scheme: "http" + * host: "bucket.aws.com" + * port: "8080" + * path: "myfile.dat" + * query: "q1=v1&q2=v2" + * } + * + * Cannot be NULL. + * + * `region` (char *) + * + * Pointer to NULL-terminated string, specifying S3 "region", + * e.g., "us-east-1". + * + * Required to authenticate. + * + * `secret_id` (char *) + * + * Pointer to NULL-terminated string for "secret" access id to S3 resource. + * + * Requred to authenticate. + * + * `signing_key` (unsigned char *) + * + * Pointer to `SHA256_DIGEST_LENGTH`-long string for "re-usable" signing + * key, generated via + * `HMAC-SHA256(HMAC-SHA256(HMAC-SHA256(HMAC-SHA256("AWS4<secret_key>", + * "<yyyyMMDD"), "<aws-region>"), "<aws-service>"), "aws4_request")` + * which may be re-used for several (up to seven (7)) days from creation? + * Computed once upon file open. + * + * Requred to authenticate. + * + * + * + * Programmer: Jacob Smith + * + *---------------------------------------------------------------------------- + */ +typedef struct { + unsigned long magic; +#ifdef H5_HAVE_ROS3_VFD + CURL *curlhandle; + size_t filesize; + char *httpverb; + parsed_url_t *purl; + char *region; + char *secret_id; + unsigned char *signing_key; +#endif /* ifdef H5_HAVE_ROS3_VFD */ +} s3r_t; +#define S3COMMS_S3R_MAGIC 0x44d8d79 + +/******************************************* + * DECLARATION OF HTTP FIELD LIST ROUTINES * + *******************************************/ + +herr_t H5FD_s3comms_hrb_node_set(hrb_node_t **L, + const char *name, + const char *value); + +/*********************************************** + * DECLARATION OF HTTP REQUEST BUFFER ROUTINES * + ***********************************************/ + +herr_t H5FD_s3comms_hrb_destroy(hrb_t **buf); + +hrb_t * H5FD_s3comms_hrb_init_request(const char *verb, + const char *resource, + const char *host); + +/************************************* + * DECLARATION OF S3REQUEST ROUTINES * + *************************************/ + +H5_DLL herr_t H5FD_s3comms_s3r_close(s3r_t *handle); + +H5_DLL size_t H5FD_s3comms_s3r_get_filesize(s3r_t *handle); + +H5_DLL s3r_t * H5FD_s3comms_s3r_open(const char url[], + const char region[], + const char id[], + const unsigned char signing_key[]); + +H5_DLL herr_t H5FD_s3comms_s3r_read(s3r_t *handle, + haddr_t offset, + size_t len, + void *dest); + +/********************************* + * DECLARATION OF OTHER ROUTINES * + *********************************/ + +H5_DLL struct tm * gmnow(void); + +herr_t H5FD_s3comms_aws_canonical_request(char *canonical_request_dest, + char *signed_headers_dest, + hrb_t *http_request); + +H5_DLL herr_t H5FD_s3comms_bytes_to_hex(char *dest, + const unsigned char *msg, + size_t msg_len, + hbool_t lowercase); + +herr_t H5FD_s3comms_free_purl(parsed_url_t *purl); + +herr_t H5FD_s3comms_HMAC_SHA256(const unsigned char *key, + size_t key_len, + const char *msg, + size_t msg_len, + char *dest); + +herr_t H5FD_s3comms_load_aws_profile(const char *name, + char *key_id_out, + char *secret_access_key_out, + char *aws_region_out); + +herr_t H5FD_s3comms_nlowercase(char *dest, + const char *s, + size_t len); + +herr_t H5FD_s3comms_parse_url(const char *str, + parsed_url_t **purl); + +herr_t H5FD_s3comms_percent_encode_char(char *repr, + const unsigned char c, + size_t *repr_len); + +H5_DLL herr_t H5FD_s3comms_signing_key(unsigned char *md, + const char *secret, + const char *region, + const char *iso8601now); + +herr_t H5FD_s3comms_tostringtosign(char *dest, + const char *req_str, + const char *now, + const char *region); + +H5_DLL herr_t H5FD_s3comms_trim(char *dest, + char *s, + size_t s_len, + size_t *n_written); + +H5_DLL herr_t H5FD_s3comms_uriencode(char *dest, + const char *s, + size_t s_len, + hbool_t encode_slash, + size_t *n_written); + + diff --git a/src/Makefile.am b/src/Makefile.am index 0eaae1a..f737d5d 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -63,8 +63,8 @@ libhdf5_la_SOURCES= H5.c H5checksum.c H5dbg.c H5system.c H5timer.c H5trace.c \ H5FA.c H5FAcache.c H5FAdbg.c H5FAdblock.c H5FAdblkpage.c H5FAhdr.c \ H5FAint.c H5FAstat.c H5FAtest.c \ H5FD.c H5FDcore.c \ - H5FDfamily.c H5FDint.c H5FDlog.c \ - H5FDmulti.c H5FDsec2.c H5FDspace.c H5FDstdio.c H5FDtest.c \ + H5FDfamily.c H5FDhdfs.c H5FDint.c H5FDlog.c H5FDs3comms.c \ + H5FDmulti.c H5FDros3.c H5FDsec2.c H5FDspace.c H5FDstdio.c H5FDtest.c \ H5FL.c H5FO.c H5FS.c H5FScache.c H5FSdbg.c H5FSint.c H5FSsection.c \ H5FSstat.c H5FStest.c \ H5G.c H5Gbtree2.c H5Gcache.c \ @@ -138,8 +138,8 @@ include_HEADERS = hdf5.h H5api_adpt.h H5overflow.h H5pubconf.h H5public.h H5vers H5Cpublic.h H5Dpublic.h \ H5Epubgen.h H5Epublic.h H5ESpublic.h H5Fpublic.h \ H5FDpublic.h H5FDcore.h H5FDdirect.h \ - H5FDfamily.h H5FDlog.h H5FDmpi.h H5FDmpio.h \ - H5FDmulti.h H5FDsec2.h H5FDstdio.h H5FDwindows.h \ + H5FDfamily.h H5FDhdfs.h H5FDlog.h H5FDmpi.h H5FDmpio.h \ + H5FDmulti.h H5FDros3.h H5FDsec2.h H5FDstdio.h H5FDwindows.h \ H5Gpublic.h H5Ipublic.h H5Lpublic.h \ H5MMpublic.h H5Opublic.h H5Ppublic.h \ H5PLextern.h H5PLpublic.h \ @@ -40,16 +40,18 @@ #include "H5Zpublic.h" /* Data filters */ /* Predefined file drivers */ -#include "H5FDcore.h" /* Files stored entirely in memory */ -#include "H5FDdirect.h" /* Linux direct I/O */ -#include "H5FDfamily.h" /* File families */ +#include "H5FDcore.h" /* Files stored entirely in memory */ +#include "H5FDdirect.h" /* Linux direct I/O */ +#include "H5FDfamily.h" /* File families */ +#include "H5FDhdfs.h" /* Hadoop HDFS */ #include "H5FDlog.h" /* sec2 driver with I/O logging (for debugging) */ -#include "H5FDmpi.h" /* MPI-based file drivers */ -#include "H5FDmulti.h" /* Usage-partitioned file family */ -#include "H5FDsec2.h" /* POSIX unbuffered file I/O */ -#include "H5FDstdio.h" /* Standard C buffered I/O */ +#include "H5FDmpi.h" /* MPI-based file drivers */ +#include "H5FDmulti.h" /* Usage-partitioned file family */ +#include "H5FDros3.h" /* R/O S3 "file" I/O */ +#include "H5FDsec2.h" /* POSIX unbuffered file I/O */ +#include "H5FDstdio.h" /* Standard C buffered I/O */ #ifdef H5_HAVE_WINDOWS -#include "H5FDwindows.h" /* Win32 I/O */ +#include "H5FDwindows.h" /* Win32 I/O */ #endif /* Virtual object layer (VOL) connectors */ diff --git a/src/libhdf5.settings.in b/src/libhdf5.settings.in index f856ebc..baa99ea 100644 --- a/src/libhdf5.settings.in +++ b/src/libhdf5.settings.in @@ -79,6 +79,8 @@ Parallel Filtered Dataset Writes: @PARALLEL_FILTERED_WRITES@ I/O filters (external): @EXTERNAL_FILTERS@ MPE: @MPE@ Direct VFD: @DIRECT_VFD@ + (Read-Only) S3 VFD: @ROS3_VFD@ + (Read-Only) HDFS VFD: @HAVE_LIBHDFS@ dmalloc: @HAVE_DMALLOC@ Packages w/ extra debug output: @INTERNAL_DEBUG_OUTPUT@ API tracing: @TRACE_API@ diff --git a/test/CMakeLists.txt b/test/CMakeLists.txt index 8333c78..cbd1901 100644 --- a/test/CMakeLists.txt +++ b/test/CMakeLists.txt @@ -267,6 +267,9 @@ set (H5_TESTS enc_dec_plist_cross_platform getname vfd + ros3 + s3comms + hdfs ntypes dangle dtransform diff --git a/test/CMakeVFDTests.cmake b/test/CMakeVFDTests.cmake index 2767bdc..7f661a5 100644 --- a/test/CMakeVFDTests.cmake +++ b/test/CMakeVFDTests.cmake @@ -18,6 +18,7 @@ # included from CMakeTEsts.cmake set (VFD_LIST + hdfs sec2 stdio core diff --git a/test/Makefile.am b/test/Makefile.am index 1b82aa4..dbee9c2 100644 --- a/test/Makefile.am +++ b/test/Makefile.am @@ -59,11 +59,11 @@ TEST_PROG= testhdf5 \ stab gheap evict_on_close farray earray btree2 fheap \ pool accum hyperslab istore bittests dt_arith page_buffer \ dtypes dsets cmpd_dset filter_fail extend direct_chunk external efc \ - objcopy links unlink twriteorder big mtime fillval mount \ - flush1 flush2 app_ref enum set_extent ttsafe enc_dec_plist \ - enc_dec_plist_cross_platform getname vfd ntypes dangle dtransform \ - reserved cross_read freespace mf vds file_image unregister \ - cache_logging cork swmr vol + objcopy links unlink twriteorder big mtime fillval mount flush1 \ + flush2 app_ref enum set_extent ttsafe enc_dec_plist \ + enc_dec_plist_cross_platform getname vfd ros3 s3comms hdfs ntypes \ + dangle dtransform reserved cross_read freespace mf vds file_image \ + unregister cache_logging cork swmr vol # List programs to be built when testing here. # error_test and err_compat are built at the same time as the other tests, but executed by testerror.sh. @@ -145,7 +145,7 @@ ttsafe_SOURCES=ttsafe.c ttsafe_dcreate.c ttsafe_error.c ttsafe_cancel.c \ ttsafe_acreate.c cache_image_SOURCES=cache_image.c genall5.c -VFD_LIST = sec2 stdio core core_paged split multi family +VFD_LIST = hdfs sec2 stdio core core_paged split multi family if DIRECT_VFD_CONDITIONAL VFD_LIST += direct endif @@ -201,8 +201,8 @@ CHECK_CLEANFILES+=accum.h5 cmpd_dset.h5 compact_dataset.h5 dataset.h5 dset_offse dtransform.h5 test_filters.h5 get_file_name.h5 tstint[1-2].h5 \ unlink_chunked.h5 btree2.h5 btree2_tmp.h5 objcopy_src.h5 objcopy_dst.h5 \ objcopy_ext.dat trefer1.h5 trefer2.h5 app_ref.h5 farray.h5 farray_tmp.h5 \ - earray.h5 earray_tmp.h5 efc[0-5].h5 log_vfd_out.log \ - new_multi_file_v16-r.h5 new_multi_file_v16-s.h5 \ + earray.h5 earray_tmp.h5 efc[0-5].h5 log_vfd_out.log log_ros3_out.log \ + log_s3comms_out.log new_multi_file_v16-r.h5 new_multi_file_v16-s.h5 \ split_get_file_image_test-m.h5 split_get_file_image_test-r.h5 \ file_image_core_test.h5.copy unregister_filter_1.h5 unregister_filter_2.h5 \ vds_virt.h5 vds_dapl.h5 vds_src_[0-1].h5 \ diff --git a/test/hdfs.c b/test/hdfs.c new file mode 100644 index 0000000..11b9918 --- /dev/null +++ b/test/hdfs.c @@ -0,0 +1,1836 @@ +/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * Read-Only HDFS Virtual File Driver (VFD) * + * Copyright (c) 2018, The HDF Group. * + * * + * All rights reserved. * + * * + * NOTICE: * + * All information contained herein is, and remains, the property of The HDF * + * Group. The intellectual and technical concepts contained herein are * + * proprietary to The HDF Group. Dissemination of this information or * + * reproduction of this material is strictly forbidden unless prior written * + * permission is obtained from The HDF Group. * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ + +/* + * Purpose: + * + * Verify behavior for Read-Only HDFS VFD. + * + * Demonstrates basic use cases and fapl interaction. + * + * Programmer: Jacob Smith <jake.smith@hdfgroup.org> + * 2018-04-23 + */ + +#include "h5test.h" /* testing utilities */ +#include "H5FDhdfs.h" /* this file driver's utilities */ + + +#define HDFS_TEST_DEBUG 0 +#define HDFS_TEST_MAX_BUF_SIZE 256 +#if 0 +#define S3_TEST_MAX_URL_SIZE 256 +#endif /* s3comms relic */ + +/***************************************************************************** + * + * FILE-LOCAL TESTING MACROS + * + * Purpose: + * + * 1) Upon test failure, goto-jump to single-location teardown in test + * function. E.g., `error:` (consistency with HDF corpus) or + * `failed:` (reflects purpose). + * >>> using "error", in part because `H5E_BEGIN_TRY` expects it. + * 2) Increase clarity and reduce overhead found with `TEST_ERROR`. + * e.g., "if(somefunction(arg, arg2) < 0) TEST_ERROR:" + * requires reading of entire line to know whether this if/call is + * part of the test setup, test operation, or a test unto itself. + * 3) Provide testing macros with optional user-supplied failure message; + * if not supplied (NULL), generate comparison output in the spirit of + * test-driven development. E.g., "expected 5 but was -3" + * User messages clarify test's purpose in code, encouraging description + * without relying on comments. + * 4) Configurable expected-actual order in generated comparison strings. + * Some prefer `VERIFY(expected, actual)`, others + * `VERIFY(actual, expected)`. Provide preprocessor ifdef switch + * to satifsy both parties, assuming one paradigm per test file. + * (One could #undef and redefine the flag through the file as desired, + * but _why_.) + * + * Provided as courtesy, per consideration for inclusion in the library + * proper. + * + * Macros: + * + * JSVERIFY_EXP_ACT - ifdef flag, configures comparison order + * FAIL_IF() - check condition + * FAIL_UNLESS() - check _not_ condition + * JSVERIFY() - long-int equality check; prints reason/comparison + * JSVERIFY_NOT() - long-int inequality check; prints + * JSVERIFY_STR() - string equality check; prints + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *****************************************************************************/ + + +/*---------------------------------------------------------------------------- + * + * ifdef flag: JSVERIFY_EXP_ACT + * + * JSVERIFY macros accept arguments as (EXPECTED, ACTUAL[, reason]) + * default, if this is undefined, is (ACTUAL, EXPECTED[, reason]) + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_EXP_ACT 1L + + +/*---------------------------------------------------------------------------- + * + * Macro: JSFAILED_AT() + * + * Purpose: + * + * Preface a test failure by printing "*FAILED*" and location to stdout + * Similar to `H5_FAILED(); AT();` from h5test.h + * + * *FAILED* at somefile.c:12 in function_name()... + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSFAILED_AT() { \ + HDprintf("*FAILED* at %s:%d in %s()...\n", __FILE__, __LINE__, FUNC); \ +} + + +/*---------------------------------------------------------------------------- + * + * Macro: FAIL_IF() + * + * Purpose: + * + * Make tests more accessible and less cluttered than + * `if (thing == otherthing()) TEST_ERROR` + * paradigm. + * + * The following lines are roughly equivalent: + * + * `if (myfunc() < 0) TEST_ERROR;` (as seen elsewhere in HDF tests) + * `FAIL_IF(myfunc() < 0)` + * + * Prints a generic "FAILED AT" line to stdout and jumps to `error`, + * similar to `TEST_ERROR` in h5test.h + * + * Programmer: Jacob Smith + * 2017-10-23 + * + *---------------------------------------------------------------------------- + */ +#define FAIL_IF(condition) \ +if (condition) { \ + JSFAILED_AT() \ + goto error; \ +} + + +/*---------------------------------------------------------------------------- + * + * Macro: FAIL_UNLESS() + * + * Purpose: + * + * TEST_ERROR wrapper to reduce cognitive overhead from "negative tests", + * e.g., "a != b". + * + * Opposite of FAIL_IF; fails if the given condition is _not_ true. + * + * `FAIL_IF( 5 != my_op() )` + * is equivalent to + * `FAIL_UNLESS( 5 == my_op() )` + * However, `JSVERIFY(5, my_op(), "bad return")` may be even clearer. + * (see JSVERIFY) + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#if 0 /* UNUSED */ +#define FAIL_UNLESS(condition) \ +if (!(condition)) { \ + JSFAILED_AT() \ + goto error; \ +} +#endif /* UNUSED */ + + +/*---------------------------------------------------------------------------- + * + * Macro: JSERR_LONG() + * + * Purpose: + * + * Print an failure message for long-int arguments. + * ERROR-AT printed first. + * If `reason` is given, it is printed on own line and newlined after + * else, prints "expected/actual" aligned on own lines. + * + * *FAILED* at myfile.c:488 in somefunc()... + * forest must be made of trees. + * + * or + * + * *FAILED* at myfile.c:488 in somefunc()... + * ! Expected 425 + * ! Actual 3 + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSERR_LONG(expected, actual, reason) { \ + JSFAILED_AT() \ + if (reason!= NULL) { \ + HDprintf("%s\n", (reason)); \ + } else { \ + HDprintf(" ! Expected %ld\n ! Actual %ld\n", \ + (long)(expected), (long)(actual)); \ + } \ +} + + +/*---------------------------------------------------------------------------- + * + * Macro: JSERR_STR() + * + * Purpose: + * + * Print an failure message for string arguments. + * ERROR-AT printed first. + * If `reason` is given, it is printed on own line and newlined after + * else, prints "expected/actual" aligned on own lines. + * + * *FAILED* at myfile.c:421 in myfunc()... + * Blue and Red strings don't match! + * + * or + * + * *FAILED* at myfile.c:421 in myfunc()... + * !!! Expected: + * this is my expected + * string + * !!! Actual: + * not what I expected at all + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSERR_STR(expected, actual, reason) { \ + JSFAILED_AT() \ + if ((reason) != NULL) { \ + HDprintf("%s\n", (reason)); \ + } else { \ + HDprintf("!!! Expected:\n%s\n!!!Actual:\n%s\n", \ + (expected), (actual)); \ + } \ +} + + + +#ifdef JSVERIFY_EXP_ACT + + +/*---------------------------------------------------------------------------- + * + * Macro: JSVERIFY() + * + * Purpose: + * + * Verify that two long integers are equal. + * If unequal, print failure message + * (with `reason`, if not NULL; expected/actual if NULL) + * and jump to `error` at end of function + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY(expected, actual, reason) \ +if ((long)(actual) != (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)) \ + goto error; \ +} /* JSVERIFY */ + + +/*---------------------------------------------------------------------------- + * + * Macro: JSVERIFY_NOT() + * + * Purpose: + * + * Verify that two long integers are _not_ equal. + * If equal, print failure message + * (with `reason`, if not NULL; expected/actual if NULL) + * and jump to `error` at end of function + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_NOT(expected, actual, reason) \ +if ((long)(actual) == (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)) \ + goto error; \ +} /* JSVERIFY_NOT */ + + +/*---------------------------------------------------------------------------- + * + * Macro: JSVERIFY_STR() + * + * Purpose: + * + * Verify that two strings are equal. + * If unequal, print failure message + * (with `reason`, if not NULL; expected/actual if NULL) + * and jump to `error` at end of function + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_STR(expected, actual, reason) \ +if (strcmp((actual), (expected)) != 0) { \ + JSERR_STR((expected), (actual), (reason)); \ + goto error; \ +} /* JSVERIFY_STR */ + + +#else +/* JSVERIFY_EXP_ACT not defined + * + * Repeats macros above, but with actual/expected parameters reversed. + */ + + +/*---------------------------------------------------------------------------- + * Macro: JSVERIFY() + * See: JSVERIFY documentation above. + * Programmer: Jacob Smith + * 2017-10-14 + *---------------------------------------------------------------------------- + */ +#define JSVERIFY(actual, expected, reason) \ +if ((long)(actual) != (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)); \ + goto error; \ +} /* JSVERIFY */ + + +/*---------------------------------------------------------------------------- + * Macro: JSVERIFY_NOT() + * See: JSVERIFY_NOT documentation above. + * Programmer: Jacob Smith + * 2017-10-14 + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_NOT(actual, expected, reason) \ +if ((long)(actual) == (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)) \ + goto error; \ +} /* JSVERIFY_NOT */ + + +/*---------------------------------------------------------------------------- + * Macro: JSVERIFY_STR() + * See: JSVERIFY_STR documentation above. + * Programmer: Jacob Smith + * 2017-10-14 + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_STR(actual, expected, reason) \ +if (strcmp((actual), (expected)) != 0) { \ + JSERR_STR((expected), (actual), (reason)); \ + goto error; \ +} /* JSVERIFY_STR */ + +#endif /* ifdef/else JSVERIFY_EXP_ACT */ + +/******************************** + * OTHER MACROS AND DEFINITIONS * + ********************************/ + +/* copied from src/hdfs.c + */ +#define MAXADDR (((haddr_t)1<<(8*sizeof(HDoff_t)-1))-1) +#define MAX_HDFS_NAMENODE_NAME 128 + +/******************************* + * FILE-LOCAL GLOBAL VARIABLES * + *******************************/ +static const char filename_missing[] = "/tmp/missing.txt"; +static const char filename_bard[] = "/tmp/t8.shakespeare.txt"; +static const char filename_raven[] = "/tmp/Poe_Raven.txt"; +static const char filename_example_h5[] = "/tmp/t.h5"; +static H5FD_hdfs_fapl_t default_fa = { + 1, /* fa version */ + "localhost", /* namenode name */ + 0, /* namenode port */ + "", /* user name */ + "", /* kerberos path */ + 1024, /* buffer size */ +}; + +/****************** + * TEST FUNCTIONS * + ******************/ + + +/*--------------------------------------------------------------------------- + * + * Function: test_fapl_config_validation() + * + * Purpose: + * + * Test data consistency of fapl configuration. + * Tests `H5FD_hdfs_validate_config` indirectly through `H5Pset_fapl_hdfs`. + * + * Return: + * + * PASSED : 0 + * FAILED : 1 + * + * Programmer: Jacob Smith + * 2018-04-25 + * + * Changes: None. + * + *--------------------------------------------------------------------------- + */ +static int +test_fapl_config_validation(void) +{ + /********************* + * test-local macros * + *********************/ + + /************************* + * test-local structures * + *************************/ + + struct testcase { + const char *msg; + herr_t expected; + H5FD_hdfs_fapl_t config; + }; + + /************************ + * test-local variables * + ************************/ + + hid_t fapl_id = -1; /* file access property list ID */ + H5FD_hdfs_fapl_t config; + H5FD_hdfs_fapl_t fa_fetch; + herr_t success = SUCCEED; + unsigned int i = 0; + unsigned int ncases = 6; /* should equal number of cases */ + struct testcase *case_ptr = NULL; /* dumb work-around for possible */ + /* dynamic cases creation because */ + /* of compiler warnings Wlarger-than */ + struct testcase cases_arr[] = { + { "default config fapl", + SUCCEED, + { 1, /* version */ + "localhost", /* namenode_name */ + 0, /* namenode_port number */ + "some_user", /* user_name */ + "", /* kerberos_ticket_cache path */ + -1, /* stream_buffer_size */ + }, + }, + { "invalid version number (2)", + FAIL, + { 2, /* version */ + "localhost", /* namenode_name */ + 0, /* namenode_port number */ + "some_user", /* user_name */ + "", /* kerberos_ticket_cache path */ + -1, /* stream_buffer_size */ + }, + }, + { "invalid version number (0)", + FAIL, + { 0, /* version */ + "localhost", /* namenode_name */ + 0, /* namenode_port number */ + "some_user", /* user_name */ + "", /* kerberos_ticket_cache path */ + -1, /* stream_buffer_size */ + }, + }, + { "nonsense kerberos path still ok?", + SUCCEED, + { 1, /* version */ + "localhost", /* namenode_name */ + 0, /* namenode_port number */ + "some_user", /* user_name */ + "pathToSomewhere", /* kerberos_ticket_cache path */ + -1, /* stream_buffer_size */ + }, + }, + { "namenode port number too high", + FAIL, + { 1, /* version */ + "localhost", /* namenode_name */ + 88000, /* namenode_port number */ + "some_user", /* user_name */ + "", /* kerberos_ticket_cache path */ + -1, /* stream_buffer_size */ + }, + }, + { "negative namenode port number", + FAIL, + { 1, /* version */ + "localhost", /* namenode_name */ + -1, /* namenode_port number */ + "some_user", /* user_name */ + "", /* kerberos_ticket_cache path */ + -1, /* stream_buffer_size */ + }, + }, + }; + + TESTING("HDFS fapl configuration validation"); + + /********* + * TESTS * + *********/ + + for (i = 0; i < ncases; i++) { + + /*--------------- + * per-test setup + *--------------- + */ + case_ptr = &cases_arr[i]; + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( fapl_id < 0 ) /* sanity-check */ + + /*----------------------------------- + * Actually test -- set fapl. + * Mute stack trace in failure cases. + *----------------------------------- + */ + H5E_BEGIN_TRY { + /* `H5FD_hdfs_validate_config(...)` is static/private + * to src/hdfs.c and cannot (and should not?) be tested directly? + * Instead, validate config through public api. + */ + success = H5Pset_fapl_hdfs(fapl_id, &case_ptr->config); + } H5E_END_TRY; + + JSVERIFY( case_ptr->expected, success, case_ptr->msg ) + + /* Make sure we can get back what we put in. + * Only valid if the fapl configuration does not result in error. + */ + if (success == SUCCEED) { + config = case_ptr->config; + JSVERIFY( SUCCEED, + H5Pget_fapl_hdfs(fapl_id, &fa_fetch), + "unable to get fapl" ) + JSVERIFY( H5FD__CURR_HDFS_FAPL_T_VERSION, + fa_fetch.version, + "invalid version number" ) + JSVERIFY( config.version, + fa_fetch.version, + "version number mismatch" ) + JSVERIFY( config.namenode_port, + fa_fetch.namenode_port, + "namenode port mismatch" ) + JSVERIFY( config.stream_buffer_size, + fa_fetch.stream_buffer_size, + "streambuffer size mismatch" ) + JSVERIFY_STR( config.namenode_name, + fa_fetch.namenode_name, + NULL ) + JSVERIFY_STR( config.user_name, + fa_fetch.user_name, + NULL ) + JSVERIFY_STR( config.kerberos_ticket_cache, + fa_fetch.kerberos_ticket_cache, + NULL ) + } + + /*----------------------------- + * per-test sanitation/teardown + *----------------------------- + */ + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + fapl_id = -1; + + } /* for each test case */ + + PASSED(); + return 0; + +error: + /*********** + * CLEANUP * + ***********/ + + if (fapl_id < 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + } + return 1; + +} /* test_fapl_config_validation */ + + +/*------------------------------------------------------------------------- + * + * Function: test_hdfs_fapl() + * + * Purpose: Tests the file handle interface for the HDFS driver. + * + * For now, test only fapl & flags. Extend as the + * work on the VFD continues. + * + * Return: Success: 0 + * Failure: 1 + * + * Programmer: Jacob Smith + * 2018-04-25 + * + * Changes: None. + * + *------------------------------------------------------------------------- + */ +static int +test_hdfs_fapl(void) +{ + /************************ + * test-local variables * + ************************/ + + hid_t fapl_id = -1; /* file access property list ID */ + hid_t driver_id = -1; /* ID for this VFD */ + unsigned long driver_flags = 0; /* VFD feature flags */ + H5FD_hdfs_fapl_t hdfs_fa_0 = { + 1, /* version*/ + "", /* node name */ + 9000, /* node port */ + "", /* username */ + "", /* kerb cache path */ + 1024, /* stream buffer size */ + }; + + TESTING("HDFS fapl "); + + /* Set property list and file name for HDFS driver. + */ + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( fapl_id < 0 ) + + FAIL_IF( FAIL == H5Pset_fapl_hdfs(fapl_id, &hdfs_fa_0) ) + + driver_id = H5Pget_driver(fapl_id); + FAIL_IF( driver_id < 0 ) + + /**************** + * Check that the VFD feature flags are correct + * SPEC MAY CHANGE + ******************/ + + FAIL_IF( H5FDdriver_query(driver_id, &driver_flags) < 0 ) + + JSVERIFY_NOT( 0, (driver_flags & H5FD_FEAT_DATA_SIEVE), + "bit(s) in `driver_flags` must align with " + "H5FD_FEAT_DATA_SIEVE" ) + + JSVERIFY( H5FD_FEAT_DATA_SIEVE, driver_flags, + "H5FD_FEAT_DATA_SIEVE should be the only supported flag") + + PASSED(); + return 0; + +error: + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + + return 1; + +} /* test_hdfs_fapl() */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_vfd_open() + * + * Purpose: + * + * Demonstrate/specify VFD-level "Open" failure cases + * + * Return: + * + * PASSED : 0 + * FAILED : 1 + * + * Programmer: Jacob Smith + * 2018-06-07 + * + *--------------------------------------------------------------------------- + */ +static int +test_vfd_open(void) +{ + /********************* + * test-local macros * + *********************/ + +/* selectors for which fapl to use in testcase */ +#define FAPL_H5P_DEFAULT -2 +#define FAPL_UNCONFIGURED -3 /* H5P_FILE_ACCESS */ +#define FAPL_HDFS -4 + +#ifdef H5_HAVE_LIBHDFS + /************************* + * test-local structures * + *************************/ + + struct test_condition { + const char *message; + const char *url; + unsigned flags; + int which_fapl; + haddr_t maxaddr; + hbool_t might_use_other_driver; + }; + + /************************ + * test-local variables * + ************************/ + + struct test_condition failing_conditions[] = { + { "default property list (H5P_DEFAULT) is invalid", + filename_bard, + H5F_ACC_RDONLY, + FAPL_H5P_DEFAULT, + MAXADDR, + TRUE, + }, + { "generic file access property list is invalid", + filename_bard, + H5F_ACC_RDONLY, + FAPL_UNCONFIGURED, + MAXADDR, + TRUE, + }, + { "filename cannot be null", + NULL, + H5F_ACC_RDONLY, + FAPL_HDFS, + MAXADDR, + FALSE, + }, + { "filename cannot be empty", + "", + H5F_ACC_RDONLY, + FAPL_HDFS, + MAXADDR, + FALSE, + }, + { "file at filename must exist", + filename_missing, + H5F_ACC_RDONLY, + FAPL_HDFS, + MAXADDR, + FALSE, + }, + { "read-write flag not supported", + filename_bard, + H5F_ACC_RDWR, + FAPL_HDFS, + MAXADDR, + FALSE, + }, + { "truncate flag not supported", + filename_bard, + H5F_ACC_TRUNC, + FAPL_HDFS, + MAXADDR, + FALSE, + }, + { "create flag not supported", + filename_bard, + H5F_ACC_CREAT, + FAPL_HDFS, + MAXADDR, + FALSE, + }, + { "EXCL flag not supported", + filename_bard, + H5F_ACC_EXCL, + FAPL_HDFS, + MAXADDR, + FALSE, + }, + { "maxaddr cannot be 0 (caught in `H5FD_open()`)", + filename_bard, + H5F_ACC_RDONLY, + FAPL_HDFS, + 0, + FALSE, + }, + }; +#endif /* H5_HAVE_LIBHDFS */ + H5FD_t *fd = NULL; + hid_t fapl_hdfs = -1; + hid_t fapl_unconfigured = -1; + unsigned i = 0; + unsigned failing_conditions_count = 10; + + TESTING("HDFS VFD-level open"); + +#ifndef H5_HAVE_LIBHDFS + SKIPPED(); + puts(" HDFS VFD is not enabled"); + fflush(stdout); + return 0; +#else + fapl_unconfigured = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( fapl_unconfigured < 0 ) + + fapl_hdfs = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( fapl_hdfs < 0 ) + FAIL_IF( FAIL == H5Pset_fapl_hdfs(fapl_hdfs, &default_fa) ) + + /********* + * TESTS * + *********/ + + /* all the test cases that will _not_ open + */ + for (i = 0; i < failing_conditions_count; i++) { + struct test_condition T = failing_conditions[i]; + hid_t fapl_id = H5P_DEFAULT; + + fd = NULL; + + if (T.which_fapl == FAPL_UNCONFIGURED) + fapl_id = fapl_unconfigured; + else if (T.which_fapl == FAPL_HDFS) + fapl_id = fapl_hdfs; + +#if HDFS_TEST_DEBUG + HDfprintf(stderr, "testing: %s\n", T.message); +#endif /* HDFS_TEST_DEBUG */ + + H5E_BEGIN_TRY { + fd = H5FDopen(T.url, T.flags, fapl_id, T.maxaddr); + } H5E_END_TRY; + if (NULL != fd) { + if (TRUE == T.might_use_other_driver && + H5FD_HDFS != fd->driver_id) + { + HDfprintf(stderr, "\n!!!!! WARNING !!!!!\n" \ + " Successful open of file on local system " \ + "with non-HDFS VFD.\n"); + JSVERIFY(SUCCEED, H5FDclose(fd), + "unable to close errant open"); + fd = NULL; + } else { + JSVERIFY(1, 0, T.message); /* print message and fail */ + } + } + } + + FAIL_IF( NULL != fd ) /* sanity check */ + +#if HDFS_TEST_DEBUG + HDfprintf(stderr, "nominal open\n"); +#endif /* HDFS_TEST_DEBUG */ + + /* finally, show that a file can be opened + */ + fd = H5FDopen( + filename_bard, + H5F_ACC_RDONLY, + fapl_hdfs, + MAXADDR); + FAIL_IF( NULL == fd ) + + /************ + * TEARDOWN * + ************/ + +#if HDFS_TEST_DEBUG + HDfprintf(stderr, "teardown...\n"); +#endif /* HDFS_TEST_DEBUG */ + + FAIL_IF( FAIL == H5FDclose(fd) ) + fd = NULL; + + FAIL_IF( FAIL == H5Pclose(fapl_hdfs) ) + fapl_hdfs = -1; + + FAIL_IF( FAIL == H5Pclose(fapl_unconfigured) ) + fapl_unconfigured = -1; + + PASSED(); + return 0; +#endif /* H5_HAVE_LIBHDFS */ + +error: + /*********** + * CLEANUP * + ***********/ + + if (fd) + (void)H5FDclose(fd); + H5E_BEGIN_TRY { + if (fapl_hdfs >= 0) + (void)H5Pclose(fapl_hdfs); + if (fapl_unconfigured >= 0) + (void)H5Pclose(fapl_unconfigured); + } H5E_END_TRY; + + return 1; + +#undef FAPL_H5P_DEFAULT +#undef FAPL_UNCONFIGURED +#undef FAPL_HDFS + +} /* test_vfd_open */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_eof_eoa() + * + * Purpose: + * + * Demonstrate behavior of get_eof, get_eoa, and set_eoa. + * + * Return: + * + * PASSED : 0 + * FAILED : 1 + * + * Programmer: Jacob Smith + * 2018-06-07 + * + *--------------------------------------------------------------------------- + */ +static int +test_eof_eoa(void) +{ + /********************* + * test-local macros * + *********************/ + + /************************* + * test-local structures * + *************************/ + + /************************ + * test-local variables * + ************************/ + + H5FD_t *fd_shakespeare = NULL; + hid_t fapl_id = -1; + + TESTING("HDFS eof/eoa gets and sets"); + +#ifndef H5_HAVE_LIBHDFS + SKIPPED(); + puts(" HDFS VFD is not enabled"); + fflush(stdout); + return 0; +#else + /********* + * SETUP * + *********/ + + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( 0 > fapl_id ) + FAIL_IF( FAIL == H5Pset_fapl_hdfs(fapl_id, &default_fa) ) + + fd_shakespeare = H5FDopen( + filename_bard, + H5F_ACC_RDONLY, + fapl_id, + HADDR_UNDEF); + FAIL_IF( NULL == fd_shakespeare ) + + /********* + * TESTS * + *********/ + + /* verify as found + */ + JSVERIFY( 5458199, H5FDget_eof(fd_shakespeare, H5FD_MEM_DEFAULT), NULL ) + JSVERIFY( H5FDget_eof(fd_shakespeare, H5FD_MEM_DEFAULT), + H5FDget_eof(fd_shakespeare, H5FD_MEM_DRAW), + "mismatch between DEFAULT and RAW memory types" ) + JSVERIFY( 0, + H5FDget_eoa(fd_shakespeare, H5FD_MEM_DEFAULT), + "EoA should be unset by H5FDopen" ) + + /* set EoA below EoF + */ + JSVERIFY( SUCCEED, + H5FDset_eoa(fd_shakespeare, H5FD_MEM_DEFAULT, 44442202), + "unable to set EoA (lower)" ) + JSVERIFY( 5458199, + H5FDget_eof(fd_shakespeare, H5FD_MEM_DEFAULT), + "EoF changed" ) + JSVERIFY( 44442202, + H5FDget_eoa(fd_shakespeare, H5FD_MEM_DEFAULT), + "EoA unchanged" ) + + /* set EoA above EoF + */ + JSVERIFY( SUCCEED, + H5FDset_eoa(fd_shakespeare, H5FD_MEM_DEFAULT, 6789012), + "unable to set EoA (higher)" ) + JSVERIFY( 5458199, + H5FDget_eof(fd_shakespeare, H5FD_MEM_DEFAULT), + "EoF changed" ) + JSVERIFY( 6789012, + H5FDget_eoa(fd_shakespeare, H5FD_MEM_DEFAULT), + "EoA unchanged" ) + + /************ + * TEARDOWN * + ************/ + + FAIL_IF( FAIL == H5FDclose(fd_shakespeare) ) + fd_shakespeare = NULL; + + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + fapl_id = -1; + + PASSED(); + return 0; +#endif /* H5_HAVE_LIBHDFS */ + +error: + /*********** + * CLEANUP * + ***********/ + + if (fd_shakespeare != NULL) { + (void)H5FDclose(fd_shakespeare); + } + if (fapl_id >= 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + } + + return 1; + +} /* test_eof_eoa */ + + +/*----------------------------------------------------------------------------- + * + * Function: test_H5FDread_without_eoa_set_fails() + * + * Purpose: + * + * Demonstrate a not-obvious constraint by the library, preventing + * file read before EoA is set + * + * Programmer: Jacob Smith + * 2018-06-08 + * + *----------------------------------------------------------------------------- + */ +static int +test_H5FDread_without_eoa_set_fails(void) +{ +#ifdef H5_HAVE_LIBHDFS + char buffer[HDFS_TEST_MAX_BUF_SIZE]; + unsigned int i = 0; +#endif /* H5_HAVE_LIBHDFS */ + H5FD_t *file_shakespeare = NULL; + hid_t fapl_id = -1; + + TESTING("HDFS VFD read-eoa temporal coupling library limitation"); + +#ifndef H5_HAVE_LIBHDFS + SKIPPED(); + puts(" HDFS VFD is not enabled"); + fflush(stdout); + return 0; +#else + /********* + * SETUP * + *********/ + + /* create HDFS fapl + */ + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( fapl_id < 0 ) + FAIL_IF( FAIL == H5Pset_fapl_hdfs(fapl_id, &default_fa) ) + + file_shakespeare = H5FDopen( + filename_bard, + H5F_ACC_RDONLY, + fapl_id, + MAXADDR); + FAIL_IF( NULL == file_shakespeare ) + + JSVERIFY( 0, H5FDget_eoa(file_shakespeare, H5FD_MEM_DEFAULT), + "EoA should remain unset by H5FDopen" ) + + /* zero buffer contents */ + for (i = 0; i < HDFS_TEST_MAX_BUF_SIZE; i++) { + buffer[i] = 0; + } + + /******** + * TEST * + ********/ + + H5E_BEGIN_TRY { /* mute stack trace on expected failure */ + JSVERIFY( FAIL, + H5FDread(file_shakespeare, + H5FD_MEM_DRAW, + H5P_DEFAULT, + 1200699, + 102, + buffer), + "cannot read before eoa is set" ) + } H5E_END_TRY; + for (i = 0; i < HDFS_TEST_MAX_BUF_SIZE; i++) { + JSVERIFY( 0, (unsigned)buffer[i], "buffer was modified by write!" ) + } + + /************ + * TEARDOWN * + ************/ + + FAIL_IF( FAIL == H5FDclose(file_shakespeare) ) + file_shakespeare = NULL; + + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + fapl_id = -1; + + PASSED(); + return 0; +#endif /* H5_HAVE_LIBHDFS */ + +error: + /*********** + * CLEANUP * + ***********/ + + if (file_shakespeare) { + (void)H5FDclose(file_shakespeare); + } + if (fapl_id >= 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + } + + return 1; + +} /* test_H5FDread_without_eoa_set_fails */ + + + +/*--------------------------------------------------------------------------- + * + * Function: test_read() + * + * Purpose: + * + * Return: + * + * PASSED : 0 + * FAILED : 1 + * + * Programmer: Jacob Smith + * 2018-06-08 + * + *--------------------------------------------------------------------------- + */ +static int +test_read(void) +{ + /********************* + * test-local macros * + *********************/ + +#ifdef H5_HAVE_LIBHDFS + /************************* + * test-local structures * + *************************/ + struct testcase { + const char *message; /* purpose of test case */ + haddr_t eoa_set; /* set file EOA to this prior to read */ + size_t addr; /* offset of read in file */ + size_t len; /* length of read in file */ + herr_t success; /* expected return value of read function */ + const char *expected; /* expected contents of buffer; failure ignores */ + }; + + /************************ + * test-local variables * + ************************/ + struct testcase cases[] = { + { "successful range-get", + 6464, + 5691, + 32, /* fancy quotes are three bytes each(?) */ + SUCCEED, + "Quoth the Raven “Nevermore.”", + }, + { "read past EOA fails (EOA < EOF < addr)", + 3000, + 4000, + 100, + FAIL, + NULL, + }, + { "read overlapping EOA fails (EOA < addr < EOF < (addr+len))", + 3000, + 8000, + 100, + FAIL, + NULL, + }, + { "read past EOA/EOF fails ((EOA==EOF) < addr)", + 6464, + 7000, + 100, + FAIL, + NULL, + }, + { "read overlapping EOA/EOF fails (addr < (EOA==EOF) < (addr+len))", + 6464, + 6400, + 100, + FAIL, + NULL, + }, + { "read between EOF and EOA fails (EOF < addr < (addr+len) < EOA)", + 8000, + 7000, + 100, + FAIL, + NULL, + }, + }; + unsigned testcase_count = 6; + unsigned test_i = 0; + struct testcase test; + herr_t open_return = FAIL; + char buffer[HDFS_TEST_MAX_BUF_SIZE]; + unsigned int i = 0; +#endif /* H5_HAVE_LIBHDFS */ + H5FD_t *file_raven = NULL; + hid_t fapl_id = -1; + + TESTING("HDFS VFD read/range-gets"); + +#ifndef H5_HAVE_LIBHDFS + SKIPPED(); + puts(" HDFS VFD is not enabled"); + fflush(stdout); + return 0; +#else + /********* + * SETUP * + *********/ + + /* create HDFS fapl + */ + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( fapl_id < 0 ) + FAIL_IF( FAIL == H5Pset_fapl_hdfs(fapl_id, &default_fa) ) + + /* zero buffer contents */ + for (i = 0; i < HDFS_TEST_MAX_BUF_SIZE; i++) { + buffer[i] = 0; + } + + /* open file + */ + file_raven = H5FDopen( + filename_raven, + H5F_ACC_RDONLY, + fapl_id, + HADDR_UNDEF); /* Demonstrate success with "automatic" value */ + FAIL_IF( NULL == file_raven ) + + JSVERIFY( 6464, H5FDget_eof(file_raven, H5FD_MEM_DEFAULT), NULL ) + + /********* + * TESTS * + *********/ + + for (test_i = 0; test_i < testcase_count; test_i++) { + + /* -------------- * + * per-test setup * + * -------------- */ + + test = cases[test_i]; + open_return = FAIL; + + FAIL_IF( HDFS_TEST_MAX_BUF_SIZE < test.len ) /* buffer too small! */ + + FAIL_IF( FAIL == + H5FDset_eoa( file_raven, H5FD_MEM_DEFAULT, test.eoa_set) ) + + /* zero buffer contents */ + for (i = 0; i < HDFS_TEST_MAX_BUF_SIZE; i++) { + buffer[i] = 0; + } + + /* ------------ * + * conduct test * + * ------------ */ + + H5E_BEGIN_TRY { + open_return = H5FDread( + file_raven, + H5FD_MEM_DRAW, + H5P_DEFAULT, + test.addr, + test.len, + buffer); + } H5E_END_TRY; + + JSVERIFY( test.success, + open_return, + test.message ) + + if (open_return == SUCCEED) { + JSVERIFY_STR( test.expected, buffer, NULL ) + } + + } /* for each testcase */ + + /************ + * TEARDOWN * + ************/ + + FAIL_IF( FAIL == H5FDclose(file_raven) ) + file_raven = NULL; + + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + fapl_id = -1; + + PASSED(); + return 0; +#endif /* H5_HAVE_LIBHDFS */ + +error: + /*********** + * CLEANUP * + ***********/ + + if (file_raven != 0) + (void)H5FDclose(file_raven); + if (fapl_id >= 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + } + + return 1; + +} /* test_read */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_noops_and_autofails() + * + * Purpose: + * + * Demonstrate the unavailable and do-nothing routines unique to + * Read-Only VFD. + * + * Return: + * + * PASSED : 0 + * FAILED : 1 + * + * Programmer: Jacob Smith + * 2017-11-06 + * + * Changes: + * + modify from S3VFD codebase to HDFS; Minor changes, mostly. + * + Jacob Smith 2018-06-08 + * + *--------------------------------------------------------------------------- + */ +static int +test_noops_and_autofails(void) +{ + /********************* + * test-local macros * + *********************/ + + /************************* + * test-local structures * + *************************/ + + /************************ + * test-local variables * + ************************/ + + hid_t fapl_id = -1; + H5FD_t *file = NULL; + const char data[36] = "The Force shall be with you, always"; + + TESTING("HDFS VFD always-fail and no-op routines"); + + /********* + * SETUP * + *********/ + +#ifndef H5_HAVE_LIBHDFS + SKIPPED(); + puts(" HDFS VFD is not enabled"); + fflush(stdout); + return 0; +#else + /* create HDFS fapl + */ + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( fapl_id < 0 ) + FAIL_IF( FAIL == H5Pset_fapl_hdfs(fapl_id, &default_fa) ) + + /* open file + */ + file = H5FDopen( + filename_bard, + H5F_ACC_RDONLY, + fapl_id, + HADDR_UNDEF); + FAIL_IF( NULL == file ) + + /********* + * TESTS * + *********/ + + /* auto-fail calls to write and truncate + */ + H5E_BEGIN_TRY { + JSVERIFY( FAIL, + H5FDwrite(file, H5FD_MEM_DRAW, H5P_DEFAULT, 1000, 35, data), + "write must fail" ) + } H5E_END_TRY; + + H5E_BEGIN_TRY { + JSVERIFY( FAIL, + H5FDtruncate(file, H5P_DEFAULT, FALSE), + "truncate must fail" ) + } H5E_END_TRY; + + H5E_BEGIN_TRY { + JSVERIFY( FAIL, + H5FDtruncate(file, H5P_DEFAULT, TRUE), + "truncate must fail (closing)" ) + } H5E_END_TRY; + + /* no-op calls to `lock()` and `unlock()` + */ + JSVERIFY( SUCCEED, + H5FDlock(file, TRUE), + "lock always succeeds; has no effect" ) + JSVERIFY( SUCCEED, + H5FDlock(file, FALSE), + NULL ) + JSVERIFY( SUCCEED, + H5FDunlock(file), + NULL ) + /* Lock/unlock with null file or similar error crashes tests. + * HDassert in calling heirarchy, `H5FD[un]lock()` and `H5FD_[un]lock()` + */ + + /************ + * TEARDOWN * + ************/ + + FAIL_IF( FAIL == H5FDclose(file) ) + file = NULL; + + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + fapl_id = -1; + + PASSED(); + return 0; +#endif /* H5_HAVE_LIBHDFS */ + +error: + /*********** + * CLEANUP * + ***********/ + + if (fapl_id >= 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + } + if (file != NULL) { + (void)H5FDclose(file); + } + + return 1; + +} /* test_noops_and_autofails*/ + + +/*--------------------------------------------------------------------------- + * + * Function: test_cmp() + * + * Purpose: + * + * Verify "file comparison" behavior. + * + * Return: + * + * PASSED : 0 + * FAILED : 1 + * + * Programmer: Jacob Smith + * 2017-11-06 + * + *--------------------------------------------------------------------------- + */ +static int +test_cmp(void) +{ +#if 0 + /********************* + * test-local macros * + *********************/ + + /************************* + * test-local structures * + *************************/ + + /************************ + * test-local variables * + ************************/ + + H5FD_t *fd_raven = NULL; + H5FD_t *fd_shakes = NULL; + H5FD_t *fd_raven_2 = NULL; + hbool_t curl_ready = FALSE; + hid_t fapl_id = -1; + + + + TESTING("HDFS cmp (comparison)"); + + if (s3_test_credentials_loaded == 0) { + SKIPPED(); + puts(" s3 credentials are not loaded"); + fflush(stdout); + return 0; + } + + /********* + * SETUP * + *********/ + + FAIL_IF( CURLE_OK != curl_global_init(CURL_GLOBAL_DEFAULT) ) + curl_ready = TRUE; + + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( 0 > fapl_id ) + JSVERIFY( SUCCEED, H5Pset_fapl_hdfs(fapl_id, &restricted_access_fa), NULL ) + + fd_raven = H5FDopen( + url_text_public, + H5F_ACC_RDONLY, + fapl_id, + HADDR_UNDEF); + FAIL_IF( NULL == fd_raven ) + + fd_shakes = H5FDopen( + url_text_restricted, + H5F_ACC_RDONLY, + fapl_id, + HADDR_UNDEF); + FAIL_IF( NULL == fd_shakes ) + + fd_raven_2 = H5FDopen( + url_text_public, + H5F_ACC_RDONLY, + fapl_id, + HADDR_UNDEF); + FAIL_IF( NULL == fd_raven_2 ) + + /********* + * TESTS * + *********/ + + JSVERIFY( 0, H5FDcmp(fd_raven, fd_raven_2), NULL ) + JSVERIFY( -1, H5FDcmp(fd_raven, fd_shakes), NULL ) + JSVERIFY( 1, H5FDcmp(fd_shakes, fd_raven_2), NULL ) + + /************ + * TEARDOWN * + ************/ + + FAIL_IF( FAIL == H5FDclose(fd_raven) ) + fd_raven = NULL; + FAIL_IF( FAIL == H5FDclose(fd_shakes) ) + fd_shakes = NULL; + FAIL_IF( FAIL == H5FDclose(fd_raven_2) ) + fd_raven_2 = NULL; + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + fapl_id = -1; + + curl_global_cleanup(); + curl_ready = FALSE; + + PASSED(); + return 0; + +error: + /*********** + * CLEANUP * + ***********/ + + if (fd_raven != NULL) (void)H5FDclose(fd_raven); + if (fd_raven_2 != NULL) (void)H5FDclose(fd_raven_2); + if (fd_shakes != NULL) (void)H5FDclose(fd_shakes); + if (TRUE == curl_ready) curl_global_cleanup(); + if (fapl_id >= 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + } + +#else + /* + JSFAILED_AT() + HDprintf("TODO\n"); + return 1; + */ + + TESTING("HDFS cmp (comparison)"); + SKIPPED(); + HDfprintf( + stderr, + " TODO: Distinct valid fapls to open the same file.\n"); + + return 0; +#endif /* s3comms relic */ + +} /* test_cmp */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_H5F_integration() + * + * Purpose: + * + * Demonstrate H5F (File interface) behavior with files on HDFS. + * + * Return: + * + * PASSED : 0 + * FAILED : 1 + * + * Programmer: Jacob Smith + * 2017-11-07 + * + * Changes: + * + modify from S3VFD codebase to HDFS; Minor changes, mostly. + * + Jacob Smith 2018-06-08 + * + *--------------------------------------------------------------------------- + */ +static int +test_H5F_integration(void) +{ + /********************* + * test-local macros * + *********************/ + + /************************* + * test-local structures * + *************************/ + + /************************ + * test-local variables * + ************************/ + + hid_t file = -1; + hid_t fapl_id = -1; + + TESTING("HDFS file access through HD5F library (H5F API)"); + + /********* + * SETUP * + *********/ + +#ifndef H5_HAVE_LIBHDFS + SKIPPED(); + puts(" HDFS VFD is not enabled"); + fflush(stdout); + return 0; +#else + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( 0 > fapl_id ) + FAIL_IF( FAIL == H5Pset_fapl_hdfs(fapl_id, &default_fa) ) + + /********* + * TESTS * + *********/ + + /* Read-Write Open access is not allowed with this file driver. + */ + H5E_BEGIN_TRY { + FAIL_IF( 0 <= H5Fopen( + filename_example_h5, + H5F_ACC_RDWR, + fapl_id) ) + } H5E_END_TRY; + + /* H5Fcreate() is not allowed with this file driver. + */ + H5E_BEGIN_TRY { + FAIL_IF( 0 <= H5Fcreate( + filename_missing, + H5F_ACC_RDONLY, + H5P_DEFAULT, + fapl_id) ) + } H5E_END_TRY; + + /* Successful open. + */ + file = H5Fopen( + filename_example_h5, + H5F_ACC_RDONLY, + fapl_id); + FAIL_IF( file < 0 ) + + /************ + * TEARDOWN * + ************/ + + FAIL_IF( FAIL == H5Fclose(file) ) + file = -1; + + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + fapl_id = -1; + + PASSED(); + return 0; +#endif /* H5_HAVE_LIBHDFS */ + +error: + /*********** + * CLEANUP * + ***********/ + +#if HDFS_TEST_DEBUG + HDprintf("\nerror!"); fflush(stdout); +#endif /* HDFS_TEST_DEBUG */ + + if (fapl_id >= 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + } + if (file > 0) { + (void)H5Fclose(file); + } + + return 1; + +} /* test_H5F_integration */ + + +/*------------------------------------------------------------------------- + * + * Function: main + * + * Purpose: Tests the basic features of Virtual File Drivers + * + * Return: Success: 0 + * Failure: 1 + * + * Programmer: Jacob Smith + * 2017-10-23 + * + *------------------------------------------------------------------------- + */ +int +main(void) +{ + int nerrors = 0; + + /****************** + * commence tests * + ******************/ + + static char hdfs_namenode_name[MAX_HDFS_NAMENODE_NAME] = ""; + const char *hdfs_namenode_name_env = NULL; + + hdfs_namenode_name_env = HDgetenv("HDFS_TEST_NAMENODE_NAME"); + if (hdfs_namenode_name_env == NULL || hdfs_namenode_name_env[0] == '\0') { + HDstrncpy(hdfs_namenode_name, "localhost", 9); + } else { + HDstrncpy(default_fa.namenode_name, hdfs_namenode_name_env, MAX_HDFS_NAMENODE_NAME); + } + + h5_reset(); + + HDprintf("Testing hdfs VFD functionality.\n"); + + nerrors += test_fapl_config_validation(); + nerrors += test_hdfs_fapl(); + nerrors += test_vfd_open(); + nerrors += test_eof_eoa(); + nerrors += test_H5FDread_without_eoa_set_fails(); + nerrors += test_read(); + nerrors += test_noops_and_autofails(); + nerrors += test_cmp(); + nerrors += test_H5F_integration(); + + if (nerrors > 0) { + HDprintf("***** %d hdfs TEST%s FAILED! *****\n", + nerrors, + nerrors > 1 ? "S" : ""); + nerrors = 1; + } else { + HDprintf("All hdfs tests passed.\n"); + } + return nerrors; /* 0 if no errors, 1 if any errors */ + +} /* main() */ + + diff --git a/test/ros3.c b/test/ros3.c new file mode 100644 index 0000000..51c1a89 --- /dev/null +++ b/test/ros3.c @@ -0,0 +1,2020 @@ +/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * Read-Only S3 Virtual File Driver (VFD) * + * Copyright (c) 2017-2018, The HDF Group. * + * * + * All rights reserved. * + * * + * NOTICE: * + * All information contained herein is, and remains, the property of The HDF * + * Group. The intellectual and technical concepts contained herein are * + * proprietary to The HDF Group. Dissemination of this information or * + * reproduction of this material is strictly forbidden unless prior written * + * permission is obtained from The HDF Group. * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ + +/* + * Purpose: + * + * Verify behavior for Read-Only S3 VFD + * at the VFL (virtual file layer) level. + * + * Demonstrates basic use cases and fapl/dxpl interaction. + * + * Programmer: Jacob Smith <jake.smith@hdfgroup.org> + * 2017-10-11 + */ + +#include "h5test.h" + +#include "H5FDprivate.h" /* Virtual File Driver utilities */ +#include "H5FDros3.h" /* this file driver's utilities */ +#include "H5FDs3comms.h" /* for loading of credentials */ + + + +/***************************************************************************** + * + * FILE-LOCAL TESTING MACROS + * + * Purpose: + * + * 1) Upon test failure, goto-jump to single-location teardown in test + * function. E.g., `error:` (consistency with HDF corpus) or + * `failed:` (reflects purpose). + * >>> using "error", in part because `H5E_BEGIN_TRY` expects it. + * 2) Increase clarity and reduce overhead found with `TEST_ERROR`. + * e.g., "if(somefunction(arg, arg2) < 0) TEST_ERROR:" + * requires reading of entire line to know whether this if/call is + * part of the test setup, test operation, or a test unto itself. + * 3) Provide testing macros with optional user-supplied failure message; + * if not supplied (NULL), generate comparison output in the spirit of + * test-driven development. E.g., "expected 5 but was -3" + * User messages clarify test's purpose in code, encouraging description + * without relying on comments. + * 4) Configurable expected-actual order in generated comparison strings. + * Some prefer `VERIFY(expected, actual)`, others + * `VERIFY(actual, expected)`. Provide preprocessor ifdef switch + * to satifsy both parties, assuming one paradigm per test file. + * (One could #undef and redefine the flag through the file as desired, + * but _why_.) + * + * Provided as courtesy, per consideration for inclusion in the library + * proper. + * + * Macros: + * + * JSVERIFY_EXP_ACT - ifdef flag, configures comparison order + * FAIL_IF() - check condition + * FAIL_UNLESS() - check _not_ condition + * JSVERIFY() - long-int equality check; prints reason/comparison + * JSVERIFY_NOT() - long-int inequality check; prints + * JSVERIFY_STR() - string equality check; prints + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *****************************************************************************/ + + +/*---------------------------------------------------------------------------- + * + * ifdef flag: JSVERIFY_EXP_ACT + * + * JSVERIFY macros accept arguments as (EXPECTED, ACTUAL[, reason]) + * default, if this is undefined, is (ACTUAL, EXPECTED[, reason]) + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_EXP_ACT 1L + + +/*---------------------------------------------------------------------------- + * + * Macro: JSFAILED_AT() + * + * Purpose: + * + * Preface a test failure by printing "*FAILED*" and location to stdout + * Similar to `H5_FAILED(); AT();` from h5test.h + * + * *FAILED* at somefile.c:12 in function_name()... + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSFAILED_AT() { \ + HDprintf("*FAILED* at %s:%d in %s()...\n", __FILE__, __LINE__, FUNC); \ +} + + +/*---------------------------------------------------------------------------- + * + * Macro: FAIL_IF() + * + * Purpose: + * + * Make tests more accessible and less cluttered than + * `if (thing == otherthing()) TEST_ERROR` + * paradigm. + * + * The following lines are roughly equivalent: + * + * `if (myfunc() < 0) TEST_ERROR;` (as seen elsewhere in HDF tests) + * `FAIL_IF(myfunc() < 0)` + * + * Prints a generic "FAILED AT" line to stdout and jumps to `error`, + * similar to `TEST_ERROR` in h5test.h + * + * Programmer: Jacob Smith + * 2017-10-23 + * + *---------------------------------------------------------------------------- + */ +#define FAIL_IF(condition) \ +if (condition) { \ + JSFAILED_AT() \ + goto error; \ +} + + +/*---------------------------------------------------------------------------- + * + * Macro: FAIL_UNLESS() + * + * Purpose: + * + * TEST_ERROR wrapper to reduce cognitive overhead from "negative tests", + * e.g., "a != b". + * + * Opposite of FAIL_IF; fails if the given condition is _not_ true. + * + * `FAIL_IF( 5 != my_op() )` + * is equivalent to + * `FAIL_UNLESS( 5 == my_op() )` + * However, `JSVERIFY(5, my_op(), "bad return")` may be even clearer. + * (see JSVERIFY) + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#if 0 /* UNUSED */ +#define FAIL_UNLESS(condition) \ +if (!(condition)) { \ + JSFAILED_AT() \ + goto error; \ +} +#endif + + +/*---------------------------------------------------------------------------- + * + * Macro: JSERR_LONG() + * + * Purpose: + * + * Print an failure message for long-int arguments. + * ERROR-AT printed first. + * If `reason` is given, it is printed on own line and newlined after + * else, prints "expected/actual" aligned on own lines. + * + * *FAILED* at myfile.c:488 in somefunc()... + * forest must be made of trees. + * + * or + * + * *FAILED* at myfile.c:488 in somefunc()... + * ! Expected 425 + * ! Actual 3 + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSERR_LONG(expected, actual, reason) { \ + JSFAILED_AT() \ + if (reason!= NULL) { \ + HDprintf("%s\n", (reason)); \ + } else { \ + HDprintf(" ! Expected %ld\n ! Actual %ld\n", \ + (long)(expected), (long)(actual)); \ + } \ +} + + +/*---------------------------------------------------------------------------- + * + * Macro: JSERR_STR() + * + * Purpose: + * + * Print an failure message for string arguments. + * ERROR-AT printed first. + * If `reason` is given, it is printed on own line and newlined after + * else, prints "expected/actual" aligned on own lines. + * + * *FAILED* at myfile.c:421 in myfunc()... + * Blue and Red strings don't match! + * + * or + * + * *FAILED* at myfile.c:421 in myfunc()... + * !!! Expected: + * this is my expected + * string + * !!! Actual: + * not what I expected at all + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSERR_STR(expected, actual, reason) { \ + JSFAILED_AT() \ + if ((reason) != NULL) { \ + HDprintf("%s\n", (reason)); \ + } else { \ + HDprintf("!!! Expected:\n%s\n!!!Actual:\n%s\n", \ + (expected), (actual)); \ + } \ +} + + + +#ifdef JSVERIFY_EXP_ACT + + +/*---------------------------------------------------------------------------- + * + * Macro: JSVERIFY() + * + * Purpose: + * + * Verify that two long integers are equal. + * If unequal, print failure message + * (with `reason`, if not NULL; expected/actual if NULL) + * and jump to `error` at end of function + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY(expected, actual, reason) \ +if ((long)(actual) != (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)) \ + goto error; \ +} /* JSVERIFY */ + + +/*---------------------------------------------------------------------------- + * + * Macro: JSVERIFY_NOT() + * + * Purpose: + * + * Verify that two long integers are _not_ equal. + * If equal, print failure message + * (with `reason`, if not NULL; expected/actual if NULL) + * and jump to `error` at end of function + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_NOT(expected, actual, reason) \ +if ((long)(actual) == (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)) \ + goto error; \ +} /* JSVERIFY_NOT */ + + +/*---------------------------------------------------------------------------- + * + * Macro: JSVERIFY_STR() + * + * Purpose: + * + * Verify that two strings are equal. + * If unequal, print failure message + * (with `reason`, if not NULL; expected/actual if NULL) + * and jump to `error` at end of function + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_STR(expected, actual, reason) \ +if (strcmp((actual), (expected)) != 0) { \ + JSERR_STR((expected), (actual), (reason)); \ + goto error; \ +} /* JSVERIFY_STR */ + + +#else +/* JSVERIFY_EXP_ACT not defined + * + * Repeats macros above, but with actual/expected parameters reversed. + */ + + +/*---------------------------------------------------------------------------- + * Macro: JSVERIFY() + * See: JSVERIFY documentation above. + * Programmer: Jacob Smith + * 2017-10-14 + *---------------------------------------------------------------------------- + */ +#define JSVERIFY(actual, expected, reason) \ +if ((long)(actual) != (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)); \ + goto error; \ +} /* JSVERIFY */ + + +/*---------------------------------------------------------------------------- + * Macro: JSVERIFY_NOT() + * See: JSVERIFY_NOT documentation above. + * Programmer: Jacob Smith + * 2017-10-14 + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_NOT(actual, expected, reason) \ +if ((long)(actual) == (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)) \ + goto error; \ +} /* JSVERIFY_NOT */ + + +/*---------------------------------------------------------------------------- + * Macro: JSVERIFY_STR() + * See: JSVERIFY_STR documentation above. + * Programmer: Jacob Smith + * 2017-10-14 + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_STR(actual, expected, reason) \ +if (strcmp((actual), (expected)) != 0) { \ + JSERR_STR((expected), (actual), (reason)); \ + goto error; \ +} /* JSVERIFY_STR */ + +#endif /* ifdef/else JSVERIFY_EXP_ACT */ + +/******************************** + * OTHER MACROS AND DEFINITIONS * + ********************************/ + +/* copied from src/ros3.c + */ +#define MAXADDR (((haddr_t)1<<(8*sizeof(HDoff_t)-1))-1) + +#ifdef H5_HAVE_ROS3_VFD +#define S3_TEST_PROFILE_NAME "ros3_vfd_test" + +#define S3_TEST_MAX_URL_SIZE 256 + +#define S3_TEST_RESOURCE_TEXT_RESTRICTED "t8.shakespeare.txt" +#define S3_TEST_RESOURCE_TEXT_PUBLIC "Poe_Raven.txt" +#define S3_TEST_RESOURCE_H5_PUBLIC "GMODO-SVM01.h5" +#define S3_TEST_RESOURCE_MISSING "missing.csv" + +static char url_text_restricted[S3_TEST_MAX_URL_SIZE] = ""; +static char url_text_public[S3_TEST_MAX_URL_SIZE] = ""; +static char url_h5_public[S3_TEST_MAX_URL_SIZE] = ""; +static char url_missing[S3_TEST_MAX_URL_SIZE] = ""; +static char s3_test_bucket_url[S3_TEST_MAX_URL_SIZE] = ""; +static hbool_t s3_test_bucket_defined = FALSE; + +/* Global variables for aws test profile. + * An attempt is made to read ~/.aws/credentials and ~/.aws/config upon test + * startup -- if unable to open either file or cannot load region, id, and key, + * tests connecting with S3 will not be run + */ +static int s3_test_credentials_loaded = 0; +static char s3_test_aws_region[16]; +static char s3_test_aws_access_key_id[64]; +static char s3_test_aws_secret_access_key[128]; + +H5FD_ros3_fapl_t restricted_access_fa = { + H5FD__CURR_ROS3_FAPL_T_VERSION, /* fapl version */ + TRUE, /* authenticate */ + "", /* aws region */ + "", /* access key id */ + ""}; /* secret access key */ + +H5FD_ros3_fapl_t anonymous_fa = { + H5FD__CURR_ROS3_FAPL_T_VERSION, + FALSE, "", "", "" }; +#endif /* H5_HAVE_ROS3_VFD */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_fapl_config_validation() + * + * Purpose: + * + * Test data consistency of fapl configuration. + * Tests `H5FD_ros3_validate_config` indirectly through `H5Pset_fapl_ros3`. + * + * Return: + * + * PASSED : 0 + * FAILED : 1 + * + * Programmer: Jacob Smith + * 2017-10-23 + * + * Changes: None. + * + *--------------------------------------------------------------------------- + */ +static int +test_fapl_config_validation(void) +{ + /********************* + * test-local macros * + *********************/ + + /************************* + * test-local structures * + *************************/ + + struct testcase { + const char *msg; + herr_t expected; + H5FD_ros3_fapl_t config; + }; + + /************************ + * test-local variables * + ************************/ + +#ifdef H5_HAVE_ROS3_VFD + hid_t fapl_id = -1; /* file access property list ID */ + H5FD_ros3_fapl_t config; + H5FD_ros3_fapl_t fa_fetch; + herr_t success = SUCCEED; + unsigned int i = 0; + unsigned int ncases = 8; /* should equal number of cases */ + struct testcase *case_ptr = NULL; /* dumb work-around for possible */ + /* dynamic cases creation because */ + /* of compiler warnings Wlarger-than */ + struct testcase cases_arr[] = { + { "non-authenticating config allows empties.\n", + SUCCEED, + { H5FD__CURR_ROS3_FAPL_T_VERSION, /* version */ + FALSE, /* authenticate */ + "", /* aws_region */ + "", /* secret_id */ + "", /* secret_key */ + }, + }, + { "authenticating config asks for populated strings.\n", + FAIL, + { H5FD__CURR_ROS3_FAPL_T_VERSION, + TRUE, + "", + "", + "", + }, + }, + { "populated strings; key is the empty string?\n", + SUCCEED, + { H5FD__CURR_ROS3_FAPL_T_VERSION, + TRUE, + "region", + "me", + "", + }, + }, + { "id cannot be empty.\n", + FAIL, + { H5FD__CURR_ROS3_FAPL_T_VERSION, + TRUE, + "", + "me", + "", + }, + }, + { "region cannot be empty.\n", + FAIL, + { H5FD__CURR_ROS3_FAPL_T_VERSION, + TRUE, + "where", + "", + "", + }, + }, + { "all strings populated.\n", + SUCCEED, + { H5FD__CURR_ROS3_FAPL_T_VERSION, + TRUE, + "where", + "who", + "thisIsA GREAT seeeecrit", + }, + }, + { "incorrect version should fail\n", + FAIL, + { 12345, + FALSE, + "", + "", + "", + }, + }, + { "non-authenticating config cares not for (de)population" + "of strings.\n", + SUCCEED, + { H5FD__CURR_ROS3_FAPL_T_VERSION, + FALSE, + "someregion", + "someid", + "somekey", + }, + }, + }; + +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("ROS3 fapl configuration validation"); + + /********* + * TESTS * + *********/ + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD not enabled"); + fflush(stdout); + return 0; +#else + + if (FALSE == s3_test_bucket_defined) { + SKIPPED(); + puts(" environment variable HDF5_ROS3_TEST_BUCKET_URL not defined"); + fflush(stdout); + return 0; + } + + for (i = 0; i < ncases; i++) { + + /*--------------- + * per-test setup + *--------------- + */ + case_ptr = &cases_arr[i]; + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( fapl_id < 0 ) /* sanity-check */ + + /*----------------------------------- + * Actually test. + * Mute stack trace in failure cases. + *----------------------------------- + */ + H5E_BEGIN_TRY { + /* `H5FD_ros3_validate_config(...)` is static/private + * to src/ros3.c and cannot (and should not?) be tested directly? + * Instead, validate config through public api. + */ + success = H5Pset_fapl_ros3(fapl_id, &case_ptr->config); + } H5E_END_TRY; + + JSVERIFY( case_ptr->expected, success, case_ptr->msg ) + + /* Make sure we can get back what we put in. + * Only valid if the fapl configuration does not result in error. + */ + if (success == SUCCEED) { + config = case_ptr->config; + JSVERIFY( SUCCEED, + H5Pget_fapl_ros3(fapl_id, &fa_fetch), + "unable to get fapl" ) + + JSVERIFY( H5FD__CURR_ROS3_FAPL_T_VERSION, + fa_fetch.version, + "invalid version number" ) + JSVERIFY( config.version, + fa_fetch.version, + "version number mismatch" ) + JSVERIFY( config.authenticate, + fa_fetch.authenticate, + "authentication flag mismatch" ) + JSVERIFY_STR( config.aws_region, + fa_fetch.aws_region, + NULL ) + JSVERIFY_STR( config.secret_id, + fa_fetch.secret_id, + NULL ) + JSVERIFY_STR( config.secret_key, + fa_fetch.secret_key, + NULL ) + } + + /*----------------------------- + * per-test sanitation/teardown + *----------------------------- + */ + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + fapl_id = -1; + + } /* for each test case */ + + PASSED(); + return 0; + +error: + /*********** + * CLEANUP * + ***********/ + + if (fapl_id < 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + } + return 1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_fapl_config_validation */ + + +/*------------------------------------------------------------------------- + * + * Function: test_ros3_fapl() + * + * Purpose: Tests the file handle interface for the ROS3 driver + * + * As the ROS3 driver is 1) read only, 2) requires access + * to an S3 server (minio for now), this test is quite + * different from the other tests. + * + * For now, test only fapl & flags. Extend as the + * work on the VFD continues. + * + * Return: Success: 0 + * Failure: 1 + * + * Programmer: John Mainzer + * 7/12/17 + * + * Changes: Test only fapl and flags. + * Jacob Smith 2017 + * + *------------------------------------------------------------------------- + */ +static int +test_ros3_fapl(void) +{ + /************************ + * test-local variables * + ************************/ + +#ifdef H5_HAVE_ROS3_VFD + hid_t fapl_id = -1; /* file access property list ID */ + hid_t driver_id = -1; /* ID for this VFD */ + unsigned long driver_flags = 0; /* VFD feature flags */ + H5FD_ros3_fapl_t ros3_fa_0 = { + H5FD__CURR_ROS3_FAPL_T_VERSION, /* version */ + FALSE, /* authenticate */ + "", /* aws_region */ + "", /* secret_id */ + "plugh", /* secret_key */ + }; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("ROS3 fapl "); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD not enabled"); + fflush(stdout); + return 0; +#else + /* Set property list and file name for ROS3 driver. + */ + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( fapl_id < 0 ) + + FAIL_IF( FAIL == H5Pset_fapl_ros3(fapl_id, &ros3_fa_0) ) + + driver_id = H5Pget_driver(fapl_id); + FAIL_IF( driver_id < 0 ) + + /**************** + * Check that the VFD feature flags are correct + * SPEC MAY CHANGE + ******************/ + + FAIL_IF( H5FDdriver_query(driver_id, &driver_flags) < 0 ) + + JSVERIFY_NOT( 0, (driver_flags & H5FD_FEAT_DATA_SIEVE), + "bit(s) in `driver_flags` must align with " + "H5FD_FEAT_DATA_SIEVE" ) + + JSVERIFY( H5FD_FEAT_DATA_SIEVE, driver_flags, + "H5FD_FEAT_DATA_SIEVE should be the only supported flag") + + PASSED(); + return 0; + +error: + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + + return 1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_ros3_fapl() */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_vfd_open() + * + * Purpose: + * + * Demonstrate/specify VFD-level "Open" failure cases + * + * Return: + * + * PASSED : 0 + * FAILED : 1 + * + * Programmer: Jacob Smith + * 1027-11-03 + * + *--------------------------------------------------------------------------- + */ +static int +test_vfd_open(void) +{ + /********************* + * test-local macros * + *********************/ + +#ifdef H5_HAVE_ROS3_VFD + +#define FAPL_H5P_DEFAULT -2 +#define FAPL_FILE_ACCESS -3 +#define FAPL_ROS3_ANON -4 + + /************************* + * test-local structures * + *************************/ + + struct test_condition { + const char *message; + const char *url; + unsigned flags; + int which_fapl; + haddr_t maxaddr; + }; + + /************************ + * test-local variables * + ************************/ + + struct test_condition tests[] = { + { "default property list (H5P_DEFAULT) is invalid", + url_text_public, + H5F_ACC_RDONLY, + FAPL_H5P_DEFAULT, + MAXADDR, + }, + { "generic file access property list is invalid", + url_text_public, + H5F_ACC_RDONLY, + FAPL_FILE_ACCESS, + MAXADDR, + }, + { "filename cannot be null", + NULL, + H5F_ACC_RDONLY, + FAPL_ROS3_ANON, + MAXADDR, + }, + { "filename cannot be empty", + "", + H5F_ACC_RDONLY, + FAPL_ROS3_ANON, + MAXADDR, + }, + { "filename must exist", + url_missing, + H5F_ACC_RDONLY, + FAPL_ROS3_ANON, + MAXADDR, + }, + { "read-write flag not supported", + url_text_public, + H5F_ACC_RDWR, + FAPL_ROS3_ANON, + MAXADDR, + }, + { "truncate flag not supported", + url_text_public, + H5F_ACC_TRUNC, + FAPL_ROS3_ANON, + MAXADDR, + }, + { "create flag not supported", + url_text_public, + H5F_ACC_CREAT, + FAPL_ROS3_ANON, + MAXADDR, + }, + { "EXCL flag not supported", + url_text_public, + H5F_ACC_EXCL, + FAPL_ROS3_ANON, + MAXADDR, + }, + { "maxaddr cannot be 0 (caught in `H5FD_open()`)", + url_text_public, + H5F_ACC_RDONLY, + FAPL_ROS3_ANON, + 0, + }, + }; + H5FD_t *fd = NULL; + hbool_t curl_ready = FALSE; + hid_t fapl_id = -1; + hid_t fapl_file_access = -1; + unsigned i = 0; + unsigned tests_count = 10; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("ROS3 VFD-level open"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD not enabled"); + fflush(stdout); + return 0; +#else + + if (FALSE == s3_test_bucket_defined) { + SKIPPED(); + puts(" environment variable HDF5_ROS3_TEST_BUCKET_URL not defined"); + fflush(stdout); + return 0; + } + + FAIL_IF( CURLE_OK != curl_global_init(CURL_GLOBAL_DEFAULT) ) + curl_ready = TRUE; + + fapl_file_access = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( fapl_file_access < 0 ) + + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( fapl_id < 0 ) + FAIL_IF( FAIL == H5Pset_fapl_ros3(fapl_id, &anonymous_fa) ) + + /********* + * TESTS * + *********/ + + /* all the test cases that will _not_ open + */ + for (i = 0; i < tests_count; i++) { + struct test_condition T = tests[i]; + hid_t _fapl_id = H5P_DEFAULT; + + fd = NULL; + + if (T.which_fapl == FAPL_FILE_ACCESS) + _fapl_id = fapl_file_access; + else if (T.which_fapl == FAPL_ROS3_ANON) + _fapl_id = fapl_id; + + H5E_BEGIN_TRY { + fd = H5FDopen(T.url, T.flags, _fapl_id, T.maxaddr); + } H5E_END_TRY; + if (NULL != fd) + JSVERIFY(1, 0, T.message); /* wrapper to print message and fail */ + } + + FAIL_IF( NULL != fd ) + + /* finally, show that a file can be opened + */ + fd = H5FDopen( + url_text_public, + H5F_ACC_RDONLY, + fapl_id, + MAXADDR); + FAIL_IF( NULL == fd ) + + /************ + * TEARDOWN * + ************/ + + FAIL_IF( FAIL == H5FDclose(fd) ) + fd = NULL; + + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + fapl_id = -1; + + FAIL_IF( FAIL == H5Pclose(fapl_file_access) ) + fapl_file_access = -1; + + curl_global_cleanup(); + curl_ready = FALSE; + + PASSED(); + return 0; + +error: + /*********** + * CLEANUP * + ***********/ + + if (fd) { + (void)H5FDclose(fd); + } + if (fapl_id >= 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + } + if (fapl_file_access >= 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_file_access); + } H5E_END_TRY; + } + if (curl_ready == TRUE) { + curl_global_cleanup(); + } + + return 1; + +#undef FAPL_FILE_ACCESS +#undef FAPL_H5P_DEFAULT +#undef FAPL_ROS3_ANON + +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_vfd_open */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_eof_eoa() + * + * Purpose: + * + * Demonstrate behavior of get_eof, get_eoa, and set_eoa. + * + * Return: + * + * PASSED : 0 + * FAILED : 1 + * + * Programmer: Jacob Smith + * 2017-11-08 + * + *--------------------------------------------------------------------------- + */ +static int +test_eof_eoa(void) +{ + /********************* + * test-local macros * + *********************/ + + /************************* + * test-local structures * + *************************/ + + /************************ + * test-local variables * + ************************/ + +#ifdef H5_HAVE_ROS3_VFD + H5FD_t *fd_shakespeare = NULL; + hbool_t curl_ready = FALSE; + hid_t fapl_id = -1; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("ROS3 eof/eoa gets and sets"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD not enabled"); + fflush(stdout); + return 0; +#else + if (s3_test_credentials_loaded == 0) { + SKIPPED(); + puts(" s3 credentials are not loaded"); + fflush(stdout); + return 0; + } + + if (FALSE == s3_test_bucket_defined) { + SKIPPED(); + puts(" environment variable HDF5_ROS3_TEST_BUCKET_URL not defined"); + fflush(stdout); + return 0; + } + + /********* + * SETUP * + *********/ + + FAIL_IF( CURLE_OK != curl_global_init(CURL_GLOBAL_DEFAULT) ) + curl_ready = TRUE; + + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( 0 > fapl_id ) + FAIL_IF( FAIL == H5Pset_fapl_ros3(fapl_id, &restricted_access_fa) ) + + fd_shakespeare = H5FDopen( + url_text_restricted, + H5F_ACC_RDONLY, + fapl_id, + HADDR_UNDEF); + FAIL_IF( NULL == fd_shakespeare ) + + /********* + * TESTS * + *********/ + + /* verify as found + */ + JSVERIFY( 5458199, H5FDget_eof(fd_shakespeare, H5FD_MEM_DEFAULT), NULL ) + JSVERIFY( H5FDget_eof(fd_shakespeare, H5FD_MEM_DEFAULT), + H5FDget_eof(fd_shakespeare, H5FD_MEM_DRAW), + "mismatch between DEFAULT and RAW memory types" ) + JSVERIFY( 0, + H5FDget_eoa(fd_shakespeare, H5FD_MEM_DEFAULT), + "EoA should be unset by H5FDopen" ) + + /* set EoA below EoF + */ + JSVERIFY( SUCCEED, + H5FDset_eoa(fd_shakespeare, H5FD_MEM_DEFAULT, 44442202), + "unable to set EoA (lower)" ) + JSVERIFY( 5458199, + H5FDget_eof(fd_shakespeare, H5FD_MEM_DEFAULT), + "EoF changed" ) + JSVERIFY( 44442202, + H5FDget_eoa(fd_shakespeare, H5FD_MEM_DEFAULT), + "EoA unchanged" ) + + /* set EoA above EoF + */ + JSVERIFY( SUCCEED, + H5FDset_eoa(fd_shakespeare, H5FD_MEM_DEFAULT, 6789012), + "unable to set EoA (higher)" ) + JSVERIFY( 5458199, + H5FDget_eof(fd_shakespeare, H5FD_MEM_DEFAULT), + "EoF changed" ) + JSVERIFY( 6789012, + H5FDget_eoa(fd_shakespeare, H5FD_MEM_DEFAULT), + "EoA unchanged" ) + + /************ + * TEARDOWN * + ************/ + + FAIL_IF( FAIL == H5FDclose(fd_shakespeare) ) + + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + fapl_id = -1; + + curl_global_cleanup(); + curl_ready = FALSE; + + PASSED(); + return 0; + +error: + /*********** + * CLEANUP * + ***********/ + + if (fd_shakespeare) (void)H5FDclose(fd_shakespeare); + if (TRUE == curl_ready) curl_global_cleanup(); + if (fapl_id >= 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + } + + return 1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_eof_eoa */ + + +/*----------------------------------------------------------------------------- + * + * Function: test_H5FDread_without_eoa_set_fails() + * + * Purpose: + * + * Demonstrate a not-obvious constraint by the library, preventing + * file read before EoA is set + * + * Programmer: Jacob Smith + * 2018-01-26 + * + *----------------------------------------------------------------------------- + */ +static int +test_H5FDread_without_eoa_set_fails(void) +{ +#ifdef H5_HAVE_ROS3_VFD + char buffer[256]; + unsigned int i = 0; + H5FD_t *file_shakespeare = NULL; + hid_t fapl_id = -1; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("ROS3 VFD read-eoa temporal coupling library limitation "); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD not enabled"); + fflush(stdout); + return 0; +#else + if (s3_test_credentials_loaded == 0) { + SKIPPED(); + puts(" s3 credentials are not loaded"); + fflush(stdout); + return 0; + } + + if (FALSE == s3_test_bucket_defined) { + SKIPPED(); + puts(" environment variable HDF5_ROS3_TEST_BUCKET_URL not defined"); + fflush(stdout); + return 0; + } + + /********* + * SETUP * + *********/ + + /* create ROS3 fapl + */ + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( fapl_id < 0 ) + FAIL_IF( FAIL == H5Pset_fapl_ros3(fapl_id, &restricted_access_fa) ) + + file_shakespeare = H5FDopen( + url_text_restricted, + H5F_ACC_RDONLY, + fapl_id, + MAXADDR); + FAIL_IF( NULL == file_shakespeare ) + + JSVERIFY( 0, H5FDget_eoa(file_shakespeare, H5FD_MEM_DEFAULT), + "EoA should remain unset by H5FDopen" ) + + for (i = 0; i < 256; i++) + buffer[i] = 0; /* zero buffer contents */ + + /******** + * TEST * + ********/ + + H5E_BEGIN_TRY { /* mute stack trace on expected failure */ + JSVERIFY( FAIL, + H5FDread(file_shakespeare, + H5FD_MEM_DRAW, + H5P_DEFAULT, + 1200699, + 102, + buffer), + "cannot read before eoa is set" ) + } H5E_END_TRY; + JSVERIFY_STR( "", buffer, "buffer should remain untouched" ) + + /************ + * TEARDOWN * + ************/ + + FAIL_IF( FAIL == H5FDclose(file_shakespeare) ) + file_shakespeare = NULL; + + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + fapl_id = -1; + + PASSED(); + return 0; + +error: + /*********** + * CLEANUP * + ***********/ + + if (file_shakespeare) { (void)H5FDclose(file_shakespeare); } + if (fapl_id >= 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + } + + return 1; + +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_H5FDread_without_eoa_set_fails */ + + + +/*--------------------------------------------------------------------------- + * + * Function: test_read() + * + * Purpose: + * + * Return: + * + * PASSED : 0 + * FAILED : 1 + * + * Programmer: Jacob Smith + * 2017-11-06 + * + *--------------------------------------------------------------------------- + */ +static int +test_read(void) +{ + /********************* + * test-local macros * + *********************/ + + /************************* + * test-local structures * + *************************/ + struct testcase { + const char *message; /* purpose of test case */ + haddr_t eoa_set; /* set file EOA to this prior to read */ + size_t addr; /* offset of read in file */ + size_t len; /* length of read in file */ + herr_t success; /* expected return value of read function */ + const char *expected; /* expected contents of buffer; failure ignores */ + }; + + /************************ + * test-local variables * + ************************/ +#ifdef H5_HAVE_ROS3_VFD + struct testcase cases[] = { + { "successful range-get", + 6464, + 5691, + 32, /* fancy quotes are three bytes each(?) */ + SUCCEED, + "Quoth the Raven “Nevermore.”", + }, + { "read past EOA fails (EOA < EOF < addr)", + 3000, + 4000, + 100, + FAIL, + NULL, + }, + { "read overlapping EOA fails (EOA < addr < EOF < (addr+len))", + 3000, + 8000, + 100, + FAIL, + NULL, + }, + { "read past EOA/EOF fails ((EOA==EOF) < addr)", + 6464, + 7000, + 100, + FAIL, + NULL, + }, + { "read overlapping EOA/EOF fails (addr < (EOA==EOF) < (addr+len))", + 6464, + 6400, + 100, + FAIL, + NULL, + }, + { "read between EOF and EOA fails (EOF < addr < (addr+len) < EOA)", + 8000, + 7000, + 100, + FAIL, + NULL, + }, + }; + unsigned testcase_count = 6; + unsigned test_i = 0; + struct testcase test; + herr_t open_return = FAIL; + char buffer[S3_TEST_MAX_URL_SIZE]; + unsigned int i = 0; + H5FD_t *file_raven = NULL; + hid_t fapl_id = -1; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("ROS3 VFD read/range-gets"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD not enabled"); + fflush(stdout); + return 0; +#else + if (s3_test_credentials_loaded == 0) { + SKIPPED(); + puts(" s3 credentials are not loaded"); + fflush(stdout); + return 0; + } + + if (FALSE == s3_test_bucket_defined) { + SKIPPED(); + puts(" environment variable HDF5_ROS3_TEST_BUCKET_URL not defined"); + fflush(stdout); + return 0; + } + + /********* + * SETUP * + *********/ + + /* create ROS3 fapl + */ + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( fapl_id < 0 ) + FAIL_IF( FAIL == H5Pset_fapl_ros3(fapl_id, &restricted_access_fa) ) + + /* open file + */ + file_raven = H5FDopen( /* will open with "authenticating" fapl */ + url_text_public, /* TODO: check return state: anon access of restricted says OK? (not NULL) */ + H5F_ACC_RDONLY, + fapl_id, + HADDR_UNDEF); /* Demonstrate success with "automatic" value */ + FAIL_IF( NULL == file_raven ) + + JSVERIFY( 6464, H5FDget_eof(file_raven, H5FD_MEM_DEFAULT), NULL ) + + /********* + * TESTS * + *********/ + + for (test_i = 0; test_i < testcase_count; test_i++) { + + /* -------------- * + * per-test setup * + * -------------- */ + + test = cases[test_i]; + open_return = FAIL; + + FAIL_IF( S3_TEST_MAX_URL_SIZE < test.len ) /* buffer too small! */ + + FAIL_IF( FAIL == + H5FD_set_eoa( file_raven, H5FD_MEM_DEFAULT, test.eoa_set) ) + + for (i = 0; i < S3_TEST_MAX_URL_SIZE; i++) /* zero buffer contents */ + buffer[i] = 0; + + /* ------------ * + * conduct test * + * ------------ */ + + H5E_BEGIN_TRY { + open_return = H5FDread( + file_raven, + H5FD_MEM_DRAW, + H5P_DEFAULT, + test.addr, + test.len, + buffer); + } H5E_END_TRY; + + JSVERIFY( test.success, + open_return, + test.message ) + if (open_return == SUCCEED) + JSVERIFY_STR( test.expected, buffer, NULL ) + + } /* for each testcase */ + + /************ + * TEARDOWN * + ************/ + + FAIL_IF( FAIL == H5FDclose(file_raven) ) + file_raven = NULL; + + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + fapl_id = -1; + + PASSED(); + return 0; + +error: + /*********** + * CLEANUP * + ***********/ + + if (file_raven) + (void)H5FDclose(file_raven); + if (fapl_id >= 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + } + + return 1; + +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_read */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_noops_and_autofails() + * + * Purpose: + * + * Demonstrate the unavailable and do-nothing routines unique to + * Read-Only VFD. + * + * Return: + * + * PASSED : 0 + * FAILED : 1 + * + * Programmer: Jacob Smith + * 2017-11-06 + * + *--------------------------------------------------------------------------- + */ +static int +test_noops_and_autofails(void) +{ + /********************* + * test-local macros * + *********************/ + + /************************* + * test-local structures * + *************************/ + + /************************ + * test-local variables * + ************************/ + +#ifdef H5_HAVE_ROS3_VFD + hbool_t curl_ready = FALSE; + hid_t fapl_id = -1; + H5FD_t *file = NULL; + const char data[36] = "The Force shall be with you, always"; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("ROS3 VFD always-fail and no-op routines"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD not enabled"); + fflush(stdout); + return 0; +#else + + if (FALSE == s3_test_bucket_defined) { + SKIPPED(); + puts(" environment variable HDF5_ROS3_TEST_BUCKET_URL not defined"); + fflush(stdout); + return 0; + } + + /********* + * SETUP * + *********/ + + FAIL_IF( CURLE_OK != curl_global_init(CURL_GLOBAL_DEFAULT) ) + curl_ready = TRUE; + + /* create ROS3 fapl + */ + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( fapl_id < 0 ) + JSVERIFY( SUCCEED, H5Pset_fapl_ros3(fapl_id, &anonymous_fa), NULL ) + + /* open file + */ + file = H5FDopen( + url_text_public, + H5F_ACC_RDONLY, + fapl_id, + HADDR_UNDEF); + FAIL_IF( NULL == file ) + + /********* + * TESTS * + *********/ + + /* auto-fail calls to write and truncate + */ + H5E_BEGIN_TRY { + JSVERIFY( FAIL, + H5FDwrite(file, H5FD_MEM_DRAW, H5P_DEFAULT, 1000, 35, data), + "write must fail" ) + } H5E_END_TRY; + + H5E_BEGIN_TRY { + JSVERIFY( FAIL, + H5FDtruncate(file, H5P_DEFAULT, FALSE), + "truncate must fail" ) + } H5E_END_TRY; + + H5E_BEGIN_TRY { + JSVERIFY( FAIL, + H5FDtruncate(file, H5P_DEFAULT, TRUE), + "truncate must fail (closing)" ) + } H5E_END_TRY; + + /* no-op calls to `lock()` and `unlock()` + */ + JSVERIFY( SUCCEED, + H5FDlock(file, TRUE), + "lock always succeeds; has no effect" ) + JSVERIFY( SUCCEED, + H5FDlock(file, FALSE), + NULL ) + JSVERIFY( SUCCEED, + H5FDunlock(file), + NULL ) + /* Lock/unlock with null file or similar error crashes tests. + * HDassert in calling heirarchy, `H5FD[un]lock()` and `H5FD_[un]lock()` + */ + + /************ + * TEARDOWN * + ************/ + + FAIL_IF( FAIL == H5FDclose(file) ) + file = NULL; + + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + fapl_id = -1; + + curl_global_cleanup(); + curl_ready = FALSE; + + PASSED(); + return 0; + +error: + /*********** + * CLEANUP * + ***********/ + + if (fapl_id >= 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + } + if (file) { (void)H5FDclose(file); } + if (curl_ready == TRUE) { curl_global_cleanup(); } + + return 1; + +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_noops_and_autofails*/ + + +/*--------------------------------------------------------------------------- + * + * Function: test_cmp() + * + * Purpose: + * + * Verify "file comparison" behavior. + * + * Return: + * + * PASSED : 0 + * FAILED : 1 + * + * Programmer: Jacob Smith + * 2017-11-06 + * + *--------------------------------------------------------------------------- + */ +static int +test_cmp(void) +{ + /********************* + * test-local macros * + *********************/ + + /************************* + * test-local structures * + *************************/ + + /************************ + * test-local variables * + ************************/ + +#ifdef H5_HAVE_ROS3_VFD + H5FD_t *fd_raven = NULL; + H5FD_t *fd_shakes = NULL; + H5FD_t *fd_raven_2 = NULL; + hbool_t curl_ready = FALSE; + hid_t fapl_id = -1; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("ROS3 cmp (comparison)"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD not enabled"); + fflush(stdout); + return 0; +#else + if (s3_test_credentials_loaded == 0) { + SKIPPED(); + puts(" s3 credentials are not loaded"); + fflush(stdout); + return 0; + } + + if (FALSE == s3_test_bucket_defined) { + SKIPPED(); + puts(" environment variable HDF5_ROS3_TEST_BUCKET_URL not defined"); + fflush(stdout); + return 0; + } + + /********* + * SETUP * + *********/ + + FAIL_IF( CURLE_OK != curl_global_init(CURL_GLOBAL_DEFAULT) ) + curl_ready = TRUE; + + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( 0 > fapl_id ) + JSVERIFY( SUCCEED, H5Pset_fapl_ros3(fapl_id, &restricted_access_fa), NULL ) + + fd_raven = H5FDopen( + url_text_public, + H5F_ACC_RDONLY, + fapl_id, + HADDR_UNDEF); + FAIL_IF( NULL == fd_raven ) + + fd_shakes = H5FDopen( + url_text_restricted, + H5F_ACC_RDONLY, + fapl_id, + HADDR_UNDEF); + FAIL_IF( NULL == fd_shakes ) + + fd_raven_2 = H5FDopen( + url_text_public, + H5F_ACC_RDONLY, + fapl_id, + HADDR_UNDEF); + FAIL_IF( NULL == fd_raven_2 ) + + /********* + * TESTS * + *********/ + + JSVERIFY( 0, H5FDcmp(fd_raven, fd_raven_2), NULL ) + JSVERIFY( -1, H5FDcmp(fd_raven, fd_shakes), NULL ) + JSVERIFY( -1, H5FDcmp(fd_shakes, fd_raven_2), NULL ) + + /************ + * TEARDOWN * + ************/ + + FAIL_IF( FAIL == H5FDclose(fd_raven) ) + fd_raven = NULL; + FAIL_IF( FAIL == H5FDclose(fd_shakes) ) + fd_shakes = NULL; + FAIL_IF( FAIL == H5FDclose(fd_raven_2) ) + fd_raven_2 = NULL; + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + fapl_id = -1; + + curl_global_cleanup(); + curl_ready = FALSE; + + PASSED(); + return 0; + +error: + /*********** + * CLEANUP * + ***********/ + + if (fd_raven != NULL) (void)H5FDclose(fd_raven); + if (fd_raven_2 != NULL) (void)H5FDclose(fd_raven_2); + if (fd_shakes != NULL) (void)H5FDclose(fd_shakes); + if (TRUE == curl_ready) curl_global_cleanup(); + if (fapl_id >= 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + } + + return 1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_cmp */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_H5F_integration() + * + * Purpose: + * + * Demonstrate S3 file-open through H5F API. + * + * Return: + * + * PASSED : 0 + * FAILED : 1 + * + * Programmer: Jacob Smith + * 2017-11-07 + * + *--------------------------------------------------------------------------- + */ +static int +test_H5F_integration(void) +{ + /********************* + * test-local macros * + *********************/ + + /************************* + * test-local structures * + *************************/ + + /************************ + * test-local variables * + ************************/ + +#ifdef H5_HAVE_ROS3_VFD + hid_t file = -1; + hid_t fapl_id = -1; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("S3 file access through HD5F library (H5F API)"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD not enabled"); + fflush(stdout); + return 0; +#else + if (s3_test_credentials_loaded == 0) { + SKIPPED(); + puts(" s3 credentials are not loaded"); + fflush(stdout); + return 0; + } + + if (FALSE == s3_test_bucket_defined) { + SKIPPED(); + puts(" environment variable HDF5_ROS3_TEST_BUCKET_URL not defined"); + fflush(stdout); + return 0; + } + + /********* + * SETUP * + *********/ + + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( 0 > fapl_id ) + FAIL_IF( FAIL == H5Pset_fapl_ros3(fapl_id, &restricted_access_fa) ) + + /********* + * TESTS * + *********/ + + /* Read-Write Open access is not allowed with this file driver. + */ + H5E_BEGIN_TRY { + FAIL_IF( 0 <= H5Fopen( + url_h5_public, + H5F_ACC_RDWR, + fapl_id) ) + } H5E_END_TRY; + + /* H5Fcreate() is not allowed with this file driver. + */ + H5E_BEGIN_TRY { + FAIL_IF( 0 <= H5Fcreate( + url_missing, + H5F_ACC_RDONLY, + H5P_DEFAULT, + fapl_id) ) + } H5E_END_TRY; + + /* Successful open. + */ + file = H5Fopen( + url_h5_public, + H5F_ACC_RDONLY, + fapl_id); + FAIL_IF( file < 0 ) + + /************ + * TEARDOWN * + ************/ + + FAIL_IF( FAIL == H5Fclose(file) ) + file = -1; + + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + fapl_id = -1; + + PASSED(); + return 0; + +error: + /*********** + * CLEANUP * + ***********/ +HDprintf("\nerror!"); fflush(stdout); + + if (fapl_id >= 0) { + H5E_BEGIN_TRY { + (void)H5Pclose(fapl_id); + } H5E_END_TRY; + } + if (file > 0) + (void)H5Fclose(file); + + return 1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_H5F_integration */ + + + + +/*------------------------------------------------------------------------- + * + * Function: main + * + * Purpose: Tests the basic features of Virtual File Drivers + * + * Return: Success: 0 + * Failure: 1 + * + * Programmer: Jacob Smith + * 2017-10-23 + * + *------------------------------------------------------------------------- + */ +int +main(void) +{ + int nerrors = 0; + +#ifdef H5_HAVE_ROS3_VFD + const char *bucket_url_env = NULL; + + /************************ + * initialize test urls * + ************************/ + + bucket_url_env = HDgetenv("HDF5_ROS3_TEST_BUCKET_URL"); + if (bucket_url_env == NULL || bucket_url_env[0] == '\0') { + HDprintf("WARNING: S3 bucket url is not defined in enviornment " \ + "variable 'HDF5_ROS3_TEST_BUCKET_URL'!\n"); + } else { + HDstrncpy(s3_test_bucket_url, bucket_url_env, S3_TEST_MAX_URL_SIZE); + s3_test_bucket_defined = TRUE; + } + + if (S3_TEST_MAX_URL_SIZE < snprintf( + url_text_restricted, + (size_t)S3_TEST_MAX_URL_SIZE, + "%s/%s", + (const char *)s3_test_bucket_url, + (const char *)S3_TEST_RESOURCE_TEXT_RESTRICTED)) + { + HDprintf("* ros3 setup failed (text_restricted) ! *\n"); + return 1; + } + if (S3_TEST_MAX_URL_SIZE < HDsnprintf( + url_text_public, + (size_t)S3_TEST_MAX_URL_SIZE, + "%s/%s", + (const char *)s3_test_bucket_url, + (const char *)S3_TEST_RESOURCE_TEXT_PUBLIC)) + { + HDprintf("* ros3 setup failed (text_public) ! *\n"); + return 1; + } + if (S3_TEST_MAX_URL_SIZE < HDsnprintf( + url_h5_public, + (size_t)S3_TEST_MAX_URL_SIZE, + "%s/%s", + (const char *)s3_test_bucket_url, + (const char *)S3_TEST_RESOURCE_H5_PUBLIC)) + { + HDprintf("* ros3 setup failed (h5_public) ! *\n"); + return 1; + } + if (S3_TEST_MAX_URL_SIZE < HDsnprintf( + url_missing, + S3_TEST_MAX_URL_SIZE, + "%s/%s", + (const char *)s3_test_bucket_url, + (const char *)S3_TEST_RESOURCE_MISSING)) + { + HDprintf("* ros3 setup failed (missing) ! *\n"); + return 1; + } + + /************************************** + * load credentials and prepare fapls * + **************************************/ + + /* "clear" profile data strings */ + s3_test_aws_access_key_id[0] = '\0'; + s3_test_aws_secret_access_key[0] = '\0'; + s3_test_aws_region[0] = '\0'; + + /* attempt to load test credentials + * if unable, certain tests will be skipped + */ + if (SUCCEED == H5FD_s3comms_load_aws_profile( + S3_TEST_PROFILE_NAME, + s3_test_aws_access_key_id, + s3_test_aws_secret_access_key, + s3_test_aws_region)) + { + s3_test_credentials_loaded = 1; + HDstrncpy(restricted_access_fa.aws_region, + (const char *)s3_test_aws_region, + H5FD__ROS3_MAX_REGION_LEN); + HDstrncpy(restricted_access_fa.secret_id, + (const char *)s3_test_aws_access_key_id, + H5FD__ROS3_MAX_SECRET_ID_LEN); + HDstrncpy(restricted_access_fa.secret_key, + (const char *)s3_test_aws_secret_access_key, + H5FD__ROS3_MAX_SECRET_KEY_LEN); + } +#endif /* H5_HAVE_ROS3_VFD */ + + /****************** + * commence tests * + ******************/ + + h5_reset(); + + HDprintf("Testing ros3 VFD functionality.\n"); + + nerrors += test_fapl_config_validation(); + nerrors += test_ros3_fapl(); + nerrors += test_vfd_open(); + nerrors += test_eof_eoa(); + nerrors += test_H5FDread_without_eoa_set_fails(); + nerrors += test_read(); + nerrors += test_noops_and_autofails(); + nerrors += test_cmp(); + nerrors += test_H5F_integration(); + + if (nerrors > 0) { + HDprintf("***** %d ros3 TEST%s FAILED! *****\n", + nerrors, + nerrors > 1 ? "S" : ""); + nerrors = 1; + } else { + HDprintf("All ros3 tests passed.\n"); + } + return nerrors; /* 0 if no errors, 1 if any errors */ + +} /* main() */ + + diff --git a/test/s3comms.c b/test/s3comms.c new file mode 100644 index 0000000..71d93c2 --- /dev/null +++ b/test/s3comms.c @@ -0,0 +1,2813 @@ +/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * Read-Only S3 Virtual File Driver (VFD) * + * Copyright (c) 2017-2018, The HDF Group. * + * * + * All rights reserved. * + * * + * NOTICE: * + * All information contained herein is, and remains, the property of The HDF * + * Group. The intellectual and technical concepts contained herein are * + * proprietary to The HDF Group. Dissemination of this information or * + * reproduction of this material is strictly forbidden unless prior written * + * permission is obtained from The HDF Group. * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ + +/* + * Purpose: Unit tests for the S3 Communications (s3comms) module. + * + * Programmer: Jacob Smith <jake.smith@hdfgroup.org> + * 2017-10-11 + */ + +#include "h5test.h" +#include "H5FDs3comms.h" +#include "H5MMprivate.h" /* memory management */ + +/***************************************************************************** + * + * FILE-LOCAL TESTING MACROS + * + * Purpose: + * + * 1) Upon test failure, goto-jump to single-location teardown in test + * function. E.g., `error:` (consistency with HDF corpus) or + * `failed:` (reflects purpose). + * >>> using "error", in part because `H5E_BEGIN_TRY` expects it. + * 2) Increase clarity and reduce overhead found with `TEST_ERROR`. + * e.g., "if(somefunction(arg, arg2) < 0) TEST_ERROR:" + * requires reading of entire line to know whether this if/call is + * part of the test setup, test operation, or a test unto itself. + * 3) Provide testing macros with optional user-supplied failure message; + * if not supplied (NULL), generate comparison output in the spirit of + * test-driven development. E.g., "expected 5 but was -3" + * User messages clarify test's purpose in code, encouraging description + * without relying on comments. + * 4) Configurable expected-actual order in generated comparison strings. + * Some prefer `VERIFY(expected, actual)`, others + * `VERIFY(actual, expected)`. Provide preprocessor ifdef switch + * to satifsy both parties, assuming one paradigm per test file. + * (One could #undef and redefine the flag through the file as desired, + * but _why_.) + * Provided as courtesy, per consideration for inclusion in the library + * proper. + * + * Macros: + * + * JSVERIFY_EXP_ACT - ifdef flag, configures comparison order + * FAIL_IF() - check condition + * FAIL_UNLESS() - check _not_ condition + * JSVERIFY() - long-int equality check; prints reason/comparison + * JSVERIFY_NOT() - long-int inequality check; prints + * JSVERIFY_STR() - string equality check; prints + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *****************************************************************************/ + + +/*---------------------------------------------------------------------------- + * + * ifdef flag: JSVERIFY_EXP_ACT + * + * JSVERIFY macros accept arguments as (EXPECTED, ACTUAL[, reason]) + * default, if this is undefined, is (ACTUAL, EXPECTED[, reason]) + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_EXP_ACT 1L + + +/*---------------------------------------------------------------------------- + * + * Macro: JSFAILED_AT() + * + * Purpose: + * + * Preface a test failure by printing "*FAILED*" and location to stdout + * Similar to `H5_FAILED(); AT();` from h5test.h + * + * *FAILED* at somefile.c:12 in function_name()... + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSFAILED_AT() { \ + HDprintf("*FAILED* at %s:%d in %s()...\n", __FILE__, __LINE__, FUNC); \ +} + + +/*---------------------------------------------------------------------------- + * + * Macro: FAIL_IF() + * + * Purpose: + * + * Make tests more accessible and less cluttered than + * `if (thing == otherthing()) TEST_ERROR` + * paradigm. + * + * The following lines are roughly equivalent: + * + * `if (myfunc() < 0) TEST_ERROR;` (as seen elsewhere in HDF tests) + * `FAIL_IF(myfunc() < 0)` + * + * Prints a generic "FAILED AT" line to stdout and jumps to `error`, + * similar to `TEST_ERROR` in h5test.h + * + * Programmer: Jacob Smith + * 2017-10-23 + * + *---------------------------------------------------------------------------- + */ +#define FAIL_IF(condition) \ +if (condition) { \ + JSFAILED_AT() \ + goto error; \ +} + + +/*---------------------------------------------------------------------------- + * + * Macro: FAIL_UNLESS() + * + * Purpose: + * + * TEST_ERROR wrapper to reduce cognitive overhead from "negative tests", + * e.g., "a != b". + * + * Opposite of FAIL_IF; fails if the given condition is _not_ true. + * + * `FAIL_IF( 5 != my_op() )` + * is equivalent to + * `FAIL_UNLESS( 5 == my_op() )` + * However, `JSVERIFY(5, my_op(), "bad return")` may be even clearer. + * (see JSVERIFY) + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define FAIL_UNLESS(condition) \ +if (!(condition)) { \ + JSFAILED_AT() \ + goto error; \ +} + + +/*---------------------------------------------------------------------------- + * + * Macro: JSERR_LONG() + * + * Purpose: + * + * Print an failure message for long-int arguments. + * ERROR-AT printed first. + * If `reason` is given, it is printed on own line and newlined after + * else, prints "expected/actual" aligned on own lines. + * + * *FAILED* at myfile.c:488 in somefunc()... + * forest must be made of trees. + * + * or + * + * *FAILED* at myfile.c:488 in somefunc()... + * ! Expected 425 + * ! Actual 3 + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSERR_LONG(expected, actual, reason) { \ + JSFAILED_AT() \ + if (reason!= NULL) { \ + HDprintf("%s\n", (reason)); \ + } else { \ + HDprintf(" ! Expected %ld\n ! Actual %ld\n", \ + (long)(expected), (long)(actual)); \ + } \ +} + + +/*---------------------------------------------------------------------------- + * + * Macro: JSERR_STR() + * + * Purpose: + * + * Print an failure message for string arguments. + * ERROR-AT printed first. + * If `reason` is given, it is printed on own line and newlined after + * else, prints "expected/actual" aligned on own lines. + * + * *FAILED* at myfile.c:421 in myfunc()... + * Blue and Red strings don't match! + * + * or + * + * *FAILED* at myfile.c:421 in myfunc()... + * !!! Expected: + * this is my expected + * string + * !!! Actual: + * not what I expected at all + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSERR_STR(expected, actual, reason) { \ + JSFAILED_AT() \ + if ((reason) != NULL) { \ + HDprintf("%s\n", (reason)); \ + } else { \ + HDprintf("!!! Expected:\n%s\n!!!Actual:\n%s\n", \ + (expected), (actual)); \ + } \ +} + +#ifdef JSVERIFY_EXP_ACT +/* VERIFY rountines with paramter order (<expected>, <actual> [, <msg> ]) + */ + + +/*---------------------------------------------------------------------------- + * + * Macro: JSVERIFY() + * + * Purpose: + * + * Verify that two long integers are equal. + * If unequal, print failure message + * (with `reason`, if not NULL; expected/actual if NULL) + * and jump to `error` at end of function + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY(expected, actual, reason) \ +if ((long)(actual) != (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)) \ + goto error; \ +} /* JSVERIFY */ + +#if 0 /* UNUSED */ + +/*---------------------------------------------------------------------------- + * + * Macro: JSVERIFY_NOT() + * + * Purpose: + * + * Verify that two long integers are _not_ equal. + * If equal, print failure message + * (with `reason`, if not NULL; expected/actual if NULL) + * and jump to `error` at end of function + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_NOT(expected, actual, reason) \ +if ((long)(actual) == (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)) \ + goto error; \ +} /* JSVERIFY_NOT */ +#endif /* JSVERIFY_NOT unused */ + + +/*---------------------------------------------------------------------------- + * + * Macro: JSVERIFY_STR() + * + * Purpose: + * + * Verify that two strings are equal. + * If unequal, print failure message + * (with `reason`, if not NULL; expected/actual if NULL) + * and jump to `error` at end of function + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_STR(expected, actual, reason) \ +if (strcmp((actual), (expected)) != 0) { \ + JSERR_STR((expected), (actual), (reason)); \ + goto error; \ +} /* JSVERIFY_STR */ + + +#else +/* JSVERIFY_EXP_ACT not defined + * + * Repeats macros above, but with actual/expected parameters reversed. + */ + + +/*---------------------------------------------------------------------------- + * Macro: JSVERIFY() + * See: JSVERIFY documentation above. + * Programmer: Jacob Smith + * 2017-10-14 + *---------------------------------------------------------------------------- + */ +#define JSVERIFY(actual, expected, reason) \ +if ((long)(actual) != (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)); \ + goto error; \ +} /* JSVERIFY */ + +#if 0 /* UNUSED */ + +/*---------------------------------------------------------------------------- + * Macro: JSVERIFY_NOT() + * See: JSVERIFY_NOT documentation above. + * Programmer: Jacob Smith + * 2017-10-14 + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_NOT(actual, expected, reason) \ +if ((long)(actual) == (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)) \ + goto error; \ +} /* JSVERIFY_NOT */ +#endif /* JSVERIFY_NOT unused */ + + +/*---------------------------------------------------------------------------- + * Macro: JSVERIFY_STR() + * See: JSVERIFY_STR documentation above. + * Programmer: Jacob Smith + * 2017-10-14 + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_STR(actual, expected, reason) \ +if (strcmp((actual), (expected)) != 0) { \ + JSERR_STR((expected), (actual), (reason)); \ + goto error; \ +} /* JSVERIFY_STR */ + +#endif /* ifdef/else JSVERIFY_EXP_ACT */ + + +#ifdef H5_HAVE_ROS3_VFD + +#define S3_TEST_PROFILE_NAME "ros3_vfd_test" + +#define S3_TEST_RESOURCE_TEXT_RESTRICTED "t8.shakespeare.txt" +#define S3_TEST_RESOURCE_TEXT_PUBLIC "Poe_Raven.txt" +#define S3_TEST_RESOURCE_MISSING "missing.csv" + +#define S3_TEST_RUN_TIMEOUT 0 /* run tests that might hang */ +#define S3_TEST_MAX_URL_SIZE 256 /* char array size */ + +/* Global variables for aws test profile. + * An attempt is made to read ~/.aws/credentials and ~/.aws/config upon test + * startup -- if unable to open either file or cannot load region, id, and key, + * tests connecting with S3 will not be run + */ +static int s3_test_credentials_loaded = 0; +static char s3_test_aws_region[16] = ""; +static char s3_test_aws_access_key_id[64] = ""; +static char s3_test_aws_secret_access_key[128] = ""; +static char s3_test_bucket_url[S3_TEST_MAX_URL_SIZE] = ""; +static hbool_t s3_test_bucket_defined = FALSE; + +#endif /* H5_HAVE_ROS3_VFD */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_macro_format_credential() + * + * Purpose: + * + * Demonstrate that the macro `S3COMMS_FORMAT_CREDENTIAL` + * performs as expected. + * + * Programmer: Jacob Smith + * 2017-09-19 + * + *---------------------------------------------------------------------------- + */ +static herr_t +test_macro_format_credential(void) +{ + /************************ + * test-local variables * + ************************/ + +#ifdef H5_HAVE_ROS3_VFD + char dest[256]; + const char access[] = "AKIAIOSFODNN7EXAMPLE"; + const char date[] = "20130524"; + const char region[] = "us-east-1"; + const char service[] = "s3"; + const char expected[] = + "AKIAIOSFODNN7EXAMPLE/20130524/us-east-1/s3/aws4_request"; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("test_macro_format_credential"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD is not enabled"); + fflush(stdout); + return 0; +#else + FAIL_IF( S3COMMS_MAX_CREDENTIAL_SIZE < + S3COMMS_FORMAT_CREDENTIAL(dest, access, date, region, service) ) + + JSVERIFY_STR( expected, dest, NULL ) + + PASSED(); + return 0; +#endif /* H5_HAVE_ROS3_VFD */ + +error: + return -1; + +} /* test_macro_format_credential */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_aws_canonical_request() + * + * Purpose: + * + * Demonstrate the construction of a Canoncial Request (and Signed Headers) + * + * Elided / not yet implemented: + * Query strings + * request "body" + * + * Programmer: Jacob Smith + * 2017-10-04 + * + *--------------------------------------------------------------------------- + */ +static herr_t +test_aws_canonical_request(void) +{ + /************************* + * test-local structures * + *************************/ + +#ifdef H5_HAVE_ROS3_VFD + struct header { + const char *name; + const char *value; + }; + + struct testcase { + const char *exp_request; + const char *exp_headers; + const char *verb; + const char *resource; + unsigned int listsize; + struct header list[5]; + }; + + /************************ + * test-local variables * + ************************/ + + struct testcase cases[] = { + { "GET\n/some/path.file\n\nhost:somebucket.someserver.somedomain\nrange:bytes=150-244\n\nhost;range\ne3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "host;range", + "GET", + "/some/path.file", + 2, + { {"Range", "bytes=150-244"}, + {"Host", "somebucket.someserver.somedomain"}, + }, + }, + { "HEAD\n/bucketpath/myfile.dat\n\nhost:place.domain\nx-amz-content-sha256:e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855\nx-amz-date:19411207T150803Z\n\nhost;x-amz-content-sha256;x-amz-date\ne3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "host;x-amz-content-sha256;x-amz-date", + "HEAD", + "/bucketpath/myfile.dat", + 3, + { {"x-amz-content-sha256", "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855"}, + {"host", "place.domain"}, + {"x-amz-date", "19411207T150803Z"}, + } + }, + { "PUT\n/\n\n\n\ne3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "", + "PUT", + "/", + 0, + {{"",""},}, /* unused; satisfies compiler */ + }, + }; /* struct testcase cases[] */ + struct testcase *C = NULL; + char cr_dest[512]; /* canonical request */ + hrb_t *hrb = NULL; /* http request buffer object */ + unsigned int i = 0; /* looping/indexing */ + unsigned int j = 0; /* looping/indexing */ + hrb_node_t *node = NULL; /* http headers list pointer */ + unsigned int n_cases = 3; + char sh_dest[64]; /* signed headers */ +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("test_aws_canonical_request"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD is not enabled"); + fflush(stdout); + return 0; +#else + for (i = 0; i < n_cases; i++) { + /* pre-test bookkeeping + */ + C = &cases[i]; + for (j = 0; j < 256; j++) { cr_dest[j] = 0; } /* zero request buffer */ + for (j = 0; j < 64; j++) { sh_dest[j] = 0; } /* zero headers buffer */ + + /* create HTTP request object with given verb, resource/path + */ + hrb = H5FD_s3comms_hrb_init_request(C->verb, + C->resource, + "HTTP/1.1"); + HDassert(hrb->body == NULL); + + /* Create headers list from test case input + */ + for (j = 0; j < C->listsize; j++) { + FAIL_IF( FAIL == + H5FD_s3comms_hrb_node_set( + &node, + C->list[j].name, + C->list[j].value)); + } + + hrb->first_header = node; + + /* test + */ + JSVERIFY( SUCCEED, + H5FD_s3comms_aws_canonical_request(cr_dest, sh_dest, hrb), + " unable to compose canonical request" ) + JSVERIFY_STR( C->exp_headers, sh_dest, NULL ) + JSVERIFY_STR( C->exp_request, cr_dest, NULL ) + + /* tear-down + */ + while (node != NULL) + FAIL_IF( FAIL == + H5FD_s3comms_hrb_node_set(&node, node->name, NULL)); + HDassert(NULL == node); + FAIL_IF( FAIL == H5FD_s3comms_hrb_destroy(&hrb)); + HDassert(NULL == hrb); + + } /* for each test case */ + + /*************** + * ERROR CASES * + ***************/ + + /* malformed hrb and/or node-list + */ + JSVERIFY( FAIL, H5FD_s3comms_aws_canonical_request(cr_dest, sh_dest, NULL), + "http request object cannot be null" ) + + hrb = H5FD_s3comms_hrb_init_request("GET", "/", "HTTP/1.1"); + JSVERIFY( FAIL, H5FD_s3comms_aws_canonical_request(NULL, sh_dest, hrb), + "canonical request destination cannot be NULL" ) + + JSVERIFY( FAIL, H5FD_s3comms_aws_canonical_request(cr_dest, NULL, hrb), + "signed headers destination cannot be null" ) + + FAIL_IF( FAIL == H5FD_s3comms_hrb_destroy(&hrb) ) + HDassert( NULL == hrb ); + + PASSED(); + return 0; + +error: + + if (node != NULL) { + while (node != NULL) + (void)H5FD_s3comms_hrb_node_set(&node, node->name, NULL); + HDassert( node == NULL ); + } + if (hrb != NULL) + (void)H5FD_s3comms_hrb_destroy(&hrb); + + return -1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_aws_canonical_request */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_bytes_to_hex + * + * Purpose: + * + * Define and verify behavior of `H5FD_s3comms_bytes_to_hex()`. + * + * Return: + * + * Success: 0 + * Failure: -1 + * + * Programmer: Jacob Smith + * 2017-09-14 + * + *--------------------------------------------------------------------------- + */ +static herr_t +test_bytes_to_hex(void) +{ + /************************* + * test-local structures * + *************************/ + + struct testcase { + const char exp[17]; /* in size * 2 + 1 for null terminator */ + const unsigned char in[8]; + size_t size; + hbool_t lower; + }; + + /************************ + * test-local variables * + ************************/ + + struct testcase cases[] = { + { "52F3000C9A", + {82,243,0,12,154}, + 5, + FALSE, + }, + { "009a0cf3005200", /* lowercase alphas */ + {0,154,12,243,0,82,0}, + 7, + TRUE, + }, + { "", + {17,63,26,56}, + 0, + FALSE, /* irrelevant */ + }, + }; + int i = 0; + int n_cases = 3; + char out[17]; + int out_off = 0; + + + + TESTING("bytes-to-hex"); + + for (i = 0; i < n_cases; i++) { + for (out_off = 0; out_off < 17; out_off++) { + out[out_off] = 0; + } + + JSVERIFY( SUCCEED, + H5FD_s3comms_bytes_to_hex(out, + cases[i].in, + cases[i].size, + cases[i].lower), + NULL ) + + JSVERIFY_STR(cases[i].exp, out, NULL) + } + + /* dest cannot be null + */ + JSVERIFY( FAIL, + H5FD_s3comms_bytes_to_hex( + NULL, + (const unsigned char *)"nada", + 5, + FALSE), + "destination cannot be null" ) + + PASSED(); + return 0; + +error: + return -1; + +} /* test_bytes_to_hex */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_hrb_init_request() + * + * Purpose: + * + * Define and verify behavior of `H5FD_s3comms_hrb_init_request()` + * + * Programmer: Jacob Smith + * 2017-09-20 + * + *--------------------------------------------------------------------------- + */ +static herr_t +test_hrb_init_request(void) +{ +#ifdef H5_HAVE_ROS3_VFD + /********************* + * test-local macros * + *********************/ + + /************************* + * test-local structures * + *************************/ + + struct testcase { + const char msg[64]; + const char *verb; + const char *resource; + const char *exp_res; + const char *version; + hbool_t ret_null; + }; + + /************************ + * test-local variables * + ************************/ + + struct testcase cases[] = { + { "get HTTP request just as we provided", + "GET", + "/path/to/some/file", + "/path/to/some/file", + "HTTP/1.1", + FALSE, + }, + { "null verb substitues to GET", + NULL, + "/MYPATH/MYFILE.tiff", + "/MYPATH/MYFILE.tiff", + "HTTP/1.1", + FALSE, + }, + { "demonstrate non-GET verb", + "HEAD", + "/MYPATH/MYFILE.tiff", + "/MYPATH/MYFILE.tiff", + "HTTP/1.1", + FALSE, + }, + { "slash prepented to resource path, if necessary", + NULL, + "MYPATH/MYFILE.tiff", + "/MYPATH/MYFILE.tiff", + NULL, + FALSE, + }, + { "null resource path causes problem", + "GET", + NULL, + NULL, + NULL, + TRUE, + }, + }; + struct testcase *C = NULL; + unsigned int i = 0; + unsigned int ncases = 5; + hrb_t *req = NULL; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("hrb_init_request"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD is not enabled"); + fflush(stdout); + return 0; +#else + for (i = 0; i < ncases; i++) { + C = &cases[i]; + req = H5FD_s3comms_hrb_init_request(C->verb, + C->resource, + C->version); + if (cases[i].ret_null == TRUE) { + FAIL_IF( req != NULL ); + } else { + FAIL_IF( req == NULL ); + JSVERIFY( S3COMMS_HRB_MAGIC, req->magic, NULL ) + if (C->verb == NULL) { + JSVERIFY_STR( "GET", req->verb, NULL ) + } else { + JSVERIFY_STR( req->verb, C->verb, NULL ) + } + JSVERIFY_STR( "HTTP/1.1", req->version, NULL ) + JSVERIFY_STR( C->exp_res, req->resource, NULL ) + FAIL_IF( req->first_header != NULL ); + FAIL_IF( req->body != NULL ); + JSVERIFY( 0, req->body_len, NULL ) + JSVERIFY( SUCCEED, H5FD_s3comms_hrb_destroy(&req), + "unable to destroy hrb_t" ) + FAIL_IF( NULL != req ); /* should annull pointer as well as free */ + } + + } /* for each testcase */ + + PASSED(); + return 0; + +error: + (void)H5FD_s3comms_hrb_destroy(&req); + + return -1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_hrb_init_request */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_hrb_node_set() + * + * Purpose: + * + * Test operations on hrb_node_t structure + * + * Programmer: Jacob Smith + * 2017-09-22 + * + *--------------------------------------------------------------------------- + */ +static herr_t +test_hrb_node_set(void) +{ +#ifdef H5_HAVE_ROS3_VFD + /************************* + * test-local structures * + *************************/ + + /* bundle of name/value representing an hrb_node_t + */ + typedef struct node_mock_t { + const char *name; + const char *value; + } node_mock_t; + + /* bundle for a testcase + * + * `message` + * purpose of the testcase + * + * `delta` + * container for name and value strings to pass into node-set function + * to to modify the list. + * + * `returned` + * expected return value of node-set function + * + * `given` + * `expected` + * string arrays representing the state of the list before and after + * modification. The number of strings must be even, with each name + * paired to a value. `NULL` terminates the list, with `{NULL}` + * representing the empty list. + */ + typedef struct testcase { + const char *message; + node_mock_t delta; + herr_t returned; + const char *given[11]; /* name/value pairs in array; NULL sentinel */ + const char *expected[11]; + } testcase; + + /************************ + * test-local variables * + ************************/ + + testcase cases[] = { + { "cannot remove node from null list", + { "Host", NULL }, + FAIL, + {NULL}, + {NULL}, + }, + { "cannot create list with NULL field name", + { NULL, "somevalue" }, + FAIL, + {NULL}, + {NULL}, + }, + { "create a new list", + { "Host", "somevalue" }, + SUCCEED, + {NULL}, + { "Host", "somevalue", + NULL, + }, + }, + { "insert new node at head list", + { "Host", "somevalue" }, + SUCCEED, + { "Range", "bytes=20-40", + NULL, + }, + { "Host", "somevalue", + "Range", "bytes=20-40", + NULL, + }, + }, + { "append new node at list end", + { "x-amz-date", "somevalue" }, + SUCCEED, + { "Range", "bytes=20-40", + NULL, + }, + { "Range", "bytes=20-40", + "x-amz-date", "somevalue", + NULL, + }, + }, + { "insert new node inside list", + { "Intermediary", "somevalue" }, + SUCCEED, + { "Host", "somehost" , + "Range", "bytes=20-40", + NULL, + }, + { "Host", "somehost", + "Intermediary", "somevalue", + "Range", "bytes=20-40", + NULL, + }, + }, + { "modify node", + { "Range", "bytes=40-80" }, + SUCCEED, + { "Host", "somehost", + "Range", "bytes=20-40", + NULL, + }, + { "Host", "somehost", + "Range", "bytes=40-80", + NULL, + }, + }, + { "modify node with new case", + { "RANGE", "bytes=40-80" }, + SUCCEED, + { "Host", "somehost", + "Range", "bytes=20-40", + NULL, + }, + { "Host", "somehost", + "RANGE", "bytes=40-80", + NULL, + }, + }, + { "cannot add node with no name", + { NULL, "bytes=40-80" }, + FAIL, + { "Host", "somehost", + NULL, + }, + { "Host", "somehost", + NULL, + }, + }, + { "add node with 'empty' name", + { "", "bytes=40-80" }, + SUCCEED, + { "Host", "somehost", + NULL, + }, + { "", "bytes=40-80", + "Host", "somehost", + NULL, + }, + }, + { "remove node from end of list", + { "Host", NULL }, + SUCCEED, + { "Date", "Thr, 25 Jan 2018", + "Host", "somehost", + NULL, + }, + { "Date", "Thr, 25 Jan 2018", + NULL, + }, + }, + { "remove node from middle of list", + { "Host", NULL }, + SUCCEED, + { "Date", "Thr, 25 Jan 2018", + "Host", "somehost", + "Range", "bytes=20-40", + NULL, + }, + { "Date", "Thr, 25 Jan 2018", + "Range", "bytes=20-40", + NULL, + }, + }, + { "remove node from start of list", + { "Date", NULL }, + SUCCEED, + { "Date", "Thr, 25 Jan 2018", + "Host", "somehost", + "Range", "bytes=20-40", + NULL, + }, + { "Host", "somehost", + "Range", "bytes=20-40", + NULL, + }, + }, + { "remove only node in list", + { "Date", NULL }, + SUCCEED, + { "Date", "Thr, 25 Jan 2018", + NULL, + }, + { NULL, + }, + }, + { "attempt to remove absent node fails", + { "Host", NULL }, + FAIL, + { "Date", "Thr, 25 Jan 2018", + "Range", "bytes=20-40", + NULL, + }, + { "Date", "Thr, 25 Jan 2018", + "Range", "bytes=20-40", + NULL, + }, + }, + { "removal is case-insensitive", + { "hOsT", NULL }, + SUCCEED, + { "Date", "Thr, 25 Jan 2018", + "Host", "somehost", + "Range", "bytes=20-40", + NULL, + }, + { "Date", "Thr, 25 Jan 2018", + "Range", "bytes=20-40", + NULL, + }, + }, + }; + unsigned testcases_count = 16; + unsigned test_i = 0; + + hrb_node_t *list = NULL; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("test_hrb_node_t"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD is not enabled"); + fflush(stdout); + return 0; +#else + for (test_i = 0; test_i < testcases_count; test_i++) { + const hrb_node_t *node = NULL; + const testcase *test = &(cases[test_i]); + unsigned mock_i = 0; + + /********* + * SETUP * + *********/ + + for (mock_i = 0; test->given[mock_i] != NULL; mock_i += 2) { + const char *name = test->given[mock_i]; + const char *valu = test->given[mock_i+1]; + + FAIL_IF( SUCCEED != + H5FD_s3comms_hrb_node_set(&list, name, valu) ) + } + /******** + * TEST * + ********/ + + /* perform modification on list + */ + JSVERIFY( test->returned, + H5FD_s3comms_hrb_node_set(&list, + test->delta.name, + test->delta.value), + test->message ) + + + /* verify resulting list + */ + node = list; + mock_i = 0; + while (test->expected[mock_i] != NULL && node != NULL) { + const char *name = test->expected[mock_i]; + const char *valu = test->expected[mock_i+1]; + + JSVERIFY_STR( name, node->name, NULL ) + JSVERIFY_STR( valu, node->value, NULL ) + + mock_i += 2; + node = node->next; + } + FAIL_IF( test->expected[mock_i] != NULL ) + FAIL_IF( node != NULL ) + + /************ + * TEARDOWN * + ************/ + + while (list != NULL) { + FAIL_IF( SUCCEED != + H5FD_s3comms_hrb_node_set(&list, list->name, NULL) ) + } + } + + PASSED(); + return 0; + +error: + while (list != NULL) + (void)H5FD_s3comms_hrb_node_set(&list, list->name, NULL); + + return -1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_hrb_node_t */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_HMAC_SHA256() + * + * Purpose: + * + * Define and verify behavior of `H5FD_s3comms_HMAC_SHA256()` + * + * Programmer: Jacob Smith + * 2017-09-19 + * + *--------------------------------------------------------------------------- + */ +static herr_t +test_HMAC_SHA256(void) +{ + + /************************* + * test-local structures * + *************************/ + +#ifdef H5_HAVE_ROS3_VFD + struct testcase { + herr_t ret; /* SUCCEED/FAIL expected from call */ + const unsigned char key[SHA256_DIGEST_LENGTH]; + size_t key_len; + const char *msg; + size_t msg_len; + const char *exp; /* not used if ret == FAIL */ + size_t dest_size; /* if 0, `dest` is not malloc'd */ + }; + + /************************ + * test-local variables * + ************************/ + + struct testcase cases[] = { + { SUCCEED, + { 0xdb, 0xb8, 0x93, 0xac, 0xc0, 0x10, 0x96, 0x49, + 0x18, 0xf1, 0xfd, 0x43, 0x3a, 0xdd, 0x87, 0xc7, + 0x0e, 0x8b, 0x0d, 0xb6, 0xbe, 0x30, 0xc1, 0xfb, + 0xea, 0xfe, 0xfa, 0x5e, 0xc6, 0xba, 0x83, 0x78, + }, + SHA256_DIGEST_LENGTH, + "AWS4-HMAC-SHA256\n20130524T000000Z\n20130524/us-east-1/s3/aws4_request\n7344ae5b7ee6c3e7e6b0fe0640412a37625d1fbfff95c48bbb2dc43964946972", + HDstrlen("AWS4-HMAC-SHA256\n20130524T000000Z\n20130524/us-east-1/s3/aws4_request\n7344ae5b7ee6c3e7e6b0fe0640412a37625d1fbfff95c48bbb2dc43964946972"), + "f0e8bdb87c964420e857bd35b5d6ed310bd44f0170aba48dd91039c6036bdb41", + SHA256_DIGEST_LENGTH * 2 + 1, /* +1 for null terminator */ + }, + { SUCCEED, + {'J','e','f','e'}, + 4, + "what do ya want for nothing?", + 28, + "5bdcc146bf60754e6a042426089575c75a003f089d2739839dec58b964ec3843", + SHA256_DIGEST_LENGTH * 2 + 1, + }, + { FAIL, + "DOESN'T MATTER", + 14, + "ALSO IRRELEVANT", + 15, + NULL, + 0, /* dest -> null, resulting in immediate error */ + }, + }; + char *dest = NULL; + int i = 0; + int n_cases = 3; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("HMAC_SHA256"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD is not enabled"); + fflush(stdout); + return 0; +#else + for (i = 0; i < n_cases; i++) { + if (cases[i].dest_size == 0) { + dest = NULL; + } else { + dest = (char *)HDmalloc(sizeof(char) * cases[i].dest_size); + HDassert(dest != NULL); + } + + JSVERIFY( cases[i].ret, + H5FD_s3comms_HMAC_SHA256( + cases[i].key, + cases[i].key_len, + cases[i].msg, + cases[i].msg_len, + dest), + cases[i].msg ); + if (cases[i].ret == SUCCEED) { +#ifdef VERBOSE + if (0 != + strncmp(cases[i].exp, + dest, + HDstrlen(cases[i].exp))) + { + /* print out how wrong things are, and then fail + */ + dest = (char *)realloc(dest, cases[i].dest_size + 1); + HDassert(dest != NULL); + dest[cases[i].dest_size] = 0; + HDfprintf(stdout, + "ERROR:\n!!! \"%s\"\n != \"%s\"\n", + cases[i].exp, + dest); + TEST_ERROR; + } +#else + /* simple pass/fail test + */ + JSVERIFY( 0, + strncmp(cases[i].exp, dest, HDstrlen(cases[i].exp)), + NULL); +#endif + } + free(dest); + } + + PASSED(); + return 0; + +error: + free(dest); + return -1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_HMAC_SHA256 */ + + +/*---------------------------------------------------------------------------- + * + * Function: test_nlowercase() + * + * Purpose: + * + * Define and verify behavior of `H5FD_s3comms_nlowercase()` + * + * Programmer: Jacob Smith + * 2017-19-18 + * + *---------------------------------------------------------------------------- + */ +static herr_t +test_nlowercase(void) +{ + /************************* + * test-local structures * + *************************/ + +#ifdef H5_HAVE_ROS3_VFD + struct testcase { + const char *in; + size_t len; + const char *exp; + }; + + /************************ + * test-local variables * + ************************/ + + /* any character after in exp on or after exp[len] is undefined. + * in this test, kept as the null character for simplicity. + */ + struct testcase cases[] = { + { "HALlEluJAh", + 6, + "hallel", + }, + { "all\0 lower", + 10, + "all\0 lower", + }, + { "to meeeeeee", + 0, + "", + }, + }; + char *dest = NULL; + int i = 0; + int n_cases = 3; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("nlowercase"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD is not enabled"); + fflush(stdout); + return 0; +#else + for (i = 0; i < n_cases; i++) { + dest = (char *)HDmalloc(sizeof(char) * 16); + + JSVERIFY( SUCCEED, + H5FD_s3comms_nlowercase(dest, + cases[i].in, + cases[i].len), + cases[i].in ) + if (cases[i].len > 0) { + JSVERIFY( 0, strncmp(dest, cases[i].exp, cases[i].len), NULL ) + } + free(dest); + } + + JSVERIFY( FAIL, + H5FD_s3comms_nlowercase(NULL, + cases[0].in, + cases[0].len), + "null distination should fail" ) + + PASSED(); + return 0; + +error: + free(dest); + return -1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_nlowercase */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_parse_url() + * + * Programmer: Jacob Smith + * 2017-11-?? + * + *--------------------------------------------------------------------------- + */ +static herr_t +test_parse_url(void) +{ + /********************* + * test-local macros * + *********************/ + + /************************* + * test-local structures * + *************************/ + +#ifdef H5_HAVE_ROS3_VFD + typedef struct { + const char *scheme; + const char *host; + const char *port; + const char *path; + const char *query; + } const_purl_t; + + struct testcase { + const char *url; + herr_t exp_ret; /* expected return; */ + /* if FAIL, `expected` is unused */ + const_purl_t expected; + const char *msg; + }; + + /************************ + * test-local variables * + ************************/ + + parsed_url_t *purl = NULL; + unsigned int i = 0; + unsigned int ncases = 15; + struct testcase cases[] = { + { NULL, + FAIL, + { NULL, NULL, NULL, NULL, NULL }, + "null url", + }, + { "", + FAIL, + { NULL, NULL, NULL, NULL, NULL }, + "empty url", + }, + { "ftp://[1000:4000:0002:2010]", + SUCCEED, + { "ftp", + "[1000:4000:0002:2010]", + NULL, + NULL, + NULL, + }, + "IPv6 ftp and empty path (root)", + }, + { "ftp://[1000:4000:0002:2010]:2040", + SUCCEED, + { "ftp", + "[1000:4000:0002:2010]", + "2040", + NULL, + NULL, + }, + "root IPv6 ftp with port", + }, + { "http://some.domain.org:9000/path/to/resource.txt", + SUCCEED, + { "http", + "some.domain.org", + "9000", + "path/to/resource.txt", + NULL, + }, + "without query", + }, + { "https://domain.me:00/file.txt?some_params unchecked", + SUCCEED, + { "https", + "domain.me", + "00", + "file.txt", + "some_params unchecked", + }, + "with query", + }, + { "ftp://domain.com/", + SUCCEED, + { "ftp", + "domain.com", + NULL, + NULL, + NULL, + }, + "explicit root w/out port", + }, + { "ftp://domain.com:1234/", + SUCCEED, + { "ftp", + "domain.com", + "1234", + NULL, + NULL, + }, + "explicit root with port", + }, + { "ftp://domain.com:1234/file?", + FAIL, + { NULL, NULL, NULL, NULL, NULL, }, + "empty query is invalid", + }, + { "ftp://:1234/file", + FAIL, + { NULL, NULL, NULL, NULL, NULL, }, + "no host", + }, + { "h&r block", + FAIL, + { NULL, NULL, NULL, NULL, NULL, }, + "no scheme (bad URL)", + }, + { "http://domain.com?a=b&d=b", + SUCCEED, + { "http", + "domain.com", + NULL, + NULL, + "a=b&d=b", + }, + "QUERY with implict PATH", + }, + { "http://[5]/path?a=b&d=b", + SUCCEED, + { "http", + "[5]", + NULL, + "path", + "a=b&d=b", + }, + "IPv6 extraction is really dumb", + }, + { "http://[1234:5678:0910:1112]:port/path", + FAIL, + { NULL, NULL, NULL, NULL, NULL, }, + "non-decimal PORT (port)", + }, + { "http://mydomain.com:01a3/path", + FAIL, + { NULL, NULL, NULL, NULL, NULL, }, + "non-decimal PORT (01a3)", + }, + }; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("url-parsing functionality"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD is not enabled"); + fflush(stdout); + return 0; +#else + /********* + * TESTS * + *********/ + + for (i = 0; i < ncases; i++) { + HDassert( purl == NULL ); + + JSVERIFY( cases[i].exp_ret, + H5FD_s3comms_parse_url(cases[i].url, &purl), + cases[i].msg ) + + if (cases[i].exp_ret == FAIL) { + /* on FAIL, `purl` should be untouched--remains NULL */ + FAIL_UNLESS( purl == NULL ) + } else { + /* on SUCCEED, `purl` should be set */ + FAIL_IF( purl == NULL ) + + if (cases[i].expected.scheme != NULL) { + FAIL_IF( NULL == purl->scheme ) + JSVERIFY_STR( cases[i].expected.scheme, + purl->scheme, + cases[i].msg ) + } else { + FAIL_UNLESS( NULL == purl->scheme ) + } + + if (cases[i].expected.host != NULL) { + FAIL_IF( NULL == purl->host ) + JSVERIFY_STR( cases[i].expected.host, + purl->host, + cases[i].msg ) + } else { + FAIL_UNLESS( NULL == purl->host ) + } + + if (cases[i].expected.port != NULL) { + FAIL_IF( NULL == purl->port ) + JSVERIFY_STR( cases[i].expected.port, + purl->port, + cases[i].msg ) + } else { + FAIL_UNLESS( NULL == purl->port ) + } + + if (cases[i].expected.path != NULL) { + FAIL_IF( NULL == purl->path ) + JSVERIFY_STR( cases[i].expected.path, + purl->path, + cases[i].msg ) + } else { + FAIL_UNLESS( NULL == purl->path ) + } + + if (cases[i].expected.query != NULL) { + FAIL_IF( NULL == purl->query ) + JSVERIFY_STR( cases[i].expected.query, + purl->query, + cases[i].msg ) + } else { + FAIL_UNLESS( NULL == purl->query ) + } + } /* if parse-url return SUCCEED/FAIL */ + + /* per-test cleanup + * well-behaved, even if `purl` is NULL + */ + FAIL_IF( FAIL == H5FD_s3comms_free_purl(purl) ) + purl = NULL; + + } /* for each testcase */ + + PASSED(); + return 0; + +error: + /*********** + * cleanup * + ***********/ + (void)H5FD_s3comms_free_purl(purl); + + return -1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_parse_url */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_percent_encode_char() + * + * Purpose: + * + * Define and verify behavior of `H5FD_s3comms_percent_encode_char()` + * + * Return: + * + * Success: 0 + * Failure: -1 + * + * Programmer: Jacob Smith + * 2017-09-14 + * + *--------------------------------------------------------------------------- + */ +static herr_t +test_percent_encode_char(void) +{ + /************************* + * test-local structures * + *************************/ + +#ifdef H5_HAVE_ROS3_VFD + struct testcase { + const char c; + const char *exp; + size_t exp_len; + }; + + /************************ + * test-local variables * + ************************/ + + struct testcase cases[] = { + {'$', "%24", 3}, /* u+0024 dollar sign */ + {' ', "%20", 3}, /* u+0020 space */ + {'^', "%5E", 3}, /* u+0094 carat */ + {'/', "%2F", 3}, /* u+002f solidus (forward slash) */ + /* {??, "%C5%8C", 6},*/ /* u+014c Latin Capital Letter O with Macron */ + /* Not included because it is multibyte "wide" character that poses */ + /* issues both in the underlying function and in being written in */ + /* this file. */ + /* {'¢', "%C2%A2", 6}, */ /* u+00a2 cent sign */ + /* above works, but complains about wide character overflow */ + /* Elide for now, until it is determined (a) unnecessary or */ + /* (b) requiring signature change to accommodate wide characters */ + {'\0', "%00", 3}, /* u+0000 null */ + }; + char dest[13]; + size_t dest_len = 0; + int i = 0; + int n_cases = 5; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("percent encode characters"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD is not enabled"); + fflush(stdout); + return 0; +#else + for (i = 0; i < n_cases; i++) { + JSVERIFY( SUCCEED, + H5FD_s3comms_percent_encode_char( + dest, + (const unsigned char)cases[i].c, + &dest_len), + NULL ) + JSVERIFY(cases[i].exp_len, dest_len, NULL ) + JSVERIFY(0, strncmp(dest, cases[i].exp, dest_len), NULL ) + JSVERIFY_STR( cases[i].exp, dest, NULL ) + } + + JSVERIFY( FAIL, + H5FD_s3comms_percent_encode_char( + NULL, + (const unsigned char)'^', + &dest_len), + NULL ) + + PASSED(); + return 0; + +error: + return -1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_percent_encode_char */ + + +/*--------------------------------------------------------------------------- + * Function: test_s3r_open() + * + * Programmer: Jacob Smith 2018-01-24 + * + * Changes: None + * + *--------------------------------------------------------------------------- + */ +static herr_t +test_s3r_get_filesize(void) +{ +#ifdef H5_HAVE_ROS3_VFD + + /************************ + * test-local variables * + ************************/ + + char url_raven[S3_TEST_MAX_URL_SIZE]; + s3r_t *handle = NULL; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("s3r_get_filesize"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD is not enabled"); + fflush(stdout); + return 0; +#else + /* setup -- compose url to target resource + */ + if (FALSE == s3_test_bucket_defined) { + SKIPPED(); + puts(" environment variable HDF5_ROS3_TEST_BUCKET_URL not defined"); + fflush(stdout); + return 0; + } + + FAIL_IF( S3_TEST_MAX_URL_SIZE < + HDsnprintf(url_raven, + S3_TEST_MAX_URL_SIZE, + "%s/%s", + s3_test_bucket_url, + S3_TEST_RESOURCE_TEXT_PUBLIC) ); + + JSVERIFY( 0, H5FD_s3comms_s3r_get_filesize(NULL), + "filesize of the null handle should be 0" ) + + handle = H5FD_s3comms_s3r_open(url_raven, NULL, NULL, NULL); + FAIL_IF( handle == NULL ) + + JSVERIFY( 6464, H5FD_s3comms_s3r_get_filesize(handle), NULL ) + + + FAIL_IF( SUCCEED != H5FD_s3comms_s3r_close(handle) ) + + PASSED(); + return 0; + +error: + if (handle != NULL) + (void)H5FD_s3comms_s3r_close(handle); + + return -1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_s3r_get_filesize */ + + +/*--------------------------------------------------------------------------- + * Function: test_s3r_open() + * + * Programmer: Jacob Smith 2018-01-?? + * + * Changes: None + * + *--------------------------------------------------------------------------- + */ +static herr_t +test_s3r_open(void) +{ +#ifdef H5_HAVE_ROS3_VFD + + /************************ + * test-local variables * + ************************/ + + char url_missing[S3_TEST_MAX_URL_SIZE]; + char url_raven[S3_TEST_MAX_URL_SIZE]; + char url_raven_badport[S3_TEST_MAX_URL_SIZE]; + char url_shakespeare[S3_TEST_MAX_URL_SIZE]; + unsigned char signing_key[SHA256_DIGEST_LENGTH]; + struct tm *now = NULL; + char iso8601now[ISO8601_SIZE]; + s3r_t *handle = NULL; + hbool_t curl_ready = FALSE; + parsed_url_t *purl = NULL; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("s3r_open"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD is not enabled"); + fflush(stdout); + return 0; +#else + if (s3_test_credentials_loaded == 0) { + SKIPPED(); + puts(" s3 credentials are not loaded"); + fflush(stdout); + return 0; + } + if (FALSE == s3_test_bucket_defined) { + SKIPPED(); + puts(" environment variable HDF5_ROS3_TEST_BUCKET_URL not defined"); + fflush(stdout); + return 0; + } + + /****************** + * PRE-TEST SETUP * + ******************/ + + FAIL_IF( S3_TEST_MAX_URL_SIZE < + HDsnprintf(url_shakespeare, + S3_TEST_MAX_URL_SIZE, + "%s/%s", + s3_test_bucket_url, + S3_TEST_RESOURCE_TEXT_RESTRICTED) ); + + FAIL_IF( S3_TEST_MAX_URL_SIZE < + HDsnprintf(url_missing, + S3_TEST_MAX_URL_SIZE, + "%s/%s", + s3_test_bucket_url, + S3_TEST_RESOURCE_MISSING) ); + + FAIL_IF( S3_TEST_MAX_URL_SIZE < + HDsnprintf(url_raven, + S3_TEST_MAX_URL_SIZE, + "%s/%s", + s3_test_bucket_url, + S3_TEST_RESOURCE_TEXT_PUBLIC) ); + + /* Set given bucket url with invalid/inactive port number for badport. + * Note, this sort of micro-management of parsed_url_t is not advised + */ + FAIL_IF( FAIL == H5FD_s3comms_parse_url(s3_test_bucket_url, &purl) ) + if (purl->port == NULL) { + purl->port = (char *)H5MM_malloc(sizeof(char) * 5); + FAIL_IF( purl->port == NULL ); + FAIL_IF( 5 < HDsnprintf(purl->port, 5, "9000") ) + } else if (strcmp(purl->port, "9000") != 0) { + FAIL_IF( 5 < HDsnprintf(purl->port, 5, "9000") ) + } else { + FAIL_IF( 5 < HDsnprintf(purl->port, 5, "1234") ) + } + FAIL_IF( S3_TEST_MAX_URL_SIZE < + HDsnprintf(url_raven_badport, + S3_TEST_MAX_URL_SIZE, + "%s://%s:%s/%s", + purl->scheme, + purl->host, + purl->port, + S3_TEST_RESOURCE_TEXT_PUBLIC) ); + + curl_global_init(CURL_GLOBAL_DEFAULT); + curl_ready = TRUE; + + now = gmnow(); + FAIL_IF( now == NULL ) + FAIL_IF( ISO8601NOW(iso8601now, now) != (ISO8601_SIZE - 1) ); + + /* It is desired to have means available to verify that signing_key + * was set successfully and to an expected value. + */ + FAIL_IF( FAIL == + H5FD_s3comms_signing_key( + signing_key, + (const char *)s3_test_aws_secret_access_key, + (const char *)s3_test_aws_region, + (const char *)iso8601now) ); + + /************************* + * OPEN NONEXISTENT FILE * + *************************/ + + /* attempt anonymously + */ + handle = H5FD_s3comms_s3r_open(url_missing, NULL, NULL, NULL); + FAIL_IF( handle != NULL ); + + /* attempt with authentication + */ + handle = H5FD_s3comms_s3r_open( + url_missing, + (const char *)s3_test_aws_region, + (const char *)s3_test_aws_access_key_id, + (const unsigned char *)signing_key); + FAIL_IF( handle != NULL ); + + /************************* + * INACTIVE PORT ON HOST * + *************************/ + +#if S3_TEST_RUN_TIMEOUT +printf("Opening on inactive port may hang for a minute; waiting for timeout\n"); + handle = H5FD_s3comms_s3r_open(url_raven_badport, NULL, NULL, NULL); + FAIL_IF( handle != NULL ); +#endif + + /******************************* + * INVALID AUTHENTICATION INFO * + *******************************/ + + /* anonymous access on restricted file + */ + handle = H5FD_s3comms_s3r_open(url_shakespeare, NULL, NULL, NULL); + FAIL_IF( handle != NULL ); + + /* passed in a bad ID + */ + handle = H5FD_s3comms_s3r_open( + url_shakespeare, + (const char *)s3_test_aws_region, + "I_MADE_UP_MY_ID", + (const unsigned char *)signing_key); + FAIL_IF( handle != NULL ); + + /* using an invalid signing key + */ + handle = H5FD_s3comms_s3r_open( + url_shakespeare, + (const char *)s3_test_aws_region, + (const char *)s3_test_aws_access_key_id, + (const unsigned char *)EMPTY_SHA256); + FAIL_IF( handle != NULL ); + + /******************************* + * SUCCESSFUL OPEN (AND CLOSE) * + *******************************/ + + /* anonymous + */ + handle = H5FD_s3comms_s3r_open(url_raven, NULL, NULL, NULL); + FAIL_IF( handle == NULL ); + JSVERIFY( 6464, H5FD_s3comms_s3r_get_filesize(handle), + "did not get expected filesize" ) + JSVERIFY( SUCCEED, + H5FD_s3comms_s3r_close(handle), + "unable to close file" ) + handle = NULL; + + /* using authentication on anonymously-accessible file? + */ + handle = H5FD_s3comms_s3r_open( + url_raven, + (const char *)s3_test_aws_region, + (const char *)s3_test_aws_access_key_id, + (const unsigned char *)signing_key); + FAIL_IF( handle == NULL ); + JSVERIFY( 6464, H5FD_s3comms_s3r_get_filesize(handle), NULL ) + JSVERIFY( SUCCEED, + H5FD_s3comms_s3r_close(handle), + "unable to close file" ) + handle = NULL; + + /* authenticating + */ + handle = H5FD_s3comms_s3r_open( + url_shakespeare, + (const char *)s3_test_aws_region, + (const char *)s3_test_aws_access_key_id, + (const unsigned char *)signing_key); + FAIL_IF( handle == NULL ); + JSVERIFY( 5458199, H5FD_s3comms_s3r_get_filesize(handle), NULL ) + JSVERIFY( SUCCEED, + H5FD_s3comms_s3r_close(handle), + "unable to close file" ) + handle = NULL; + + + + curl_global_cleanup(); + curl_ready = FALSE; + + FAIL_IF( FAIL == H5FD_s3comms_free_purl(purl) ) + purl = NULL; + + PASSED(); + return 0; +error: + /*********** + * cleanup * + ***********/ + + if (handle != NULL) + H5FD_s3comms_s3r_close(handle); + if (purl != NULL) + H5FD_s3comms_free_purl(purl); + if (curl_ready == TRUE) + curl_global_cleanup(); + + return -1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_s3r_open */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_s3r_read() + * + * Purpose: + * + * Specify and demonstrate the use and life cycle of an S3 Request handle + * `s3r_t`, through its related functions. + * + * H5FD_s3comms_s3r_open + * H5FD_s3comms_s3r_getsize << called by open() _only_ + * H5FD_s3comms_s3r_read << called by getsize(), multiple times working + * H5FD_s3comms_s3r_close + * + * Shows most basic curl interation. + * + * Programmer: Jacob Smith + * 2017-10-06 + * + *--------------------------------------------------------------------------- + */ +static herr_t +test_s3r_read(void) +{ +#ifdef H5_HAVE_ROS3_VFD + +#define S3COMMS_TEST_BUFFER_SIZE 256 + + /************************ + * test-local variables * + ************************/ + + char url_raven[S3_TEST_MAX_URL_SIZE]; + char buffer[S3COMMS_TEST_BUFFER_SIZE]; + s3r_t *handle = NULL; + hbool_t curl_ready = FALSE; + unsigned int i = 0; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("test_s3r_read"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD is not enabled"); + fflush(stdout); + return 0; +#else + /* + * initial setup + */ + if (FALSE == s3_test_bucket_defined) { + SKIPPED(); + puts(" environment variable HDF5_ROS3_TEST_BUCKET_URL not defined"); + fflush(stdout); + return 0; + } + + curl_global_init(CURL_GLOBAL_DEFAULT); + curl_ready = TRUE; + FAIL_IF( S3_TEST_MAX_URL_SIZE < + HDsnprintf(url_raven, + S3_TEST_MAX_URL_SIZE, + "%s/%s", + s3_test_bucket_url, + S3_TEST_RESOURCE_TEXT_PUBLIC) ); + + for (i = 0; i < S3COMMS_TEST_BUFFER_SIZE; i++) + buffer[i] = '\0'; + + /* open file + */ + handle = H5FD_s3comms_s3r_open(url_raven, NULL, NULL, NULL); + FAIL_IF( handle == NULL ) + JSVERIFY( 6464, H5FD_s3comms_s3r_get_filesize(handle), NULL ) + + for (i = 0; i < S3COMMS_TEST_BUFFER_SIZE; i++) + buffer[i] = '\0'; + + /********************** + * read start of file * + **********************/ + + JSVERIFY( SUCCEED, + H5FD_s3comms_s3r_read( + handle, + (haddr_t)0, + (size_t)118, + buffer), + NULL ) + JSVERIFY_STR ( + "Once upon a midnight dreary, while I pondered, weak and weary,\n" \ + "Over many a quaint and curious volume of forgotten lore", + buffer, + NULL ) + + for (i = 0; i < S3COMMS_TEST_BUFFER_SIZE; i++) + buffer[i] = '\0'; + + /************************ + * read arbitrary range * + ************************/ + + JSVERIFY( SUCCEED, + H5FD_s3comms_s3r_read( + handle, + (haddr_t)2540, + (size_t)54, + buffer), + NULL ) + JSVERIFY_STR( "the grave and stern decorum of the countenance it wore", + buffer, + NULL ) + + for (i = 0; i < S3COMMS_TEST_BUFFER_SIZE; i++) + buffer[i] = '\0'; + + /********************** + * read one character * + **********************/ + + JSVERIFY(SUCCEED, + H5FD_s3comms_s3r_read( + handle, + (haddr_t)2540, + (size_t)1, + buffer), + NULL ) + JSVERIFY_STR( "t", buffer, NULL ) + + + for (i = 0; i < S3COMMS_TEST_BUFFER_SIZE; i++) + buffer[i] = '\0'; + + /*************** + * read to EoF * + ***************/ + + JSVERIFY( SUCCEED, + H5FD_s3comms_s3r_read( + handle, + (haddr_t)6370, + (size_t)0, + buffer), + NULL ) + JSVERIFY( 0, + strncmp(buffer, + "And my soul from out that shadow that lies floating on the floor\nShall be lifted—nevermore!\n", + 94), + buffer ) + + for (i = 0; i < S3COMMS_TEST_BUFFER_SIZE; i++) + buffer[i] = '\0'; + + /***************** + * read past eof * + *****************/ + + JSVERIFY( FAIL, + H5FD_s3comms_s3r_read( + handle, + (haddr_t)6400, + (size_t)100, /* 6400+100 > 6464 */ + buffer), + NULL ) + JSVERIFY( 0, strcmp("", buffer), NULL ) + + /************************ + * read starts past eof * + ************************/ + + JSVERIFY( FAIL, + H5FD_s3comms_s3r_read( + handle, + (haddr_t)1200699, /* 1200699 > 6464 */ + (size_t)100, + buffer), + NULL ) + JSVERIFY( 0, strcmp("", buffer), NULL ) + + /********************** + * read starts on eof * + **********************/ + + JSVERIFY( FAIL, + H5FD_s3comms_s3r_read( + handle, + (haddr_t)6464, + (size_t)0, + buffer), + NULL ) + JSVERIFY( 0, strcmp("", buffer), NULL ) + + /************* + * TEAR DOWN * + *************/ + + JSVERIFY( SUCCEED, + H5FD_s3comms_s3r_close(handle), + "unable to close file" ) + handle = NULL; + + curl_global_cleanup(); + curl_ready = FALSE; + + PASSED(); + return 0; + +error: + /*********** + * cleanup * + ***********/ + + if (handle != NULL) + H5FD_s3comms_s3r_close(handle); + + if (curl_ready == TRUE) + curl_global_cleanup(); + + return -1; + +#undef S3COMMS_TEST_BUFFER_SIZE +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_s3r_read */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_signing_key() + * + * Purpose: + * + * Define and verify behavior of `H5FD_s3comms_signing_key()` + * + * More test cases would be a very good idea. + * + * Programmer: Jacob Smith + * 2017-09-18 + * + *--------------------------------------------------------------------------- + */ +static herr_t +test_signing_key(void) +{ +#ifdef H5_HAVE_ROS3_VFD + /************************* + * test-local structures * + *************************/ + + struct testcase { + const char *region; + const char *secret_key; + const char *when; + unsigned char exp[SHA256_DIGEST_LENGTH]; + }; + + /************************ + * test-local variables * + ************************/ + + struct testcase cases[] = { + { "us-east-1", + "wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY", + "20130524T000000Z", + { 0xdb, 0xb8, 0x93, 0xac, 0xc0, 0x10, 0x96, 0x49, + 0x18, 0xf1, 0xfd, 0x43, 0x3a, 0xdd, 0x87, 0xc7, + 0x0e, 0x8b, 0x0d, 0xb6, 0xbe, 0x30, 0xc1, 0xfb, + 0xea, 0xfe, 0xfa, 0x5e, 0xc6, 0xba, 0x83, 0x78, + }, + }, + }; + int i = 0; + unsigned char *key = NULL; + int ncases = 1; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("signing_key"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD is not enabled"); + fflush(stdout); + return 0; +#else + for (i = 0; i < ncases; i++) { + key = (unsigned char *)HDmalloc(sizeof(unsigned char) * \ + SHA256_DIGEST_LENGTH); + HDassert(key != NULL); + + JSVERIFY( SUCCEED, + H5FD_s3comms_signing_key( + key, + cases[i].secret_key, + cases[i].region, + cases[i].when), + NULL ) + + JSVERIFY( 0, + strncmp((const char *)cases[i].exp, + (const char *)key, + SHA256_DIGEST_LENGTH), + cases[i].exp ) + + free(key); + key = NULL; + } + + + /*************** + * ERROR CASES * + ***************/ + + key = (unsigned char *)HDmalloc(sizeof(unsigned char) * \ + SHA256_DIGEST_LENGTH); + HDassert(key != NULL); + + JSVERIFY( FAIL, + H5FD_s3comms_signing_key( + NULL, + cases[0].secret_key, + cases[0].region, + cases[0].when), + "destination cannot be NULL" ) + + JSVERIFY( FAIL, + H5FD_s3comms_signing_key( + key, + NULL, + cases[0].region, + cases[0].when), + "secret key cannot be NULL" ) + + JSVERIFY( FAIL, + H5FD_s3comms_signing_key( + key, + cases[0].secret_key, + NULL, + cases[0].when), + "aws region cannot be NULL" ) + + JSVERIFY( FAIL, + H5FD_s3comms_signing_key( + key, + cases[0].secret_key, + cases[0].region, + NULL), + "time string cannot be NULL" ) + + free(key); + key = NULL; + + PASSED(); + return 0; + +error: + if (key != NULL) { + free(key); + } + + return -1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_signing_key */ + + +/*--------------------------------------------------------------------------- + * + * Function: test_tostringtosign() + * + * Purpose: + * + * Verify that we can get the "string to sign" from a Canonical Request and + * related information. + * + * Demonstrate failure cases. + * + * Return: + * + * Success: 0 + * Failure: -1 + * + * Programmer: Jacob Smith + * 2017-09-13 + * + *--------------------------------------------------------------------------- + */ +static herr_t +test_tostringtosign(void) +{ +#ifdef H5_HAVE_ROS3_VFD + /************************ + * test-local variables * + ************************/ + + const char canonreq[] = "GET\n/test.txt\n\nhost:examplebucket.s3.amazonaws.com\nrange:bytes=0-9\nx-amz-content-sha256:e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855\nx-amz-date:20130524T000000Z\n\nhost;range;x-amz-content-sha256;x-amz-date\ne3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855"; + const char iso8601now[] = "20130524T000000Z"; + const char region[] = "us-east-1"; + char s2s[512]; +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("s3comms tostringtosign"); + +#ifndef H5_HAVE_ROS3_VFD + SKIPPED(); + puts(" ROS3 VFD is not enabled"); + fflush(stdout); + return 0; +#else + JSVERIFY( SUCCEED, + H5FD_s3comms_tostringtosign(s2s, canonreq, iso8601now, region), + "unable to create string to sign" ) + + JSVERIFY_STR( "AWS4-HMAC-SHA256\n20130524T000000Z\n20130524/us-east-1/s3/aws4_request\n7344ae5b7ee6c3e7e6b0fe0640412a37625d1fbfff95c48bbb2dc43964946972", + s2s, NULL ) + + JSVERIFY( FAIL, + H5FD_s3comms_tostringtosign(s2s, NULL, iso8601now, region), + "canonical request string cannot be NULL" ) + + JSVERIFY( FAIL, + H5FD_s3comms_tostringtosign(s2s, canonreq, NULL, region), + "time string cannot be NULL" ) + + JSVERIFY( FAIL, + H5FD_s3comms_tostringtosign(s2s, canonreq, iso8601now, NULL), + "aws region cannot be NULL" ) + + PASSED(); + return 0; + +error : + return -1; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_tostringtosign */ + + +/*---------------------------------------------------------------------------- + * + * Function: test_trim() + * + * Purpose: + * + * Define and verify behavior of `H5FD_s3comms_trim()`. + * + * Programmer: Jacob Smith + * 2017-09-14 + * + *---------------------------------------------------------------------------- + */ +static herr_t +test_trim(void) +{ + /************************* + * test-local structures * + *************************/ + + struct testcase { + const char *in; + size_t in_len; + const char *exp; + size_t exp_len; + }; + + /************************ + * test-local variables * + ************************/ + + struct testcase cases[] = { + { "block string", + 12, + "block string", + 12, + }, + { " \n\r \t", + 6, + "", + 0, + }, + { " \twhite b4", + 10, + "white b4", + 8, + }, + { "white after\r\n ", + 15, + "white after", + 11, + }, + { " on\nends\t", + 9, + "on\nends", + 7, + }, + }; + char dest[32]; + size_t dest_len = 0; + int i = 0; + int n_cases = 5; + char *str = NULL; + + + + TESTING("s3comms trim"); + + for (i = 0; i < n_cases; i++) { + HDassert(str == NULL); + str = (char *)HDmalloc(sizeof(char) * cases[i].in_len); + HDassert(str != NULL); + HDstrncpy(str, cases[i].in, cases[i].in_len); + + JSVERIFY( SUCCEED, + H5FD_s3comms_trim(dest, str, cases[i].in_len, &dest_len), + NULL ) + JSVERIFY( cases[i].exp_len, dest_len, cases[i].in ) + if (dest_len > 0) { + JSVERIFY( 0, strncmp(cases[i].exp, dest, dest_len), + cases[i].exp ) + } + free(str); + str = NULL; + } + + JSVERIFY( SUCCEED, H5FD_s3comms_trim(dest, NULL, 3, &dest_len), + "should not fail when trimming a null string" ); + JSVERIFY( 0, dest_len, "trimming NULL string writes 0 characters" ) + + HDassert(str == NULL); + str = (char *)HDmalloc(sizeof(char *) * 11); + HDassert(str != NULL); + memcpy(str, "some text ", 11); /* string with null terminator */ + JSVERIFY( FAIL, H5FD_s3comms_trim(NULL, str, 10, &dest_len), + "destination for trim cannot be NULL" ); + free(str); + str = NULL; + + PASSED(); + return 0; + +error: + if (str != NULL) + free(str); + return -1; + +} /* test_trim */ + + +/*---------------------------------------------------------------------------- + * + * Function: test_uriencode() + * + * Purpose: + * + * Define and verify behavior of `H5FD_s3comms_uriencode()`. + * + * Programmer: Jacob Smith + * 2017-09-14 + * + *---------------------------------------------------------------------------- + */ +static herr_t +test_uriencode(void) +{ + /************************* + * test-local structures * + *************************/ + + struct testcase { + const char *str; + size_t s_len; + hbool_t encode_slash; + const char *expected; + }; + + /************************ + * test-local variables * + ************************/ + + struct testcase cases[] = { + { "/path/to/resource.jpg", + 21, + FALSE, + "/path/to/resource.jpg", + }, + { "/path/to/resource.jpg", + 21, + TRUE, + "%2Fpath%2Fto%2Fresource.jpg", + }, + { "string got_spaa ces", + 20, + TRUE, + "string%20got_spaa%20%20ces", + }, + { "sp ac~es/and-sl ash.encoded", + 27, + TRUE, + "sp%20ac~es%2Fand-sl%20ash.encoded", + }, + { "sp ac~es/and-sl ash.unencoded", + 29, + FALSE, + "sp%20ac~es/and-sl%20ash.unencoded", + }, + { "/path/to/resource.txt", + 0, + FALSE, + "", + + } + }; + char *dest = NULL; + size_t dest_written = 0; + int i = 0; + int ncases = 6; + size_t str_len = 0; + + + + TESTING("s3comms uriencode") + + for (i = 0; i < ncases; i++) { + str_len = cases[i].s_len; + dest = (char *)HDmalloc(sizeof(char) * str_len * 3 + 1); + FAIL_IF( dest == NULL ) + + JSVERIFY( SUCCEED, + H5FD_s3comms_uriencode( + dest, + cases[i].str, + str_len, + cases[i].encode_slash, + &dest_written), + NULL ); + JSVERIFY( HDstrlen(cases[i].expected), + dest_written, + NULL ) + JSVERIFY( 0, + strncmp(dest, cases[i].expected, dest_written), + cases[i].expected ); + + free(dest); + dest = NULL; + } + + /*************** + * ERROR CASES * + ***************/ + + dest = (char *)HDmalloc(sizeof(char) * 15); + HDassert(dest != NULL); + + JSVERIFY( FAIL, + H5FD_s3comms_uriencode(NULL, "word$", 5, false, &dest_written), + "destination cannot be NULL" ); + JSVERIFY( FAIL, + H5FD_s3comms_uriencode(dest, NULL, 5, false, &dest_written), + "source string cannot be NULL" ); + + free(dest); + dest = NULL; + + PASSED(); + return 0; + +error: + if (dest != NULL) { + free(dest); + } + return -1; + +} /* test_uriencode */ + + + + +/*------------------------------------------------------------------------- + * Function: main() + * + * Purpose: + * + * Run unit tests for S3 Communications (s3comms). + * + * Return: + * + * Success: 0 + * Failure: 1 + * + * Programmer: Jacob Smith + * 2017-10-12 + * + *------------------------------------------------------------------------- + */ +int +main(void) +{ + int nerrors = 0; +#ifdef H5_HAVE_ROS3_VFD + const char *bucket_url_env = NULL; +#endif /* H5_HAVE_ROS3_VFD */ + + h5_reset(); + + HDprintf("Testing S3Communications functionality.\n"); + +#ifdef H5_HAVE_ROS3_VFD + + /* "clear" profile data strings */ + s3_test_aws_access_key_id[0] = '\0'; + s3_test_aws_secret_access_key[0] = '\0'; + s3_test_aws_region[0] = '\0'; + s3_test_bucket_url[0] = '\0'; + +/* TODO: unit/regression test for H5FD_s3comms_load_aws_profile() + * requires a few test files and/or manipulation of default path + */ + /* attempt to load test credentials + * if unable, certain tests will be skipped + */ + if (SUCCEED == H5FD_s3comms_load_aws_profile( + S3_TEST_PROFILE_NAME, + s3_test_aws_access_key_id, + s3_test_aws_secret_access_key, + s3_test_aws_region)) + { + s3_test_credentials_loaded = 1; + } + + bucket_url_env = HDgetenv("HDF5_ROS3_TEST_BUCKET_URL"); + if (bucket_url_env == NULL || bucket_url_env[0] == '\0') { + HDprintf("WARNING: S3 bucket url is not defined in enviornment " \ + "variable 'HDF5_ROS3_TEST_BUCKET_URL'!\n"); + } else { + HDstrncpy(s3_test_bucket_url, bucket_url_env, S3_TEST_MAX_URL_SIZE); + s3_test_bucket_defined = TRUE; + } + +#endif /* H5_HAVE_ROS3_VFD */ + + /* tests ordered rougly by dependence */ + nerrors += test_macro_format_credential() < 0 ? 1 : 0; + nerrors += test_trim() < 0 ? 1 : 0; + nerrors += test_nlowercase() < 0 ? 1 : 0; + nerrors += test_uriencode() < 0 ? 1 : 0; + nerrors += test_percent_encode_char() < 0 ? 1 : 0; + nerrors += test_bytes_to_hex() < 0 ? 1 : 0; + nerrors += test_HMAC_SHA256() < 0 ? 1 : 0; + nerrors += test_signing_key() < 0 ? 1 : 0; + nerrors += test_hrb_node_set() < 0 ? 1 : 0; + nerrors += test_hrb_init_request() < 0 ? 1 : 0; + nerrors += test_parse_url() < 0 ? 1 : 0; + nerrors += test_aws_canonical_request() < 0 ? 1 : 0; + nerrors += test_tostringtosign() < 0 ? 1 : 0; + nerrors += test_s3r_open() < 0 ? 1 : 0; + nerrors += test_s3r_get_filesize() < 0 ? 1 : 0; + nerrors += test_s3r_read() < 0 ? 1 : 0; + + if(nerrors) { + HDprintf("***** %d S3comms TEST%s FAILED! *****\n", + nerrors, + nerrors > 1 ? "S" : ""); + return 1; + } /* end if */ + + HDprintf("All S3comms tests passed.\n"); + + return 0; +} /* end main() */ + @@ -58,6 +58,7 @@ const char *FILENAME[] = { "stdio_file", /*7*/ "windows_file", /*8*/ "new_multi_file_v16",/*9*/ + "ro_s3_file6", /*10*/ NULL }; @@ -66,7 +67,7 @@ const char *FILENAME[] = { #define COMPAT_BASENAME "family_v16_" #define MULTI_COMPAT_BASENAME "multi_file_v16" - + /*------------------------------------------------------------------------- * Function: test_sec2 * @@ -178,7 +179,7 @@ error: return -1; } /* end test_sec2() */ - + /*------------------------------------------------------------------------- * Function: test_core * @@ -534,7 +535,7 @@ error: return -1; } /* end test_core() */ - + /*------------------------------------------------------------------------- * Function: test_direct * @@ -754,7 +755,7 @@ error: #endif /*H5_HAVE_DIRECT*/ } - + /*------------------------------------------------------------------------- * Function: test_family_opens * @@ -835,7 +836,7 @@ error: } /* end test_family_opens() */ #pragma GCC diagnostic pop - + /*------------------------------------------------------------------------- * Function: test_family * @@ -1017,7 +1018,7 @@ error: return -1; } - + /*------------------------------------------------------------------------- * Function: test_family_compat * @@ -1129,7 +1130,7 @@ error: } /* end test_family_compat() */ #pragma GCC diagnostic pop - + /*------------------------------------------------------------------------- * Function: test_multi_opens * @@ -1170,7 +1171,7 @@ test_multi_opens(char *fname) } /* end test_multi_opens() */ #pragma GCC diagnostic pop - + /*------------------------------------------------------------------------- * Function: test_multi * @@ -1404,7 +1405,7 @@ error: return FAIL; } /* end test_multi() */ - + /*------------------------------------------------------------------------- * Function: test_multi_compat * @@ -1578,7 +1579,7 @@ error: return -1; } - + /*------------------------------------------------------------------------- * Function: test_log * @@ -1689,7 +1690,7 @@ error: return -1; } - + /*------------------------------------------------------------------------- * Function: test_stdio * @@ -1794,7 +1795,7 @@ error: } - + /*------------------------------------------------------------------------- * Function: test_windows * @@ -1916,7 +1917,146 @@ error: } /* end test_windows() */ - +/*------------------------------------------------------------------------- + * Function: test_ros3 + * + * Purpose: Tests the file handle interface for the ROS3 driver + * + * As the ROS3 driver is 1) read only, 2) requires access + * to an S3 server (minio for now), this test is quite + * different from the other tests. + * + * For now, test only fapl & flags. Extend as the + * work on the VFD continues. + * + * Return: Success: 0 + * Failure: -1 + * + * Programmer: John Mainzer + * 7/12/17 + * + *------------------------------------------------------------------------- + */ +static herr_t +test_ros3(void) +{ + hid_t fid = -1; /* file ID */ + hid_t fapl_id = -1; /* file access property list ID */ + hid_t fapl_id_out = -1; /* from H5Fget_access_plist */ + hid_t driver_id = -1; /* ID for this VFD */ + unsigned long driver_flags = 0; /* VFD feature flags */ + char filename[1024]; /* filename */ + void *os_file_handle = NULL; /* OS file handle */ + hsize_t file_size; /* file size */ + H5FD_ros3_fapl_t test_ros3_fa; + H5FD_ros3_fapl_t ros3_fa_0 = + { + /* version = */ H5FD__CURR_ROS3_FAPL_T_VERSION, + /* authenticate = */ FALSE, + /* aws_region = */ "", + /* secret_id = */ "", + /* secret_key = */ "plugh", + }; + + TESTING("ROS3 file driver"); + + /* Set property list and file name for ROS3 driver. */ + if((fapl_id = H5Pcreate(H5P_FILE_ACCESS)) < 0) + TEST_ERROR; + + if(H5Pset_fapl_ros3(fapl_id, &ros3_fa_0) < 0) + TEST_ERROR; + + /* verify that the ROS3 FAPL entry is set as expected */ + if(H5Pget_fapl_ros3(fapl_id, &test_ros3_fa) < 0) + TEST_ERROR; + + /* need a macro to compare instances of H5FD_ros3_fapl_t */ + if((test_ros3_fa.version != ros3_fa_0.version) || + (test_ros3_fa.authenticate != ros3_fa_0.authenticate) || + (strcmp(test_ros3_fa.aws_region, ros3_fa_0.aws_region) != 0) || + (strcmp(test_ros3_fa.secret_id, ros3_fa_0.secret_id) != 0) || + (strcmp(test_ros3_fa.secret_key, ros3_fa_0.secret_key) != 0)) + TEST_ERROR; + + h5_fixname(FILENAME[10], fapl_id, filename, sizeof(filename)); + + /* Check that the VFD feature flags are correct */ + if ((driver_id = H5Pget_driver(fapl_id)) < 0) + TEST_ERROR; + + if (H5FDdriver_query(driver_id, &driver_flags) < 0) + TEST_ERROR; + + if(!(driver_flags & H5FD_FEAT_DATA_SIEVE)) TEST_ERROR + + /* Check for extra flags not accounted for above */ + if(driver_flags != (H5FD_FEAT_DATA_SIEVE)) + TEST_ERROR + + /* can't create analogs of the following tests until the + * ROS3 driver is up and running in a minimal fashion. + * Comment them out until we get to them. + */ +#if 0 + if((fid = H5Fcreate(filename, H5F_ACC_TRUNC, H5P_DEFAULT, fapl_id)) < 0) + TEST_ERROR; + + /* Retrieve the access property list... */ + if((fapl_id_out = H5Fget_access_plist(fid)) < 0) + TEST_ERROR; + + /* Check that the driver is correct */ + if(H5FD_ROS3 != H5Pget_driver(fapl_id_out)) + TEST_ERROR; + + /* ...and close the property list */ + if(H5Pclose(fapl_id_out) < 0) + TEST_ERROR; + + /* Check that we can get an operating-system-specific handle from + * the library. + */ + if(H5Fget_vfd_handle(fid, H5P_DEFAULT, &os_file_handle) < 0) + TEST_ERROR; + if(os_file_handle == NULL) + FAIL_PUTS_ERROR("NULL os-specific vfd/file handle was returned from H5Fget_vfd_handle"); + + + /* There is no garantee the size of metadata in file is constant. + * Just try to check if it's reasonable. + * + * Currently it should be around 2 KB. + */ + if(H5Fget_filesize(fid, &file_size) < 0) + TEST_ERROR; + if(file_size < 1 * KB || file_size > 4 * KB) + FAIL_PUTS_ERROR("suspicious file size obtained from H5Fget_filesize"); + + /* Close and delete the file */ + if(H5Fclose(fid) < 0) + TEST_ERROR; + h5_delete_test_file(FILENAME[0], fapl_id); + + /* Close the fapl */ + if(H5Pclose(fapl_id) < 0) + TEST_ERROR; +#endif + + PASSED(); + return 0; + +error: + H5E_BEGIN_TRY { + H5Pclose(fapl_id); + H5Pclose(fapl_id_out); + H5Fclose(fid); + } H5E_END_TRY; + return -1; +} /* end test_ros3() */ + + + /*------------------------------------------------------------------------- * Function: main * @@ -1949,6 +2089,7 @@ main(void) nerrors += test_log() < 0 ? 1 : 0; nerrors += test_stdio() < 0 ? 1 : 0; nerrors += test_windows() < 0 ? 1 : 0; + nerrors += test_ros3() < 0 ? 1 : 0; if(nerrors) { HDprintf("***** %d Virtual File Driver TEST%s FAILED! *****\n", diff --git a/tools/CMakeLists.txt b/tools/CMakeLists.txt index 400039e..aa09aa6 100644 --- a/tools/CMakeLists.txt +++ b/tools/CMakeLists.txt @@ -9,4 +9,11 @@ add_subdirectory (src) #-- Add the tests if (BUILD_TESTING) add_subdirectory (test) + +# -------------------------------------------------------------------- +# If S3 or HDFS enabled, then we need to test the tools library +# -------------------------------------------------------------------- + if (HDF5_ENABLE_ROS3_VFD OR HDF5_ENABLE_HDFS) + add_subdirectory (libtest) + endif () endif () diff --git a/tools/lib/h5tools_utils.c b/tools/lib/h5tools_utils.c index e7e017f..a3cd7d9 100644 --- a/tools/lib/h5tools_utils.c +++ b/tools/lib/h5tools_utils.c @@ -21,6 +21,10 @@ #include "H5private.h" #include "h5trav.h" +#ifdef H5_HAVE_ROS3_VFD +#include "H5FDros3.h" +#endif + /* global variables */ unsigned h5tools_nCols = 80; /* ``get_option'' variables */ @@ -97,7 +101,7 @@ parallel_print(const char* format, ...) HDva_end(ap); } - + /*------------------------------------------------------------------------- * Function: error_msg * @@ -122,7 +126,7 @@ error_msg(const char *fmt, ...) HDva_end(ap); } - + /*------------------------------------------------------------------------- * Function: warn_msg * @@ -161,7 +165,7 @@ help_ref_msg(FILE *output) HDfprintf(output, "see the <%s> entry in the 'HDF5 Reference Manual'.\n",h5tools_getprogname()); } - + /*------------------------------------------------------------------------- * Function: get_option * @@ -322,7 +326,229 @@ get_option(int argc, const char **argv, const char *opts, const struct long_opti return opt_opt; } - + +/***************************************************************************** + * + * Function: parse_tuple() + * + * Purpose: + * + * Create array of pointers to strings, identified as elements in a tuple + * of arbitrary length separated by provided character. + * ("tuple" because "nple" looks strange) + * + * * Receives pointer to start of tuple sequence string, '('. + * * Attempts to separate elements by token-character `sep`. + * * If the separator character is preceded by a backslash '\', + * the backslash is deleted and the separator is included in the + * element string as any other character. + * * To end an element with a backslash, escape the backslash, e.g. + * "(myelem\\,otherelem) -> {"myelem\", "otherelem"} + * * In all other cases, a backslash appearing not as part of "\\" or + * "\<sep>" digraph will be included berbatim. + * * Last two characters in the string MUST be ")\0". + * + * * Generates a copy of the input string `start`, (src..")\0"), replacing + * separators and close-paren with null charaters. + * * This string is allocated at runtime and should be freed when done. + * * Generates array of char pointers, and directs start of each element + * (each pointer) into this copy. + * * Each tuple element points to the start of its string (substring) + * and ends with a null terminator. + * * This array is allocated at runtime and should be freed when done. + * * Reallocates and expands elements array during parsing. + * * Initially allocated for 2 (plus one null entry), and grows by + * powers of 2. + * * The final 'slot' in the element array (elements[nelements], e.g.) + * always points to NULL. + * * The number of elements found and stored are passed out through pointer + * to unsigned, `nelems`. + * + * Return: + * + * FAIL If malformed--does not look like a tuple "(...)" + * or major error was encountered while parsing. + * or + * SUCCEED String looks properly formed "(...)" and no major errors. + * + * Stores number of elements through pointer `nelems`. + * Stores list of pointers to char (first char in each element + * string) through pointer `ptrs_out`. + * NOTE: `ptrs_out[nelems] == NULL` should be true. + * NOTE: list is malloc'd by function, and should be freed + * when done. + * Stores "source string" for element pointers through `cpy_out`. + * NOTE: Each element substring is null-terminated. + * NOTE: There may be extra characters after the last element + * (past its null terminator), but is guaranteed to + * be null-terminated. + * NOTE: `cpy_out` string is malloc'd by function, + * and should be freed when done. + * + * Programmer: Jacob Smith + * 2017-11-10 + * + * Changes: None. + * + ***************************************************************************** + */ +herr_t +parse_tuple(const char *start, + int sep, + char **cpy_out, + unsigned *nelems, + char ***ptrs_out) +{ + char *elem_ptr = NULL; + char *dest_ptr = NULL; + unsigned elems_count = 0; + char **elems = NULL; /* more like *elems[], but complier... */ + char **elems_re = NULL; /* temporary pointer, for realloc */ + char *cpy = NULL; + herr_t ret_value = SUCCEED; + unsigned init_slots = 2; + + + + /***************** + * SANITY-CHECKS * + *****************/ + + /* must start with "(" + */ + if (start[0] != '(') { + ret_value = FAIL; + goto done; + } + + /* must end with ")" + */ + while (start[elems_count] != '\0') { + elems_count++; + } + if (start[elems_count - 1] != ')') { + ret_value = FAIL; + goto done; + } + + elems_count = 0; + + + + /*********** + * PREPARE * + ***********/ + + /* create list + */ + elems = (char **)HDmalloc(sizeof(char *) * (init_slots + 1)); + if (elems == NULL) { ret_value = FAIL; goto done; } /* CANTALLOC */ + + /* create destination string + */ + start++; /* advance past opening paren '(' */ + cpy = (char *)HDmalloc(sizeof(char) * (HDstrlen(start))); /* no +1; less '(' */ + if (cpy == NULL) { ret_value = FAIL; goto done; } /* CANTALLOC */ + + /* set pointers + */ + dest_ptr = cpy; /* start writing copy here */ + elem_ptr = cpy; /* first element starts here */ + elems[elems_count++] = elem_ptr; /* set first element pointer into list */ + + + + /********* + * PARSE * + *********/ + + while (*start != '\0') { + /* For each character in the source string... + */ + if (*start == '\\') { + /* Possibly an escape digraph. + */ + if ((*(start + 1) == '\\') || + (*(start + 1) == sep) ) + { + /* Valid escape digraph of "\\" or "\<sep>". + */ + start++; /* advance past escape char '\' */ + *(dest_ptr++) = *(start++); /* Copy subsequent char */ + /* and advance pointers. */ + } else { + /* Not an accepted escape digraph. + * Copy backslash character. + */ + *(dest_ptr++) = *(start++); + } + } else if (*start == sep) { + /* Non-escaped separator. + * Terminate elements substring in copy, record element, advance. + * Expand elements list if appropriate. + */ + *(dest_ptr++) = 0; /* Null-terminate elem substring in copy */ + /* and advance pointer. */ + start++; /* Advance src pointer past separator. */ + elem_ptr = dest_ptr; /* Element pointer points to start of first */ + /* character after null sep in copy. */ + elems[elems_count++] = elem_ptr; /* Set elem pointer in list */ + /* and increment count. */ + + /* Expand elements list, if necessary. + */ + if (elems_count == init_slots) { + init_slots *= 2; + elems_re = (char **)realloc(elems, sizeof(char *) * \ + (init_slots + 1)); + if (elems_re == NULL) { + /* CANTREALLOC */ + ret_value = FAIL; + goto done; + } + elems = elems_re; + } + } else if (*start == ')' && *(start + 1) == '\0') { + /* Found terminal, non-escaped close-paren. Last element. + * Write null terminator to copy. + * Advance source pointer to gently break from loop. + * Requred to prevent ")" from always being added to last element. + */ + start++; + } else { + /* Copy character into destination. Advance pointers. + */ + *(dest_ptr++) = *(start++); + } + } + *dest_ptr = '\0'; /* Null-terminate destination string. */ + elems[elems_count] = NULL; /* Null-terminate elements list. */ + + + + /******************** + * PASS BACK VALUES * + ********************/ + + *ptrs_out = elems; + *nelems = elems_count; + *cpy_out = cpy; + +done: + if (ret_value == FAIL) { + /* CLEANUP */ + if (cpy) free(cpy); + if (elems) free(elems); + } + + return ret_value; + +} /* parse_tuple */ + + + + + /*------------------------------------------------------------------------- * Function: indentation * @@ -344,7 +570,7 @@ indentation(unsigned x) } } - + /*------------------------------------------------------------------------- * Function: print_version * @@ -362,7 +588,7 @@ print_version(const char *progname) ((const char *)H5_VERS_SUBRELEASE)[0] ? "-" : "", H5_VERS_SUBRELEASE); } - + /*------------------------------------------------------------------------- * Function: init_table * @@ -384,7 +610,7 @@ init_table(table_t **tbl) *tbl = table; } - + /*------------------------------------------------------------------------- * Function: free_table * @@ -408,7 +634,7 @@ free_table(table_t *table) } #ifdef H5DUMP_DEBUG - + /*------------------------------------------------------------------------- * Function: dump_table * @@ -429,7 +655,7 @@ dump_table(char* tablename, table_t *table) table->objs[u].displayed, table->objs[u].recorded); } - + /*------------------------------------------------------------------------- * Function: dump_tables * @@ -447,7 +673,7 @@ dump_tables(find_objs_t *info) } #endif /* H5DUMP_DEBUG */ - + /*------------------------------------------------------------------------- * Function: search_obj * @@ -470,7 +696,7 @@ search_obj(table_t *table, haddr_t objno) return NULL; } - + /*------------------------------------------------------------------------- * Function: find_objs_cb * @@ -546,7 +772,7 @@ find_objs_cb(const char *name, const H5O_info_t *oinfo, const char *already_seen return ret_value; } - + /*------------------------------------------------------------------------- * Function: init_objs * @@ -591,7 +817,7 @@ done: return ret_value; } - + /*------------------------------------------------------------------------- * Function: add_obj * @@ -622,7 +848,7 @@ add_obj(table_t *table, haddr_t objno, const char *objname, hbool_t record) table->objs[u].displayed = 0; } - + #ifndef H5_HAVE_TMPFILE /*------------------------------------------------------------------------- * Function: tmpfile @@ -841,3 +1067,266 @@ done: return ret_value; } + +/*---------------------------------------------------------------------------- + * + * Function: h5tools_populate_ros3_fapl() + * + * Purpose: + * + * Set the values of a ROS3 fapl configuration object. + * + * If the values pointer is NULL, sets fapl target `fa` to a default + * (valid, current-version, non-authenticating) fapl config. + * + * If `values` pointer is _not_ NULL, expects `values` to contain at least + * three non-null pointers to null-terminated strings, corresponding to: + * { aws_region, + * secret_id, + * secret_key, + * } + * If all three strings are empty (""), the default fapl will be default. + * Both aws_region and secret_id values must be both empty or both + * populated. If + * Only secret_key is allowed to be empty (the empty string, ""). + * All values are checked against overflow as defined in the ros3 vfd + * header file; if a value overruns the permitted space, FAIL is returned + * and the function aborts without resetting the fapl to values initially + * present. + * + * Return: + * + * 0 (failure) if... + * * Read-Only S3 VFD is not enabled. + * * NULL fapl pointer: (NULL, {...} ) + * * Warning: In all cases below, fapl will be set as "default" + * before error occurs. + * * NULL value strings: (&fa, {NULL?, NULL? NULL?, ...}) + * * Incomplete fapl info: + * * empty region, non-empty id, key either way + * * (&fa, {"", "...", "?"}) + * * empty id, non-empty region, key either way + * * (&fa, {"...", "", "?"}) + * * "non-empty key and either id or region empty + * * (&fa, {"", "", "...") + * * (&fa, {"", "...", "...") + * * (&fa, {"...", "", "...") + * * Any string would overflow allowed space in fapl definition. + * or + * 1 (success) + * * Sets components in fapl_t pointer, copying strings as appropriate. + * * "Default" fapl (valid version, authenticate->False, empty strings) + * * `values` pointer is NULL + * * (&fa, NULL) + * * first three strings in `values` are empty ("") + * * (&fa, {"", "", "", ...} + * * Authenticating fapl + * * region, id, and optional key provided + * * (&fa, {"...", "...", ""}) + * * (&fa, {"...", "...", "..."}) + * + * Programmer: Jacob Smith + * 2017-11-13 + * + * Changes: None. + * + *---------------------------------------------------------------------------- + */ +int +h5tools_populate_ros3_fapl(H5FD_ros3_fapl_t *fa, + const char **values) +{ +#ifndef H5_HAVE_ROS3_VFD + return 0; +#else + int show_progress = 0; /* set to 1 for debugging */ + int ret_value = 1; /* 1 for success, 0 for failure */ + /* e.g.? if (!populate()) { then failed } */ + + if (show_progress) { + HDprintf("called h5tools_populate_ros3_fapl\n"); + } + + if (fa == NULL) { + if (show_progress) { + HDprintf(" ERROR: null pointer to fapl_t\n"); + } + ret_value = 0; + goto done; + } + + if (show_progress) { + HDprintf(" preset fapl with default values\n"); + } + fa->version = H5FD__CURR_ROS3_FAPL_T_VERSION; + fa->authenticate = FALSE; + *(fa->aws_region) = '\0'; + *(fa->secret_id) = '\0'; + *(fa->secret_key) = '\0'; + + /* sanity-check supplied values + */ + if (values != NULL) { + if (values[0] == NULL) { + if (show_progress) { + HDprintf(" ERROR: aws_region value cannot be NULL\n"); + } + ret_value = 0; + goto done; + } + if (values[1] == NULL) { + if (show_progress) { + HDprintf(" ERROR: secret_id value cannot be NULL\n"); + } + ret_value = 0; + goto done; + } + if (values[2] == NULL) { + if (show_progress) { + HDprintf(" ERROR: secret_key value cannot be NULL\n"); + } + ret_value = 0; + goto done; + } + + /* if region and ID are supplied (key optional), write to fapl... + * fail if value would overflow + */ + if (*values[0] != '\0' && + *values[1] != '\0') + { + if (HDstrlen(values[0]) > H5FD__ROS3_MAX_REGION_LEN) { + if (show_progress) { + HDprintf(" ERROR: aws_region value too long\n"); + } + ret_value = 0; + goto done; + } + HDmemcpy(fa->aws_region, values[0], + (HDstrlen(values[0]) + 1)); + if (show_progress) { + HDprintf(" aws_region set\n"); + } + + + if (HDstrlen(values[1]) > H5FD__ROS3_MAX_SECRET_ID_LEN) { + if (show_progress) { + HDprintf(" ERROR: secret_id value too long\n"); + } + ret_value = 0; + goto done; + } + HDmemcpy(fa->secret_id, + values[1], + (HDstrlen(values[1]) + 1)); + if (show_progress) { + HDprintf(" secret_id set\n"); + } + + if (HDstrlen(values[2]) > H5FD__ROS3_MAX_SECRET_KEY_LEN) { + if (show_progress) { + HDprintf(" ERROR: secret_key value too long\n"); + } + ret_value = 0; + goto done; + } + HDmemcpy(fa->secret_key, + values[2], + (HDstrlen(values[2]) + 1)); + if (show_progress) { + HDprintf(" secret_key set\n"); + } + + fa->authenticate = TRUE; + if (show_progress) { + HDprintf(" set to authenticate\n"); + } + + } else if (*values[0] != '\0' || + *values[1] != '\0' || + *values[2] != '\0') + { + if (show_progress) { + HDprintf( + " ERROR: invalid assortment of empty/non-empty values\n" + ); + } + ret_value = 0; + goto done; + } + } /* values != NULL */ + +done: + return ret_value; +#endif /* H5_HAVE_ROS3_VFD */ + +} /* h5tools_populate_ros3_fapl */ + + +/*----------------------------------------------------------------------------- + * + * Function: h5tools_set_configured_fapl + * + * Purpose: prepare fapl_id with the given property list, according to + * VFD prototype. + * + * Return: 0 on failure, 1 on success + * + * Programmer: Jacob Smith + * 2018-05-21 + * + * Changes: None. + * + *----------------------------------------------------------------------------- + */ +int +h5tools_set_configured_fapl(hid_t fapl_id, + const char vfd_name[], + void *fapl_t_ptr) +{ + int ret_value = 1; + + if (fapl_id < 0) { + return 0; + } + + if (!strcmp("", vfd_name)) { + goto done; + +#ifdef H5_HAVE_ROS3_VFD + } else if (!strcmp("ros3", vfd_name)) { + if ((fapl_id == H5P_DEFAULT) || + (fapl_t_ptr == NULL) || + (FAIL == H5Pset_fapl_ros3( + fapl_id, + (H5FD_ros3_fapl_t *)fapl_t_ptr))) + { + ret_value = 0; + goto done; + } +#endif /* H5_HAVE_ROS3_VFD */ + +#ifdef H5_HAVE_LIBHDFS + } else if (!strcmp("hdfs", vfd_name)) { + if ((fapl_id == H5P_DEFAULT) || + (fapl_t_ptr == NULL) || + (FAIL == H5Pset_fapl_hdfs( + fapl_id, + (H5FD_hdfs_fapl_t *)fapl_t_ptr))) + { + ret_value = 0; + goto done; + } +#endif /* H5_HAVE_LIBHDFS */ + + } else { + ret_value = 0; /* unrecognized fapl type "name" */ + } + +done: + return ret_value; + +} /* h5tools_set_configured_fapl() */ + + + diff --git a/tools/lib/h5tools_utils.h b/tools/lib/h5tools_utils.h index 4c2bf1e..1c6ba2a 100644 --- a/tools/lib/h5tools_utils.h +++ b/tools/lib/h5tools_utils.h @@ -123,6 +123,11 @@ H5TOOLS_DLLVAR unsigned h5tools_nCols; /*max number of columns for H5TOOLS_DLL void indentation(unsigned); H5TOOLS_DLL void print_version(const char *progname); H5TOOLS_DLL void parallel_print(const char* format, ... ); +H5TOOLS_DLL herr_t parse_tuple(const char *start, + int sep, + char **cpy_out, + unsigned *nelems, + char ***ptrs_out); H5TOOLS_DLL void error_msg(const char *fmt, ...); H5TOOLS_DLL void warn_msg(const char *fmt, ...); H5TOOLS_DLL void help_ref_msg(FILE *output); @@ -174,6 +179,11 @@ H5TOOLS_DLL void h5tools_setprogname(const char*progname); H5TOOLS_DLL int h5tools_getstatus(void); H5TOOLS_DLL void h5tools_setstatus(int d_status); H5TOOLS_DLL int h5tools_getenv_update_hyperslab_bufsize(void); +H5TOOLS_DLL int h5tools_set_configured_fapl(hid_t fapl_id, + const char vfd_name[], + void *fapl_t_ptr); +H5TOOLS_DLL int h5tools_populate_ros3_fapl(H5FD_ros3_fapl_t *fa, + const char **values); #ifdef __cplusplus } #endif diff --git a/tools/libtest/CMakeLists.txt b/tools/libtest/CMakeLists.txt new file mode 100644 index 0000000..f5e0aa6 --- /dev/null +++ b/tools/libtest/CMakeLists.txt @@ -0,0 +1,21 @@ +cmake_minimum_required (VERSION 3.10) +project (HDF5_TOOLS_LIBTEST C) + +#----------------------------------------------------------------------------- +# Add the h5tools_utils test executables +#----------------------------------------------------------------------------- +add_executable (h5tools_utils ${HDF5_TOOLS_LIBTEST_SOURCE_DIR}/h5tools_utils.c) +target_include_directories(h5tools_utils PRIVATE "${HDF5_TOOLS_DIR}/lib;${HDF5_SRC_DIR};${HDF5_BINARY_DIR};$<$<BOOL:${HDF5_ENABLE_PARALLEL}>:${MPI_C_INCLUDE_DIRS}>") +TARGET_C_PROPERTIES (h5tools_utils STATIC) +target_link_libraries (h5tools_utils PRIVATE ${HDF5_TOOLS_LIB_TARGET} ${HDF5_LIB_TARGET}) +set_target_properties (h5tools_utils PROPERTIES FOLDER tools) + +if (BUILD_SHARED_LIBS) + add_executable (h5tools_utils-shared ${HDF5_TOOLS_LIBTEST_SOURCE_DIR}/h5tools_utils.c) + target_include_directories(h5tools_utils-shared PRIVATE "${HDF5_TOOLS_DIR}/lib;${HDF5_SRC_DIR};${HDF5_BINARY_DIR};$<$<BOOL:${HDF5_ENABLE_PARALLEL}>:${MPI_C_INCLUDE_DIRS}>") + TARGET_C_PROPERTIES (h5tools_utils-shared SHARED) + target_link_libraries (h5tools_utils-shared PRIVATE ${HDF5_TOOLS_LIBSH_TARGET} ${HDF5_LIBSH_TARGET}) + set_target_properties (h5tools_utils-shared PROPERTIES FOLDER tools) +endif () + +include (CMakeTests.cmake) diff --git a/tools/libtest/CMakeTests.cmake b/tools/libtest/CMakeTests.cmake new file mode 100644 index 0000000..403969d --- /dev/null +++ b/tools/libtest/CMakeTests.cmake @@ -0,0 +1,49 @@ +# +# Copyright by The HDF Group. +# All rights reserved. +# +# This file is part of HDF5. The full HDF5 copyright notice, including +# terms governing use, modification, and redistribution, is contained in +# the COPYING file, which can be found at the root of the source code +# distribution tree, or in https://support.hdfgroup.org/ftp/HDF5/releases. +# If you do not have access to either file, you may request a copy from +# help@hdfgroup.org. +# + +############################################################################## +############################################################################## +### T E S T I N G ### +############################################################################## +############################################################################## + + +############################################################################## +############################################################################## +### T H E T E S T S M A C R O S ### +############################################################################## +############################################################################## + + macro (ADD_H5_TEST resultfile resultcode) + add_test ( + NAME H5LIBTEST-${resultfile}-clear-objects + COMMAND ${CMAKE_COMMAND} + -E remove + ${resultfile}.out + ${resultfile}.out.err + ) + if (NOT "${last_test}" STREQUAL "") + set_tests_properties (H5LIBTEST-${resultfile}-clear-objects PROPERTIES DEPENDS ${last_test}) + endif () + add_test (NAME H5LIBTEST-${resultfile} COMMAND $<TARGET_FILE:h5tools_utils> ${ARGN}) + if (NOT "${resultcode}" STREQUAL "0") + set_tests_properties (H5LIBTEST-${resultfile} PROPERTIES WILL_FAIL "true") + endif () + set_tests_properties (H5LIBTEST-${resultfile} PROPERTIES DEPENDS H5LIBTEST-${resultfile}-clear-objects) + endmacro () + +############################################################################## +############################################################################## +### T H E T E S T S ### +############################################################################## +############################################################################## + ADD_H5_TEST (h5tools_utils-default 0) diff --git a/tools/libtest/Makefile.am b/tools/libtest/Makefile.am new file mode 100644 index 0000000..5aa72b8 --- /dev/null +++ b/tools/libtest/Makefile.am @@ -0,0 +1,34 @@ +# +# Read-Only S3 Virtual File Driver (VFD) +# Copyright (c) 2017-2018, The HDF Group. +# +# All rights reserved. +# +# NOTICE: +# All information contained herein is, and remains, the property of The HDF +# Group. The intellectual and technical concepts contained herein are +# proprietary to The HDF Group. Dissemination of this information or +# reproduction of this material is strictly forbidden unless prior written +# permission is obtained from The HDF Group. +## +## Makefile.am +## Run automake to generate a Makefile.in from this file. +# +# HDF5 Library Makefile(.in) +# + +include $(top_srcdir)/config/commence.am + +# Include src and tools/lib directories +AM_CPPFLAGS+=-I$(top_srcdir)/src -I$(top_srcdir)/tools/lib + +# All programs depend on the hdf5 and h5tools libraries +LDADD=$(LIBH5TOOLS) $(LIBHDF5) + + +# main target +bin_PROGRAMS=h5tools_utils +# check_PROGRAMS=$(TEST_PROG) + + +include $(top_srcdir)/config/conclude.am diff --git a/tools/libtest/h5tools_utils.c b/tools/libtest/h5tools_utils.c new file mode 100644 index 0000000..56e8a01 --- /dev/null +++ b/tools/libtest/h5tools_utils.c @@ -0,0 +1,1296 @@ +/* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * Copyright (c) 2017-2018, The HDF Group. * + * * + * All rights reserved. * + * * + * NOTICE: * + * All information contained herein is, and remains, the property of The HDF * + * Group. The intellectual and technical concepts contained herein are * + * proprietary to The HDF Group. Dissemination of this information or * + * reproduction of this material is strictly forbidden unless prior written * + * permission is obtained from The HDF Group. * + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */ + +/* + * Purpose: unit-test functionality of the routines in `tools/lib/h5tools_utils` + * + * Jacob Smith 2017-11-10 + */ + +#include "hdf5.h" +#include "H5private.h" +#include "h5tools_utils.h" +/* #include "h5test.h" */ /* linking failure */ + +#define UTIL_TEST_DEBUG 0 + +#ifndef _H5TEST_ + +#define AT() fprintf(stdout, " at %s:%d in %s()...\n", \ + __FILE__, __LINE__, FUNC); + +#define FAILED(msg) { \ + fprintf(stdout, "*FAILED*"); AT() \ + if (msg == NULL) { \ + fprintf(stdout,"(NULL)\n"); \ + } else { \ + fprintf(stdout, "%s\n", msg); \ + } \ + fflush(stdout); \ +} + +#define TESTING(msg) { \ + fprintf(stdout, "TESTING %-62s", (msg)); \ + fflush(stdout); \ +} + +#define PASSED() { \ + fprintf(stdout, " PASSED\n"); \ + fflush(stdout); \ +} + +#endif /* ifndef _H5TEST_ */ + +#ifndef __js_test__ + +#define __js_test__ 1L + +/***************************************************************************** + * + * FILE-LOCAL TESTING MACROS + * + * Purpose: + * + * 1. Upon test failure, goto-jump to single-location teardown in test + * function. E.g., `error:` (consistency with HDF corpus) or + * `failed:` (reflects purpose). + * >>> using "error", in part because `H5E_BEGIN_TRY` expects it. + * 2. Increase clarity and reduce overhead found with `TEST_ERROR`. + * e.g., "if(somefunction(arg, arg2) < 0) TEST_ERROR:" + * requires reading of entire line to know whether this if/call is + * part of the test setup, test operation, or a test unto itself. + * 3. Provide testing macros with optional user-supplied failure message; + * if not supplied (NULL), generate comparison output in the spirit of + * test-driven development. E.g., "expected 5 but was -3" + * User messages clarify test's purpose in code, encouraging description + * without relying on comments. + * 4. Configurable expected-actual order in generated comparison strings. + * Some prefer `VERIFY(expected, actual)`, others + * `VERIFY(actual, expected)`. Provide preprocessor ifdef switch + * to satifsy both parties, assuming one paradigm per test file. + * (One could #undef and redefine the flag through the file as desired, + * but _why_.) + * + * Provided as courtesy, per consideration for inclusion in the library + * proper. + * + * Macros: + * + * JSVERIFY_EXP_ACT - ifdef flag, configures comparison order + * FAIL_IF() - check condition + * FAIL_UNLESS() - check _not_ condition + * JSVERIFY() - long-int equality check; prints reason/comparison + * JSVERIFY_NOT() - long-int inequality check; prints + * JSVERIFY_STR() - string equality check; prints + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *****************************************************************************/ + + +/*---------------------------------------------------------------------------- + * + * ifdef flag: JSVERIFY_EXP_ACT + * + * JSVERIFY macros accept arguments as (EXPECTED, ACTUAL[, reason]) + * default, if this is undefined, is (ACTUAL, EXPECTED[, reason]) + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_EXP_ACT 1L + + +/*---------------------------------------------------------------------------- + * + * Macro: JSFAILED_AT() + * + * Purpose: + * + * Preface a test failure by printing "*FAILED*" and location to stdout + * Similar to `H5_FAILED(); AT();` from h5test.h + * + * *FAILED* at somefile.c:12 in function_name()... + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSFAILED_AT() { \ + HDprintf("*FAILED* at %s:%d in %s()...\n", __FILE__, __LINE__, FUNC); \ +} + + +/*---------------------------------------------------------------------------- + * + * Macro: FAIL_IF() + * + * Purpose: + * + * Make tests more accessible and less cluttered than + * `if (thing == otherthing()) TEST_ERROR` + * paradigm. + * + * The following lines are roughly equivalent: + * + * `if (myfunc() < 0) TEST_ERROR;` (as seen elsewhere in HDF tests) + * `FAIL_IF(myfunc() < 0)` + * + * Prints a generic "FAILED AT" line to stdout and jumps to `error`, + * similar to `TEST_ERROR` in h5test.h + * + * Programmer: Jacob Smith + * 2017-10-23 + * + *---------------------------------------------------------------------------- + */ +#define FAIL_IF(condition) \ +if (condition) { \ + JSFAILED_AT() \ + goto error; \ +} + + +/*---------------------------------------------------------------------------- + * + * Macro: FAIL_UNLESS() + * + * Purpose: + * + * TEST_ERROR wrapper to reduce cognitive overhead from "negative tests", + * e.g., "a != b". + * + * Opposite of FAIL_IF; fails if the given condition is _not_ true. + * + * `FAIL_IF( 5 != my_op() )` + * is equivalent to + * `FAIL_UNLESS( 5 == my_op() )` + * However, `JSVERIFY(5, my_op(), "bad return")` may be even clearer. + * (see JSVERIFY) + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define FAIL_UNLESS(condition) \ +if (!(condition)) { \ + JSFAILED_AT() \ + goto error; \ +} + + +/*---------------------------------------------------------------------------- + * + * Macro: JSERR_LONG() + * + * Purpose: + * + * Print an failure message for long-int arguments. + * ERROR-AT printed first. + * If `reason` is given, it is printed on own line and newlined after + * else, prints "expected/actual" aligned on own lines. + * + * *FAILED* at myfile.c:488 in somefunc()... + * forest must be made of trees. + * + * or + * + * *FAILED* at myfile.c:488 in somefunc()... + * ! Expected 425 + * ! Actual 3 + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSERR_LONG(expected, actual, reason) { \ + JSFAILED_AT() \ + if (reason!= NULL) { \ + HDprintf("%s\n", (reason)); \ + } else { \ + HDprintf(" ! Expected %ld\n ! Actual %ld\n", \ + (long)(expected), (long)(actual)); \ + } \ +} + + +/*---------------------------------------------------------------------------- + * + * Macro: JSERR_STR() + * + * Purpose: + * + * Print an failure message for string arguments. + * ERROR-AT printed first. + * If `reason` is given, it is printed on own line and newlined after + * else, prints "expected/actual" aligned on own lines. + * + * *FAILED* at myfile.c:421 in myfunc()... + * Blue and Red strings don't match! + * + * or + * + * *FAILED* at myfile.c:421 in myfunc()... + * !!! Expected: + * this is my expected + * string + * !!! Actual: + * not what I expected at all + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSERR_STR(expected, actual, reason) { \ + JSFAILED_AT() \ + if ((reason) != NULL) { \ + HDprintf("%s\n", (reason)); \ + } else { \ + HDprintf("!!! Expected:\n%s\n!!!Actual:\n%s\n", \ + (expected), (actual)); \ + } \ +} + +#ifdef JSVERIFY_EXP_ACT + + +/*---------------------------------------------------------------------------- + * + * Macro: JSVERIFY() + * + * Purpose: + * + * Verify that two long integers are equal. + * If unequal, print failure message + * (with `reason`, if not NULL; expected/actual if NULL) + * and jump to `error` at end of function + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY(expected, actual, reason) \ +if ((long)(actual) != (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)) \ + goto error; \ +} /* JSVERIFY */ + + +/*---------------------------------------------------------------------------- + * + * Macro: JSVERIFY_NOT() + * + * Purpose: + * + * Verify that two long integers are _not_ equal. + * If equal, print failure message + * (with `reason`, if not NULL; expected/actual if NULL) + * and jump to `error` at end of function + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_NOT(expected, actual, reason) \ +if ((long)(actual) == (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)) \ + goto error; \ +} /* JSVERIFY_NOT */ + + +/*---------------------------------------------------------------------------- + * + * Macro: JSVERIFY_STR() + * + * Purpose: + * + * Verify that two strings are equal. + * If unequal, print failure message + * (with `reason`, if not NULL; expected/actual if NULL) + * and jump to `error` at end of function + * + * Programmer: Jacob Smith + * 2017-10-24 + * + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_STR(expected, actual, reason) \ +if (strcmp((actual), (expected)) != 0) { \ + JSERR_STR((expected), (actual), (reason)); \ + goto error; \ +} /* JSVERIFY_STR */ + + +#else /* JSVERIFY_EXP_ACT not defined */ + /* Repeats macros above, but with actual/expected parameters reversed. */ + + +/*---------------------------------------------------------------------------- + * Macro: JSVERIFY() + * See: JSVERIFY documentation above. + * Programmer: Jacob Smith + * 2017-10-14 + *---------------------------------------------------------------------------- + */ +#define JSVERIFY(actual, expected, reason) \ +if ((long)(actual) != (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)); \ + goto error; \ +} /* JSVERIFY */ + + +/*---------------------------------------------------------------------------- + * Macro: JSVERIFY_NOT() + * See: JSVERIFY_NOT documentation above. + * Programmer: Jacob Smith + * 2017-10-14 + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_NOT(actual, expected, reason) \ +if ((long)(actual) == (long)(expected)) { \ + JSERR_LONG((expected), (actual), (reason)) \ + goto error; \ +} /* JSVERIFY_NOT */ + + +/*---------------------------------------------------------------------------- + * Macro: JSVERIFY_STR() + * See: JSVERIFY_STR documentation above. + * Programmer: Jacob Smith + * 2017-10-14 + *---------------------------------------------------------------------------- + */ +#define JSVERIFY_STR(actual, expected, reason) \ +if (strcmp((actual), (expected)) != 0) { \ + JSERR_STR((expected), (actual), (reason)); \ + goto error; \ +} /* JSVERIFY_STR */ + +#endif /* ifdef/else JSVERIFY_EXP_ACT */ + +#endif /* __js_test__ */ + +/* if > 0, be very verbose when performing tests */ +#define H5TOOLS_UTILS_TEST_DEBUG 0 + +/******************/ +/* TEST FUNCTIONS */ +/******************/ + + +/*---------------------------------------------------------------------------- + * + * Function: test_parse_tuple() + * + * Purpose: + * + * Provide unit tests and specification for the `parse_tuple()` function. + * + * Return: + * + * 0 Tests passed. + * 1 Tests failed. + * + * Programmer: Jacob Smith + * 2017-11-11 + * + * Changes: None. + * + *---------------------------------------------------------------------------- + */ +static unsigned +test_parse_tuple(void) +{ + /************************* + * TEST-LOCAL STRUCTURES * + *************************/ + + struct testcase { + const char *test_msg; /* info about test case */ + const char *in_str; /* input string */ + int sep; /* separator "character" */ + herr_t exp_ret; /* expected SUCCEED / FAIL */ + unsigned exp_nelems; /* expected number of elements */ + /* (no more than 7!) */ + const char *exp_elems[7]; /* list of elements (no more than 7!) */ + }; + + /****************** + * TEST VARIABLES * + ******************/ + + struct testcase cases[] = { + { "bad start", + "words(before)", + ';', + FAIL, + 0, + {NULL}, + }, + { "tuple not closed", + "(not ok", + ',', + FAIL, + 0, + {NULL}, + }, + { "empty tuple", + "()", + '-', + SUCCEED, + 1, + {""}, + }, + { "no separator", + "(stuff keeps on going)", + ',', + SUCCEED, + 1, + {"stuff keeps on going"}, + }, + { "4-ple, escaped seperator", + "(elem0,elem1,el\\,em2,elem3)", /* "el\,em" */ + ',', + SUCCEED, + 4, + {"elem0", "elem1", "el,em2", "elem3"}, + }, + { "5-ple, escaped escaped separator", + "(elem0,elem1,el\\\\,em2,elem3)", + ',', + SUCCEED, + 5, + {"elem0", "elem1", "el\\", "em2", "elem3"}, + }, + { "escaped non-comma separator", + "(5-2-7-2\\-6-2)", + '-', + SUCCEED, + 5, + {"5","2","7","2-6","2"}, + }, + { "embedded close-paren", + "(be;fo)re)", + ';', + SUCCEED, + 2, + {"be", "fo)re"}, + }, + { "embedded non-escaping backslash", + "(be;fo\\re)", + ';', + SUCCEED, + 2, + {"be", "fo\\re"}, + }, + { "double close-paren at end", + "(be;fore))", + ';', + SUCCEED, + 2, + {"be", "fore)"}, + }, + { "empty elements", + "(;a1;;a4;)", + ';', + SUCCEED, + 5, + {"", "a1", "", "a4", ""}, + }, + { "nested tuples with different separators", + "((4,e,a);(6,2,a))", + ';', + SUCCEED, + 2, + {"(4,e,a)","(6,2,a)"}, + }, + { "nested tuples with same separators", + "((4,e,a),(6,2,a))", + ',', + SUCCEED, + 6, + {"(4","e","a)","(6","2","a)"}, + }, + { "real-world use case", + "(us-east-2,AKIAIMC3D3XLYXLN5COA,ugs5aVVnLFCErO/8uW14iWE3K5AgXMpsMlWneO/+)", + ',', + SUCCEED, + 3, + {"us-east-2", + "AKIAIMC3D3XLYXLN5COA", + "ugs5aVVnLFCErO/8uW14iWE3K5AgXMpsMlWneO/+"}, + } + }; + struct testcase tc; + unsigned n_tests = 14; + unsigned i = 0; + unsigned count = 0; + unsigned elem_i = 0; + char **parsed = NULL; + char *cpy = NULL; + herr_t success = TRUE; + hbool_t show_progress = FALSE; + + + + TESTING("arbitrary-count tuple parsing"); + +#if H5TOOLS_UTILS_TEST_DEBUG > 0 + show_progress = TRUE; +#endif /* H5TOOLS_UTILS_TEST_DEBUG */ + + /********* + * TESTS * + *********/ + + for (i = 0; i < n_tests; i++) { + + /* SETUP + */ + HDassert(parsed == NULL); + HDassert(cpy == NULL); + tc = cases[i]; + if (show_progress == TRUE) { + printf("testing %d: %s...\n", i, tc.test_msg); + } + + /* VERIFY + */ + success = parse_tuple(tc.in_str, tc.sep, + &cpy, &count, &parsed); + + JSVERIFY( tc.exp_ret, success, "function returned incorrect value" ) + JSVERIFY( tc.exp_nelems, count, NULL ) + if (success == SUCCEED) { + FAIL_IF( parsed == NULL ) + for (elem_i = 0; elem_i < count; elem_i++) { + JSVERIFY_STR( tc.exp_elems[elem_i], parsed[elem_i], NULL ) + } + /* TEARDOWN */ + HDassert(parsed != NULL); + HDassert(cpy != NULL); + free(parsed); + parsed = NULL; + free(cpy); + cpy = NULL; + } else { + FAIL_IF( parsed != NULL ) + } /* if parse_tuple() == SUCCEED or no */ + + } /* for each testcase */ + + PASSED(); + return 0; + +error: + /*********** + * CLEANUP * + ***********/ + + if (parsed != NULL) free(parsed); + if (cpy != NULL) free(cpy); + + return 1; + +} /* test_parse_tuple */ + + +/*---------------------------------------------------------------------------- + * + * Function: test_populate_ros3_fa() + * + * Purpose: Verify behavior of `populate_ros3_fa()` + * + * Return: 0 if test passes + * 1 if failure + * + * Programmer: Jacob Smith + * 2017-11-13 + * + * Changes: None + * + *---------------------------------------------------------------------------- + */ +static unsigned +test_populate_ros3_fa(void) +{ +#ifdef H5_HAVE_ROS3_VFD + /************************* + * TEST-LOCAL STRUCTURES * + *************************/ + + /************************ + * TEST-LOCAL VARIABLES * + ************************/ + + hbool_t show_progress = FALSE; + int bad_version = 0xf87a; /* arbitrarily wrong version number */ +#endif /* H5_HAVE_ROS3_VFD */ + + TESTING("programmatic ros3 fapl population"); + +#ifndef H5_HAVE_ROS3_VFD + puts(" -SKIP-"); + puts(" Read-Only S3 VFD not enabled"); + fflush(stdout); + return 0; +#else +#if H5TOOLS_UTILS_TEST_DEBUG > 0 + show_progress = TRUE; +#endif /* H5TOOLS_UTILS_TEST_DEBUG */ + + HDassert(bad_version != H5FD__CURR_ROS3_FAPL_T_VERSION); + + /********* + * TESTS * + *********/ + + /* NULL fapl config pointer fails + */ + { + const char *values[] = {"x", "y", "z"}; + + if (show_progress) { HDprintf("NULL fapl pointer\n"); } + + JSVERIFY( 0, h5tools_populate_ros3_fapl(NULL, values), + "fapl pointer cannot be null" ) + } + + /* NULL values pointer yields default fapl + */ + { + H5FD_ros3_fapl_t fa = {bad_version, TRUE, "u", "v", "w"}; + + if (show_progress) { HDprintf("NULL values pointer\n"); } + + JSVERIFY( 1, h5tools_populate_ros3_fapl(&fa, NULL), + "NULL values pointer yields \"default\" fapl" ) + JSVERIFY( H5FD__CURR_ROS3_FAPL_T_VERSION, fa.version, NULL ) + JSVERIFY( FALSE, fa.authenticate, NULL ) + JSVERIFY_STR( "", fa.aws_region, NULL ) + JSVERIFY_STR( "", fa.secret_id, NULL ) + JSVERIFY_STR( "", fa.secret_key, NULL ) + } + + /* all-empty values + * yields default fapl + */ + { + H5FD_ros3_fapl_t fa = {bad_version, TRUE, "u", "v", "w"}; + const char *values[] = {"", "", ""}; + + if (show_progress) { HDprintf("all empty values\n"); } + + JSVERIFY( 1, h5tools_populate_ros3_fapl(&fa, values), + "empty values yields \"default\" fapl" ) + JSVERIFY( H5FD__CURR_ROS3_FAPL_T_VERSION, fa.version, NULL ) + JSVERIFY( FALSE, fa.authenticate, NULL ) + JSVERIFY_STR( "", fa.aws_region, NULL ) + JSVERIFY_STR( "", fa.secret_id, NULL ) + JSVERIFY_STR( "", fa.secret_key, NULL ) + } + + /* successfully set fapl with values + * excess value is ignored + */ + { + H5FD_ros3_fapl_t fa = {bad_version, FALSE, "a", "b", "c"}; + const char *values[] = {"x", "y", "z", "a"}; + + if (show_progress) { HDprintf("successful full set\n"); } + + JSVERIFY( 1, h5tools_populate_ros3_fapl(&fa, values), + "four values" ) + JSVERIFY( H5FD__CURR_ROS3_FAPL_T_VERSION, fa.version, NULL ) + JSVERIFY( TRUE, fa.authenticate, NULL ) + JSVERIFY_STR( "x", fa.aws_region, NULL ) + JSVERIFY_STR( "y", fa.secret_id, NULL ) + JSVERIFY_STR( "z", fa.secret_key, NULL ) + } + + /* NULL region + * yeilds default fapl + */ + { + H5FD_ros3_fapl_t fa = {bad_version, FALSE, "a", "b", "c"}; + const char *values[] = {NULL, "y", "z", NULL}; + + if (show_progress) { HDprintf("NULL region\n"); } + + JSVERIFY( 0, h5tools_populate_ros3_fapl(&fa, values), + "could not fill fapl" ) + JSVERIFY( H5FD__CURR_ROS3_FAPL_T_VERSION, fa.version, NULL ) + JSVERIFY( FALSE, fa.authenticate, NULL ) + JSVERIFY_STR( "", fa.aws_region, NULL ) + JSVERIFY_STR( "", fa.secret_id, NULL ) + JSVERIFY_STR( "", fa.secret_key, NULL ) + } + + /* empty region + * yeilds default fapl + */ + { + H5FD_ros3_fapl_t fa = {bad_version, FALSE, "a", "b", "c"}; + const char *values[] = {"", "y", "z", NULL}; + + if (show_progress) { HDprintf("empty region; non-empty id, key\n"); } + + JSVERIFY( 0, h5tools_populate_ros3_fapl(&fa, values), + "could not fill fapl" ) + JSVERIFY( H5FD__CURR_ROS3_FAPL_T_VERSION, fa.version, NULL ) + JSVERIFY( FALSE, fa.authenticate, NULL ) + JSVERIFY_STR( "", fa.aws_region, NULL ) + JSVERIFY_STR( "", fa.secret_id, NULL ) + JSVERIFY_STR( "", fa.secret_key, NULL ) + } + + /* region overflow + * yeilds default fapl + */ + { + H5FD_ros3_fapl_t fa = {bad_version, FALSE, "a", "b", "c"}; + const char *values[] = { + "somewhere over the rainbow not too high " \ + "there is another rainbow bounding some darkened sky", + "y", + "z"}; + + if (show_progress) { HDprintf("region overflow\n"); } + + HDassert(strlen(values[0]) > H5FD__ROS3_MAX_REGION_LEN); + + JSVERIFY( 0, h5tools_populate_ros3_fapl(&fa, values), + "could not fill fapl" ) + JSVERIFY( H5FD__CURR_ROS3_FAPL_T_VERSION, fa.version, NULL ) + JSVERIFY( FALSE, fa.authenticate, NULL ) + JSVERIFY_STR( "", fa.aws_region, NULL ) + JSVERIFY_STR( "", fa.secret_id, NULL ) + JSVERIFY_STR( "", fa.secret_key, NULL ) + } + + /* NULL id + * yields default fapl + */ + { + H5FD_ros3_fapl_t fa = {bad_version, FALSE, "a", "b", "c"}; + const char *values[] = {"x", NULL, "z", NULL}; + + if (show_progress) { HDprintf("NULL id\n"); } + + JSVERIFY( 0, h5tools_populate_ros3_fapl(&fa, values), + "could not fill fapl" ) + JSVERIFY( H5FD__CURR_ROS3_FAPL_T_VERSION, fa.version, NULL ) + JSVERIFY( FALSE, fa.authenticate, NULL ) + JSVERIFY_STR( "", fa.aws_region, NULL ) + JSVERIFY_STR( "", fa.secret_id, NULL ) + JSVERIFY_STR( "", fa.secret_key, NULL ) + } + + /* empty id (non-empty region, key) + * yeilds default fapl + */ + { + H5FD_ros3_fapl_t fa = {bad_version, FALSE, "a", "b", "c"}; + const char *values[] = {"x", "", "z", NULL}; + + if (show_progress) { HDprintf("empty id; non-empty region and key\n"); } + + JSVERIFY( 0, h5tools_populate_ros3_fapl(&fa, values), + "could not fill fapl" ) + JSVERIFY( H5FD__CURR_ROS3_FAPL_T_VERSION, fa.version, NULL ) + JSVERIFY( FALSE, fa.authenticate, NULL ) + JSVERIFY_STR( "", fa.aws_region, NULL ) + JSVERIFY_STR( "", fa.secret_id, NULL ) + JSVERIFY_STR( "", fa.secret_key, NULL ) + } + + /* id overflow + * partial set: region + */ + { + H5FD_ros3_fapl_t fa = {bad_version, FALSE, "a", "b", "c"}; + const char *values[] = { + "x", + "Why is it necessary to solve the problem? " \ + "What benefits will you receive by solving the problem? " \ + "What is the unknown? " \ + "What is it you don't yet understand? " \ + "What is the information you have? " \ + "What isn't the problem? " \ + "Is the information insufficient, redundant, or contradictory? " \ + "Should you draw a diagram or figure of the problem? " \ + "What are the boundaries of the problem? " \ + "Can you separate the various parts of the problem?", + "z"}; + + if (show_progress) { HDprintf("id overflow\n"); } + + HDassert(strlen(values[1]) > H5FD__ROS3_MAX_SECRET_ID_LEN); + + JSVERIFY( 0, h5tools_populate_ros3_fapl(&fa, values), + "could not fill fapl" ) + JSVERIFY( H5FD__CURR_ROS3_FAPL_T_VERSION, fa.version, NULL ) + JSVERIFY( FALSE, fa.authenticate, NULL ) + JSVERIFY_STR( "x", fa.aws_region, NULL ) + JSVERIFY_STR( "", fa.secret_id, NULL ) + JSVERIFY_STR( "", fa.secret_key, NULL ) + } + + /* NULL key + * yields default fapl + */ + { + H5FD_ros3_fapl_t fa = {bad_version, FALSE, "a", "b", "c"}; + const char *values[] = {"x", "y", NULL, NULL}; + + if (show_progress) { HDprintf("NULL key\n"); } + + JSVERIFY( 0, h5tools_populate_ros3_fapl(&fa, values), + "could not fill fapl" ) + JSVERIFY( H5FD__CURR_ROS3_FAPL_T_VERSION, fa.version, NULL ) + JSVERIFY( FALSE, fa.authenticate, NULL ) + JSVERIFY_STR( "", fa.aws_region, NULL ) + JSVERIFY_STR( "", fa.secret_id, NULL ) + JSVERIFY_STR( "", fa.secret_key, NULL ) + } + + /* empty key (non-empty region, id) + * yeilds authenticating fapl + */ + { + H5FD_ros3_fapl_t fa = {bad_version, FALSE, "a", "b", "c"}; + const char *values[] = {"x", "y", "", NULL}; + + if (show_progress) { HDprintf("empty key; non-empty region and id\n"); } + + JSVERIFY( 1, h5tools_populate_ros3_fapl(&fa, values), + "could not fill fapl" ) + JSVERIFY( H5FD__CURR_ROS3_FAPL_T_VERSION, fa.version, NULL ) + JSVERIFY( TRUE, fa.authenticate, NULL ) + JSVERIFY_STR( "x", fa.aws_region, NULL ) + JSVERIFY_STR( "y", fa.secret_id, NULL ) + JSVERIFY_STR( "", fa.secret_key, NULL ) + } + + /* empty key, region (non-empty id) + * yeilds default fapl + */ + { + H5FD_ros3_fapl_t fa = {bad_version, FALSE, "a", "b", "c"}; + const char *values[] = {"", "y", "", NULL}; + + if (show_progress) { HDprintf("empty key and region; non-empty id\n"); } + + JSVERIFY( 0, h5tools_populate_ros3_fapl(&fa, values), + "could not fill fapl" ) + JSVERIFY( H5FD__CURR_ROS3_FAPL_T_VERSION, fa.version, NULL ) + JSVERIFY( FALSE, fa.authenticate, NULL ) + JSVERIFY_STR( "", fa.aws_region, NULL ) + JSVERIFY_STR( "", fa.secret_id, NULL ) + JSVERIFY_STR( "", fa.secret_key, NULL ) + } + + /* empty key, id (non-empty region) + * yeilds default fapl + */ + { + H5FD_ros3_fapl_t fa = {bad_version, FALSE, "a", "b", "c"}; + const char *values[] = {"x", "", "", NULL}; + + if (show_progress) { HDprintf("empty key and id; non-empty region\n"); } + + JSVERIFY( 0, h5tools_populate_ros3_fapl(&fa, values), + "could not fill fapl" ) + JSVERIFY( H5FD__CURR_ROS3_FAPL_T_VERSION, fa.version, NULL ) + JSVERIFY( FALSE, fa.authenticate, NULL ) + JSVERIFY_STR( "", fa.aws_region, NULL ) + JSVERIFY_STR( "", fa.secret_id, NULL ) + JSVERIFY_STR( "", fa.secret_key, NULL ) + } + + /* key overflow + * partial set: region, id + */ + { + H5FD_ros3_fapl_t fa = {bad_version, FALSE, "a", "b", "c"}; + const char *values[] = { + "x", + "y", + "Why is it necessary to solve the problem? " \ + "What benefits will you receive by solving the problem? " \ + "What is the unknown? " \ + "What is it you don't yet understand? " \ + "What is the information you have? " \ + "What isn't the problem? " \ + "Is the information insufficient, redundant, or contradictory? " \ + "Should you draw a diagram or figure of the problem? " \ + "What are the boundaries of the problem? " \ + "Can you separate the various parts of the problem?"}; + + if (show_progress) { HDprintf("key overflow\n"); } + + HDassert(strlen(values[2]) > H5FD__ROS3_MAX_SECRET_KEY_LEN); + + JSVERIFY( 0, h5tools_populate_ros3_fapl(&fa, values), + "could not fill fapl" ) + JSVERIFY( H5FD__CURR_ROS3_FAPL_T_VERSION, fa.version, NULL ) + JSVERIFY( FALSE, fa.authenticate, NULL ) + JSVERIFY_STR( "x", fa.aws_region, NULL ) + JSVERIFY_STR( "y", fa.secret_id, NULL ) + JSVERIFY_STR( "", fa.secret_key, NULL ) + } + + /* use case + */ + { + H5FD_ros3_fapl_t fa = {0, 0, "", "", ""}; + const char *values[] = { + "us-east-2", + "AKIAIMC3D3XLYXLN5COA", + "ugs5aVVnLFCErO/8uW14iWE3K5AgXMpsMlWneO/+" + }; + JSVERIFY( 1, + h5tools_populate_ros3_fapl(&fa, values), + "unable to set use case" ) + JSVERIFY( 1, fa.version, "version check" ) + JSVERIFY( 1, fa.authenticate, "should authenticate" ) + } + + PASSED(); + return 0; + +error : + /*********** + * CLEANUP * + ***********/ + + return 1; + +#endif /* H5_HAVE_ROS3_VFD */ + +} /* test_populate_ros3_fa */ + + +/*---------------------------------------------------------------------------- + * + * Function: test_set_configured_fapl() + * + * Purpose: Verify `h5tools_set_configured_fapl()` with ROS3 VFD + * + * Return: 0 if test passes + * 1 if failure + * + * Programmer: Jacob Smith + * 2018-07-12 + * + * Changes: None + * + *---------------------------------------------------------------------------- + */ +static unsigned +test_set_configured_fapl(void) +{ +#define UTIL_TEST_NOFAPL 1 +#define UTIL_TEST_DEFAULT 2 +#define UTIL_TEST_CREATE 3 + + /************************* + * TEST-LOCAL STRUCTURES * + *************************/ + typedef struct testcase { + const char message[88]; + int expected; + int fapl_choice; + const char vfdname[12]; + void *conf_fa; + } testcase; + + typedef struct other_fa_t { + int a; + int b; + int c; + } other_fa_t; + + /************************ + * TEST-LOCAL VARIABLES * + ************************/ + + hid_t fapl_id = -1; + other_fa_t wrong_fa = {0x432, 0xf82, 0x9093}; + H5FD_ros3_fapl_t ros3_anon_fa = {1, FALSE, "", "", ""}; + H5FD_ros3_fapl_t ros3_auth_fa = { + 1, /* fapl version */ + TRUE, /* authenticate */ + "us-east-1", /* aws region */ + "12345677890abcdef", /* simulate access key ID */ + "oiwnerwe9u0234nJw0-aoj+dsf", /* simulate secret key */ + }; + H5FD_hdfs_fapl_t hdfs_fa = { + 1, /* fapl version */ + "", /* namenode name */ + 0, /* namenode port */ + "", /* kerberos ticket cache */ + "", /* user name */ + 2048, /* stream buffer size */ + }; + unsigned n_cases = 7; /* number of common testcases */ + testcase cases[] = { + { "(common) should fail: no fapl id", + 0, + UTIL_TEST_NOFAPL, + "", + NULL, + }, + { "(common) should fail: no fapl id (with struct)", + 0, + UTIL_TEST_NOFAPL, + "", + &wrong_fa, + }, + { "(common) H5P_DEFAULT with no struct should succeed", + 1, + UTIL_TEST_DEFAULT, + "", + NULL, + }, + { "(common) H5P_DEFAULT with (ignored) struct should succeed", + 1, + UTIL_TEST_DEFAULT, + "", + &wrong_fa, + }, + { "(common) provided fapl entry should not fail", + 1, + UTIL_TEST_CREATE, + "", + NULL, + }, + { "(common) provided fapl entry should not fail; ignores struct", + 1, + UTIL_TEST_CREATE, + "", + &wrong_fa, + }, + { "(common) should fail: unrecoginzed vfd name", + 0, + UTIL_TEST_DEFAULT, + "unknown", + NULL, + }, + +#ifdef H5_HAVE_ROS3_VFD + /* WARNING: add number of ROS3 test cases after array definition + */ + { "(ROS3) should fail: no fapl id, no struct", + 0, + UTIL_TEST_NOFAPL, + "ros3", + NULL, + }, + { "(ROS3) should fail: no fapl id", + 0, + UTIL_TEST_NOFAPL, + "ros3", + &ros3_anon_fa, + }, + { "(ROS3) should fail: no struct", + 0, + UTIL_TEST_CREATE, + "ros3", + NULL, + }, + { "(ROS3) successful set", + 1, + UTIL_TEST_CREATE, + "ros3", + &ros3_anon_fa, + }, + { "(ROS3) should fail: attempt to set DEFAULT fapl", + 0, + UTIL_TEST_DEFAULT, + "ros3", + &ros3_anon_fa, + }, +#endif /* H5_HAVE_ROS3_VFD */ + +#ifdef H5_HAVE_LIBHDFS + /* WARNING: add number of HDFS test cases after array definition + */ + { "(HDFS) should fail: no fapl id, no struct", + 0, + UTIL_TEST_NOFAPL, + "hdfs", + NULL, + }, + { "(HDFS) should fail: no fapl id", + 0, + UTIL_TEST_NOFAPL, + "hdfs", + &hdfs_fa, + }, + { "(HDFS) should fail: no struct", + 0, + UTIL_TEST_CREATE, + "hdfs", + NULL, + }, + { "(HDFS) successful set", + 1, + UTIL_TEST_CREATE, + "hdfs", + &hdfs_fa, + }, + { "(HDFS) should fail: attempt to set DEFAULT fapl", + 0, + UTIL_TEST_DEFAULT, + "hdfs", + &hdfs_fa, + }, +#endif /* H5_HAVE_LIBHDFS */ + + }; /* testcases `cases` array */ + +#ifdef H5_HAVE_ROS3_VFD + n_cases += 5; +#endif /* H5_HAVE_ROS3_VFD */ + +#ifdef H5_HAVE_LIBHDFS + n_cases += 5; +#endif /* H5_HAVE_LIBHDFS */ + + TESTING("programmatic fapl set"); + + for (unsigned i = 0; i < n_cases; i++) { + int result; + testcase C = cases[i]; + + fapl_id = -1; + +#if UTIL_TEST_DEBUG + HDfprintf(stderr, "setup test %d\t%s\n", i, C.message); fflush(stderr); +#endif /* UTIL_TEST_DEBUG */ + + /* per-test setup */ + if (C.fapl_choice == UTIL_TEST_DEFAULT) { + fapl_id = H5P_DEFAULT; + } else if (C.fapl_choice == UTIL_TEST_CREATE) { + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + FAIL_IF( fapl_id < 0 ) + } + +#if UTIL_TEST_DEBUG + HDfprintf(stderr, "before test\n"); fflush(stderr); +#endif /* UTIL_TEST_DEBUG */ + + /* test */ + result = h5tools_set_configured_fapl( + fapl_id, + C.vfdname, + C.conf_fa); + JSVERIFY( result, C.expected, C.message ) + +#if UTIL_TEST_DEBUG + HDfprintf(stderr, "after test\n"); fflush(stderr); +#endif /* UTIL_TEST_DEBUG */ + + /* per-test-teardown */ + if (fapl_id > 0) { + FAIL_IF( FAIL == H5Pclose(fapl_id) ) + } + fapl_id = -1; + +#if UTIL_TEST_DEBUG + HDfprintf(stderr, "after cleanup\n"); fflush(stderr); +#endif /* UTIL_TEST_DEBUG */ + + } + +#if UTIL_TEST_DEBUG + HDfprintf(stderr, "after loop\n"); fflush(stderr); +#endif /* UTIL_TEST_DEBUG */ + + PASSED(); + return 0; + +error : + /*********** + * CLEANUP * + ***********/ + +#if UTIL_TEST_DEBUG + HDfprintf(stderr, "ERROR\n"); fflush(stderr); +#endif /* UTIL_TEST_DEBUG */ + + if (fapl_id > 0) { + (void)H5Pclose(fapl_id); + } + + return 1; + +#undef UTIL_TEST_NOFAPL +#undef UTIL_TEST_DEFAULT +#undef UTIL_TEST_CREATE +} /* test_set_configured_fapl */ + + +/*---------------------------------------------------------------------------- + * + * Function: main() + * + * Purpose: Run all test functions. + * + * Return: 0 iff all test pass + * 1 iff any failures + * + * Programmer: Jacob Smith + * 2017-11-10 + * + * Changes: None. + * + *---------------------------------------------------------------------------- + */ +int +main(void) +{ + unsigned nerrors = 0; + +#ifdef _H5TEST_ + h5reset(); /* h5test? */ +#endif /* _H5TEST_ */ + + HDfprintf(stdout, "Testing h5tools_utils corpus.\n"); + + nerrors += test_parse_tuple(); + nerrors += test_populate_ros3_fa(); + nerrors += test_set_configured_fapl(); + + if (nerrors > 0) { + HDfprintf(stdout, "***** %d h5tools_utils TEST%s FAILED! *****\n", + nerrors, + nerrors > 1 ? "S" : ""); + nerrors = 1; + } else { + HDfprintf(stdout, "All h5tools_utils tests passed\n"); + } + + return (int)nerrors; + +} /* main */ + + diff --git a/tools/src/h5dump/h5dump.c b/tools/src/h5dump/h5dump.c index b9e37e8..a824197 100644 --- a/tools/src/h5dump/h5dump.c +++ b/tools/src/h5dump/h5dump.c @@ -24,6 +24,23 @@ static int doxml = 0; static int useschema = 1; static const char *xml_dtd_uri = NULL; +static H5FD_ros3_fapl_t ros3_fa = { + 1, /* version */ + false, /* authenticate */ + "", /* aws region */ + "", /* access key id */ + "", /* secret access key */ +}; + +static H5FD_hdfs_fapl_t hdfs_fa = { + 1, /* fapl version */ + "localhost", /* namenode name */ + 0, /* namenode port */ + "", /* kerberos ticket cache */ + "", /* user name */ + 2048, /* stream buffer size */ +}; + /* module-scoped variables for XML option */ #define DEFAULT_XSD "http://www.hdfgroup.org/HDF5/XML/schema/HDF5-File.xsd" #define DEFAULT_DTD "http://www.hdfgroup.org/HDF5/XML/DTD/HDF5-File.dtd" @@ -188,6 +205,8 @@ static struct long_options l_opts[] = { { "any_path", require_arg, 'N' }, { "vds-view-first-missing", no_arg, 'v' }, { "vds-gap-size", require_arg, 'G' }, + { "s3-cred", require_arg, '$' }, + { "hdfs-attrs", require_arg, '#' }, { NULL, 0, '\0' } }; @@ -241,6 +260,16 @@ usage(const char *prog) PRINTVALSTREAM(rawoutstream, " -b B, --binary=B Binary file output, of form B\n"); PRINTVALSTREAM(rawoutstream, " -O F, --ddl=F Output ddl text into file F\n"); PRINTVALSTREAM(rawoutstream, " Use blank(empty) filename F to suppress ddl display\n"); + PRINTVALSTREAM(rawoutstream, " --s3-cred=<cred> Supply S3 authentication information to \"ros3\" vfd.\n"); + PRINTVALSTREAM(rawoutstream, " <cred> :: \"(<aws-region>,<access-id>,<access-key>)\"\n"); + PRINTVALSTREAM(rawoutstream, " If absent or <cred> -> \"(,,)\", no authentication.\n"); + PRINTVALSTREAM(rawoutstream, " Has no effect is filedriver is not `ros3'.\n"); + PRINTVALSTREAM(rawoutstream, " --hdfs-attrs=<attrs> Supply configuration information for HDFS file access.\n"); + PRINTVALSTREAM(rawoutstream, " For use with \"--filedriver=hdfs\"\n"); + PRINTVALSTREAM(rawoutstream, " <attrs> :: (<namenode name>,<namenode port>,\n"); + PRINTVALSTREAM(rawoutstream, " <kerberos cache path>,<username>,\n"); + PRINTVALSTREAM(rawoutstream, " <buffer size>)\n"); + PRINTVALSTREAM(rawoutstream, " Any absent attribute will use a default value.\n"); PRINTVALSTREAM(rawoutstream, "--------------- Object Options ---------------\n"); PRINTVALSTREAM(rawoutstream, " -a P, --attribute=P Print the specified attribute\n"); PRINTVALSTREAM(rawoutstream, " If an attribute name contains a slash (/), escape the\n"); @@ -1282,6 +1311,126 @@ end_collect: hand = NULL; h5tools_setstatus(EXIT_SUCCESS); goto done; + + case '$': +#ifndef H5_HAVE_ROS3_VFD + error_msg("Read-Only S3 VFD not enabled.\n"); + h5tools_setstatus(EXIT_FAILURE); + goto done; +#else + /* s3 credential */ + { + char **s3_cred = NULL; + char *s3_cred_string = NULL; + const char *ccred[3]; + unsigned nelems = 0; + if ( FAIL == + parse_tuple(opt_arg, ',', + &s3_cred_string, &nelems, &s3_cred)) + { + error_msg("unable to parse malformed s3 credentials\n"); + usage(h5tools_getprogname()); + free_handler(hand, argc); + hand= NULL; + h5tools_setstatus(EXIT_FAILURE); + goto done; + } + if (nelems != 3) { + error_msg("s3 credentials expects 3 elements\n"); + usage(h5tools_getprogname()); + free_handler(hand, argc); + hand= NULL; + h5tools_setstatus(EXIT_FAILURE); + goto done; + } + ccred[0] = (const char *)s3_cred[0]; + ccred[1] = (const char *)s3_cred[1]; + ccred[2] = (const char *)s3_cred[2]; + if (0 == h5tools_populate_ros3_fapl(&ros3_fa, ccred)) { + error_msg("Invalid S3 credentials\n"); + usage(h5tools_getprogname()); + free_handler(hand, argc); + hand= NULL; + h5tools_setstatus(EXIT_FAILURE); + goto done; + } + HDfree(s3_cred); + HDfree(s3_cred_string); + } /* s3 credential block */ + break; +#endif /* H5_HAVE_ROS3_VFD */ + + case '#': +#ifndef H5_HAVE_LIBHDFS + error_msg("HDFS VFD is not enabled.\n"); + goto error; +#else + { + /* read hdfs properties tuple and store values in `hdfs_fa` + */ + unsigned nelems = 0; + char *props_src = NULL; + char **props = NULL; + unsigned long k = 0; + if (FAIL == parse_tuple( + (const char *)opt_arg, + ',', + &props_src, + &nelems, + &props)) + { + error_msg("unable to parse hdfs properties tuple\n"); + goto error; + } + /* sanity-check tuple count + */ + if (nelems != 5) { + h5tools_setstatus(EXIT_FAILURE); + goto error; + } + /* Populate fapl configuration structure with given + * properties. + * WARNING: No error-checking is done on length of input + * strings... Silent overflow is possible, albeit + * unlikely. + */ + if (strncmp(props[0], "", 1)) { + HDstrncpy(hdfs_fa.namenode_name, + (const char *)props[0], + HDstrlen(props[0])); + } + if (strncmp(props[1], "", 1)) { + k = strtoul((const char *)props[1], NULL, 0); + if (errno == ERANGE) { + h5tools_setstatus(EXIT_FAILURE); + goto error; + } + hdfs_fa.namenode_port = (int32_t)k; + } + if (strncmp(props[2], "", 1)) { + HDstrncpy(hdfs_fa.kerberos_ticket_cache, + (const char *)props[2], + HDstrlen(props[2])); + } + if (strncmp(props[3], "", 1)) { + HDstrncpy(hdfs_fa.user_name, + (const char *)props[3], + HDstrlen(props[3])); + } + if (strncmp(props[4], "", 1)) { + k = strtoul((const char *)props[4], NULL, 0); + if (errno == ERANGE) { + h5tools_setstatus(EXIT_FAILURE); + goto error; + } + hdfs_fa.stream_buffer_size = (int32_t)k; + } + HDfree(props); + HDfree(props_src); + } +#endif /* H5_HAVE_LIBHDFS */ + break; + case '?': default: usage(h5tools_getprogname()); @@ -1354,6 +1503,7 @@ main(int argc, const char *argv[]) { hid_t fid = -1; hid_t gid = -1; + hid_t fapl_id = H5P_DEFAULT; H5E_auto2_t func; H5E_auto2_t tools_func; H5O_info_t oi; @@ -1440,10 +1590,60 @@ main(int argc, const char *argv[]) /* Initialize indexing options */ h5trav_set_index(sort_by, sort_order); + if (driver != NULL) { + void *conf_fa = NULL; + + if (!strcmp(driver, "ros3")) { +#ifndef H5_HAVE_ROS3_VFD + error_msg("Read-Only S3 VFD not enabled.\n"); + h5tools_setstatus(EXIT_FAILURE); + goto done; +#else + conf_fa = (void *)&ros3_fa; +#endif /* H5_HAVE_ROS3_VFD */ + } else if (!HDstrcmp(driver, "hdfs")) { +#ifndef H5_HAVE_LIBHDFS + error_msg("HDFS VFD is not enabled.\n"); + h5tools_setstatus(EXIT_FAILURE); + goto done; +#else + conf_fa = (void *)&hdfs_fa; +#endif /* H5_HAVE_LIBHDFS */ + } + + if (conf_fa != NULL) { + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + if (fapl_id < 0) { + error_msg("unable to create fapl entry\n"); + h5tools_setstatus(EXIT_FAILURE); + goto done; + } + if (0 == h5tools_set_configured_fapl( + fapl_id, + driver, /* guaranteed "ros3" or "hdfs" */ + conf_fa)) /* appropriate to driver */ + { + error_msg("unable to set fapl\n"); + h5tools_setstatus(EXIT_FAILURE); + goto done; + } + } + } /* driver defined */ + while(opt_ind < argc) { fname = HDstrdup(argv[opt_ind++]); - fid = h5tools_fopen(fname, H5F_ACC_RDONLY, H5P_DEFAULT, driver, NULL, 0); + if (fapl_id != H5P_DEFAULT) { + fid = H5Fopen(fname, H5F_ACC_RDONLY, fapl_id); + } else { + fid = h5tools_fopen( + fname, + H5F_ACC_RDONLY, + H5P_DEFAULT, + driver, + NULL, + 0); + } if (fid < 0) { error_msg("unable to open file \"%s\"\n", fname); @@ -1624,6 +1824,11 @@ done: /* Free tables for objects */ table_list_free(); + if (fapl_id != H5P_DEFAULT && 0 < H5Pclose(fapl_id)) { + error_msg("Can't close fapl entry\n"); + h5tools_setstatus(EXIT_FAILURE); + } + if(fid >=0) if (H5Fclose(fid) < 0) h5tools_setstatus(EXIT_FAILURE); @@ -1645,127 +1850,7 @@ done: H5Eset_auto2(H5E_DEFAULT, func, edata); leave(h5tools_getstatus()); -} - -/*------------------------------------------------------------------------- - * Function: h5_fileaccess - * - * Purpose: Returns a file access template which is the default template - * but with a file driver set according to the constant or - * environment variable HDF5_DRIVER - * - * Return: Success: A file access property list - * - * Failure: -1 - * - * Programmer: Robb Matzke - * Thursday, November 19, 1998 - * - * Modifications: - * - *------------------------------------------------------------------------- - */ -hid_t -h5_fileaccess(void) -{ - static const char *multi_letters = "msbrglo"; - const char *val = NULL; - const char *name; - char s[1024]; - hid_t fapl = -1; - - /* First use the environment variable, then the constant */ - val = HDgetenv("HDF5_DRIVER"); -#ifdef HDF5_DRIVER - if (!val) val = HDF5_DRIVER; -#endif - - if ((fapl=H5Pcreate(H5P_FILE_ACCESS))<0) return -1; - if (!val || !*val) return fapl; /*use default*/ - - HDstrncpy(s, val, sizeof s); - s[sizeof(s)-1] = '\0'; - if (NULL==(name=HDstrtok(s, " \t\n\r"))) return fapl; - - if (!HDstrcmp(name, "sec2")) { - /* Unix read() and write() system calls */ - if (H5Pset_fapl_sec2(fapl)<0) return -1; - } - else if (!HDstrcmp(name, "stdio")) { - /* Standard C fread() and fwrite() system calls */ - if (H5Pset_fapl_stdio(fapl)<0) return -1; - } - else if (!HDstrcmp(name, "core")) { - /* In-core temporary file with 1MB increment */ - if (H5Pset_fapl_core(fapl, 1024*1024, FALSE)<0) return -1; - } - else if (!HDstrcmp(name, "split")) { - /* Split meta data and raw data each using default driver */ - if (H5Pset_fapl_split(fapl, "-m.h5", H5P_DEFAULT, "-r.h5", H5P_DEFAULT) < 0) - return -1; - } - else if (!HDstrcmp(name, "multi")) { - /* Multi-file driver, general case of the split driver */ - H5FD_mem_t memb_map[H5FD_MEM_NTYPES]; - hid_t memb_fapl[H5FD_MEM_NTYPES]; - const char *memb_name[H5FD_MEM_NTYPES]; - char sv[H5FD_MEM_NTYPES][1024]; - haddr_t memb_addr[H5FD_MEM_NTYPES]; - H5FD_mem_t mt; - - HDmemset(memb_map, 0, sizeof memb_map); - HDmemset(memb_fapl, 0, sizeof memb_fapl); - HDmemset(memb_name, 0, sizeof memb_name); - HDmemset(memb_addr, 0, sizeof memb_addr); - - if(HDstrlen(multi_letters)==H5FD_MEM_NTYPES) { - for (mt=H5FD_MEM_DEFAULT; mt<H5FD_MEM_NTYPES; H5_INC_ENUM(H5FD_mem_t,mt)) { - memb_fapl[mt] = H5P_DEFAULT; - memb_map[mt] = mt; - sprintf(sv[mt], "%%s-%c.h5", multi_letters[mt]); - memb_name[mt] = sv[mt]; - memb_addr[mt] = (haddr_t)MAX(mt - 1, 0) * (HADDR_MAX / 10); - } - } - else { - error_msg("Bad multi_letters list\n"); - return FAIL; - } - - if (H5Pset_fapl_multi(fapl, memb_map, memb_fapl, memb_name, memb_addr, FALSE) < 0) - return -1; - } - else if (!HDstrcmp(name, "family")) { - hsize_t fam_size = 100*1024*1024; /*100 MB*/ - - /* Family of files, each 1MB and using the default driver */ - if ((val=HDstrtok(NULL, " \t\n\r"))) - fam_size = (hsize_t)(HDstrtod(val, NULL) * 1024*1024); - if (H5Pset_fapl_family(fapl, fam_size, H5P_DEFAULT)<0) - return -1; - } - else if (!HDstrcmp(name, "log")) { - long log_flags = H5FD_LOG_LOC_IO; - - /* Log file access */ - if ((val = HDstrtok(NULL, " \t\n\r"))) - log_flags = HDstrtol(val, NULL, 0); - - if (H5Pset_fapl_log(fapl, NULL, (unsigned)log_flags, 0) < 0) - return -1; - } - else if (!HDstrcmp(name, "direct")) { - /* Substitute Direct I/O driver with sec2 driver temporarily because - * some output has sec2 driver as the standard. */ - if (H5Pset_fapl_sec2(fapl)<0) return -1; - } - else { - /* Unknown driver */ - return -1; - } - - return fapl; -} +} /* main */ /*------------------------------------------------------------------------- @@ -1813,3 +1898,4 @@ add_prefix(char **prfx, size_t *prfx_len, const char *name) HDstrcat(HDstrcat(*prfx, "/"), name); } /* end add_prefix */ + diff --git a/tools/src/h5ls/h5ls.c b/tools/src/h5ls/h5ls.c index 4bc1526..c81da1e 100644 --- a/tools/src/h5ls/h5ls.c +++ b/tools/src/h5ls/h5ls.c @@ -158,7 +158,7 @@ static hbool_t print_int_type(h5tools_str_t *buffer, hid_t type, int ind); static hbool_t print_float_type(h5tools_str_t *buffer, hid_t type, int ind); static herr_t visit_obj(hid_t file, const char *oname, iter_t *iter); - + /*------------------------------------------------------------------------- * Function: usage * @@ -216,6 +216,15 @@ usage (void) PRINTVALSTREAM(rawoutstream, " -V, --version Print version number and exit\n"); PRINTVALSTREAM(rawoutstream, " --vfd=DRIVER Use the specified virtual file driver\n"); PRINTVALSTREAM(rawoutstream, " -x, --hexdump Show raw data in hexadecimal format\n"); + PRINTVALSTREAM(rawoutstream, " --s3-cred=C Supply S3 authentication information to \"ros3\" vfd.\n"); + PRINTVALSTREAM(rawoutstream, " Accepts tuple of \"(<aws-region>,<access-id>,<access-key>)\".\n"); + PRINTVALSTREAM(rawoutstream, " If absent or C->\"(,,)\", defaults to no-authentication.\n"); + PRINTVALSTREAM(rawoutstream, " Has no effect if vfd flag not set to \"ros3\".\n"); + PRINTVALSTREAM(rawoutstream, " --hdfs-attrs=A Supply configuration information to Hadoop VFD.\n"); + PRINTVALSTREAM(rawoutstream, " Accepts tuple of (<namenode name>,<namenode port>,\n"); + PRINTVALSTREAM(rawoutstream, " ...<kerberos cache path>,<username>,<buffer size>)\n"); + PRINTVALSTREAM(rawoutstream, " If absent or A == '(,,,,)', all default values are used.\n"); + PRINTVALSTREAM(rawoutstream, " Has no effect if vfd flag is not 'hdfs'.\n"); PRINTVALSTREAM(rawoutstream, "\n"); PRINTVALSTREAM(rawoutstream, " file/OBJECT\n"); PRINTVALSTREAM(rawoutstream, " Each object consists of an HDF5 file name optionally followed by a\n"); @@ -237,7 +246,7 @@ usage (void) PRINTVALSTREAM(rawoutstream, " Replaced by --enable-error-stack.\n"); } - + /*------------------------------------------------------------------------- * Function: print_string @@ -315,7 +324,7 @@ print_string(h5tools_str_t *buffer, const char *s, hbool_t escape_spaces) return nprint; } - + /*------------------------------------------------------------------------- * Function: print_obj_name * @@ -364,7 +373,7 @@ print_obj_name(h5tools_str_t *buffer, const iter_t *iter, const char *oname, return TRUE; } - + /*------------------------------------------------------------------------- * Function: print_native_type * @@ -489,7 +498,7 @@ print_native_type(h5tools_str_t *buffer, hid_t type, int ind) return TRUE; } - + /*------------------------------------------------------------------------- * Function: print_ieee_type * @@ -527,7 +536,7 @@ print_ieee_type(h5tools_str_t *buffer, hid_t type, int ind) return TRUE; } - + /*------------------------------------------------------------------------- * Function: print_precision * @@ -619,7 +628,7 @@ print_precision(h5tools_str_t *buffer, hid_t type, int ind) } } - + /*------------------------------------------------------------------------- * Function: print_int_type * @@ -693,7 +702,7 @@ print_int_type(h5tools_str_t *buffer, hid_t type, int ind) return TRUE; } - + /*------------------------------------------------------------------------- * Function: print_float_type * @@ -807,7 +816,7 @@ print_float_type(h5tools_str_t *buffer, hid_t type, int ind) return TRUE; } - + /*------------------------------------------------------------------------- * Function: print_cmpd_type * @@ -860,7 +869,7 @@ print_cmpd_type(h5tools_str_t *buffer, hid_t type, int ind) return TRUE; } - + /*------------------------------------------------------------------------- * Function: print_enum_type * @@ -985,7 +994,7 @@ print_enum_type(h5tools_str_t *buffer, hid_t type, int ind) return TRUE; } - + /*------------------------------------------------------------------------- * Function: print_string_type * @@ -1086,7 +1095,7 @@ print_string_type(h5tools_str_t *buffer, hid_t type, int H5_ATTR_UNUSED ind) return TRUE; } - + /*------------------------------------------------------------------------- * Function: print_reference_type * @@ -1124,7 +1133,7 @@ print_reference_type(h5tools_str_t *buffer, hid_t type, int H5_ATTR_UNUSED ind) return TRUE; } - + /*------------------------------------------------------------------------- * Function: print_opaque_type * @@ -1160,7 +1169,7 @@ print_opaque_type(h5tools_str_t *buffer, hid_t type, int ind) return TRUE; } - + /*------------------------------------------------------------------------- * Function: print_vlen_type * @@ -1190,7 +1199,7 @@ print_vlen_type(h5tools_str_t *buffer, hid_t type, int ind) return TRUE; } - + /*--------------------------------------------------------------------------- * Purpose: Print information about an array type * @@ -1237,7 +1246,7 @@ print_array_type(h5tools_str_t *buffer, hid_t type, int ind) return TRUE; } - + /*------------------------------------------------------------------------- * Function: print_bitfield_type * @@ -1345,7 +1354,7 @@ print_type(h5tools_str_t *buffer, hid_t type, int ind) (unsigned long)H5Tget_size(type), (unsigned)data_class); } - + /*------------------------------------------------------------------------- * Function: dump_dataset_values * @@ -1475,7 +1484,7 @@ dump_dataset_values(hid_t dset) PRINTVALSTREAM(rawoutstream, "\n"); } - + /*------------------------------------------------------------------------- * Function: list_attr * @@ -1662,7 +1671,7 @@ list_attr(hid_t obj, const char *attr_name, const H5A_info_t H5_ATTR_UNUSED *ain return 0; } - + /*------------------------------------------------------------------------- * Function: dataset_list1 * @@ -1727,7 +1736,7 @@ dataset_list1(hid_t dset) return 0; } - + /*------------------------------------------------------------------------- * Function: dataset_list2 * @@ -1962,7 +1971,7 @@ dataset_list2(hid_t dset, const char H5_ATTR_UNUSED *name) return 0; } /* end dataset_list2() */ - + /*------------------------------------------------------------------------- * Function: datatype_list2 * @@ -2004,7 +2013,7 @@ datatype_list2(hid_t type, const char H5_ATTR_UNUSED *name) return 0; } - + /*------------------------------------------------------------------------- * Function: list_obj * @@ -2160,7 +2169,7 @@ done: } /* end list_obj() */ - + /*------------------------------------------------------------------------- * Function: list_lnk * @@ -2354,7 +2363,7 @@ done: return 0; } /* end list_lnk() */ - + /*------------------------------------------------------------------------- * Function: visit_obj * @@ -2434,7 +2443,7 @@ done: return retval; } - + /*------------------------------------------------------------------------- * Function: get_width * @@ -2550,7 +2559,7 @@ out: return ret; } - + /*------------------------------------------------------------------------- * Function: leave * @@ -2573,7 +2582,7 @@ leave(int ret) HDexit(ret); } - + /*------------------------------------------------------------------------- * Function: main * @@ -2602,6 +2611,26 @@ main(int argc, const char *argv[]) char drivername[50]; const char *preferred_driver = NULL; int err_exit = 0; + hid_t fapl_id = H5P_DEFAULT; + + /* default "anonymous" s3 configuration */ + H5FD_ros3_fapl_t ros3_fa = { + 1, /* fapl version */ + false, /* authenticate */ + "", /* aws region */ + "", /* access key id */ + "", /* secret access key */ + }; + + /* "default" HDFS configuration */ + H5FD_hdfs_fapl_t hdfs_fa = { + 1, /* fapl version */ + "localhost", /* namenode name */ + 0, /* namenode port */ + "", /* kerberos ticket cache */ + "", /* user name */ + 2048, /* stream buffer size */ + }; h5tools_setprogname(PROGRAMNAME); h5tools_setstatus(EXIT_SUCCESS); @@ -2701,6 +2730,185 @@ main(int argc, const char *argv[]) usage(); leave(EXIT_FAILURE); } + + } else if (!HDstrncmp(argv[argno], "--s3-cred=", (size_t)10)) { +#ifndef H5_HAVE_ROS3_VFD + HDfprintf(rawerrorstream, + "Error: Read-Only S3 VFD is not enabled\n\n"); + usage(); + leave(EXIT_FAILURE); +#else + unsigned nelems = 0; + char *start = NULL; + char *s3cred_src = NULL; + char **s3cred = NULL; + char const *ccred[3]; + /* try to parse s3 credentials tuple + */ + start = strchr(argv[argno], '='); + if (start == NULL) { + HDfprintf(rawerrorstream, + "Error: Unable to parse null credentials tuple\n" + " For anonymous access, omit \"--s3-cred\" and use" + "only \"--vfd=ros3\"\n\n"); + usage(); + leave(EXIT_FAILURE); + } + start++; + if (FAIL == + parse_tuple((const char *)start, ',', + &s3cred_src, &nelems, &s3cred)) + { + HDfprintf(rawerrorstream, + "Error: Unable to parse S3 credentials\n\n"); + usage(); + leave(EXIT_FAILURE); + } + /* sanity-check tuple count + */ + if (nelems != 3) { + HDfprintf(rawerrorstream, + "Error: Invalid S3 credentials\n\n"); + usage(); + leave(EXIT_FAILURE); + } + ccred[0] = (const char *)s3cred[0]; + ccred[1] = (const char *)s3cred[1]; + ccred[2] = (const char *)s3cred[2]; + if (0 == h5tools_populate_ros3_fapl(&ros3_fa, ccred)) { + HDfprintf(rawerrorstream, + "Error: Invalid S3 credentials\n\n"); + usage(); + leave(EXIT_FAILURE); + } + HDfree(s3cred); + HDfree(s3cred_src); +#endif /* H5_HAVE_ROS3_VFD */ + + } else if (!HDstrncmp(argv[argno], "--hdfs-attrs=", (size_t)13)) { +#ifndef H5_HAVE_LIBHDFS + PRINTVALSTREAM(rawoutstream, "The HDFS VFD is not enabled.\n"); + leave(EXIT_FAILURE); +#else + /* Parse received configuration data and set fapl config struct + */ + + hbool_t _debug = FALSE; + unsigned nelems = 0; + char const *start = NULL; + char *props_src = NULL; + char **props = NULL; + unsigned long k = 0; + + /* try to parse tuple + */ + if (_debug) { + HDfprintf(stderr, "configuring hdfs...\n"); + } + start = argv[argno]+13; /* should never segfault: worst case of */ + if (*start != '(') /* null-termintor after '='. */ + { + if (_debug) { + HDfprintf(stderr, " no tuple.\n"); + } + usage(); + leave(EXIT_FAILURE); + } + if (FAIL == + parse_tuple((const char *)start, ',', + &props_src, &nelems, &props)) + { + HDfprintf(stderr, + " unable to parse tuple.\n"); + usage(); + leave(EXIT_FAILURE); + } + + /* sanity-check tuple count + */ + if (nelems != 5) { + HDfprintf(stderr, + " expected 5-ple, got `%d`\n", + nelems); + usage(); + leave(EXIT_FAILURE); + } + if (_debug) { + HDfprintf(stderr, + " got hdfs-attrs tuple: `(%s,%s,%s,%s,%s)`\n", + props[0], + props[1], + props[2], + props[3], + props[4]); + } + + /* Populate fapl configuration structure with given properties. + * WARNING: No error-checking is done on length of input strings... + * Silent overflow is possible, albeit unlikely. + */ + if (HDstrncmp(props[0], "", 1)) { + if (_debug) { + HDfprintf(stderr, + " setting namenode name: %s\n", + props[0]); + } + HDstrncpy(hdfs_fa.namenode_name, + (const char *)props[0], + HDstrlen(props[0])); + } + if (HDstrncmp(props[1], "", 1)) { + k = strtoul((const char *)props[1], NULL, 0); + if (errno == ERANGE) { + HDfprintf(stderr, + " supposed port number wasn't.\n"); + leave(EXIT_FAILURE); + } + if (_debug) { + HDfprintf(stderr, + " setting namenode port: %lu\n", + k); + } + hdfs_fa.namenode_port = (int32_t)k; + } + if (HDstrncmp(props[2], "", 1)) { + if (_debug) { + HDfprintf(stderr, + " setting kerb cache path: %s\n", + props[2]); + } + HDstrncpy(hdfs_fa.kerberos_ticket_cache, + (const char *)props[2], + HDstrlen(props[2])); + } + if (HDstrncmp(props[3], "", 1)) { + if (_debug) { + HDfprintf(stderr, + " setting username: %s\n", + props[3]); + } + HDstrncpy(hdfs_fa.user_name, + (const char *)props[3], + HDstrlen(props[3])); + } + if (HDstrncmp(props[4], "", 1)) { + k = HDstrtoul((const char *)props[4], NULL, 0); + if (errno == ERANGE) { + HDfprintf(stderr, + " supposed buffersize number wasn't.\n"); + leave(EXIT_FAILURE); + } + if (_debug) { + HDfprintf(stderr, + " setting stream buffer size: %lu\n", + k); + } + hdfs_fa.stream_buffer_size = (int32_t)k; + } + HDfree(props); + HDfree(props_src); +#endif /* H5_HAVE_LIBHDFS */ + } else if('-'!=argv[argno][1]) { /* Single-letter switches */ for(s = argv[argno] + 1; *s; s++) { @@ -2772,6 +2980,7 @@ main(int argc, const char *argv[]) } /* end switch */ } /* end for */ } else { + HDfprintf(stderr, "Unknown argument: %s\n", argv[argno]); usage(); leave(EXIT_FAILURE); } @@ -2791,6 +3000,49 @@ main(int argc, const char *argv[]) leave(EXIT_FAILURE); } + if (preferred_driver) { + void *conf_fa = NULL; + + if (!HDstrcmp(preferred_driver, "ros3")) { +#ifndef H5_HAVE_ROS3_VFD + HDfprintf(rawerrorstream, + "Error: Read-Only S3 VFD not enabled.\n\n"); + usage(); + leave(EXIT_FAILURE); +#else + conf_fa = (void *)&ros3_fa; +#endif /* H5_HAVE_ROS3_VFD */ + + } else if (!HDstrcmp(preferred_driver, "hdfs")) { +#ifndef H5_HAVE_LIBHDFS + PRINTVALSTREAM(rawoutstream, "The HDFS VFD is not enabled.\n"); + leave(EXIT_FAILURE); +#else + conf_fa = (void *)&hdfs_fa; +#endif /* H5_HAVE_LIBHDFS */ + } + + if (conf_fa != NULL) { + HDassert(fapl_id == H5P_DEFAULT); + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + if (fapl_id < 0) { + HDfprintf(rawerrorstream, + "Error: Unable to create fapl entry\n\n"); + leave(EXIT_FAILURE); + } + if (0 == h5tools_set_configured_fapl( + fapl_id, + preferred_driver, + conf_fa)) + { + HDfprintf(rawerrorstream, + "Error: Unable to set fapl\n\n"); + usage(); + leave(EXIT_FAILURE); + } + } + } /* preferred_driver defined */ + /* Turn off HDF5's automatic error printing unless you're debugging h5ls */ if(!show_errors_g) H5Eset_auto2(H5E_DEFAULT, NULL, NULL); @@ -2820,7 +3072,12 @@ main(int argc, const char *argv[]) file = -1; while(fname && *fname) { - file = h5tools_fopen(fname, H5F_ACC_RDONLY, H5P_DEFAULT, preferred_driver, drivername, sizeof drivername); + if (fapl_id != H5P_DEFAULT) { + file = H5Fopen(fname, H5F_ACC_RDONLY, fapl_id); + } + else { + file = h5tools_fopen(fname, H5F_ACC_RDONLY, H5P_DEFAULT, preferred_driver, drivername, sizeof drivername); + } if(file >= 0) { if(verbose_g) @@ -2933,6 +3190,14 @@ main(int argc, const char *argv[]) err_exit = 1; } /* end while */ + if (fapl_id != H5P_DEFAULT) { + if (0 < H5Pclose(fapl_id)) { + HDfprintf(rawerrorstream, + "Error: Unable to set close fapl entry\n\n"); + leave(EXIT_FAILURE); + } + } + if (err_exit) leave(EXIT_FAILURE); else diff --git a/tools/src/h5stat/h5stat.c b/tools/src/h5stat/h5stat.c index 450f731..59038a0 100644 --- a/tools/src/h5stat/h5stat.c +++ b/tools/src/h5stat/h5stat.c @@ -74,7 +74,7 @@ typedef struct iter_t { ohdr_info_t group_ohdr_info; /* Object header information for groups */ hsize_t max_attrs; /* Maximum attributes from a group */ - unsigned long *num_small_attrs; /* Size of small attributes tracked */ + unsigned long *num_small_attrs; /* Size of small attributes tracked */ unsigned attr_nbins; /* Number of bins for attribute counts */ unsigned long *attr_bins; /* Pointer to array of bins for attribute counts */ @@ -118,6 +118,29 @@ typedef struct iter_t { } iter_t; +static const char *drivername = ""; + +/* default "anonymous" s3 configuration + */ +static H5FD_ros3_fapl_t ros3_fa = { + 1, /* fapl version */ + false, /* authenticate */ + "", /* aws region */ + "", /* access key id */ + "", /* secret access key */ +}; + +/* default HDFS access configuration + */ +static H5FD_hdfs_fapl_t hdfs_fa = { + 1, /* fapl version */ + "localhost", /* namenode name */ + 0, /* namenode port */ + "", /* kerberos ticket cache */ + "", /* user name */ + 2048, /* stream buffer size */ +}; + static int display_all = TRUE; /* Enable the printing of selected statistics */ @@ -146,7 +169,7 @@ struct handler_t { char **obj; }; -static const char *s_opts ="Aa:Ddm:EFfhGgl:sSTO:V"; +static const char *s_opts ="Aa:Ddm:EFfhGgl:sSTO:Vw:"; /* e.g. "filemetadata" has to precede "file"; "groupmetadata" has to precede "group" etc. */ static struct long_options l_opts[] = { {"help", no_arg, 'h'}, @@ -246,6 +269,8 @@ static struct long_options l_opts[] = { { "summ", no_arg, 'S' }, { "sum", no_arg, 'S' }, { "su", no_arg, 'S' }, + { "s3-cred", require_arg, 'w' }, + { "hdfs-attrs", require_arg, 'H' }, { NULL, 0, '\0' } }; @@ -257,7 +282,7 @@ leave(int ret) } - + /*------------------------------------------------------------------------- * Function: usage * @@ -295,9 +320,19 @@ static void usage(const char *prog) HDfprintf(stdout, " -s, --freespace Print free space information\n"); HDfprintf(stdout, " -S, --summary Print summary of file space information\n"); HDfprintf(stdout, " --enable-error-stack Prints messages from the HDF5 error stack as they occur\n"); + HDfprintf(stdout, " --s3-cred=<cred> Access file on S3, using provided credential\n"); + HDfprintf(stdout, " <cred> :: (region,id,key)\n"); + HDfprintf(stdout, " If <cred> == \"(,,)\", no authentication is used.\n"); + HDfprintf(stdout, " --hdfs-attrs=<attrs> Access a file on HDFS with given configuration\n"); + HDfprintf(stdout, " attributes.\n"); + HDfprintf(stdout, " <attrs> :: (<namenode name>,<namenode port>,\n"); + HDfprintf(stdout, " <kerberos cache path>,<username>,\n"); + HDfprintf(stdout, " <buffer size>)\n"); + HDfprintf(stdout, " If an attribute is empty, a default value will be\n"); + HDfprintf(stdout, " used.\n"); } - + /*------------------------------------------------------------------------- * Function: ceil_log10 * @@ -324,7 +359,7 @@ ceil_log10(unsigned long x) return ret; } /* ceil_log10() */ - + /*------------------------------------------------------------------------- * Function: attribute_stats * @@ -374,7 +409,7 @@ attribute_stats(iter_t *iter, const H5O_info_t *oi) return 0; } /* end attribute_stats() */ - + /*------------------------------------------------------------------------- * Function: group_stats * @@ -456,7 +491,7 @@ done: return ret_value; } /* end group_stats() */ - + /*------------------------------------------------------------------------- * Function: dataset_stats * @@ -647,7 +682,7 @@ done: return ret_value; } /* end dataset_stats() */ - + /*------------------------------------------------------------------------- * Function: datatype_stats * @@ -679,7 +714,7 @@ done: return ret_value; } /* end datatype_stats() */ - + /*------------------------------------------------------------------------- * Function: obj_stats * @@ -735,7 +770,7 @@ done: return ret_value; } /* end obj_stats() */ - + /*------------------------------------------------------------------------- * Function: lnk_stats * @@ -833,7 +868,7 @@ freespace_stats(hid_t fid, iter_t *iter) return 0; } /* end freespace_stats() */ - + /*------------------------------------------------------------------------- * Function: hand_free * @@ -862,7 +897,7 @@ hand_free(struct handler_t *hand) } /* end if */ } /* end hand_free() */ - + /*------------------------------------------------------------------------- * Function: parse_command_line * @@ -1014,6 +1049,119 @@ parse_command_line(int argc, const char *argv[], struct handler_t **hand_ret) } /* end if */ break; + case 'w': +#ifndef H5_HAVE_ROS3_VFD + error_msg("Read-Only S3 VFD not enabled.\n"); + goto error; +#else + { + char *cred_str = NULL; + unsigned nelems = 0; + char **cred = NULL; + char const *ccred[3]; + + if (FAIL == parse_tuple((const char *)opt_arg, ',', + &cred_str, &nelems, &cred)) { + error_msg("Unable to parse s3 credential\n"); + goto error; + } + if (nelems != 3) { + error_msg("s3 credential must have three elements\n"); + goto error; + } + ccred[0] = (const char *)cred[0]; + ccred[1] = (const char *)cred[1]; + ccred[2] = (const char *)cred[2]; + if (0 == + h5tools_populate_ros3_fapl(&ros3_fa, ccred)) + { + error_msg("Unable to set ros3 fapl config\n"); + goto error; + } + HDfree(cred); + HDfree(cred_str); + } /* parse s3-cred block */ + drivername = "ros3"; + break; +#endif /* H5_HAVE_ROS3_VFD */ + + case 'H': +#ifndef H5_HAVE_LIBHDFS + error_msg("HDFS VFD is not enabled.\n"); + goto error; +#else + { + unsigned nelems = 0; + char *props_src = NULL; + char **props = NULL; + unsigned long k = 0; + if (FAIL == parse_tuple( + (const char *)opt_arg, + ',', + &props_src, + &nelems, + &props)) + { + error_msg("unable to parse hdfs properties tuple\n"); + goto error; + } + /* sanity-check tuple count + */ + if (nelems != 5) { + char str[64] = ""; + sprintf(str, + "expected 5 elements in hdfs properties tuple " + "but found %u\n", + nelems); + HDfree(props); + HDfree(props_src); + error_msg(str); + goto error; + } + /* Populate fapl configuration structure with given + * properties. + * TODO/WARNING: No error-checking is done on length of + * input strings... Silent overflow is possible, + * albeit unlikely. + */ + if (strncmp(props[0], "", 1)) { + HDstrncpy(hdfs_fa.namenode_name, + (const char *)props[0], + HDstrlen(props[0])); + } + if (strncmp(props[1], "", 1)) { + k = strtoul((const char *)props[1], NULL, 0); + if (errno == ERANGE) { + error_msg("supposed port number wasn't.\n"); + goto error; + } + hdfs_fa.namenode_port = (int32_t)k; + } + if (strncmp(props[2], "", 1)) { + HDstrncpy(hdfs_fa.kerberos_ticket_cache, + (const char *)props[2], + HDstrlen(props[2])); + } + if (strncmp(props[3], "", 1)) { + HDstrncpy(hdfs_fa.user_name, + (const char *)props[3], + HDstrlen(props[3])); + } + if (strncmp(props[4], "", 1)) { + k = strtoul((const char *)props[4], NULL, 0); + if (errno == ERANGE) { + error_msg("supposed buffersize number wasn't.\n"); + goto error; + } + hdfs_fa.stream_buffer_size = (int32_t)k; + } + HDfree(props); + HDfree(props_src); + drivername = "hdfs"; + } + break; +#endif /* H5_HAVE_LIBHDFS */ + default: usage(h5tools_getprogname()); goto error; @@ -1040,7 +1188,7 @@ error: return -1; } - + /*------------------------------------------------------------------------- * Function: iter_free * @@ -1105,7 +1253,7 @@ iter_free(iter_t *iter) } /* end if */ } /* end iter_free() */ - + /*------------------------------------------------------------------------- * Function: print_file_info * @@ -1137,7 +1285,7 @@ print_file_info(const iter_t *iter) return 0; } /* print_file_info() */ - + /*------------------------------------------------------------------------- * Function: print_file_metadata * @@ -1197,7 +1345,7 @@ print_file_metadata(const iter_t *iter) return 0; } /* print_file_metadata() */ - + /*------------------------------------------------------------------------- * Function: print_group_info * @@ -1254,7 +1402,7 @@ print_group_info(const iter_t *iter) return 0; } /* print_group_info() */ - + /*------------------------------------------------------------------------- * Function: print_group_metadata * @@ -1281,7 +1429,7 @@ print_group_metadata(const iter_t *iter) return 0; } /* print_group_metadata() */ - + /*------------------------------------------------------------------------- * Function: print_dataset_info * @@ -1368,7 +1516,7 @@ print_dataset_info(const iter_t *iter) return 0; } /* print_dataset_info() */ - + /*------------------------------------------------------------------------- * Function: print_dataset_metadata * @@ -1397,7 +1545,7 @@ print_dset_metadata(const iter_t *iter) return 0; } /* print_dset_metadata() */ - + /*------------------------------------------------------------------------- * Function: print_dset_dtype_meta * @@ -1438,7 +1586,7 @@ print_dset_dtype_meta(const iter_t *iter) return 0; } /* print_dset_dtype_meta() */ - + /*------------------------------------------------------------------------- * Function: print_attr_info * @@ -1487,7 +1635,7 @@ print_attr_info(const iter_t *iter) return 0; } /* print_attr_info() */ - + /*------------------------------------------------------------------------- * Function: print_freespace_info * @@ -1537,7 +1685,7 @@ print_freespace_info(const iter_t *iter) return 0; } /* print_freespace_info() */ - + /*------------------------------------------------------------------------- * Function: print_storage_summary * @@ -1601,7 +1749,7 @@ print_storage_summary(const iter_t *iter) return 0; } /* print_storage_summary() */ - + /*------------------------------------------------------------------------- * Function: print_file_statistics * @@ -1648,7 +1796,7 @@ print_file_statistics(const iter_t *iter) if(display_summary) print_storage_summary(iter); } /* print_file_statistics() */ - + /*------------------------------------------------------------------------- * Function: print_object_statistics * @@ -1671,7 +1819,7 @@ print_object_statistics(const char *name) printf("Object name %s\n", name); } /* print_object_statistics() */ - + /*------------------------------------------------------------------------- * Function: print_statistics * @@ -1697,7 +1845,7 @@ print_statistics(const char *name, const iter_t *iter) print_file_statistics(iter); } /* print_statistics() */ - + /*------------------------------------------------------------------------- * Function: main * @@ -1718,6 +1866,7 @@ main(int argc, const char *argv[]) void *edata; void *tools_edata; struct handler_t *hand = NULL; + hid_t fapl_id = H5P_DEFAULT; h5tools_setprogname(PROGRAMNAME); h5tools_setstatus(EXIT_SUCCESS); @@ -1738,6 +1887,45 @@ main(int argc, const char *argv[]) if(parse_command_line(argc, argv, &hand) < 0) goto done; + /* if drivername is not null, probably need to set the fapl */ + if (HDstrcmp(drivername, "")) { + void *conf_fa = NULL; + + if (!HDstrcmp(drivername, "ros3")) { +#ifndef H5_HAVE_ROS3_VFD + error_msg("Read-Only S3 VFD not enabled.\n\n"); + goto done; +#else + conf_fa = (void *)&ros3_fa; +#endif /* H5_HAVE_ROS3_VFD */ + + } else if (!HDstrcmp(drivername, "hdfs")) { +#ifndef H5_HAVE_LIBHDFS + error_msg("HDFS VFD not enabled.\n\n"); + goto done; +#else + conf_fa = (void *)&hdfs_fa; +#endif /* H5_HAVE_LIBHDFS */ + } + + if (conf_fa != NULL) { + HDassert(fapl_id == H5P_DEFAULT); + fapl_id = H5Pcreate(H5P_FILE_ACCESS); + if (fapl_id < 0) { + error_msg("Unable to create fapl entry\n"); + goto done; + } + if (1 > h5tools_set_configured_fapl( + fapl_id, + drivername, + conf_fa)) + { + error_msg("Unable to set fapl\n"); + goto done; + } + } + } /* drivername set */ + fname = argv[opt_ind]; if(enable_error_stack > 0) { @@ -1752,7 +1940,7 @@ main(int argc, const char *argv[]) printf("Filename: %s\n", fname); - fid = H5Fopen(fname, H5F_ACC_RDONLY, H5P_DEFAULT); + fid = H5Fopen(fname, H5F_ACC_RDONLY, fapl_id); if(fid < 0) { error_msg("unable to open file \"%s\"\n", fname); h5tools_setstatus(EXIT_FAILURE); @@ -1833,6 +2021,13 @@ done: /* Free iter structure */ iter_free(&iter); + if (fapl_id != H5P_DEFAULT) { + if (0 < H5Pclose(fapl_id)) { + error_msg("unable to close fapl entry\n"); + h5tools_setstatus(EXIT_FAILURE); + } + } + if(fid >= 0 && H5Fclose(fid) < 0) { error_msg("unable to close file \"%s\"\n", fname); h5tools_setstatus(EXIT_FAILURE); diff --git a/tools/test/h5stat/testfiles/h5stat_help1.ddl b/tools/test/h5stat/testfiles/h5stat_help1.ddl index 01e39af..2ba7772 100644 --- a/tools/test/h5stat/testfiles/h5stat_help1.ddl +++ b/tools/test/h5stat/testfiles/h5stat_help1.ddl @@ -23,3 +23,13 @@ Usage: h5stat [OPTIONS] file -s, --freespace Print free space information -S, --summary Print summary of file space information --enable-error-stack Prints messages from the HDF5 error stack as they occur + --s3-cred=<cred> Access file on S3, using provided credential + <cred> :: (region,id,key) + If <cred> == "(,,)", no authentication is used. + --hdfs-attrs=<attrs> Access a file on HDFS with given configuration + attributes. + <attrs> :: (<namenode name>,<namenode port>, + <kerberos cache path>,<username>, + <buffer size>) + If an attribute is empty, a default value will be + used. diff --git a/tools/test/h5stat/testfiles/h5stat_help2.ddl b/tools/test/h5stat/testfiles/h5stat_help2.ddl index 01e39af..2ba7772 100644 --- a/tools/test/h5stat/testfiles/h5stat_help2.ddl +++ b/tools/test/h5stat/testfiles/h5stat_help2.ddl @@ -23,3 +23,13 @@ Usage: h5stat [OPTIONS] file -s, --freespace Print free space information -S, --summary Print summary of file space information --enable-error-stack Prints messages from the HDF5 error stack as they occur + --s3-cred=<cred> Access file on S3, using provided credential + <cred> :: (region,id,key) + If <cred> == "(,,)", no authentication is used. + --hdfs-attrs=<attrs> Access a file on HDFS with given configuration + attributes. + <attrs> :: (<namenode name>,<namenode port>, + <kerberos cache path>,<username>, + <buffer size>) + If an attribute is empty, a default value will be + used. diff --git a/tools/test/h5stat/testfiles/h5stat_nofile.ddl b/tools/test/h5stat/testfiles/h5stat_nofile.ddl index 01e39af..2ba7772 100644 --- a/tools/test/h5stat/testfiles/h5stat_nofile.ddl +++ b/tools/test/h5stat/testfiles/h5stat_nofile.ddl @@ -23,3 +23,13 @@ Usage: h5stat [OPTIONS] file -s, --freespace Print free space information -S, --summary Print summary of file space information --enable-error-stack Prints messages from the HDF5 error stack as they occur + --s3-cred=<cred> Access file on S3, using provided credential + <cred> :: (region,id,key) + If <cred> == "(,,)", no authentication is used. + --hdfs-attrs=<attrs> Access a file on HDFS with given configuration + attributes. + <attrs> :: (<namenode name>,<namenode port>, + <kerberos cache path>,<username>, + <buffer size>) + If an attribute is empty, a default value will be + used. diff --git a/tools/testfiles/h5dump-help.txt b/tools/testfiles/h5dump-help.txt index 19de76f..95dfc3b 100644 --- a/tools/testfiles/h5dump-help.txt +++ b/tools/testfiles/h5dump-help.txt @@ -12,6 +12,16 @@ usage: h5dump [OPTIONS] files -b B, --binary=B Binary file output, of form B -O F, --ddl=F Output ddl text into file F Use blank(empty) filename F to suppress ddl display + --s3-cred=<cred> Supply S3 authentication information to "ros3" vfd. + <cred> :: "(<aws-region>,<access-id>,<access-key>)" + If absent or <cred> -> "(,,)", no authentication. + Has no effect is filedriver is not `ros3'. + --hdfs-attrs=<attrs> Supply configuration information for HDFS file access. + For use with "--filedriver=hdfs" + <attrs> :: (<namenode name>,<namenode port>, + <kerberos cache path>,<username>, + <buffer size>) + Any absent attribute will use a default value. --------------- Object Options --------------- -a P, --attribute=P Print the specified attribute If an attribute name contains a slash (/), escape the diff --git a/tools/testfiles/help-1.ls b/tools/testfiles/help-1.ls index 491f696..396bed3 100644 --- a/tools/testfiles/help-1.ls +++ b/tools/testfiles/help-1.ls @@ -37,6 +37,15 @@ usage: h5ls [OPTIONS] file[/OBJECT] [file[/[OBJECT]...] -V, --version Print version number and exit --vfd=DRIVER Use the specified virtual file driver -x, --hexdump Show raw data in hexadecimal format + --s3-cred=C Supply S3 authentication information to "ros3" vfd. + Accepts tuple of "(<aws-region>,<access-id>,<access-key>)". + If absent or C->"(,,)", defaults to no-authentication. + Has no effect if vfd flag not set to "ros3". + --hdfs-attrs=A Supply configuration information to Hadoop VFD. + Accepts tuple of (<namenode name>,<namenode port>, + ...<kerberos cache path>,<username>,<buffer size>) + If absent or A == '(,,,,)', all default values are used. + Has no effect if vfd flag is not 'hdfs'. file/OBJECT Each object consists of an HDF5 file name optionally followed by a diff --git a/tools/testfiles/help-2.ls b/tools/testfiles/help-2.ls index 491f696..396bed3 100644 --- a/tools/testfiles/help-2.ls +++ b/tools/testfiles/help-2.ls @@ -37,6 +37,15 @@ usage: h5ls [OPTIONS] file[/OBJECT] [file[/[OBJECT]...] -V, --version Print version number and exit --vfd=DRIVER Use the specified virtual file driver -x, --hexdump Show raw data in hexadecimal format + --s3-cred=C Supply S3 authentication information to "ros3" vfd. + Accepts tuple of "(<aws-region>,<access-id>,<access-key>)". + If absent or C->"(,,)", defaults to no-authentication. + Has no effect if vfd flag not set to "ros3". + --hdfs-attrs=A Supply configuration information to Hadoop VFD. + Accepts tuple of (<namenode name>,<namenode port>, + ...<kerberos cache path>,<username>,<buffer size>) + If absent or A == '(,,,,)', all default values are used. + Has no effect if vfd flag is not 'hdfs'. file/OBJECT Each object consists of an HDF5 file name optionally followed by a diff --git a/tools/testfiles/help-3.ls b/tools/testfiles/help-3.ls index 491f696..396bed3 100644 --- a/tools/testfiles/help-3.ls +++ b/tools/testfiles/help-3.ls @@ -37,6 +37,15 @@ usage: h5ls [OPTIONS] file[/OBJECT] [file[/[OBJECT]...] -V, --version Print version number and exit --vfd=DRIVER Use the specified virtual file driver -x, --hexdump Show raw data in hexadecimal format + --s3-cred=C Supply S3 authentication information to "ros3" vfd. + Accepts tuple of "(<aws-region>,<access-id>,<access-key>)". + If absent or C->"(,,)", defaults to no-authentication. + Has no effect if vfd flag not set to "ros3". + --hdfs-attrs=A Supply configuration information to Hadoop VFD. + Accepts tuple of (<namenode name>,<namenode port>, + ...<kerberos cache path>,<username>,<buffer size>) + If absent or A == '(,,,,)', all default values are used. + Has no effect if vfd flag is not 'hdfs'. file/OBJECT Each object consists of an HDF5 file name optionally followed by a diff --git a/tools/testfiles/pbits/tnofilename-with-packed-bits.ddl b/tools/testfiles/pbits/tnofilename-with-packed-bits.ddl index 19de76f..95dfc3b 100644 --- a/tools/testfiles/pbits/tnofilename-with-packed-bits.ddl +++ b/tools/testfiles/pbits/tnofilename-with-packed-bits.ddl @@ -12,6 +12,16 @@ usage: h5dump [OPTIONS] files -b B, --binary=B Binary file output, of form B -O F, --ddl=F Output ddl text into file F Use blank(empty) filename F to suppress ddl display + --s3-cred=<cred> Supply S3 authentication information to "ros3" vfd. + <cred> :: "(<aws-region>,<access-id>,<access-key>)" + If absent or <cred> -> "(,,)", no authentication. + Has no effect is filedriver is not `ros3'. + --hdfs-attrs=<attrs> Supply configuration information for HDFS file access. + For use with "--filedriver=hdfs" + <attrs> :: (<namenode name>,<namenode port>, + <kerberos cache path>,<username>, + <buffer size>) + Any absent attribute will use a default value. --------------- Object Options --------------- -a P, --attribute=P Print the specified attribute If an attribute name contains a slash (/), escape the diff --git a/tools/testfiles/pbits/tpbitsIncomplete.ddl b/tools/testfiles/pbits/tpbitsIncomplete.ddl index 19de76f..95dfc3b 100644 --- a/tools/testfiles/pbits/tpbitsIncomplete.ddl +++ b/tools/testfiles/pbits/tpbitsIncomplete.ddl @@ -12,6 +12,16 @@ usage: h5dump [OPTIONS] files -b B, --binary=B Binary file output, of form B -O F, --ddl=F Output ddl text into file F Use blank(empty) filename F to suppress ddl display + --s3-cred=<cred> Supply S3 authentication information to "ros3" vfd. + <cred> :: "(<aws-region>,<access-id>,<access-key>)" + If absent or <cred> -> "(,,)", no authentication. + Has no effect is filedriver is not `ros3'. + --hdfs-attrs=<attrs> Supply configuration information for HDFS file access. + For use with "--filedriver=hdfs" + <attrs> :: (<namenode name>,<namenode port>, + <kerberos cache path>,<username>, + <buffer size>) + Any absent attribute will use a default value. --------------- Object Options --------------- -a P, --attribute=P Print the specified attribute If an attribute name contains a slash (/), escape the diff --git a/tools/testfiles/pbits/tpbitsLengthExceeded.ddl b/tools/testfiles/pbits/tpbitsLengthExceeded.ddl index 19de76f..95dfc3b 100644 --- a/tools/testfiles/pbits/tpbitsLengthExceeded.ddl +++ b/tools/testfiles/pbits/tpbitsLengthExceeded.ddl @@ -12,6 +12,16 @@ usage: h5dump [OPTIONS] files -b B, --binary=B Binary file output, of form B -O F, --ddl=F Output ddl text into file F Use blank(empty) filename F to suppress ddl display + --s3-cred=<cred> Supply S3 authentication information to "ros3" vfd. + <cred> :: "(<aws-region>,<access-id>,<access-key>)" + If absent or <cred> -> "(,,)", no authentication. + Has no effect is filedriver is not `ros3'. + --hdfs-attrs=<attrs> Supply configuration information for HDFS file access. + For use with "--filedriver=hdfs" + <attrs> :: (<namenode name>,<namenode port>, + <kerberos cache path>,<username>, + <buffer size>) + Any absent attribute will use a default value. --------------- Object Options --------------- -a P, --attribute=P Print the specified attribute If an attribute name contains a slash (/), escape the diff --git a/tools/testfiles/pbits/tpbitsLengthPositive.ddl b/tools/testfiles/pbits/tpbitsLengthPositive.ddl index 19de76f..95dfc3b 100644 --- a/tools/testfiles/pbits/tpbitsLengthPositive.ddl +++ b/tools/testfiles/pbits/tpbitsLengthPositive.ddl @@ -12,6 +12,16 @@ usage: h5dump [OPTIONS] files -b B, --binary=B Binary file output, of form B -O F, --ddl=F Output ddl text into file F Use blank(empty) filename F to suppress ddl display + --s3-cred=<cred> Supply S3 authentication information to "ros3" vfd. + <cred> :: "(<aws-region>,<access-id>,<access-key>)" + If absent or <cred> -> "(,,)", no authentication. + Has no effect is filedriver is not `ros3'. + --hdfs-attrs=<attrs> Supply configuration information for HDFS file access. + For use with "--filedriver=hdfs" + <attrs> :: (<namenode name>,<namenode port>, + <kerberos cache path>,<username>, + <buffer size>) + Any absent attribute will use a default value. --------------- Object Options --------------- -a P, --attribute=P Print the specified attribute If an attribute name contains a slash (/), escape the diff --git a/tools/testfiles/pbits/tpbitsMaxExceeded.ddl b/tools/testfiles/pbits/tpbitsMaxExceeded.ddl index 19de76f..95dfc3b 100644 --- a/tools/testfiles/pbits/tpbitsMaxExceeded.ddl +++ b/tools/testfiles/pbits/tpbitsMaxExceeded.ddl @@ -12,6 +12,16 @@ usage: h5dump [OPTIONS] files -b B, --binary=B Binary file output, of form B -O F, --ddl=F Output ddl text into file F Use blank(empty) filename F to suppress ddl display + --s3-cred=<cred> Supply S3 authentication information to "ros3" vfd. + <cred> :: "(<aws-region>,<access-id>,<access-key>)" + If absent or <cred> -> "(,,)", no authentication. + Has no effect is filedriver is not `ros3'. + --hdfs-attrs=<attrs> Supply configuration information for HDFS file access. + For use with "--filedriver=hdfs" + <attrs> :: (<namenode name>,<namenode port>, + <kerberos cache path>,<username>, + <buffer size>) + Any absent attribute will use a default value. --------------- Object Options --------------- -a P, --attribute=P Print the specified attribute If an attribute name contains a slash (/), escape the diff --git a/tools/testfiles/pbits/tpbitsOffsetExceeded.ddl b/tools/testfiles/pbits/tpbitsOffsetExceeded.ddl index 19de76f..95dfc3b 100644 --- a/tools/testfiles/pbits/tpbitsOffsetExceeded.ddl +++ b/tools/testfiles/pbits/tpbitsOffsetExceeded.ddl @@ -12,6 +12,16 @@ usage: h5dump [OPTIONS] files -b B, --binary=B Binary file output, of form B -O F, --ddl=F Output ddl text into file F Use blank(empty) filename F to suppress ddl display + --s3-cred=<cred> Supply S3 authentication information to "ros3" vfd. + <cred> :: "(<aws-region>,<access-id>,<access-key>)" + If absent or <cred> -> "(,,)", no authentication. + Has no effect is filedriver is not `ros3'. + --hdfs-attrs=<attrs> Supply configuration information for HDFS file access. + For use with "--filedriver=hdfs" + <attrs> :: (<namenode name>,<namenode port>, + <kerberos cache path>,<username>, + <buffer size>) + Any absent attribute will use a default value. --------------- Object Options --------------- -a P, --attribute=P Print the specified attribute If an attribute name contains a slash (/), escape the diff --git a/tools/testfiles/pbits/tpbitsOffsetNegative.ddl b/tools/testfiles/pbits/tpbitsOffsetNegative.ddl index 19de76f..95dfc3b 100644 --- a/tools/testfiles/pbits/tpbitsOffsetNegative.ddl +++ b/tools/testfiles/pbits/tpbitsOffsetNegative.ddl @@ -12,6 +12,16 @@ usage: h5dump [OPTIONS] files -b B, --binary=B Binary file output, of form B -O F, --ddl=F Output ddl text into file F Use blank(empty) filename F to suppress ddl display + --s3-cred=<cred> Supply S3 authentication information to "ros3" vfd. + <cred> :: "(<aws-region>,<access-id>,<access-key>)" + If absent or <cred> -> "(,,)", no authentication. + Has no effect is filedriver is not `ros3'. + --hdfs-attrs=<attrs> Supply configuration information for HDFS file access. + For use with "--filedriver=hdfs" + <attrs> :: (<namenode name>,<namenode port>, + <kerberos cache path>,<username>, + <buffer size>) + Any absent attribute will use a default value. --------------- Object Options --------------- -a P, --attribute=P Print the specified attribute If an attribute name contains a slash (/), escape the diff --git a/tools/testfiles/textlinksrc-nodangle-1.ls b/tools/testfiles/textlinksrc-nodangle-1.ls index 491f696..396bed3 100644 --- a/tools/testfiles/textlinksrc-nodangle-1.ls +++ b/tools/testfiles/textlinksrc-nodangle-1.ls @@ -37,6 +37,15 @@ usage: h5ls [OPTIONS] file[/OBJECT] [file[/[OBJECT]...] -V, --version Print version number and exit --vfd=DRIVER Use the specified virtual file driver -x, --hexdump Show raw data in hexadecimal format + --s3-cred=C Supply S3 authentication information to "ros3" vfd. + Accepts tuple of "(<aws-region>,<access-id>,<access-key>)". + If absent or C->"(,,)", defaults to no-authentication. + Has no effect if vfd flag not set to "ros3". + --hdfs-attrs=A Supply configuration information to Hadoop VFD. + Accepts tuple of (<namenode name>,<namenode port>, + ...<kerberos cache path>,<username>,<buffer size>) + If absent or A == '(,,,,)', all default values are used. + Has no effect if vfd flag is not 'hdfs'. file/OBJECT Each object consists of an HDF5 file name optionally followed by a diff --git a/tools/testfiles/tgroup-1.ls b/tools/testfiles/tgroup-1.ls index 491f696..396bed3 100644 --- a/tools/testfiles/tgroup-1.ls +++ b/tools/testfiles/tgroup-1.ls @@ -37,6 +37,15 @@ usage: h5ls [OPTIONS] file[/OBJECT] [file[/[OBJECT]...] -V, --version Print version number and exit --vfd=DRIVER Use the specified virtual file driver -x, --hexdump Show raw data in hexadecimal format + --s3-cred=C Supply S3 authentication information to "ros3" vfd. + Accepts tuple of "(<aws-region>,<access-id>,<access-key>)". + If absent or C->"(,,)", defaults to no-authentication. + Has no effect if vfd flag not set to "ros3". + --hdfs-attrs=A Supply configuration information to Hadoop VFD. + Accepts tuple of (<namenode name>,<namenode port>, + ...<kerberos cache path>,<username>,<buffer size>) + If absent or A == '(,,,,)', all default values are used. + Has no effect if vfd flag is not 'hdfs'. file/OBJECT Each object consists of an HDF5 file name optionally followed by a |