summaryrefslogtreecommitdiffstats
path: root/Tests/Module/GenerateExportHeader
Commit message (Collapse)AuthorAgeFilesLines
* Use string(APPEND) in TestsDaniel Pfeifer2016-07-271-2/+2
| | | | | | | Automate with: find Tests -type f -print0 | xargs -0 perl -i -0pe \ 's/set\(([a-zA-Z0-9_]+)(\s+)"\$\{\1\}([^"])/string(APPEND \1\2"\3/g'
* Tests: Run clang-format on GenerateExportHeader codeBrad King2016-06-291-1/+1
|
* GenerateExportHeader: Add option to specify custom contentBrad King2016-06-282-0/+7
|
* Revise C++ coding style using clang-formatKitware Robot2016-05-169-77/+118
| | | | | | | | | | | | | Run the `Utilities/Scripts/clang-format.bash` script to update all our C++ code to a new style defined by `.clang-format`. Use `clang-format` version 3.8. * If you reached this commit for a line in `git blame`, re-run the blame operation starting at the parent of this commit to see older history for the content. * See the parent commit for instructions to rebase a change across this style transition commit.
* GenerateExportHeader: Do not define DEFINE_NO_DEPRECATED (#16022)Andreas Schuh2016-03-1714-42/+56
| | | | | | Previously we allowed this definition to persist outside our header. This would cause conflicts across multiple such headers because the name was always the same. Fix this by avoiding the definition altogether.
* Tests: fix Module.GenerateExportHeader building on Clang/C2Mariusz PluciƄski2016-03-103-1/+93
|
* Tests: fix GenerateExportHeader directory definitionsBen Boeckel2016-02-122-9/+6
| | | | | | There's no need to stringify the values, but instead just pass in strings. The core problem is that the path may have tokens which are replaced by the preprocessor which causes an invalid path to be used.
* GenerateExportHeader: Work around buggy std::getline behavior in BCB5.James Johnston2015-12-311-0/+12
|
* GenerateExportHeader: Use dllexport/dllimport on CygwinTony Kelman2015-08-271-2/+2
| | | | | | | Cygwin sits on top of Windows and so can use explicit symbol export and import markup too. Co-Author: Yaakov Selkowitz <yselkowitz@cygwin.com>
* HP aCC: record compiler flag for Ansi C++98 support for version 3.80 onwardsRolf Eike Beer2014-04-251-0/+9
| | | | | | Old versions of aCC need a special compiler flag to get full C++98 template support as e.g. CMake itself or the Complex and ComplexOneConfig tests need. The same versions need a special flag to get a proper C++ library, too.
* Merge some GenerateExportHeader subtests.Stephen Kelly2013-12-1515-143/+28
|
* Speed up the GenerateExportHeader unit test (#14453).Stephen Kelly2013-12-1517-183/+567
| | | | | | | | Instead of running many small tests with many cmake projects, simply compare the generated export header against a reference. Remove the helper macros and the try_compiles which are duplicates of the library build tests.
* Merge topic 'VISIBILITY_PRESET-property'Brad King2013-06-055-0/+48
|\ | | | | | | | | | | cd1fa53 Add a COMPILE_OPTION for a VISIBILITY_INLINES_HIDDEN target property. 0e9f4bc Introduce target property <LANG>_VISIBILITY_PRESET
| * Add a COMPILE_OPTION for a VISIBILITY_INLINES_HIDDEN target property.Stephen Kelly2013-06-021-0/+4
| | | | | | | | | | | | | | This corresponds to the g++ and clang++ option -fvisibility-inlines-hidden on linux. On Windows with MinGW, this corresponds to -fno-keep-inline-dllexport. That option is not supported by clang currently.
| * Introduce target property <LANG>_VISIBILITY_PRESETStephen Kelly2013-06-025-0/+44
| | | | | | | | | | | | This is initialized by CMAKE_<LANG>_VISIBILITY_PRESET. The target property is used as the operand to the -fvisibility= compile option with GNU compilers and clang.
* | Merge topic 'geh-cleanup-identifiers'Brad King2013-05-245-0/+44
|\ \ | | | | | | | | | | | | 7e24997 GenerateExportHeader: Generate only C identifiers as defines
| * | GenerateExportHeader: Generate only C identifiers as definesStephen Kelly2013-05-245-0/+44
| |/ | | | | | | | | | | | | | | | | | | | | | | | | | | | | The variables in this module are used to configure a header file with defines whose name depends on the name of the target. As valid names of targets may be invalid for use as defines, convert the names of the defines used to C identifiers first. This is already done in C++ code for the DEFINE_SYMBOL property. This is not as simple as ensuring that the BASE_NAME is a C identifier, because most of the define names are configurable, and because use of a BASE_NAME which is not a C identifier, such as 4square can become a C identifier by specifying a prefix in the generate_export_header macro.
* | Merge topic 'geh-add-missing-test'Brad King2013-05-241-0/+1
|\ \ | | | | | | | | | | | | 74f24b2 Tests/Module/GenerateExportHeader: Test exported free-function
| * | Tests/Module/GenerateExportHeader: Test exported free-functionStephen Kelly2013-05-241-0/+1
| |/
* | GenerateExportHeader: Add newlines to separate the compiler output.Stephen Kelly2013-05-231-2/+2
|/ | | | Make the error logs more readable.
* GenerateExportHeader test: add newlines before end of fileRolf Eike Beer2012-01-175-5/+5
|
* Build each library only once instead of once for each test.Stephen Kelly2011-10-094-5/+58
|
* Remove unused define.Stephen Kelly2011-10-091-8/+0
|
* Add some more unit tests.Stephen Kelly2011-09-0210-0/+114
|
* Fix Compiler id variable name.Stephen Kelly2011-08-241-1/+1
|
* Disable the tests for Watcom.Stephen Kelly2011-08-241-0/+15
| | | | They are timing out and showing up on the dashboard.
* Quote paths in case there is a space in one of them.Stephen Kelly2011-08-241-3/+3
|
* Possibly fix test on HPUX.Stephen Kelly2011-08-231-1/+1
|
* Disable testing of deprecated macros.Stephen Kelly2011-08-143-19/+19
| | | | | It is not possible to test for it sanely across multiple ancient compilers supported by cmake.
* Try to error on deprecated on Intel and SunCC.Stephen Kelly2011-08-131-7/+23
|
* Perform the -Werror test only once.Stephen Kelly2011-08-131-3/+6
| | | | This way, the output of the test should be visible on CDash.
* Add some messaging output to make remaining issues findable.Stephen Kelly2011-08-121-0/+4
|
* Test -Werror instead of enabling it per compiler.Stephen Kelly2011-08-121-1/+3
| | | | Maybe the Intel compiler supports it too.
* Expect the tests to pass if hidden visibilty is not enabled.Stephen Kelly2011-08-122-9/+27
|
* Don't change the expected build result based on the platform.Stephen Kelly2011-08-111-5/+1
| | | | Hopefully the compiler feature tests will cover those cases.
* Test for compiler features, instead of for specific platforms.Stephen Kelly2011-08-111-2/+2
|
* Split the deprecated available check from setting macro values.Stephen Kelly2011-08-113-3/+3
|
* Don't export methods on already exported classes.Stephen Kelly2011-08-113-10/+1
| | | | This is always a mistake.
* Remember to surround the other deprecated test in the Borland check.Stephen Kelly2011-08-101-1/+4
|
* Export deprecated free methods too.Stephen Kelly2011-08-102-2/+2
| | | | That way they can be linked to even after being deprecated.
* Another attempt at fixing Borland.Stephen Kelly2011-08-103-4/+15
| | | | | Apparently it still needs to be exported, but only declspec(deprecated) does not work.
* Use the correct project name compiletest not compilefailStephen Kelly2011-08-101-1/+1
| | | | Should fix windows VS solution based builds.
* Only set the COMPILER_HAS_HIDDEN_VISIBILITY if GCC >= 4.2Stephen Kelly2011-08-101-16/+1
| | | | | | Hearsay has it that before that version it didn't work properly. Hopefully this will fix more dashboard builds.
* Set the CMAKE_RUNTIME_OUTPUT_DIRECTORY for windows builds.Stephen Kelly2011-08-101-0/+4
|
* Only run the failure tests with gcc >= 4.2Stephen Kelly2011-08-101-1/+16
|
* Fix typo and tests failing as a result.Stephen Kelly2011-08-101-1/+1
|
* Fix tests with clang.Stephen Kelly2011-08-101-4/+5
|
* Test expected no-op instead of aborting the build.Stephen Kelly2011-08-101-7/+4
| | | | Aborting causes make clean to fail if the executable is not created.
* Short-circuit the tests on unsupported compilers.Stephen Kelly2011-08-101-0/+6
|
* Add a newline at the end of the file.Stephen Kelly2011-08-091-1/+1
| | | | Hopefully fix the test execution reported on CDash.