summaryrefslogtreecommitdiffstats
path: root/generic/tclIO.c
Commit message (Collapse)AuthorAgeFilesLines
* * generic/tclBasic.c:dgp2004-10-061-5/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | * generic/tclBinary.c: * generic/tclCmdAH.c: * generic/tclCmdIL.c: * generic/tclCmdMZ.c: * generic/tclCompExpr.c: * generic/tclDictObj.c: * generic/tclEncoding.c: * generic/tclExecute.c: * generic/tclFCmd.c: * generic/tclHistory.c: * generic/tclIndexObj.c: * generic/tclInterp.c: * generic/tclIO.c: * generic/tclIOCmd.c: * generic/tclNamesp.c: * generic/tclObj.c: * generic/tclPkg.c: * generic/tclResult.c: * generic/tclScan.c: * generic/tclTimer.c: * generic/tclTrace.c: * generic/tclUtil.c: * generic/tclVar.c: It is a poor practice to directly set or append to the value of the objResult of an interp, because that value might be shared, and in that circumstance a Tcl_Panic() will be the result. Searched for example of this practice and replaced with safer alternatives, often using the Tcl_AppendResult() routine that dkf just rehabilitated.
* Fix [Bug 1025359]; make sure wide seeks don't lose errorsdkf2004-09-101-5/+5
|
* * generic/tclIO.h (CHANNEL_INCLOSE): New flag. Set inandreas_kupries2004-07-151-1/+45
| | | | | | | | | | | | | | | | | | | | | | * generic/tclIO.c (Tcl_UnregisterChannel): 'Tcl_Close' while the * generic/tclIO.c (Tcl_Close): close callbacks are run. Checked in 'Tcl_Close' and 'Tcl_Unregister' to prevent recursive call of 'close' in the close-callbacks. This is a possible error made by implementors of virtual filesystems based on 'tclvfs', thinking that they have to close the channel in the close handler for the filesystem. * generic/tclIO.c: * generic/tclIO.h: * Not reverting, but #ifdef'ing the changes from May 19, 2004 out of the core. This removes the ***POTENTIAL INCOMPATIBILITY*** for channel drivers it introduced. This has become possible due to Expect gaining a BlockModeProc and now handling blockingg and non-blocking modes correctly. Thus [SF Tcl Bug 943274] is still fixed if a recent enough version of Expect is used. * doc/CrtChannel.3: Added warning about usage of a channel without a BlockModeProc.
* * generic/tclIO.c (Tcl_SetChannelOption): Took out the rangedavygrvy2004-06-011-9/+1
| | | | | | | verification on -buffersize so original behavior for silently ignoring bad settings is returned. The only difference now is the behavior difference between atoi and Tcl_GetInt which gains the ability to error for 'not a number' and can understand hexidecimal.
* * generic/tclIO.c (Tcl_SetChannelOption): Invalid settingsdavygrvy2004-06-011-3/+14
| | | | | | for -buffersize weren't being reported as errors and were blindly ignored. Now reports conversion errors to an int and checks ranges for validity. atoi() swapped for Tcl_GetInt().
* * tclIO.c: Fixed [SF Tcl Bug 943274]. This is the same problem asandreas_kupries2004-05-191-6/+89
| | | | | | | | | | | | | * tclIO.h: [SF Tcl Bug 462317], see ChangeLog entry 2001-09-26. The fix done at that time is incomplete. It is possible to get around it if the actual read operation is defered and not executed in the event handler itself. Instead of tracking if we are in an read caused by a synthesized fileevent we now track if the OS has delivered a true event = actual data and bypass the driver if a read finds that there is no actual data waiting. The flag is cleared by a short or full read.
* * generic/tclIO.c (Tcl_SetChannelOption): Fixed [SF Tcl Bugandreas_kupries2004-04-231-1/+10
| | | | | 930851]. When changing the eofchar we have to zap the related flags to prevent them from prematurely aborting the next read.
* Patch 922727 committed. Implements three changes:dgp2004-04-061-2/+1
| | | | | | | | | | | | | | | | | | | | | | | * generic/tclInt.h: Reworked the Tcl header files into a clean * unix/tclUnixPort.h: hierarchy where tcl.h < tclPort.h < tclInt.h * win/tclWinInt.h: and every C source file should #include * win/tclWinPort.h: at most one of those files to satisfy its declaration needs. tclWinInt.h and tclWinPort.h also better organized so that tclWinPort.h includes the Windows implementation of cross-platform declarations, while tclWinInt.h makes declarations that are available on Windows only. * generic/tclBinary.c (TCL_NO_MATH): Deleted the generic/tclMath.h * generic/tclMath.h (removed): header file. The internal Tcl * macosx/Makefile (PRIVATE_HEADERS): header, tclInt.h, has a * win/tcl.dsp: #include <math.h> directly, and file external to Tcl needing libm should do the same. * win/Makefile.in (WIN_OBJS): Deleted the win/tclWinMtherr.c file. * win/makefile.bc (TCLOBJS): It's a vestige from matherr() days * win/makefile.vc (TCLOBJS): gone by. * win/tcl.dsp: * win/tclWinMtherr.c (removed):
* fixed typos in documentation and commentsvincentdarley2004-03-091-4/+4
|
* * generic/tclIO.c (Tcl_Ungets): fixed improper filling of the channel buffer.davygrvy2004-02-021-3/+2
| | | | [Bug 405995]
* All uses of 'panic' (the macro) changeddavygrvy2003-12-241-12/+12
| | | | | | | to 'Tcl_Panic' (the function). The #define of panic in tcl.h clearly states it is deprecated in the comments. [Patch 865264]
* * The changes below fix SF bugs [593810], and [718045].andreas_kupries2003-04-221-1/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | * generic/tclIO.c (Tcl_CutChannel, Tcl_SpliceChannel): Invoke TclpCutSockChannel and TclpSpliceSockChannel. * generic/tclInt.h: Declare TclpCutSockChannel and TclpSpliceSockChannel. * unix/tclUnixSock.c (TclpCutSockChannel, TclpSpliceSockChannel): Dummy functions, on unix the sockets are _not_ handled specially. * mac/tclMacSock.c (TclpCutSockChannel, TclpSpliceSockChannel): * win/tclWinSock.c (TclpCutSockChannel, TclpSpliceSockChannel): New functions to handle socket specific cut/splice operations: auto-initi of socket system for thread on splice, management of the module internal per-thread list of sockets, management of association of sockets with HWNDs for event notification. * win/tclWinSock.c (NewSocketInfo): Extended initialization assignments to cover all items of the structure. During debugging of the new code mentioned above I found that two fileds could contain bogus data. * win/tclWinFile.c: Added #undef HAVE_NO_FINDEX_ENUMS before definition because when compiling in debug mode the compiler complains about a redefinition, and this warning is also treated as an error.
* * generic/tclIO.c (UpdateInterest): When dropping interest inandreas_kupries2003-04-111-1/+44
| | | | | | | TCL_READABLE now dropping interest in TCL_EXCEPTION too. This fixes a bug where Expect detects eof on a file prematurely on solaris 2.6 and higher. A much more complete explanation is in the code itself (40 lines of comments for a one-line change :)
* * generic/tclIO.c (Tcl_Seek, Tcl_OutputBuffered):mdejong2003-03-061-3/+13
| | | | | | | | | | | | | | | | | If there is data buffered in the statePtr->curOutPtr member then set the BUFFER_READY flag in Tcl_Seek. This is needed so that the next call to FlushChannel will write any buffered bytes before doing the seek. The existing code would set the BUFFER_READY flag inside the Tcl_OutputBuffered function. This was a programming error made when Tcl_OutputBuffered was originally created in CVS revision 1.35. The setting of the BUFFER_READY flag should not have been included in the Tcl_OutputBuffered function. * generic/tclTest.c (TestChannelCmd): Use the Tcl_InputBuffered and Tcl_OutputBuffered util methods to query the amount of buffered input and output.
* * generic/tclIO.c (Tcl_Flush): Compare themdejong2003-03-061-2/+3
| | | | | | | | | | nextAdded member of the ChannelBuffer to the nextRemoved member to determine if any output has been buffered. The previous check against the value 0 seems to have just been a coding error. See other methods like Tcl_OutputBuffered for examples where nextAdded is compared to nextRemoved to find the number of bytes buffered.
* * generic/tclIO.c (Tcl_GetsObj): Check thatmdejong2003-03-061-2/+2
| | | | | | | | | | the eol pointer has not gone past the end of the string when in auto translation mode and the INPUT_SAW_CR flag is set. The previous code worked because the end of string value \0 was being compared to \n, this patch just skips that pointless check.
* * generic/tclIO.c (WriteBytes, WriteChars,mdejong2003-03-061-4/+15
| | | | | | | | | Tcl_GetsObj, ReadBytes): Rework calls to TranslateOutputEOL to make it clear that a boolean value is being returned. Add some comments in an effort to make the code more clear. This patch makes no functional changes.
* * generic/tclIO.c (Tcl_SetChannelOption):mdejong2003-03-061-5/+3
| | | | | | | | | | | | | Invoke the Tcl_SetChannelBufferSize method as a result of changing the -buffersize option to fconfigure. The previous implementation used some inlined code that reset the buffer size to the default size instead of ignoring the request as implemented in Tcl_SetChannelBufferSize. * tests/io.test: Update test case so that it actually checks the implementation of Tcl_SetChannelBufferSize.
* (HaveVersion): correctly decl statichobbs2003-02-191-2/+2
|
* * generic/tclIO.c (Tcl_GetsObj): Minor changemdejong2003-02-171-2/+2
| | | | | | so that eol is only assigned at the top of the TCL_TRANSLATE_AUTO case block. The other cases assign eol so this does not change any functionality.
* * generic/tclIO.c (Tcl_CutChannel, Tcl_SpliceChannel):mdejong2003-01-261-1/+5
| | | | | | | | | | | | | | | | Invoke TclpCutFileChannel and TclpSpliceFileChannel. * generic/tclInt.h: Declare TclpCutFileChannel and TclpSpliceFileChannel. * unix/tclUnixChan.c (FileCloseProc, TclpOpenFileChannel, Tcl_MakeFileChannel, TclpCutFileChannel, TclpSpliceFileChannel): Implement thread load data cut and splice for file channels. This avoids an invalid memory ref when compiled with -DDEPRECATED. * win/tclWinChan.c (FileCloseProc, TclpCutFileChannel, TclpSpliceFileChannel): Implement thread load data cut and splice for file channels. This avoids an invalid memory ref that was showing up in the thread extension.
* stringObj unicode Tcl_SetObjLength fixesvincentdarley2003-01-241-1/+7
|
* * tests/io.test:andreas_kupries2002-07-301-2/+16
| | | | | | | | | * generic/tclIO.c (WriteChars): Added flag to break out of loop if nothing of the input is consumed at all, to prevent infinite looping of called with a non-UTF-8 string. Fixes Bug 584603 (partially). Added new test "io-60.1". Might need additional changes to Tcl_Main so that unprintable results are printed as binary data.
* TIP#91 implementation; makes old style channels binary compatible withdkf2002-05-241-53/+141
| | | | new TIP#72-enabled Tcl. See http://purl.org/tcl/tip/91 for details.
* * generic/tclEncoding.c (EscapeFromUtfProc):hobbs2002-04-181-26/+59
| | | | | | * generic/tclIO.c (WriteChars, Tcl_Close): corrected the handling of outputting end escapes for escape-based encodings. [Bug #526524] (yamamoto)
* * generic/tclEncoding.c: Fix typo in comment.mdejong2002-03-111-16/+18
| | | | | | | | * generic/tclIO.c (DoReadChars, ReadBytes, ReadChars): Use NULL value instead of pointer set to NULL to make things more clear. Reorder arguments so that they match the function signatures. Cleanup little typos and add more descriptive comment.
* * tests/encoding.test: added encoding-23.* testshobbs2002-03-021-4/+11
| | | | | | | * generic/tclIO.c (FilterInputBytes): reset the TCL_ENCODING_START flags in the ChannelState when using 'gets'. [Bug #523988] Also reduced the value of ENCODING_LINESIZE from 30 to 20 as this seems to improve the performance of 'gets' according to tclbench.
* TIP#72 implementation. See ChangeLog for details.dkf2002-02-151-34/+69
| | | | | This version builds clean on Solaris/SPARC, with GCC and CC, both with and without threads and both in 32-bit and 64-bit mode.
* * Sought out and eliminated instances of CONST-casting that are nodgp2002-01-261-3/+3
| | | | longer needed after the TIP 27 effort.
* Make -eofchar and -translation options read only formdejong2002-01-251-3/+11
| | | | | | | | | | | | | | | server sockets. [Bug 496733] * generic/tclIO.c (Tcl_GetChannelOption, Tcl_SetChannelOption): Instead of returning nothing for the -translation option on a server socket, always return "auto". Return the empty string enclosed in quotes for the -eofchar option on a server socket. Fixup -eofchar usage message so that it matches the implementation. * tests/io.test: Add -eofchar tests and -translation tests to ensure options are read only on server sockets. * tests/socket.test: Update tests to account for -eofchar and -translation option changes.
* * Updated interfaces of generic/tclEncoding, generic/tclFilename.c,dgp2002-01-251-5/+5
| | | | | | | | | | | generic/tclIOUtil.c, generic/tclPipe.c, generic/tclResult.c, generic/tclUtil.c, generic/tclVar.c and mac/tclMacResource.c according to TIP 27. Tcl_TranslateFileName rewritten as wrapper around VFS-aware version. Updated callers. ***POTENTIAL INCOMPATIBILITY*** Includes source incompatibilities: argv arguments of Tcl_Concat, Tcl_JoinPath, Tcl_OpenCommandChannel, Tcl_Merge; argvPtr arguments of Tcl_SplitList and Tcl_SplitPath.
* * generic/tclIO.c (WriteChars): Fix for SF #506297, reported byandreas_kupries2002-01-211-1/+19
| | | | | | | | | | | | | | Martin Forssen <ruric@users.sourceforge.net>. The encoding chosen in the script exposing the bug writes out three intro characters when TCL_ENCODING_START is set, but does not consume any input as TCL_ENCODING_END is cleared. As some output was generated the enclosing loop calls UtfToExternal again, again with START set. Three more characters in the out and still no use of input ... To break this infinite loop we remove TCL_ENCODING_START from the set of flags after the first call (no condition is required, the later calls remove an unset flag, which is a no-op). This causes the subsequent calls to UtfToExternal to consume and convert the actual input.
* * More TIP 27 updates in tclIOUtil.c and tclIndexObj.c that weredgp2002-01-171-2/+2
| | | | | | | overlooked before. Updated callers. ***POTENTIAL INCOMPATIBILITY*** Includes a source incompatibility in the tablePtr arguments of the Tcl_GetIndexFromObj* routines.
* * Updated APIs in generic/tclUtf.c and generic/tclRegexp.c accordingdgp2002-01-171-2/+2
| | | | to the guidelines of TIP 27. Updated callers.
* * Updated APIs in generic/tclIOUtil.c and generic/tclPosixStr.cdgp2002-01-151-2/+9
| | | | according to the guidelines of TIP 27. Updated callers. [Patch 499196]
* * Updated APIs in the file generic/tclIO.c according to the guidelinesdgp2002-01-151-34/+36
| | | | | | | | | | | of TIP 27. Several minor documentation corrections as well. * Updated channel driver interface according to the guidelines of TIP 27. See also [Bug 500348]. * Moved Tcl_EolTranslation enum declaration from generic/tcl.h to generic/tclInt.h (renamed to TclEolTranslation). It is not used anywhere in Tcl's public interface.
* * Applied #219311 on behalf of Rolf Schroedterandreas_kupries2001-12-171-14/+29
| | | | | <schroedter@users.sourceforge.net> to prevent fcopy on serial ports from flooding the event queue.
* * generic/tclIO.c (Tcl_GetsObj): Applied patch for bug #491341 asandreas_kupries2001-12-121-4/+4
| | | | | provided by Don Porter <dgp@users.sourceforge.net>. Fixes assumption of having an empty Tcl_Obj to work with.
* See ChangeLogandreas_kupries2001-11-071-7/+17
|
* Undo of mistaken commit. Sorry!dgp2001-10-161-2/+2
|
* * Added test to demonstrate memory corruption problems. [Bug 219393].dgp2001-10-161-2/+2
|
* Added necessary casts to ClientData on a pair of calls to Tcl_PreserveKevin B Kenny2001-09-281-3/+3
| | | | and Tcl_Release.
* * Added Tcl_Preserve()dgp2001-09-281-2/+3
| | | | | and Tcl_Release() to fix segfault introduced by the 2001-09-26 changes. [Bug 465494]
* * The changes below fix [Bug #462317] where Expect tried to readandreas_kupries2001-09-271-16/+42
| | | | | | | | | | | | | | | | | | | | | | | more than was in the buffers and then blocked in the OS call as its pty channel driver provides no blockmodeproc through which the OS could be notified of blocking-behaviour. Because of this the general I/O core has to take more care than usual to preserve the semantics of non-blocking channels. * generic/tclIO.c (Tcl_ReadRaw): Do not read from the driver if the channel is non-blocking and the fileevent causing the read was generated by a timer. We do not know if there is data available from the OS. Instead of going to the OS for more and potentially blocking we simply signal EWOULDBLOCK to the higher levels to cause the system to wait for true fileevents. (GetInput): Same as before. (ChannelTimerProc): Added set and clear of CHANNEL_TIMER_FEV. * generic/tclIO.h (CHANNEL_TIMER_FEV): New flag for channels. Is set if a fileevent was generated by a timer, the channel is not blocking and the driver did not provide a blockmodeproc. In that case the I/O core has to be especially careful about going to the driver for more data.
* Changes due to TIP#49 "Tcl_OutputBuffered" from Rolf Schroedterdkf2001-09-061-46/+47
|
* fs updatevincentdarley2001-08-231-4/+8
|
* Changes from TIP#17 "Redo Tcl's filesystem"vincentdarley2001-07-311-43/+179
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The following files were impacted. * doc/Access.3: * doc/FileSystem.3: * doc/OpenFileChnl.3: * doc/file.n: * doc/glob.n: * generic/tcl.decls: * generic/tcl.h: * generic/tclCmdAH.c: * generic/tclCmdIL.c: * generic/tclCmdMZ.c: * generic/tclDate.c: * generic/tclDecls.h: * generic/tclEncoding.c: * generic/tclFCmd.c: * generic/tclFileName.c: * generic/tclGetDate.y: * generic/tclIO.c: * generic/tclIOCmd.c: * generic/tclIOUtil.c: * generic/tclInt.decls: * generic/tclInt.h: * generic/tclIntDecls.h: * generic/tclLoad.c: * generic/tclStubInit.c: * generic/tclTest.c: * generic/tclUtil.c: * library/init.tcl: * mac/tclMacFCmd.c: * mac/tclMacFile.c: * mac/tclMacInit.c: * mac/tclMacPort.h: * mac/tclMacResource.c: * mac/tclMacTime.c: * tests/cmdAH.test: * tests/event.test: * tests/fCmd.test: * tests/fileName.test: * tests/io.test: * tests/ioCmd.test: * tests/proc-old.test: * tests/registry.test: * tests/unixFCmd.test: * tests/winDde.test: * tests/winFCmd.test: * unix/mkLinks: * unix/tclUnixFCmd.c: * unix/tclUnixFile.c: * unix/tclUnixInit.c: * unix/tclUnixPipe.c: * win/tclWinFCmd.c: * win/tclWinFile.c: * win/tclWinInit.c: * win/tclWinPipe.c
* 2001-07-18 Andreas Kupries <andreas_kupries@users.sourceforge.net>andreas_kupries2001-07-181-1/+25
| | | | | | | * generic/tclIO.c: Aftermath to [SF #427196]. Squash empty buffers if they are smaller than the requested buffersize, to prevent reusage of old buffers and to honor changes in the requested buffersize made by the user.
* 2001-07-17 Andreas Kupries <andreas_kupries@users.sourceforge.net>andreas_kupries2001-07-171-2/+16
| | | | | | | | | | * generic/tclIO.c (GetInput): Fixed [SF #427196]. Memory was overwritten because a buffer was used after a change of the requested buffersize together with that requested buffersize and not its actual size, which was smaller. Note that the continous reuse of the smaller buffer negatively impacts performance. The system never allocates a buffer with the newly requested bigger buffersize.
* * Note that "tclbench" (see project "tcllib") was extended withandreas_kupries2001-05-191-22/+172
| | | | | | | | | | | | | | | | | | performance benchmarks for [fcopy] too. * doc/fcopy.n: Updated to reflect the extended behaviour of 'fcopy'. * tests/io.test: Added tests 'io-52.9', 'io-52.10' and 'io-52.11' to test the handling of encodings by 'fcopy' / 'TclCopychannel' [Bug #209210]. * generic/tclIO.c: Split of both 'Tcl_ReadChars' and 'Tcl_WriteChars' into a public error checking and an internal working part. The public functions now use the new internal ones. The new functions are 'DoReadChars' and 'DoWriteChars'. Extended 'CopyData' to use the new functions 'DoXChars' when required by the encodings on the input and output channels [Bug #209210].