summaryrefslogtreecommitdiffstats
path: root/generic
Commit message (Collapse)AuthorAgeFilesLines
* more cross-testing of dictionary-powered compression; describe package ↵dkf2012-06-091-43/+179
| | | | configuration
* tidy up, fix testdkf2012-06-091-27/+38
|
* compressing transforms now work with dictionaries, even if rawdkf2012-06-071-0/+16
|
* merge trunkdkf2012-06-061-1/+10
|\
| * Declare that Tcl provides the zlib 2.0 package.dkf2012-06-061-1/+10
| |
* | making the -dictionary option work with streamsdkf2012-06-061-10/+38
| |
* | more tests, more failures, more docsdkf2012-06-061-2/+13
| |
* | more test tinkeringdkf2012-06-051-13/+19
| |
* | fix broken testsdkf2012-06-051-1/+1
| |
* | merge trunkdkf2012-06-053-5/+1
|\ \ | |/
| * Remove unused variable NULL_CONTEXT.dgp2012-06-051-2/+0
| |
| * Remove unused variable mantDIGIT.dgp2012-06-051-3/+0
| |\
| | * Remove unused variable mantDIGIT.dgp2012-06-051-3/+0
| | |
| * | bug-3530536jan.nijtmans2012-06-011-0/+1
| | |
* | | merge trunkdkf2012-05-3110-521/+329
|\ \ \ | |/ /
| * | Factor out constant that must be the same in multiple places. Prompted by adkf2012-05-255-17/+17
| | | | | | | | | conversation with Alexandre Ferrieux.
| * | fix genStubs.tcl for OSXjan.nijtmans2012-05-253-16/+16
| |\ \ | | |/
| | * fix genStubs.tcljan.nijtmans2012-05-253-16/+16
| | |\
| * | \ Take cygwin handling of X11 into accountjan.nijtmans2012-05-243-4/+18
| |\ \ \ | | |/ / | | | | | | | | Implement TclpIsAtty, Cygwin only doc/dde.n: doc fix
| | * | Take cygwin handling of X11 into accountjan.nijtmans2012-05-244-26/+45
| | |\ \ | | | |/ | | | | | | | | Implement TclpIsAtty, Cygwin only doc/dde.n: doc fix
| | | * Take cygwin handling of X11 into accountjan.nijtmans2012-05-243-4/+20
| | | | | | | | | | | | | | | | Implement TclpIsAtty, Cygwin only doc/dde.n: doc fix
| * | | Revert most of [3caedf05df], since when we let cygwin share the win32 stub ↵jan.nijtmans2012-05-245-513/+307
| |\ \ \ | | |/ / | | | | | | | | | | | | | | | | table this is no longer necessary implement TclpInetNtoa for win32 Let cygwin share stub table with win32
| | * | fix merge problem on Linuxjan.nijtmans2012-05-235-793/+383
| | |\ \
| | | * | MAC_OSX_TK -> MAC_OSX_TCLjan.nijtmans2012-05-223-16/+16
| | | | |
| | | * | Revert most of [fcc5957e59], since when we let cygwin share the win32 stub ↵jan.nijtmans2012-05-225-803/+393
| | | |\ \ | | |/ / / | | | | _ | | | | | | | | | | | | table this is no longer necessary implement TclpInetNtoa for win32 Let cygwin share stub table with win32
| | | * Revert most of [ae92de6078], since when we let cygwin share the win32 stub ↵jan.nijtmans2012-05-225-509/+250
| | | | | | | | | | | | | | | | | | | | | | | | table this is no longer necessary implement TclpInetNtoa for win32 Let cygwin share stub table with win32
* | | | merge trunkdkf2012-05-2410-270/+576
|\ \ \ \ | |/ / /
| * | | minor: Cleaning up code, adding comments.dkf2012-05-241-118/+187
| | | |
| * | | * generic/tclZlib.c (ZlibTransformInput): [Bug 3525907]: Ensure thatdkf2012-05-232-63/+269
| |\ \ \ | | | | | | | | | | | | | | | | | | | | decompressed input is flushed through the transform correctly when the input stream gets to the end. Thanks to Alexandre Ferrieux and Andreas Kupries for their work on this.
| | * | | Use zero-delays instead of finite ones when posting fileevents, because (1) ↵bug_3525907ferrieux2012-05-222-16/+2
| | | | | | | | | | | | | | | | | | | | they should be zero in the first place (see UpdateInterest) and (2) ever-reposting cases will be detected with greater ease.
| | * | | Brought bugfix branch uptodate with head development.andreask2012-05-175-40/+52
| | |\ \ \
| | * | | | Fix for bug 3525907. Reworked TransformInput() entirely, tightened use of ↵andreask2012-05-172-51/+257
| | | | | | | | | | | | | | | | | | | | | | | | timed events, and added code handling special situations like EAGAIN, parent eof, etc.
| * | | | | 3528601 Fix broken --disable-threads build.dgp2012-05-221-0/+4
| | | | | |
| * | | | | When using Tcl_SetObjLength() calls to grow and shrink the objPtr->bytes buffer,dgp2012-05-212-3/+3
| |\ \ \ \ \ | | | |_|/ / | | |/| | | | | | | | | | | | | | | care must be taken that the value cannot possibly become pure Unicode. Calling Tcl_AppendToObj() has the possibility of making such a conversion. Bug found while valgrinding the trunk.
| | * | | | When using Tcl_SetObjLength() calls to grow and shrink the objPtr->bytes buffer,dgp2012-05-212-3/+3
| | |\ \ \ \ | | | | |_|/ | | | |/| | | | | | | | | | | | | | care must be taken that the value cannot possibly become pure Unicode. Calling Tcl_AppendToObj() has the possibility of making such a conversion. Bug found while valgrinding the trunk.
| | | * | | When using Tcl_SetObjLength() calls to grow and shrink the objPtr->bytes buffer,dgp2012-05-212-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | care must be taken that the value cannot possibly become pure Unicode. Calling Tcl_AppendToObj() has the possibility of making such a conversion. Bug found while valgrinding the trunk.
| * | | | | * generic/tclOOBasic.c (TclOO_Class_Constructor): [Bug 2023112]: Cutdkf2012-05-205-153/+174
| | |_|_|/ | |/| | | | | | | | | | | | | | | | | | the amount of hackiness in class constructors, and refactor some of the error message handling from [oo::define] to be saner in the face of odd happenings.
* | | | | typofixdkf2012-05-181-1/+1
| | | | |
* | | | | merge trunkdkf2012-05-175-40/+52
|\ \ \ \ \ | |/ / / /
| * | | | * generic/tclCmdMZ.c (Tcl_SwitchObjCmd): [Bug 3106532]: Correcteddkf2012-05-171-2/+6
| |\ \ \ \ | | |/ / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | resulting indexes from -indexvar option to be usable with [string range]; this was always the intention (and is consistent with [regexp -indices] too). ***POTENTIAL INCOMPATIBILITY*** Uses of [switch -regexp -indexvar] that previously compensated for the wrong offsets (by subtracting 1 from the end indices) now do not need to do so as the value is correct.
| | * | | * generic/tclCmdMZ.c (Tcl_SwitchObjCmd): [Bug 3106532]: Correcteddkf2012-05-171-2/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | resulting indexes from -indexvar option to be usable with [string range]; this was always the intention (and is consistent with [regexp -indices] too). ***POTENTIAL INCOMPATIBILITY*** Uses of [switch -regexp -indexvar] that previously compensated for the wrong offsets (by subtracting 1 from the end indices) now do not need to do so as the value is correct.
| * | | | [Bug 3445787]: Improve the compatibility of safe interpreters' version ofdkf2012-05-161-2/+40
| | | | | | | | | | | | | | | 'file' with that of unsafe interpreters.
| * | | | revert part of [8f9c16848b] move Tcl_MacOSXOpenBundleResources and ↵jan.nijtmans2012-05-153-36/+6
| |\ \ \ \ | | |/ / / | | | | / | | |_|/ | |/| | Tcl_MacOSXOpenVersionedBundleResources from the UNIX stub table back to its own osx stub table, instead let cygwin make use of the win32 stub table
| | * | Revert part of [5caa06383d] move Tcl_MacOSXOpenBundleResources and ↵jan.nijtmans2012-05-153-46/+6
| | |\ \ | | | |/ | | | | | | | | Tcl_MacOSXOpenVersionedBundleResources from the UNIX stub table back to its own osx stub table, instead let cygwin make use of the win32 stub table
| | | * revert part of [9922ac54e8]: remove Tcl_MacOSXOpenBundleResources and ↵jan.nijtmans2012-05-143-44/+43
| | | | | | | | | | | | | | | | Tcl_MacOSXOpenVersionedBundleResources from the UNIX stub table, instead let cygwin make use of the win32 stub table
* | | | merge trunkdkf2012-05-132-19/+162
|\ \ \ \ | |/ / /
| * | | Undone part of change [32d93a8414], keeping [chan postevent] synchronous for ↵andreask2012-05-091-16/+15
| | | | | | | | | | | | | | | | owner == handler.
| * | | * generic/tclIORChan.c [Bug 3522560]: Fixed the crash, enabled theandreask2012-05-092-16/+160
| | | | | | | | | | | | | | | | | | | | | | | | | | | | test case. Modified [chan postevent] to properly inject the event(s) into the owner thread's event queue for execution in the correct context. Renamed the ForwardOpTo...Thread() function to match with our terminology.
| * | | Properly close nonblocking channels even when not flushing them.ferrieux2012-05-071-3/+3
| | | |
* | | | updated C API to be more focused on supporting just some operationsdkf2012-05-104-18/+59
| | | |