summaryrefslogtreecommitdiffstats
path: root/tests/zlib.test
Commit message (Collapse)AuthorAgeFilesLines
* [25842c161] Prevent zero-length compress actions in [zlib] streaming API.dkf2016-02-031-0/+6
|
* Bump to zlib 2.0.1.dgp2014-08-221-1/+1
|
* Refactored much management of the BLOCKED and EOF flags into ChanRead() thendgp2014-05-131-1/+1
| | | | | | | began repairing some of the logic about them. Tests iogt-2.* now fail because they've been crafted as experiments recording the fine detail of reflected channel driver calls, and fixing the management of channel flags is changing that. zlib-8.5 also needed adjustment to reflect that an EOF set must come with an empty string read when flags are functioning properly.
* [98c8b3ec12] Use constraint to work around failing test; it's just a bug in thedkf2013-10-101-1/+8
| | | | underlying library, and one that appears fixed in later versions. Not our fault if some OSes don't update. (It's also in a very esoteric spot.)
* [98c8b3ec12] Make test fail in less catastrophic manner.dkf2013-09-041-1/+4
|
* typojan.nijtmans2013-09-021-8/+8
|
* Correction to comment in re key buffer size.bug_3603553dkf2013-02-111-1/+1
|
* [Bug 3603553]: Ensure that data gets written to the underlying stream bydkf2013-02-111-0/+19
| | | | | | | compressing transforms when the amount of data to be written is one buffer's-worth; problem was particularly likely to occur when compressing large quantities of not-very-compressible data. Many thanks to Piera Poggio (vampiera) for reporting.
* TIP 400 suffered from the same segfaulting flaw as 3595576.dgp2012-12-131-0/+14
| | | Segfaulting test and fix committed.
* verify zlib package presence and versiondkf2012-06-111-0/+3
|
* more cross-testing of dictionary-powered compression; describe package ↵dkf2012-06-091-2/+56
| | | | configuration
* tidy up, fix testdkf2012-06-091-4/+4
|
* compressing transforms now work with dictionaries, even if rawdkf2012-06-071-0/+16
|
* making the -dictionary option work with streamsdkf2012-06-061-2/+2
|
* more tests, more failures, more docsdkf2012-06-061-9/+59
|
* more test tinkeringdkf2012-06-051-3/+6
|
* better test that dictionaries workdkf2012-06-051-5/+7
|
* merge trunkdkf2012-06-051-0/+16
|\
| * [Bug 3530536]: zlib-7.4 fails on IRIX64jan.nijtmans2012-06-011-0/+16
| |\
| | * two more testcases, showing that only the "deflate" and "inflate" streams ↵bug_3530536jan.nijtmans2012-06-011-0/+16
| |/ | | | | | | don't update the checksum, other streams do.
* | merge trunkdkf2012-05-241-5/+23
|\ \ | |/
| * Add test showing both loss of bytes and empty-fileevent frenzy.ferrieux2012-05-201-0/+18
| |
* | merge trunkdkf2012-05-171-0/+16
|\ \ | |/
| * Cancel the timeout timers! If this isn't done, lingering timers from dgp2012-05-171-0/+16
| | | | | | | | early tests can trigger false timeouts of later tests (since they are all using a common variable name).
* | first actual test of doing something with a compression dictionarydkf2012-05-011-0/+15
| |
* | Differentiate what options may be set by format type.dkf2012-04-291-0/+20
|/
* * generic/tclZlib.c (ZlibTransformSetOption): [Bug 3517696]: Makedkf2012-04-151-0/+12
| | | flushing work correctly in a pushed compressing channel transform.
* Generalize previous fix to all of zlib.test; also, tidy up [chan configure] ↵ferrieux2011-11-111-23/+23
| | | | flags across zlib.test.
* Use nonblocking writes in single-threaded IO tests to avoid deadlocks when ↵ferrieux2011-11-111-1/+1
| | | | going beyond OS buffers [Bug 3428756].
* [Bug 3390073]: Return the correct length of written data for a compressingdkf2011-08-121-0/+32
| | | | transform, ensuring that buffers are written exactly once instead of multiple times or not at all (producing an invalid file).
* Now that we're no longer using SCM based on RCS, the RCS Keyword linesdgp2011-03-021-2/+0
| | | cause more harm than good. Purged them (except in zlib files).
* [Bug 2818131] further tests added to cover use of mismatched zlib algorithmspatthoyts2010-02-261-2/+76
| | | | | | Using zlib gzip to write and inflate to read from a channel where gets was also used for reading data was causing a crash. This has been fixed by Andreas' last commit.
* ZlibTransformClose may be called with a NULL interpreter during finalization andpatthoyts2009-07-101-36/+168
| | | | | Tcl_SetChannelError requires a list. Added some tests to ensure error propagation from the zlib library to the interp.
* [Bug #2818131] Added tests and fixed a typo that broke zlib push for deflate ↵patthoyts2009-07-091-6/+94
| | | | format.
* Additional tests for zlib stacked channel with fcopy in various modes.patthoyts2009-07-051-1/+120
|
* Make [zlib push] work with [fcopy].dkf2009-07-051-1/+29
|
* Fix bug 2662434dkf2009-03-041-1/+4
|
* fix 64bit test failuresdas2008-12-211-5/+5
|
* Compressing and decompressing channel transformation support.dkf2008-12-181-6/+31
| | | | Note that there may be "quality-of-implementation" issues left...
* Added constraintsdkf2008-12-141-18/+20
|
* Many improvements to docsdkf2008-12-131-2/+2
|
* Added test for gzip header accessdkf2008-12-121-1/+6
|
* Added basic test suite for zlibdkf2008-12-111-0/+104