summaryrefslogtreecommitdiff
path: root/smmap
Commit message (Collapse)AuthorAgeFilesLines
* Added assertion in spot that might help tracking down the issue when it happensgitpython-issue-251-debuggingSebastian Thiel2015-01-301-0/+1
| | | | [skip ci]
* All tests work, bumped versionSebastian Thiel2015-01-074-4/+13
|
* Initial attempt to fix resource usageSebastian Thiel2015-01-074-49/+46
| | | | | Reference counting is now done manually, but it seems that things can still go wrong at least during testing
* Restore compatibility to python 3.0 to 3.4v0.8.5Sebastian Thiel2015-01-062-11/+28
|
* Fixed python 3 performance regressionv0.8.4Sebastian Thiel2015-01-062-3/+4
| | | | It makes the difference between tests in 110s, or 11s
* Applied autopep8Sebastian Thiel2015-01-049-149/+161
| | | | autopep8 -v -j 8 --max-line-length 120 --in-place --recursive
* Drop semicolon at end of statementAntoine Musso2014-11-161-2/+2
| | | | Fix pep8: E703 statement ends with a semicolon
* pep8 lintingAntoine Musso2014-11-169-260/+258
| | | | | | | | | | | | | | | E201 whitespace after '(' E203 whitespace before ',' E221 multiple spaces before operator E225 missing whitespace around operator E227 missing whitespace around bitwise or shift operator E231 missing whitespace after ',' E251 unexpected spaces around keyword / parameter equals W291 trailing whitespace W293 blank line contains whitespace E302 expected 2 blank lines, found 1 E303 too many blank lines (3) W391 blank line at end of file
* Fixed incorrect usage of memoryview. It's not getting faster thoughSebastian Thiel2014-11-131-1/+1
| | | | [ skip ci ]
* Fixed a few typos and major linter errorsSebastian Thiel2014-11-137-26/+33
|
* Initial improvements to get rid of the performance regression in py3.Sebastian Thiel2014-11-123-22/+22
| | | | | Byte buffer concatenations are considerably slower here for some reason. Also there was no need for the memorybuffer.
* warnings fixesYoan Blanc2014-07-252-57/+66
|
* Change / to // (integer division) in test_buf.pyMarc Abramowitz2014-06-161-3/+4
| | | | | This fixes (the last!) test failure in Python 3, which uses "true division" for /
* Make __getitem__ handle slice for Python 3Marc Abramowitz2014-06-161-0/+2
| | | | | Python 3 doesn't have __getslice__ instead it uses __getitem__ with a slice object.
* Merge pull request #13 from Byron/use_bytes_instead_of_strMarc Abramowitz2014-06-161-1/+7
|\ | | | | Use bytes() instead of str()
| * Use bytes() instead of str()Marc Abramowitz2014-06-161-1/+7
| | | | | | | | | | bytes() is more accurate and is actually correct in Python 3, whereas str() is incorrect in Python 3, because it's a Unicode string.
* | Fix typo: "optimial" => "optimal"Marc Abramowitz2014-06-161-1/+1
|/
* Fix 2 instances of "containnig" => "containing"Marc Abramowitz2014-06-162-2/+2
|
* Deal with lack of `buffer` in py3Marc Abramowitz2014-06-163-1/+12
|
* Change / to // (integer division) in several placesMarc Abramowitz2014-06-153-5/+5
| | | | | This fixes a bunch of bugs and test failures in Python 3, which uses "true division" for /
* Delay importing sys.getrefcount until neededMarc Abramowitz2014-06-132-3/+2
| | | | | | This makes it possibly to at least install on PyPy Addresses: GH-4 ("pypy compatibility")
* _need_compat_layer is not required in python 3.David Black2014-05-161-1/+1
| | | | Signed-off-by: David Black <dblack@atlassian.com>
* instead of writing a string to a buffer api (in python 3) write a buffer.David Black2014-05-161-1/+1
| | | | Signed-off-by: David Black <dblack@atlassian.com>
* Initial work for supporting python 3 (>= 3.3).David Black2014-05-168-30/+39
| | | | Signed-off-by: David Black <dblack@atlassian.com>
* tabs to spacesSebastian Thiel2014-02-099-1455/+1455
|
* Bumped version to 0.8.2v0.8.2Sebastian Thiel2012-01-181-1/+1
|
* Added some more in-code comments to clarify why that exception is caughtSebastian Thiel2012-01-181-0/+5
|
* Workaround for #1Kenneth Reitz2011-07-081-5/+8
|
* updated changelog, bumped versionv0.8.1Sebastian Thiel2011-07-051-1/+1
|
* Fixed possible bug as a method was called using an old signature. Apparently ↵Sebastian Thiel2011-07-051-2/+1
| | | | this code branch never ran in the tests
* Removed possibly invalid documentation tagsSebastian Thiel2011-06-133-27/+35
|
* It turned out the :note: docstring was not supported. Now all documentation ↵Sebastian Thiel2011-06-131-15/+24
| | | | is being generated
* Finished tutorial section, umproved capabilities of the buffer ↵Sebastian Thiel2011-06-134-3/+100
| | | | implementation to be more pythonic. Unfortunately, not all docs build yet because of some typical sphinx issue that results in an error which doesn't at all tell what the culprit actually is
* Fixed up configuration to create api documentation for the code. Improved ↵Sebastian Thiel2011-06-132-3/+6
| | | | some markup to be valid for sphinx.
* Added option to help test cases to succeed on windows. Its for test systems onlySebastian Thiel2011-06-131-3/+28
|
* A tiny win32 fix, once againSebastian Thiel2011-06-101-3/+4
|
* Merge branch 'single'Sebastian Thiel2011-06-105-275/+400
|\
| * Some more adjustments to make it work in all python versionsSebastian Thiel2011-06-102-3/+7
| |
| * Finally the mem manager and buffer tests run with the static one tooSebastian Thiel2011-06-103-109/+131
| |
| * test are running, once again, but not yet complete regarding the static managerSebastian Thiel2011-06-102-93/+70
| |
| * Implemented static memory manager, for now without testSebastian Thiel2011-06-103-15/+99
| |
| * Changed design of memory managers to support different implementations. ↵Sebastian Thiel2011-06-103-98/+134
| | | | | | | | Currently there is a non-implemented static version, as well as the previous sliding window version.
| * moved code from cursor into manager, as it belongs there. This is a design ↵Sebastian Thiel2011-06-103-120/+122
|/ | | | error inherited from c++, but actually it makes overrides a bit harder, or lets say, less native, as the sliding mechanics where implemented in a class which is just the handle to a memory map in the end, which doesn't have to care about its allocation
* Changed names to be more descriptive, hopefully. This opens op the option to ↵Sebastian Thiel2011-06-106-53/+53
| | | | implement such a manager differently, without the sliding window mechanics, which would be quite simple and not much better than a map of mmaps in the end
* Made all types available in root packageSebastian Thiel2011-06-101-0/+4
|
* Implemented __len__ method in buffer, including small test. This has its ↵Sebastian Thiel2011-06-102-3/+28
| | | | caveats, but should be fine for responsible clients
* Fixed missing ALLOCATIONGRANULARIY in python <2.6. Python is as unportable ↵Sebastian Thiel2011-06-102-2/+9
| | | | as ever across versions with simple functionality
* Fixed some mapping issues on windows. Fixed some tests to deal with the very ↵Sebastian Thiel2011-06-105-23/+24
| | | | different granularity
* Added tests for the fd case. It shows that this is measurably faster than ↵Sebastian Thiel2011-06-104-138/+157
| | | | the string path version because there is less system overhead. Good to know actally
* System can now deal with file descriptors as input, but it still requires ↵Sebastian Thiel2011-06-103-25/+63
| | | | some more testing. Also fds need to remain open to be usable for new mapped regions
View Lorry Controller Status