| Age | Commit message (Collapse) | Author |
|---|
|
|
|
Problem: Python 2 tests don't have the same output everywhere.
Solution: Make the Python 2 tests more portable. (ZyX)
|
|
Problem: Python 2: loading modules doesn't work well.
Solution: Fix the code. Add more tests. (ZyX)
|
|
Problem: Loading Python modules is not tested.
Solution: Enable commented-out tests, add missing files. (ZyX)
|
|
Problem: Not easy to load Python modules.
Solution: Search "python2", "python3" and "pythonx" directories in
'runtimepath' for Python modules. (ZyX)
|
|
Problem: New regexp engine fails on "\(\<command\)\@<=.*"
Solution: Fix rule for postponing match. Further tune estimating whether
postponing works better. Add test.
|
|
Problem: MS-DOS: "make test" uses external rmdir command.
Solution: Rename "rmdir" to "rd". (Taro Muraoka)
|
|
Problem: New regexp engine: Matching plain text could be faster.
Solution: Detect a plain text match and handle it specifically. Add
vim_regfree().
|
|
Problem: Running test 49 takes a long time.
Solution: Don't have it grep all files.
|
|
Problem: New regexp engine: \%> not supported.
Solution: Implement \%>.
|
|
Problem: New regexp engine: \%[abc] not supported.
Solution: Implement \%[abc]. Add tests.
|
|
Problem: Many users don't like the cursor line number when 'relativenumber'
is set.
Solution: Have four combinations with 'number' and 'relativenumber'.
(Christian Brabandt)
|
|
Problem: New regexp engine: \%'m not supported.
Solution: Implement \%'m. Add tests.
|
|
Problem: New regexp engine: \%V not supported.
Solution: Implement \%V. Add tests.
|
|
Problem: nfa_recognize_char_class() implementation is inefficient.
Solution: Use bits in an int instead of chars in a string. (Dominique Pelle)
|
|
Problem: New regexp engine does not handle "~".
Solution: Add support for "~".
|
|
Problem: Python: a few more memory problems.
Solution: Add and remove Py_XDECREF(). (ZyX)
|
|
Problem: Python: Changing directory with os.chdir() causes problems for
Vim's notion of directories.
Solution: Add vim.chdir() and vim.fchdir(). (ZyX)
|
|
Problem: Python: popitem() was not defined in a standard way.
Solution: Remove the argument from popitem(). (ZyX)
|
|
Problem: New regexp engine: When a sub expression is empty \1 skips a
character.
Solution: Make \1 try the current position when the match is emtpy.
|
|
Problem: Tests 86 and 87 fail on MS-Windows. (Ken Takata)
Solution: Fix platform-specific stuff. (ZyX)
|
|
Problem: New regexp engine: \@<= and \@<! are not implemented.
Solution: Implement look-behind matching. Fix off-by-one error in old
regexp engine.
|
|
Problem: A leading star is not seen as a normal char when \{} follows.
Solution: Save and restore the parse state properly.
|
|
Problem: Old regexp engine accepts illegal range, new one doesn't.
Solution: Also accept the illegal range with the new engine.
|
|
Problem: New regexp engine: Non-greedy multi doesn't work.
Solution: Implement \{-}.
|
|
Problem: New regexp engine: only accepts up to \{,10}.
Solution: Remove upper limit. Remove dead code with NFA_PLUS.
|
|
Problem: New regexp engine: Does not support \%^ and \%$.
Solution: Support matching start and end of file.
|
|
Problem: New regexp engine: Problem with \@= matching.
Solution: Save and restore nfa_match.
|
|
Problem: Test 86 fails.
Solution: Comment out the parts that don't work. Make it pass on 32 bit
systems.
|
|
Problem: Test 87 fails.
Solution: Fix the test for Python 3.3. (ZyX) Make it pass on 32 bit systems.
|
|
Problem: New regexp engine: \@! doesn't work.
Solution: Implement the negated version of \@=.
|
|
Problem: New regexp engine: \@= and \& don't work.
Solution: Make these items work. Add column info to logging.
|
|
Problem: New regexp engine may run out of states.
Solution: Allocate states dynamically. Also make the test report errors.
|
|
Problem: New regexp engine: backreferences don't work correctly.
Solution: Add every possible start/end position on the state stack.
|
|
Problem: Vim crashes in Python tests. Compiler warning for unused function.
Solution: Disable the tests for now. Move the function.
|
|
Problem: Python: Script is auto-loaded on function creation.
Solution: Python patch 27. (ZyX)
|
|
Problem: Python: Insufficient exception and error testing.
Solution: Python patch 25. (ZyX)
|
|
Problem: Python: Function is not standard.
Solution: Python patch 22: make Function subclassable. (ZyX)
|
|
Problem: Python: List is not standard.
Solution: Python patch 21: Add standard methods and fields. (ZyX)
|
|
Problem: Python: Dictionary is not standard.
Solution: Python patch 20: Add standard methods and fields. (ZyX)
|
|
Problem: Python: not enough compatibilty.
Solution: Python patch 16: Make OutputWritelines support any sequence object
(ZyX) Note: tests fail
|
|
Problem: Negated collection does not match newline.
Solution: Handle newline differently. (Hiroshi Shirosaki)
|
|
Problem: Can't build without the +autocmd feature. (Elimar Riesebieter)
Solution: Fix use of buf and curbuf.
|
|
Problem: Python: dir() does not work properly.
Solution: Python patch 8. Add __dir__ method to all objects with custom
tp_getattr supplemented by __members__ attribute for at least
python-2* versions. __members__ is not mentioned in python-3*
dir() output even if it is accessible. (ZyX)
|
|
Problem: Python: No {Buffer,TabPage,Window}.valid attributes.
Solution: Python patch 5: add .valid (ZyX)
|
|
Problem: Python: can't assign to vim.Buffer.name.
Solution: Python patch 3. (ZyX)
|
|
Problem: New regexp engine does not support \%23c, \%<23c and the like.
Solution: Implement them. (partly by Yasuhiro Matsumoto)
|
|
Problem: Look-behind matching is very slow on long lines.
Solution: Add a byte limit to how far back an attempt is made.
|
|
Problem: "\1" .. "\9" are not supported in the new regexp engine.
Solution: Implement them. Add a few more tests.
|
|
Problem: "\ze" is not supported by the new regexp engine.
Solution: Make "\ze" work.
|
