Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Prep 4.1.0 release.4.1.0 | Tres Seaver | 2014-12-29 | 2 | -2/+3 |
| | |||||
* | Note blocker for PyPy3 support. | Tres Seaver | 2014-12-29 | 3 | -0/+10 |
| | |||||
* | Add support for Python 3.4 | Tres Seaver | 2014-12-29 | 5 | -5/+17 |
| | |||||
* | Garden ReST. | Tres Seaver | 2014-12-29 | 2 | -27/+28 |
| | | | | [ci skip] | ||||
* | Speed up Travis builds via 'sudo: false' | Tres Seaver | 2014-12-20 | 1 | -0/+1 |
| | |||||
* | Update to latest bootstrap. | Tres Seaver | 2014-12-20 | 1 | -186/+98 |
| | |||||
* | svb | Tres Seaver | 2014-01-13 | 2 | -1/+6 |
| | |||||
* | Prep 4.0.0 release.4.0.0 | Tres Seaver | 2014-01-13 | 2 | -3/+3 |
| | |||||
* | Rewrite the print testing so it works in Python 3.3. | Maurits van Rees | 2014-01-13 | 1 | -2/+10 |
| | |||||
* | Catch warnings in test_translate_existing. | Maurits van Rees | 2014-01-13 | 2 | -12/+19 |
| | | | | Otherwise a warning is printed each time you run the tests. | ||||
* | Minor pep8 for long lines. | Maurits van Rees | 2014-01-13 | 2 | -7/+14 |
| | |||||
* | Merge branch 'master' into maurits-gettext | Maurits van Rees | 2014-01-11 | 1 | -7/+7 |
|\ | |||||
| * | Try to work around build failure on Travis | Marius Gedminas | 2013-09-26 | 1 | -7/+7 |
| | | | | | | | | | | | | | | | | | | | | | | I can reproduce that failure locally, using tox. Only python 2.7 fails, Python 2.6 and 3.3 (and PyPy) are fine. The failure is a difference in filenames (relative versus absolute) in test_sa4.html. I've tried digging, but gave up. The failure appears if I run tox -e py27. It doesn't appear if I run tox -e py27 --develop. So let's tell Travis to use the latter, to at least catch new regressions. | ||||
* | | Fix warning of msgid with different default for Python 3. | Maurits van Rees | 2013-09-21 | 1 | -2/+8 |
| | | |||||
* | | Fix possible UnicodeDecodeError in warning when msgid already exists. | Maurits van Rees | 2013-09-21 | 3 | -3/+32 |
|/ | | | | | | | | This happens when comparing 'Read more…' with 'Read moreā¦'. That should not give a warning, because they are really the same, but that would need a fix somewhere else, in the code that gets the value of an i18n:attribute in a different way than an i18n:translate. Added test for the proper printing of the warning. | ||||
* | modernize bootstrap.py | Adam Groszer | 2013-07-06 | 1 | -53/+165 |
| | |||||
* | 'Updated Travis YAML.' | Stephan Richter | 2013-03-04 | 1 | -0/+12 |
| | |||||
* | Mention one rather important change in the changelog | Marius Gedminas | 2013-02-26 | 2 | -0/+5 |
| | |||||
* | Back to development: 4.0.0a2 | Marius Gedminas | 2013-02-15 | 2 | -1/+7 |
| | |||||
* | Preparing release 4.0.0a14.0.0a1 | Marius Gedminas | 2013-02-15 | 2 | -3/+3 |
| | |||||
* | Rename *.txt to *.rst for prettier Github rendering | Marius Gedminas | 2013-02-12 | 6 | -3/+3 |
| | |||||
* | Update bootstrap.py to latest v2 bootstrap | Marius Gedminas | 2013-02-12 | 1 | -78/+116 |
| | | | | (downloaded from http://downloads.buildout.org/2/bootstrap.py) | ||||
* | Add a .gitignore | Marius Gedminas | 2013-02-12 | 1 | -0/+5 |
| | |||||
* | Fix occasional \r\r\n newlines with XML test files on Windows | Marius Gedminas | 2013-02-11 | 2 | -8/+23 |
| | | | | | The test suite now passes on both Windows and Unix, with both newline styles (CRLF and LF). | ||||
* | Make runtest.py use optparse. | Marius Gedminas | 2013-02-11 | 3 | -54/+74 |
| | | | | | Make python -m zope.tal.runtest work even when your current working directory is not src/zope/tal/ | ||||
* | Use optparse instead of getopt. | Marius Gedminas | 2013-02-11 | 1 | -80/+41 |
| | |||||
* | Whoops, another use of the obsolete setpath hack. | Marius Gedminas | 2013-02-11 | 1 | -3/+0 |
| | |||||
* | Make it possible to filter FileTestCases by filename. | Marius Gedminas | 2013-02-09 | 1 | -0/+5 |
| | | | | | Example: bin/test -t test04.xml | ||||
* | This should make the tests pass on Windows. | Marius Gedminas | 2013-02-08 | 1 | -0/+5 |
| | | | | Does anybody care about Windows? | ||||
* | Explain the - parsing difference (it's the stdlib) | Marius Gedminas | 2013-02-07 | 1 | -2/+2 |
| | |||||
* | Add PyPy support because it's not hard | Marius Gedminas | 2013-02-07 | 4 | -5/+11 |
| | |||||
* | We've reached Py3K: add classifiers, enable tox | Marius Gedminas | 2013-02-07 | 3 | -3/+6 |
| | |||||
* | Towards Py3K: - inside attribute values | Marius Gedminas | 2013-02-07 | 1 | -5/+7 |
| | | | | | | | | I've no idea why this case is handled differently on Python 2 and Python 3. There's precedent for different escaping on different Python versions: before 2.6 and since 2.6. Speaking of which, let's drop the "before 2.6" case. | ||||
* | Towards Py3K: Python code inside templates | Marius Gedminas | 2013-02-07 | 1 | -9/+9 |
| | |||||
* | Towards Py3K: use html.escape instead of cgi.escape | Marius Gedminas | 2013-02-07 | 2 | -6/+14 |
| | | | | | | | | There's one difference -- html.escape also escapes single quotes ('), while cgi.escape doesn't. html.escape doesn't exist on Python 2.x, while cgi.escape emits a deprecation warning on Python 3.x. To get unified behavior across all Python versions I had to inline the escape function in zope.tal.taldefs.quote(). | ||||
* | Towards Py3K: builtins (part 2) | Marius Gedminas | 2013-02-07 | 1 | -0/+7 |
| | |||||
* | Towards Py3K: use io.StringIO | Marius Gedminas | 2013-02-07 | 5 | -6/+34 |
| | | | | | | | While Python 2.6+ also has io.StringIO(), using it causes numerous test failures -- because io.StringIO() insists on unicode, and our tests are full of (ASCII-only) native string literals. For this reason I decided to keep using StringIO.StringIO or cStringIO.StringIO on Python 2.x. | ||||
* | Towards Py3K: xml.parsers.expat wants byte strings (part 2) | Marius Gedminas | 2013-02-07 | 1 | -1/+2 |
| | | | | Also avoid ResourceWarning by explicitly closing the file. | ||||
* | Make attribute ordering deterministic | Marius Gedminas | 2013-02-07 | 3 | -6/+6 |
| | |||||
* | Better test failure reporting in FileTestCase | Marius Gedminas | 2013-02-07 | 2 | -11/+22 |
| | | | | | | | | | Previously the diff would be printed to stdout in the middle of a sea of dots, and *then* you'd get test failure assertions mentioning only the filename (and on Python 3.3 you'd also get a gratuitous SystemExit traceback in addition to the assertion failure) Now you get a sane single assertion failure with the diff inside it. | ||||
* | Replace ndiff.py with stdlib's difflib | Marius Gedminas | 2013-02-07 | 2 | -673/+2 |
| | |||||
* | Remove the ancient setpath.py hack | Marius Gedminas | 2013-02-07 | 2 | -49/+0 |
| | | | | | | | | | | | | | | Now you can debug failures by doing, e.g. .tox/py27/bin/python -m zope.tal.driver -s src/zope/tal/tests/input/test24.html .tox/py33/bin/python -m zope.tal.driver -s src/zope/tal/tests/input/test24.html and comparing the emitted bytecode. Note: if you want to use bin/py, you must also use -- to avoid a bug in buildout's interpreter wrapper script: bin/py -m zope.tal.driver -- -s src/zope/tal/tests/input/test24.html | ||||
* | Towards Py3K: avoid deprecated assertion spellings | Marius Gedminas | 2013-02-07 | 4 | -20/+20 |
| | |||||
* | Towards Py3K: xml.parsers.expat wants byte strings | Marius Gedminas | 2013-02-07 | 1 | -13/+13 |
| | |||||
* | Missing errno import | Marius Gedminas | 2013-02-07 | 1 | -0/+1 |
| | |||||
* | Unused imports | Marius Gedminas | 2013-02-07 | 2 | -2/+0 |
| | |||||
* | Towards Py3K: range() returns an iterator | Marius Gedminas | 2013-02-07 | 1 | -1/+1 |
| | |||||
* | Towards Py3K: use f(*args) instead of apply(f, args) | Marius Gedminas | 2013-02-07 | 1 | -1/+1 |
| | |||||
* | Towards Py3K: builtins | Marius Gedminas | 2013-02-07 | 4 | -0/+24 |
| | |||||
* | Towards Py3K: stdlib reorganization | Marius Gedminas | 2013-02-07 | 2 | -3/+15 |
| | | | | | | | StringIO is omitted from this checkin: we can't use io.StringIO nor io.BytesIO, because zope.tal likes to write both byte strings and Unicode objects into the same stream. This will require careful untangling. |