WebKit Bugzilla
New
Browse
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
RESOLVED FIXED
137338
Remove workarounds for versions of Python before 2.7
https://bugs.webkit.org/show_bug.cgi?id=137338
Summary
Remove workarounds for versions of Python before 2.7
Éva Balázsfalvi
Reported
2014-10-02 02:55:30 PDT
Remove workarounds for versions of Python before 2.7
Attachments
Patch
(10.68 KB, patch)
2014-10-02 03:08 PDT
,
Éva Balázsfalvi
no flags
Details
Formatted Diff
Diff
Patch
(8.73 KB, patch)
2014-10-03 06:21 PDT
,
Éva Balázsfalvi
no flags
Details
Formatted Diff
Diff
Show Obsolete
(1)
View All
Add attachment
proposed patch, testcase, etc.
Éva Balázsfalvi
Comment 1
2014-10-02 03:08:17 PDT
Created
attachment 239098
[details]
Patch
WebKit Commit Bot
Comment 2
2014-10-03 03:14:40 PDT
Comment on
attachment 239098
[details]
Patch Clearing flags on attachment: 239098 Committed
r174264
: <
http://trac.webkit.org/changeset/174264
>
WebKit Commit Bot
Comment 3
2014-10-03 03:14:43 PDT
All reviewed patches have been landed. Closing bug.
WebKit Commit Bot
Comment 4
2014-10-03 05:11:39 PDT
Re-opened since this is blocked by
bug 137392
Carlos Alberto Lopez Perez
Comment 5
2014-10-03 05:25:24 PDT
This seems to have broken run-webkit-tests on all the bots (except EFL it seems, not sure): *
https://build.webkit.org/builders/Apple%20Mavericks%20Debug%20WK1%20%28Tests%29/builds/8010
*
https://build.webkit.org/builders/Apple%20Mavericks%20Debug%20WK2%20%28Tests%29/builds/7162
*
https://build.webkit.org/builders/Apple%20MountainLion%20Debug%20WK2%20%28Tests%29/builds/19455
*
https://build.webkit.org/builders/Apple%20MountainLion%20Release%20WK1%20%28Tests%29/builds/22775
*
https://build.webkit.org/builders/Apple%20MountainLion%20Release%20WK2%20%28Tests%29/builds/21216
*
https://build.webkit.org/builders/Apple%20Win%207%20Debug%20%28Tests%29/builds/62998
*
https://build.webkit.org/builders/GTK%20Linux%2064-bit%20Release%20%28Tests%29/builds/3203
On this log from the GTK bot (at the bottom) there is a python traceback that could give some hints about what is wrong:
https://build.webkit.org/builders/GTK%20Linux%2064-bit%20Release%20%28Tests%29/builds/3203/steps/layout-test/logs/stdio
Carlos Alberto Lopez Perez
Comment 6
2014-10-03 05:27:47 PDT
run-perf-tests seems to be also broken. Locally I get: $ Tools/Scripts/run-perf-tests --platform gtk --release -2 Interactive/window-resize.html Running 1 tests Running Interactive/window-resize.html (1 of 1) Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Last character read from DRT stdout line was not a newline! This indicates either a NRWT or DRT bug. Traceback (most recent call last): File "Tools/Scripts/run-perf-tests", line 40, in <module> sys.exit(PerfTestsRunner().run()) File "/home/clopez/webkit/webkit/Tools/Scripts/webkitpy/performance_tests/perftestsrunner.py", line 205, in run unexpected = self._run_tests_set(sorted(list(tests), key=lambda test: test.test_name())) File "/home/clopez/webkit/webkit/Tools/Scripts/webkitpy/performance_tests/perftestsrunner.py", line 372, in _run_tests_set metrics = test.run(self._options.time_out_ms) File "/home/clopez/webkit/webkit/Tools/Scripts/webkitpy/performance_tests/perftest.py", line 132, in run if not self._run_with_driver(driver, time_out_ms): File "/home/clopez/webkit/webkit/Tools/Scripts/webkitpy/performance_tests/perftest.py", line 179, in _run_with_driver self._filter_output(output) File "/home/clopez/webkit/webkit/Tools/Scripts/webkitpy/performance_tests/perftest.py", line 273, in _filter_output output.text = '\n'.join([line for line in re.split('\n', output.text) if not self._should_ignore_line(self._lines_to_ignore_in_parser_result, line)]) TypeError: sequence item 0: expected string, bytearray found
Csaba Osztrogonác
Comment 7
2014-10-03 05:44:13 PDT
Comment on
attachment 239098
[details]
Patch View in context:
https://bugs.webkit.org/attachment.cgi?id=239098&action=review
> Tools/Scripts/webkitpy/port/driver.py:534 > - self.content = str() # FIXME: Should be bytearray() once we require Python 2.6. > + self.content = bytearray()
This one ...
> Tools/Scripts/webkitpy/port/server_process.py:99 > - self._output = str() # bytesarray() once we require Python 2.6 > - self._error = str() # bytesarray() once we require Python 2.6 > + self._output = bytearray() > + self._error = bytearray()
... or this one broke the tests. Éva, could you check what happened?
Éva Balázsfalvi
Comment 8
2014-10-03 06:21:38 PDT
Created
attachment 239203
[details]
Patch Took out the str() to bytearray() modification which seemingly made the layout tests fail. I'll investigate what caused the problem later.
Carlos Alberto Lopez Perez
Comment 9
2014-10-03 06:35:01 PDT
(In reply to
comment #8
)
> Created an attachment (id=239203) [details] > Patch > > Took out the str() to bytearray() modification which seemingly made the layout tests fail. I'll investigate what caused the problem later.
I tested this patch, and both run-webkit-test and run-perf-test seems to work as expected on the GTK port.
Csaba Osztrogonác
Comment 10
2014-10-03 07:26:51 PDT
Comment on
attachment 239203
[details]
Patch Let's do it without bytearray(), r=me.
WebKit Commit Bot
Comment 11
2014-10-03 08:03:32 PDT
Comment on
attachment 239203
[details]
Patch Clearing flags on attachment: 239203 Committed
r174266
: <
http://trac.webkit.org/changeset/174266
>
WebKit Commit Bot
Comment 12
2014-10-03 08:03:35 PDT
All reviewed patches have been landed. Closing bug.
Csaba Osztrogonác
Comment 13
2014-10-08 02:28:58 PDT
Comment on
attachment 239203
[details]
Patch View in context:
https://bugs.webkit.org/attachment.cgi?id=239203&action=review
> Tools/Scripts/webkitpy/common/system/autoinstall.py:-248 > - # This is a replacement for ZipFile.extractall(), which is > - # available in Python 2.6 but not in earlier versions. > - # NOTE: The version in 2.6.1 (which shipped on Snow Leopard) is broken! > - def _extract_all(self, zip_file, target_dir): > - for name in zip_file.namelist(): > - path = os.path.join(target_dir, name) > - if not os.path.basename(path): > - # Then the path ends in a slash, so it is a directory. > - os.makedirs(path) > - continue > - > - try: > - # We open this file w/o encoding, as we're reading/writing > - # the raw byte-stream from the zip file. > - outfile = open(path, 'wb') > - except IOError: > - # Not all zip files seem to list the directories explicitly, > - # so try again after creating the containing directory. > - _log.debug("Got IOError: retrying after creating directory...") > - dirname = os.path.dirname(path) > - os.makedirs(dirname) > - outfile = open(path, 'wb') > - > - try: > - outfile.write(zip_file.read(name)) > - finally: > - outfile.close() > -
Unfortunately it broke autoinstalling python-irclib-0.4.8.zip with the stock python on - Ubuntu 12.04: python 2.7.3 - Ubuntu 14.04: python 2.7.6 - OS X 10.9: python 2.7.5 I think we should add this workaround back with proper comment.
> Tools/Scripts/webkitpy/common/system/autoinstall.py:235 > - self._extract_all(zip_file, scratch_dir) > + zip_file.extractall(zip_file, scratch_dir)
extractall raises the following exception: Auto-installing package: irclib.py Traceback (most recent call last): File "/Users/oszi/WebKit/Tools/Scripts/webkitpy/test/main.py", line 234, in <module> sys.exit(main()) File "/Users/oszi/WebKit/Tools/Scripts/webkitpy/test/main.py", line 72, in main return not tester.run() File "/Users/oszi/WebKit/Tools/Scripts/webkitpy/test/main.py", line 122, in run return self._run_tests(names) File "/Users/oszi/WebKit/Tools/Scripts/webkitpy/test/main.py", line 132, in _run_tests autoinstall_everything() File "/Users/oszi/WebKit/Tools/Scripts/webkitpy/thirdparty/__init__.py", line 166, in autoinstall_everything getattr(_hook, method)() File "/Users/oszi/WebKit/Tools/Scripts/webkitpy/thirdparty/__init__.py", line 150, in _install_irc url_subpath="irclib.py") File "/Users/oszi/WebKit/Tools/Scripts/webkitpy/common/system/autoinstall.py", line 473, in install raise Exception(message) Exception: Error auto-installing the irclib.py package to: "/Users/oszi/WebKit/Tools/Scripts/webkitpy/thirdparty/autoinstalled/irc/irclib.py" --> Inner message: "There is no item named '/' in the archive"
Csaba Osztrogonác
Comment 14
2014-10-08 08:49:59 PDT
added back in
http://trac.webkit.org/changeset/174448
Note
You need to
log in
before you can comment on or make changes to this bug.
Top of Page
Format For Printing
XML
Clone This Bug