848ae584b88bbc99672cbb87ea2a253b6127fbb4
[WebKit-https.git] / Tools / ChangeLog
1 2012-01-26  Dirk Pranke  <dpranke@chromium.org>
2
3         webkitpy: re-land cleanup of test scaffolding
4         https://bugs.webkit.org/show_bug.cgi?id=77153
5
6         Reviewed by Eric Seidel.
7
8         This is an attempt to reland the change in bug 76973 / 
9         r105935. My earlier attempt to manipulate PYTHONPATH
10         seems to only work some of the time, for reasons that
11         elude me. Directly adding Tools/Scripts to sys.path from
12         inside the test script appears to work (more?) reliably.
13
14         Also, it seemed like the files didn't actually get deleted
15         the last time; I don't know why.
16
17         * Scripts/test-webkitpy:
18         * Scripts/webkitpy/common/system/executive_unittest.py:
19         (command_line):
20         (ExecutiveTest):
21         (ExecutiveTest.test_run_command_args_type):
22         (ExecutiveTest.test_run_command_with_unicode):
23         (ExecutiveTest.test_running_pids):
24         (main):
25         * Scripts/webkitpy/common/system/fileutils.py: Removed.
26         * Scripts/webkitpy/test/cat.py: Removed.
27         * Scripts/webkitpy/test/cat_unittest.py: Removed.
28         * Scripts/webkitpy/test/echo.py: Removed.
29         * Scripts/webkitpy/test/echo_unittest.py: Removed.
30
31 2012-01-26  Ojan Vafai  <ojan@chromium.org>
32
33         run-webkit-tests calls "nm" when it doesn't need to
34         https://bugs.webkit.org/show_bug.cgi?id=77143
35
36         Reviewed by Dirk Pranke.
37
38         If we are not running any test that can be skipped by checking
39         the DRT feature list or looking at the list of symbols, avoid calling
40         out to DRT or calling "nm".
41
42         This saves ~700ms on my Mac Pro. In cases where you still have to call out
43         to "nm" it's only a couple ms slower.
44
45         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
46         (TestExpectations.__init__):
47         * Scripts/webkitpy/layout_tests/port/base.py:
48         (Port.skipped_tests):
49         * Scripts/webkitpy/layout_tests/port/webkit.py:
50         (WebKitPort._should_check_for_missing_features):
51         (WebKitPort):
52         (WebKitPort._should_check_for_missing_symbols):
53         (WebKitPort._skipped_tests_for_unsupported_features):
54         (WebKitPort.skipped_layout_tests):
55         (WebKitPort.skipped_tests):
56         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
57         (WebKitPortTest.test_skipped_directories_for_symbols):
58         (test_skipped_directories_for_features):
59         (test_skipped_directories_for_features_no_matching_tests_in_test_list):
60         (test_skipped_layout_tests):
61
62 2012-01-26  Benjamin Poulain  <bpoulain@apple.com>
63
64         Support a suffix on ChangeLog filenames based on a configuration file
65         https://bugs.webkit.org/show_bug.cgi?id=76956
66
67         Reviewed by Kentaro Hara.
68
69         Modify the tools used to generate, edit and resolve ChangeLogs to support
70         the addition of a suffix to the ChangeLog filename.
71
72         This patch makes it easier to support alternate ChangeLog files in branches.
73
74         * Scripts/VCSUtils.pm:
75         (changeLogSuffix):
76         (changeLogFileName):
77         * Scripts/commit-log-editor:
78         * Scripts/prepare-ChangeLog:
79         (getLatestChangeLogs):
80         (generateNewChangeLogs):
81         (generateFileList):
82         * Scripts/resolve-ChangeLogs:
83         (findChangeLog):
84
85 2012-01-26  Sheriff Bot  <webkit.review.bot@gmail.com>
86
87         Unreviewed, rolling out r106038.
88         http://trac.webkit.org/changeset/106038
89         https://bugs.webkit.org/show_bug.cgi?id=77142
90
91         Caused a bunch of skipped tests to not be skipped. (Requested
92         by ojan on #webkit).
93
94         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
95         (TestExpectationParser):
96         (TestExpectationParser.parse):
97         (TestExpectationParser._parse_line):
98         (TestExpectationParser._collect_matching_tests):
99         (TestExpectations.__init__):
100         (TestExpectations._add_skipped_tests):
101         * Scripts/webkitpy/layout_tests/port/base.py:
102         (Port.test_dirs):
103         (Port.normalize_test_name):
104         (Port.update_baseline):
105         (Port.layout_tests_dir):
106         (Port.relative_perf_test_filename):
107         (Port.abspath_for_test):
108
109 2012-01-25  Ojan Vafai  <ojan@chromium.org>
110
111         Parsing test_expecations.txt + Skipped lists takes too long
112         https://bugs.webkit.org/show_bug.cgi?id=77059
113
114         Reviewed by Dirk Pranke.
115
116         This saves ~100ms on the Apple Mac port.
117         -memoize a bunch of path methods.
118         -Avoid doing multiple disk accesses per line.
119         -Parse the skipped list directly instead of turning it into a test_expecations.txt
120         formatting string and parsing that.
121
122         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
123         (TestExpectationParser):
124         (TestExpectationParser.parse_skipped_tests):
125         (TestExpectationParser._parse_line):
126         (TestExpectationParser._collect_matching_tests):
127         (TestExpectations.__init__):
128         (TestExpectations._add_skipped_tests):
129         * Scripts/webkitpy/layout_tests/port/base.py:
130         (Port):
131         (Port.test_isfile):
132         (Port.normalize_test_name):
133         (Port.layout_tests_dir):
134         (Port.abspath_for_test):
135
136 2012-01-25  Ojan Vafai  <ojan@chromium.org>
137
138         Only store the SVN revision in the summarized results if we're on a builder
139         https://bugs.webkit.org/show_bug.cgi?id=76976
140
141         Reviewed by Dirk Pranke.
142
143         This shaves another ~130ms off the single test run-webkit-tests runtime.
144
145         * Scripts/webkitpy/layout_tests/controllers/manager.py:
146         (summarize_results):
147         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
148         (ResultSummaryTest.get_result):
149         (ResultSummaryTest):
150         (ResultSummaryTest.get_result_summary):
151         (ResultSummaryTest.get_unexpected_results):
152         (ResultSummaryTest.test_no_svn_revision):
153         (ResultSummaryTest.test_svn_revision):
154         * Scripts/webkitpy/layout_tests/port/base.py:
155         (Port.use_trac_links_in_results_html):
156
157 2012-01-26  Luciano Wolf  <luciano.wolf@openbossa.org>
158
159         [Qt][WK2] fast/loader tests failed after r94178
160         https://bugs.webkit.org/show_bug.cgi?id=67388
161
162         Reviewed by Simon Hausmann.
163
164         Applying the same fix used for Qt-WebKit1 from bug #67254 -
165         [Qt][DRT] Normalize file:///tmp/LayoutTests in LayoutTestController::pathToLocalResource()
166         https://bugs.webkit.org/show_bug.cgi?id=67254
167         Translate file:///tmp/LayoutTests/* urls into the repository LayoutTests directory, which is
168         derived from the running location of DumpRenderTree binary.
169
170         * WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
171         (WTR::LayoutTestController::pathToLocalResource):
172
173 2012-01-26  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
174
175         Implement overridePreference for boolean preferences in WebKitTestRunner
176         https://bugs.webkit.org/show_bug.cgi?id=77033
177
178         Reviewed by Alexey Proskuryakov.
179
180         For now overridePreference supports boolean values.
181
182         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
183         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
184         (WTR::LayoutTestController::setXSSAuditorEnabled):
185         (WTR::LayoutTestController::overridePreference):
186         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
187         * WebKitTestRunner/TestController.cpp:
188         (WTR::TestController::resetStateToConsistentValues): Reset the overrides
189         explicitly and properly disable the use of PageCache when resetting the
190         state. Tests that use PageCache can now enable it via overridePreference.
191
192 2012-01-26  Alexander Færøy  <alexander.faeroy@nokia.com>
193
194         [Qt] Support opening multiple URL's from the command line in the Qt MiniBrowser
195         https://bugs.webkit.org/show_bug.cgi?id=77100
196
197         Reviewed by Simon Hausmann.
198
199         This patch makes the MiniBrowser read multiple URL's from the command
200         line such that there will be one browser window opened for each URL.
201
202         * MiniBrowser/qt/MiniBrowserApplication.cpp:
203         (MiniBrowserApplication::handleUserOptions):
204
205 2012-01-26  Alexander Færøy  <alexander.faeroy@nokia.com>
206
207         [Qt] The Qt MiniBrowser crashes when given multiple URL's
208         https://bugs.webkit.org/show_bug.cgi?id=77009
209
210         Reviewed by Simon Hausmann.
211
212         This patch fixes a crash in BrowserWindow::newWindow() where a
213         BrowserWindow-object is constructed with a 0 pointer as the
214         WindowOptions parameter which leads to a crash.
215
216         * MiniBrowser/qt/BrowserWindow.cpp:
217         (BrowserWindow::BrowserWindow):
218         (BrowserWindow::newWindow):
219         * MiniBrowser/qt/BrowserWindow.h:
220
221 2012-01-26  Jesus Sanchez-Palencia  <jesus.palencia@openbossa.org>
222
223         [Qt] Use ICU if available
224         https://bugs.webkit.org/show_bug.cgi?id=76821
225
226         Reviewed by Simon Hausmann.
227
228         Adding unicode defines for a Qt 5 based build.
229
230         * qmake/mkspecs/features/features.prf:
231
232 2012-01-26  Nándor Huszka  <huszka.nandor@stud.u-szeged.hu>
233
234         [WK2] WebKitTestRunner needs layoutTestController.setGeolocationPermission
235         https://bugs.webkit.org/show_bug.cgi?id=42545
236
237         Reviewed by Zoltan Herczeg.
238
239         Implement the setGeolocationPermission method.
240
241         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
242         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
243         (WTR::LayoutTestController::setGeolocationPermission):
244         (WTR):
245         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
246         (LayoutTestController):
247
248 2012-01-26  Balazs Kelemen  <kbalazs@webkit.org>
249
250         [NRWT] Support --ignore-metrics
251         https://bugs.webkit.org/show_bug.cgi?id=76278
252
253         Reviewed by Tony Chang.
254
255         Port the feature from ORWT.
256
257         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
258         (SingleTestRunner._run_compare_test):
259         * Scripts/webkitpy/layout_tests/port/driver.py:
260         (DriverOutput):
261         (DriverOutput.strip_metrics):
262         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
263         (DriverOutputTest):
264         (DriverOutputTest.test_strip_metrics):
265         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
266         (_set_up_derived_options):
267         (parse_args):
268
269 2012-01-26  Sheriff Bot  <webkit.review.bot@gmail.com>
270
271         Unreviewed, rolling out r105935 and r105954.
272         http://trac.webkit.org/changeset/105935
273         http://trac.webkit.org/changeset/105954
274         https://bugs.webkit.org/show_bug.cgi?id=77080
275
276         test-webkitpy is still broken (Requested by abarth on
277         #webkit).
278
279         * Scripts/test-webkitpy:
280         * Scripts/webkitpy/common/system/executive_unittest.py:
281         (never_ending_command):
282         (ExecutiveTest.test_run_command_args_type):
283         (ExecutiveTest.test_run_command_with_unicode):
284         (ExecutiveTest.test_running_pids):
285         * Scripts/webkitpy/common/system/fileutils.py:
286         (make_stdout_binary):
287         * Scripts/webkitpy/test/cat.py:
288         (command_arguments):
289         (main):
290         * Scripts/webkitpy/test/cat_unittest.py:
291         (CatTest):
292         (CatTest.assert_cat):
293         (CatTest.test_basic):
294         (CatTest.test_no_newline):
295         (CatTest.test_unicode):
296         (CatTest.test_as_command):
297         * Scripts/webkitpy/test/echo.py:
298         (command_arguments):
299         (main):
300         * Scripts/webkitpy/test/echo_unittest.py:
301         (EchoTest):
302         (EchoTest.test_basic):
303         (EchoTest.test_no_newline):
304         (EchoTest.test_unicode):
305         (EchoTest.test_argument_order):
306         (EchoTest.test_empty_arguments):
307         (EchoTest.test_no_arguments):
308         (EchoTest.test_as_command):
309
310 2012-01-26  Gabor Rapcsanyi  <rgabor@webkit.org>
311
312         [GTK] Detect cross-compilation in webkitdirs.pm to not generate gtkdoc
313         https://bugs.webkit.org/show_bug.cgi?id=76138
314
315         Reviewed by Martin Robinson.
316
317         * Scripts/webkitdirs.pm:
318         (isCrossCompilation):
319         (buildAutotoolsProject):
320
321 2012-01-25  Adam Barth  <abarth@webkit.org>
322
323         Kentaro is a reviewer now.  Yay!
324
325         * Scripts/webkitpy/common/config/committers.py:
326
327 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
328
329         nrwt: running w/ no arguments fails
330         https://bugs.webkit.org/show_bug.cgi?id=77064
331
332         Unreviewed, build fix.
333
334         This change fixes a regression introduced in r105944
335         where running without --platform <foo> or one of the other
336         platform switches would cause a null deref :(.
337
338         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
339         (main):
340
341 2012-01-25  Eric Seidel  <eric@webkit.org>
342
343         Automatically CC someone from Apple on any changes to Obj-C public API
344         https://bugs.webkit.org/show_bug.cgi?id=77053
345
346         Reviewed by Timothy Hatcher.
347
348         This has long been de-facto practice, seems now that we have
349         fancy watchlist support, we should just make the tools
350         do this automatically.
351
352         * Scripts/webkitpy/common/config/watchlist:
353
354 2012-01-25  Antoine Labour  <piman@chromium.org>
355
356         Updating status in committers.py.
357
358         Unreviewed.
359
360         * Scripts/webkitpy/common/config/committers.py:
361
362 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
363
364         test-webkitpy: executive_unittest fails if PYTHONPATH isn't set
365         https://bugs.webkit.org/show_bug.cgi?id=77058
366
367         Reviewed by Eric Seidel.
368
369         r105935 broke executive_unittest by changing which script we
370         were running and not ensuring that the PYTHONPATH was set up
371         properly.
372
373         This change also fixes a regression introduced recently where
374         test-webkitpy wasn't figuring out the path to Tools/Scripts
375         right, either (so running ./test-webkitpy would fail).
376
377         * Scripts/test-webkitpy:
378         * Scripts/webkitpy/common/system/executive_unittest.py:
379         (script_dir):
380         (ExecutiveTest.setUp):
381         (ExecutiveTest):
382         (ExecutiveTest.tearDown):
383
384 2012-01-25  Adam Barth  <abarth@webkit.org>
385
386         Disable tests on the mac-ews while we sort out hte stability issues.
387
388         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
389         (MacEWS):
390
391 2012-01-23  Dirk Pranke  <dpranke@chromium.org>
392
393         investigate use of 'mac' and 'win' as fully-specified port names for the apple ports
394         https://bugs.webkit.org/show_bug.cgi?id=76475
395
396         Reviewed by Eric Seidel.
397
398         This patch cleans up the internal usage of port names for the
399         Apple mac and Apple win ports so that 'mac' and 'win are never
400         considered "fully specified" port names: they are still legal
401         input to the PortFactory.get() routine, but only if run on the given
402         platform (i.e. trying to PortFactory.get('mac') on windows or
403         linux will return an error, since we don't know which version of
404         the port is desired.
405
406         This also cleans up handling of the webkit2 variants, so that
407         the full port names consistently follow the
408         <operating_system>-<version>[-<wk2>] convention.
409
410         Lastly this patch adds some assertions and removes some
411         try/catch logic to catch more programming and usage errors;
412         previously mac-tiger might've been allowed -- and it would've
413         translated into using just the baselines in platform/mac -- but
414         now it should fail.
415
416         * Scripts/webkitpy/layout_tests/port/apple.py:
417         (ApplePort.determine_full_port_name):
418         (ApplePort.__init__):
419         (ApplePort):
420         (ApplePort._port_name_with_version):
421         * Scripts/webkitpy/layout_tests/port/builders.py:
422         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
423         (FactoryTest.test_mac):
424         (FactoryTest.test_win):
425         * Scripts/webkitpy/layout_tests/port/mac.py:
426         (MacPort.baseline_search_path):
427         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
428         (MacTest.test_versions):
429         * Scripts/webkitpy/layout_tests/port/win.py:
430         (WinPort.baseline_search_path):
431         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
432         (WinTest.test_versions):
433
434 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
435
436         webkitpy: print nicer errors while linting expectations files, remove redundant tests
437         https://bugs.webkit.org/show_bug.cgi?id=76955
438
439         Reviewed by Adam Barth.
440
441         This change cleans up the errors that are printed so that the
442         are more "quickfix" compatible (path:lineno).
443
444         Also, there were a bunch of redundant tests between
445         layout_tests.models.test_expectations_unittest and
446         style.checkers.test_expectations_unittest. The latter should
447         just expect the former to do most of the testing.
448
449         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
450         (TestExpectationParser._parse_modifiers):
451         (TestExpectations._report_errors):
452         (TestExpectations._add_skipped_tests):
453         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
454         (test_parse_error_nonfatal):
455         (SemanticTests.test_bad_bugid):
456         * Scripts/webkitpy/style/checkers/test_expectations.py:
457         (TestExpectationsChecker.__init__):
458         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
459         (TestExpectationsTestCase.assert_lines_lint):
460         (TestExpectationsTestCase.test_valid_expectations):
461         (TestExpectationsTestCase.test_invalid_expectations):
462         (TestExpectationsTestCase.test_tab):
463
464 2012-01-25  Eric Seidel  <eric@webkit.org>
465
466         webkit-patch apply-* should always continue after failures
467         https://bugs.webkit.org/show_bug.cgi?id=77057
468
469         Reviewed by Adam Barth.
470
471         As far as I can tell there is only one potential drawback to always
472         forcing: that if you're somehow in the wrong directory it will create new
473         directories for new files.  Since webkit-patch always cd's to the root
474         it seems that's not a drawback.  The drawback of not using --force for
475         svn-apply, is that it will stop after the first failure, which is rarely
476         (if ever) the desired behavior.  This just removes the force option
477         (which was strangely hidden behind --non-interactive).  This should
478         make for a better user experiance.
479
480         * Scripts/webkitpy/common/checkout/checkout.py:
481         (Checkout.apply_patch):
482         * Scripts/webkitpy/common/checkout/checkout_mock.py:
483         (MockCheckout.apply_patch):
484         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
485         (CheckoutTest.test_chromium_deps):
486         (CheckoutTest):
487         (CheckoutTest.test_apply_patch):
488         * Scripts/webkitpy/tool/commands/download_unittest.py:
489         (DownloadCommandsTest._default_options):
490         * Scripts/webkitpy/tool/steps/applypatch.py:
491         (ApplyPatch.options):
492         (ApplyPatch.run):
493         * Scripts/webkitpy/tool/steps/options.py:
494         (Options):
495
496 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
497
498         nrwt: should be able to run --platform test interactively
499         https://bugs.webkit.org/show_bug.cgi?id=76959
500
501         Reviewed by Adam Barth.
502
503         As part of refactoring the Port interfaces, in r103254 I made
504         passing a Host object to the Port mandatory; previously TestPort
505         objects would create their own (mock) hosts. However,
506         new-run-webkit-tests always passed a real Host, not a MockHost,
507         and so you couldn't run new-run-webkit-tests --platform test
508         interactively to debug test failures.
509
510         This change fixes that by creating a MockHost instead of a real
511         host if you say --platform test*.
512
513         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
514         (main):
515
516 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
517
518         run-webkit-tests --lint-test-files should lint all the ports by default
519         https://bugs.webkit.org/show_bug.cgi?id=76749
520
521         Reviewed by Ojan Vafai.
522
523         Currently run-webkit-tests --lint-test-files will only lint the
524         test_expectations for the single port determined by the command
525         line options. This is not obvious and can produce unintended results
526         if you want to lint the chromium port (probably the normal case
527         for using --lint-test-files) but don't specify a port name or
528         --chromium.
529
530         I think we should probably lint *all* of the ports if no port
531         name is supplied. This change implements that, and also moves
532         the linting logic out of the Manager class, since this really
533         has nothing to do with Managers. It has little to do with
534         run-webkit-tests, but splitting it into a different script would
535         introduce compatibility issues. It might make sense to do that
536         anyway, in a separate patch and combining that with cleaning up
537         the style checker to share more code.
538
539         * Scripts/webkitpy/layout_tests/controllers/manager.py:
540         (Manager._strip_test_dir_prefix):
541         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
542         (lint):
543         (run):
544         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
545         (LintTest.test_all_configurations): added.
546         (LintTest.test_lint_test_files): added.
547         (LintTest.test_lint_test_files__errors): added.
548         (MainTest.test_lint_test_files): removed.
549         (MainTest.test_lint_test_files__errors): removed.
550
551 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
552
553         webkitpy: clean up a bunch of test scaffolding
554         https://bugs.webkit.org/show_bug.cgi?id=76973
555
556         Reviewed by Eric Seidel.
557
558         There were several helper modules in the webkitpy.test package
559         that appear to only be used by executive_unittest.py. I am
560         rolling them all into that file to make that clearer.
561
562         * Scripts/webkitpy/common/system/executive_unittest.py:
563         (ExecutiveTest.test_run_command_args_type):
564         (ExecutiveTest.test_run_command_with_unicode):
565         (ExecutiveTest.test_running_pids):
566         (command_line):
567         (main):
568         * Scripts/webkitpy/common/system/fileutils.py: Removed.
569         * Scripts/webkitpy/test/cat.py: Removed.
570         * Scripts/webkitpy/test/cat_unittest.py: Removed.
571         * Scripts/webkitpy/test/echo.py: Removed.
572         * Scripts/webkitpy/test/echo_unittest.py: Removed.
573
574 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
575
576         webkitpy: clean up a bunch of test scaffolding
577         https://bugs.webkit.org/show_bug.cgi?id=76973
578
579         Reviewed by Eric Seidel.
580
581         There were several helper modules in the webkitpy.test package
582         that appear to only be used by executive_unittest.py. I am
583         rolling them all into that file to make that clearer.
584
585         * Scripts/webkitpy/common/system/executive_unittest.py:
586         (ExecutiveTest.test_run_command_args_type):
587         (ExecutiveTest.test_run_command_with_unicode):
588         (ExecutiveTest.test_running_pids):
589         (command_line):
590         (main):
591         * Scripts/webkitpy/common/system/fileutils.py: Removed.
592         * Scripts/webkitpy/test/cat.py: Removed.
593         * Scripts/webkitpy/test/cat_unittest.py: Removed.
594         * Scripts/webkitpy/test/echo.py: Removed.
595         * Scripts/webkitpy/test/echo_unittest.py: Removed.
596
597 2012-01-25  Eric Seidel  <eric@webkit.org>
598
599         Remove Python 2.5 support from WebKit
600         https://bugs.webkit.org/show_bug.cgi?id=71593
601
602         Reviewed by Adam Barth.
603
604         Just removing support from this one core file.
605         I'll commit the rest of the (large) change once we've
606         gone 24 hours and seen that no bots have exploded.
607         We can roll out this one change if we break bots
608         and then roll in the rest once those bots are fixed.
609
610         * Scripts/webkitpy/common/system/executive.py:
611         (Executive.cpu_count):
612
613 2012-01-25  Hajime Morita  <morrita@google.com>
614
615         ENABLE_SHADOW_DOM should be available via build-webkit --shadow-dom
616         https://bugs.webkit.org/show_bug.cgi?id=76863
617
618         Reviewed by Dimitri Glazkov.
619
620         Added a feature flag.
621
622         * Scripts/build-webkit:
623         * qmake/mkspecs/features/features.prf:
624
625 2012-01-25  Martin Robinson  <mrobinson@igalia.com>
626
627         [WK2] Navigation events not triggered by a mouse event claim to be triggered by the left mouse button during policy decisions
628         https://bugs.webkit.org/show_bug.cgi?id=76344
629
630         Reviewed by Sam Weinig.
631
632         Add a test verifying that the mouse button during policy decisions for
633         events not triggered by mouse events is NoButton.
634
635         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
636         (TestWebKitAPI::decidePolicyForNavigationAction): Added test.
637         (TestWebKitAPI::decidePolicyForNewWindowAction): Added test.
638
639 2012-01-25  Ryosuke Niwa  <rniwa@webkit.org>
640
641         run-perf-tests should timeout sooner
642         https://bugs.webkit.org/show_bug.cgi?id=77026
643
644         Reviewed by Adam Barth.
645
646         Bump off the timeout from 10 minutes to 4 minutes.
647
648         Also use float instead of str in test result values to compress the size of json files bots upload.
649
650         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
651         (PerfTestsRunner._parse_args):
652         (PerfTestsRunner._process_chromium_style_test_result):
653         (PerfTestsRunner._process_parser_test_result):
654         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
655         (test_run_test_set_for_parser_tests):
656         (test_run_test_set_with_json_output):
657         (test_run_test_set_with_json_source):
658
659 2012-01-25  Enrica Casucci  <enrica@apple.com>
660
661         WebView should implement typingAttributes methods to work correctly with the Inspector bar in Mail.
662         https://bugs.webkit.org/show_bug.cgi?id=76951
663         <rdar://problem/9325158>
664
665         Reviewed by Alexey Proskuryakov.
666
667         * TestWebKitAPI/Tests/mac/InspectorBar.mm: Modified to reflect the changes to WebView
668         and WebHTMLView.
669         (TestWebKitAPI::TEST):
670
671 2012-01-23  MORITA Hajime  <morrita@google.com>
672
673         [Refactoring][Internals] Should have InternalSettings
674         https://bugs.webkit.org/show_bug.cgi?id=76424
675
676         Reviewed by Kent Tamura.
677
678         * GNUmakefile.am: Added so new files.
679
680 2012-01-25  Takashi Toyoshima  <toyoshim@chromium.org>
681
682         webkitpy: new-run-webkit-websocket is broken
683         https://bugs.webkit.org/show_bug.cgi?id=76999
684
685         Reviewed by Dirk Pranke.
686
687         r105674 broke new-run-webkit-websocket, since options.chromium
688         doesn't exist any more. new-run-webkit-httpd was also broken
689         and already fixed at r105835.
690         * Scripts/new-run-webkit-websocketserver:
691         (main):
692
693 2012-01-25  Nándor Huszka  <huszka.nandor@stud.u-szeged.hu>
694
695         [WK2] WebKitTestRunner needs layoutTestController.setPopupBlockingEnabled
696         https://bugs.webkit.org/show_bug.cgi?id=63458
697
698         Reviewed by Adam Roben.
699
700         Implement the setPopupBlockingEnabled method.
701
702         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
703         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
704         (WTR::LayoutTestController::setPopupBlockingEnabled):
705         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
706
707 2012-01-24  Balazs Kelemen  <kbalazs@webkit.org>
708
709         [Qt] Cleanup testfonts handling
710         https://bugs.webkit.org/show_bug.cgi?id=76068
711
712         Reviewed by Simon Hausmann.
713
714         Factor test font initialization into it's own file  so test
715         font users can share the implementation by adding the file
716         to their build.
717         Use fonts.conf from the test font directory to not require the
718         current directory to be the WebKit root.
719         Fix the -use-test-fonts option of QtTestBrowser by initializing
720         the fonts earlier, before it is too late (namely before calling
721         show on the main window).
722
723         * DumpRenderTree/qt/DumpRenderTree.pro:
724         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
725         (WebCore::DumpRenderTree::open):
726         * DumpRenderTree/qt/DumpRenderTreeQt.h:
727         * DumpRenderTree/qt/QtInitializeTestFonts.cpp: Copied from Tools/WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp.
728         (WebKit::initializeTestFonts):
729         * DumpRenderTree/qt/QtInitializeTestFonts.h: Added.
730         * DumpRenderTree/qt/fonts.conf: Removed. No need for this since
731         we will use the one from the test fonts directory.
732         * DumpRenderTree/qt/main.cpp:
733         (main): Fix the ifdefs to make it clear that the code to
734         set up the Dpi has not been ported to Qt5 yet.
735         * QtTestBrowser/QtTestBrowser.pro:
736         * QtTestBrowser/launcherwindow.h:
737         (WindowOptions::WindowOptions):
738         * QtTestBrowser/main.cpp:
739         (launcherMain):
740         (LauncherApplication::handleUserOptions):
741         * WebKitTestRunner/InjectedBundle/Target.pri:
742         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
743         (WTR::activateFonts): Remove the code to set up the Dpi
744         because it does not work with Qt5.
745
746 2012-01-25  Simon Hausmann  <simon.hausmann@nokia.com>
747
748         [Qt] build-webkit --qt requires a git-svn or SVN checkout.
749
750         Reviewed by Tor Arne Vestø.
751
752         Don't require it by making svnRevisionForDirectory return "unknown"
753         instead of bailing out with an error.
754
755         * Scripts/VCSUtils.pm:
756         (svnRevisionForDirectory):
757
758 2012-01-25  Adam Barth  <abarth@webkit.org>
759
760         Temporarily disable running the unit tests on the Chromium EWS.  These
761         tests don't work properly (see
762         https://bugs.webkit.org/show_bug.cgi?id=76979) and are currently
763         blocking the EWS.
764
765         * Scripts/webkitpy/common/config/ports.py:
766         (ChromiumPort):
767
768 2012-01-25  Adam Barth  <abarth@webkit.org>
769
770         Update results.html version of PixelZoomer
771         https://bugs.webkit.org/show_bug.cgi?id=76965
772
773         Reviewed by Ojan Vafai.
774
775         Improve readability as suggested by Ojan.
776
777         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/pixelzoomer.js:
778
779 2012-01-25  Carlos Garcia Campos  <cgarcia@igalia.com>
780
781         [GTK] Page clients don't need to be GObjects anymore in WebKit2 GTK+ API
782         https://bugs.webkit.org/show_bug.cgi?id=76899
783
784         Reviewed by Martin Robinson.
785
786         * gtk/generate-gtkdoc:
787         (get_webkit2_options): WebKitWebLoaderClient has been renamed to
788         WebKitLoaderClient.
789
790 2012-01-24  Ryosuke Niwa  <rniwa@webkit.org>
791
792         Build fix. Trigger Chromium Mac Release (Perf) on new commits.
793
794         * BuildSlaveSupport/build.webkit.org-config/config.json:
795
796 2012-01-23  Ryosuke Niwa  <rniwa@webkit.org>
797
798         WebKit needs public performance tests running on public bots
799         https://bugs.webkit.org/show_bug.cgi?id=10266
800
801         Reviewed by Adam Barth.
802
803         Completes the initial implementation of WebKit Perf bots. Add a build step for run-perf-test
804         and make google-mac-4 run it, and upload the results to webkit-perf.webkit.org.
805
806         * BuildSlaveSupport/build.webkit.org-config/config.json:
807         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
808         (RunAndUploadPerfTests):
809         (RunAndUploadPerfTests.start):
810         (UploadPerfTestResults):
811         (BuildAndPerfTestFactory):
812         (BuildAndPerfTestFactory.__init__):
813
814 2012-01-24  Ojan Vafai  <ojan@chromium.org>
815
816         Initializing the browser property of the Bugzilla class takes too long
817         https://bugs.webkit.org/show_bug.cgi?id=76960
818
819         Reviewed by Adam Barth.
820
821         Lazily initialize it. This saves ~150ms on a no-op run-webkit-tests call
822         on my Mac Pro.
823
824         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
825         (Bugzilla.__init__):
826         (Bugzilla._get_browser):
827         (Bugzilla):
828         (Bugzilla._set_browser):
829
830 2012-01-24  Mark Rowe  <mrowe@apple.com>
831
832         Death to JavaScriptGlue.
833
834         Rubber-stamped by Geoffrey Garen.
835
836         * BuildSlaveSupport/build-launcher-app:
837         (copyNightlyLauncher):
838         * Scripts/build-webkit:
839         * Scripts/do-file-rename:
840         * Scripts/do-webcore-rename:
841         * Scripts/webkitpy/common/config/build.py:
842         (_should_file_trigger_build):
843         * Scripts/webkitpy/common/config/build_unittest.py: Update the test to use another Mac-specific
844         path in conjunction with a cross-platform path.
845         (ShouldBuildTest):
846
847 2012-01-24  Kentaro Hara  <haraken@chromium.org>
848
849         webkitperl/prepare-ChangeLog_unittest/parser_unittests.pl should test stdout and stderr
850         https://bugs.webkit.org/show_bug.cgi?id=75836
851
852         Reviewed by David Kilzer.
853
854         Previously parser_unittests.pl can just test the returned value of
855         get_function_line_ranges() in prepare-ChangeLog. This patch fixes
856         parser_unittests.pl so that it also tests the stdout and stderr of
857         get_function_line_ranges().
858
859         This patch adds css_unittests_warning.css, which contains test cases
860         for warning messages of the CSS parser of prepare-ChangeLog.
861
862         * Scripts/webkitperl/prepare-ChangeLog_unittest/parser_unittests.pl:
863         (captureOutput): Captures the stdout and stderr of a given method.
864         (convertAbsolutepathToWebKitPath): Converts absolute paths in a given
865         string to relative paths.
866
867         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/css_unittests_warning-expected.txt: Added.
868         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/css_unittests_warning.css: Added.
869         (element1):
870         (element2):
871         (element3):
872         (element4):
873         ():
874         (element5):
875         (element6):
876
877         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests-expected.txt:
878         Just updated the test results, i.e. added stdout and stderr results.
879         No essential changes in the test results.
880         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/css_unittests-expected.txt: Ditto.
881         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/java_unittests-expected.txt: Ditto.
882         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/javascript_unittests-expected.txt: Ditto.
883         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/perl_unittests-expected.txt: Ditto.
884         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/python_unittests-expected.txt: Ditto.
885
886 2012-01-24  Dirk Pranke  <dpranke@chromium.org>
887
888         webkitpy: new-run-webkit-httpd is broken
889         https://bugs.webkit.org/show_bug.cgi?id=76963
890
891         Reviewed by Adam Barth.
892
893         r105674 broke new-run-webkit-httpd, since options.chromium
894         doesn't exist any more. We didn't catch this because we don't
895         run the integration tests by default anywhere :(.
896
897         * Scripts/new-run-webkit-httpd:
898         (run):
899         (main):
900
901 2012-01-24  Adam Barth  <abarth@webkit.org>
902
903         garden-o-matic pixel zoomer should let you look at every pixel of the image
904         https://bugs.webkit.org/show_bug.cgi?id=76962
905
906         Reviewed by Ojan Vafai.
907
908         Previously, you couldn't get up to the right and bottom pixels.
909
910         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/pixelzoomer.js:
911
912 2012-01-24  Enrica Casucci  <enrica@apple.com>
913
914         WebHTMLView should implement typingAttributes methods to work correctly with the Inspector bar in Mail.
915         https://bugs.webkit.org/show_bug.cgi?id=76951
916         <rdar://problem/9325158>
917
918         Added test to verify that WebHTMLView supports the typingAttributes method and that it
919         works as expected.
920         
921         Reviewed by Alexey Proskuryakov.
922
923         * TestWebKitAPI/Tests/mac/InspectorBar.mm:
924         (TestWebKitAPI::TEST):
925
926 2012-01-24  James Robinson  <jamesr@chromium.org>
927
928         Put a space between the trac link and ellipsis in sheriffbot rollout message so it linkifies better in some IRC clients
929         https://bugs.webkit.org/show_bug.cgi?id=76884
930
931         Reviewed by Adam Barth.
932
933         * Scripts/webkitpy/tool/bot/irc_command.py:
934         (Rollout.execute):
935         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
936         (SheriffIRCBotTest.test_rollout):
937         (SheriffIRCBotTest.test_revert):
938         (SheriffIRCBotTest.test_multi_rollout):
939         (SheriffIRCBotTest.test_rollout_with_r_in_svn_revision):
940         (SheriffIRCBotTest.test_multi_rollout_with_r_in_svn_revision):
941         (SheriffIRCBotTest.test_rollout_invalidate_reason):
942         (test_multi_rollout_invalidate_reason):
943
944 2012-01-24  Kentaro Hara  <haraken@chromium.org>
945
946         The cpp parser of prepare-ChangeLog cannot detect a change in classes and namespaces
947         https://bugs.webkit.org/show_bug.cgi?id=75531
948
949         Reviewed by David Kilzer.
950
951         Previously, the cpp parser of prepare-ChangeLog could not detect a change
952         outside methods. Consider the following cpp file.
953
954             namespace N {
955             int a;     // this change does not appear on ChangeLog.
956             class C {
957                 int b;     // this change does not appear on ChangeLog.
958                 void f()
959                 {
960                     int c;     // this change appears on ChangeLog.
961                 }
962                 int d;     // this change does not appear on ChangeLog.
963             };
964             int e;     // this change does not appear on ChangeLog.
965             };
966
967         The previous prepare-ChangeLog outputs just methods in which a change is found:
968             (N::C::f):
969
970         This patch fixes prepare-ChangeLog so that it outputs namespaces, classes
971         and methods in which a change is found:
972             (N):
973             (N::C):
974             (N::C::f):
975
976         * Scripts/prepare-ChangeLog:
977         (get_function_line_ranges_for_cpp): Modified as described above.
978
979         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests.cpp: Added test cases.
980         (Class104):
981         (Class105):
982         (Class106):
983         (Class106::func32):
984         (Class106::func33):
985         (NameSpace3):
986         (NameSpace4):
987         (NameSpace5):
988         (NameSpace6):
989         (Class107):
990         (NameSpace5::NameSpace6::Class107::func34):
991         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests-expected.txt:
992
993 2012-01-24  Kentaro Hara  <haraken@chromium.org>
994
995         prepare-ChangeLog outputs warnings for top-level { ... } blocks
996         https://bugs.webkit.org/show_bug.cgi?id=75943
997
998         Reviewed by David Kilzer.
999
1000         Currently, prepare-ChangeLog outputs warnings for top-level { ... } blocks,
1001         because it does not distinguish '}' at the end of a subroutine
1002         with '}' at the end of a if or for statement.
1003
1004         foo.pl:
1005             if (1) {
1006             }
1007             for (@v) {
1008             }
1009
1010         output:
1011             nested functions found at top-level at foo.pl:2
1012             nested functions found at top-level at foo.pl:4
1013
1014         This patch fixes the bug.
1015
1016         * Scripts/prepare-ChangeLog:
1017         (get_function_line_ranges_for_perl):
1018         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/perl_unittests.pl:
1019         Added test cases. No change in the test results.
1020
1021 2012-01-24  Dirk Pranke  <dpranke@chromium.org>
1022
1023         r105674 broke check-webkit-style for chromium's test_expectations.
1024
1025         Unreviewed, build fix.
1026         
1027         We now need to pass in a platform name of 'chromium' if we want
1028         the chromium port; options.chromium is no longer recognized. I
1029         will follow this patch up with a separate patch that has a test,
1030         but this at least fixes the build.
1031
1032         * Scripts/webkitpy/style/checkers/test_expectations.py:
1033         (TestExpectationsChecker._determine_port_from_exepectations_path):
1034
1035 2012-01-20  Ojan Vafai  <ojan@chromium.org>
1036
1037         check-webkit-style of the chromium test_expectations.txt file takes too long
1038         https://bugs.webkit.org/show_bug.cgi?id=76745
1039
1040         Reviewed by Dimitri Glazkov.
1041
1042         When in lint mode, have TestExpectations test all configurations instead
1043         of looping over each configuration. This also has the benefit of making
1044         the error output considerably more concise.
1045
1046         Also, got rid of the double-printing of errors when linting through check-webkit-style.
1047
1048         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1049         (Manager.lint):
1050         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1051         (TestExpectations._report_errors):
1052         (TestExpectations._add_expectations):
1053         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1054         (test_parse_error_nonfatal):
1055         (test_error_on_different_platform):
1056         * Scripts/webkitpy/style/checkers/test_expectations.py:
1057         (TestExpectationsChecker.check_test_expectations):
1058         (TestExpectationsChecker.check):
1059         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
1060         (TestExpectationsTestCase.test_determine_port_from_exepectations_path):
1061         (TestExpectationsTestCase.assert_lines_lint):
1062
1063 2012-01-24  Ryosuke Niwa  <rniwa@webkit.org>
1064
1065         Another build fix attempt after r105543.
1066
1067         * Scripts/webkitpy/common/net/file_uploader.py:
1068         (FileUploader.upload_as_multipart_form_data):
1069
1070 2012-01-24  Ryosuke Niwa  <rniwa@webkit.org>
1071
1072         Build fix attempt after r105543.
1073
1074         * Scripts/webkitpy/common/net/file_uploader.py:
1075         (FileUploader.upload_as_multipart_form_data):
1076
1077 2012-01-24  Vincent Scheib  <scheib@chromium.org>
1078
1079         [Chromium] Add Pointer Lock test hooks and mock implementation to DumpRenderTree
1080         https://bugs.webkit.org/show_bug.cgi?id=76411
1081
1082         Reviewed by Adam Barth.
1083
1084         Breaking up https://bugs.webkit.org/show_bug.cgi?id=75762
1085         into a series of patches, this change adds a trivial
1086         pointer lock implementation to the Chromium port's DumpRenderTree
1087         and exposes LayoutTestController hooks to force failure conditions.
1088
1089         * DumpRenderTree/chromium/LayoutTestController.cpp:
1090         (LayoutTestController::LayoutTestController):
1091         (LayoutTestController::didLosePointerLock):
1092         (LayoutTestController::setPointerLockWillFailAsynchronously):
1093         (LayoutTestController::setPointerLockWillFailSynchronously):
1094         * DumpRenderTree/chromium/LayoutTestController.h:
1095         * DumpRenderTree/chromium/WebViewHost.cpp:
1096         (WebViewHost::requestPointerLock):
1097         (WebViewHost::requestPointerUnlock):
1098         (WebViewHost::isPointerLocked):
1099         (WebViewHost::didCompletePointerLock):
1100         (WebViewHost::didNotCompletePointerLock):
1101         (WebViewHost::didLosePointerLock):
1102         (WebViewHost::reset):
1103         * DumpRenderTree/chromium/WebViewHost.h:
1104         (WebViewHost::setPointerLockWillFailAsynchronously):
1105         (WebViewHost::setPointerLockWillFailSynchronously):
1106
1107 2012-01-24  Mario Sanchez Prada  <msanchez@igalia.com>
1108
1109         Unreviewed, skip WK2 API test timing out on some GTK bots.
1110
1111         WebKit2APITests/TestDownloads is consistently timing out on the
1112         32bit release and 64bit debug bots.
1113
1114         * Scripts/run-gtk-tests:
1115         (TestRunner): Skip WebKit2APITests/TestDownloads.
1116
1117 2012-01-24  Andras Becsi  <andras.becsi@nokia.com>
1118
1119         [Qt] Fix the build with the newes Qt5 hashes
1120         https://bugs.webkit.org/show_bug.cgi?id=76657
1121
1122         Reviewed by Simon Hausmann.
1123
1124         * MiniBrowser/qt/MiniBrowserApplication.cpp:
1125         (MiniBrowserApplication::MiniBrowserApplication): Disable the automatic touch->mouse event synthesis
1126         for now to prevent an infinite loop where Qt always creates mouse events out of our touch mock events.
1127
1128 2012-01-24  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1129
1130         [Qt] Distinguish between Cygwin and cmd.exe in the way we set env variables
1131
1132         We can't use win32-msvc* to decide whether or not to use (set FOO=bar), as
1133         building with MinGW inside cmd.exe will require the same construct. Instead
1134         we assume that when the first character in PWD is a '/' we're running under
1135         Cygwin, and use the normal unix way of setting environment variables.
1136
1137         Reviewed by Simon Hausmann.
1138
1139         * qmake/mkspecs/features/default_pre.prf:
1140
1141 2012-01-24  Simon Hausmann  <simon.hausmann@nokia.com>
1142
1143         Some qmake build files must include Qt5's "quick" module.
1144         https://bugs.webkit.org/show_bug.cgi?id=76671
1145
1146         Unreviewed build fix.
1147
1148         * MiniBrowser/qt/MiniBrowser.pro:
1149         * WebKitTestRunner/Target.pri:
1150
1151 2012-01-24  Andras Becsi  <andras.becsi@nokia.com>
1152
1153         [Qt] Fix the event timestamps for touch mocking
1154
1155         Reviewed by Simon Hausmann.
1156
1157         * MiniBrowser/qt/MiniBrowserApplication.cpp:
1158         (MiniBrowserApplication::sendTouchEvent): We should not set QEvent::None as the
1159         timestamp argument in the QWindowSystemInterface::handleTouchEvent call.
1160
1161 2012-01-24  Carlos Garcia Campos  <cgarcia@igalia.com>
1162
1163         [GTK] Implement DownloadClient in WebKit2 GTK+ API
1164         https://bugs.webkit.org/show_bug.cgi?id=72952
1165
1166         Reviewed by Martin Robinson.
1167
1168         * gtk/generate-gtkdoc:
1169         (get_webkit2_options): Ignore WebKitDownloadClient.
1170
1171 2012-01-23  Zan Dobersek  <zandobersek@gmail.com>
1172
1173         [GTK] editing/deleting/5408255.html results are incorrect
1174         https://bugs.webkit.org/show_bug.cgi?id=53644
1175
1176         Reviewed by Martin Robinson.
1177
1178         WEBKIT_TOP_LEVEL environment variable is now set directly in either
1179         WebKitTestRunner or DumpRenderTree through usage of a compilation-time
1180         macro. This way both tools can be run outside the test harness without
1181         the need to manually set the environment variable.
1182
1183         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1184         (getTopLevelPath):
1185         * GNUmakefile.am:
1186         * Scripts/webkitpy/layout_tests/port/gtk.py:
1187         (GtkPort.setup_environ_for_server):
1188         * WebKitTestRunner/GNUmakefile.am:
1189         * WebKitTestRunner/InjectedBundle/gtk/InjectedBundleGtk.cpp:
1190         (WTR::InjectedBundle::platformInitialize):
1191
1192 2012-01-23  Dmitry Lomov  <dslomov@google.com>
1193
1194         [Chromium] Implement layoutTestController.workerThreadCount in DRT
1195         https://bugs.webkit.org/show_bug.cgi?id=74653.
1196
1197         Reviewed by Darin Fisher.
1198
1199         * DumpRenderTree/chromium/LayoutTestController.cpp:
1200         (LayoutTestController::LayoutTestController):
1201         (LayoutTestController::workerThreadCount):
1202         * DumpRenderTree/chromium/LayoutTestController.h:
1203
1204 2012-01-23  Dirk Pranke  <dpranke@chromium.org>
1205
1206         nrwt: make --chromium work like --qt
1207         https://bugs.webkit.org/show_bug.cgi?id=76875
1208
1209         Reviewed by Adam Barth.
1210
1211         --chromium used to have to be handled differently from --qt
1212         due to the way the PortFactory was implemented; there's not
1213         really a good reason for that any more so this patch makes
1214         things slightly more consistent and eliminates the
1215         options.chromium flag (--chromium is now truly a synonym for
1216         --platform chromium).
1217
1218         * Scripts/webkitpy/layout_tests/port/factory.py:
1219         (PortFactory._default_port):
1220         (PortFactory.get):
1221         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
1222         (FactoryTest.setUp):
1223         (FactoryTest.test_chromium_mac):
1224         (FactoryTest.test_chromium_linux):
1225         (FactoryTest.test_chromium_win):
1226         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1227         (parse_args):
1228
1229 2012-01-23  Dirk Pranke  <dpranke@chromium.org>
1230
1231         run-webkit-tests needs to propagate --chromium
1232         https://bugs.webkit.org/show_bug.cgi?id=76870
1233
1234         Reviewed by Eric Seidel.
1235
1236         run-webkit-tests removes '--chromium' argument from @ARGV when
1237         determining which port to run, which means that that doesn't
1238         propagate to new-run-webkit-tests. That's bad (and is handled
1239         for the other ports by re-adding the flag, but apparently we're
1240         just now noticing for Chromium).
1241
1242         * Scripts/run-webkit-tests:
1243
1244 2012-01-23  Brian Weinstein  <bweinstein@apple.com>
1245
1246         More build fixing after r105646.
1247
1248         * TestWebKitAPI/Tests/WTF/RedBlackTree.cpp:
1249         (TestWebKitAPI::TestNode::key):
1250
1251 2012-01-23  Gavin Barraclough  <barraclough@apple.com>
1252
1253         Unreviewed build fix - r105646 broke this.
1254
1255         * TestWebKitAPI/Tests/WTF/RedBlackTree.cpp:
1256         (TestWebKitAPI::TestNode::TestNode):
1257         (TestWebKitAPI::TestNode::key):
1258         (TestWebKitAPI::RedBlackTreeTest::assertEqual):
1259         (TestWebKitAPI::RedBlackTreeTest::assertSameValuesForKey):
1260         (TestWebKitAPI::RedBlackTreeTest::testDriver):
1261
1262 2012-01-23  Sheriff Bot  <webkit.review.bot@gmail.com>
1263
1264         Unreviewed, rolling out r105658.
1265         http://trac.webkit.org/changeset/105658
1266         https://bugs.webkit.org/show_bug.cgi?id=76883
1267
1268         We want this eventually, but not right at this moment
1269         (Requested by abarth on #webkit).
1270
1271         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1272
1273 2012-01-23  Adam Barth  <abarth@webkit.org>
1274
1275         garden-o-matic should support Chromium Mac Lion
1276         https://bugs.webkit.org/show_bug.cgi?id=76880
1277
1278         Reviewed by Eric Seidel.
1279
1280         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1281
1282 2012-01-23  Emil A Eklund  <eae@chromium.org>
1283
1284         check-webkit-style whitespace/operators triggers on overloaded division operator
1285         https://bugs.webkit.org/show_bug.cgi?id=76650
1286
1287         Reviewed by Darin Adler.
1288
1289         * Scripts/webkitpy/style/checkers/cpp.py:
1290         (check_spacing):
1291         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1292         (CppStyleTest.test_operator_methods):
1293
1294 2012-01-23  Ryosuke Niwa  <rniwa@webkit.org>
1295
1296         run-perf-tests should report server-side errors
1297         https://bugs.webkit.org/show_bug.cgi?id=76802
1298
1299         Reviewed by Tony Chang.
1300
1301         Report errors whenever server's response doesn't match "OK".
1302
1303         * Scripts/webkitpy/common/net/file_uploader.py:
1304         (FileUploader.upload_single_text_file):
1305         (FileUploader.upload_as_multipart_form_data):
1306         (FileUploader._upload_data.callback):
1307         (FileUploader):
1308         (FileUploader._upload_data):
1309         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1310         (PerfTestsRunner._upload_json):
1311         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1312         (test_upload_json):
1313         (test_upload_json.MockFileUploader.upload_single_text_file):
1314
1315 2012-01-22  Ryosuke Niwa  <rniwa@webkit.org>
1316
1317         run-perf-tests ignore Skipped list on chromium
1318         https://bugs.webkit.org/show_bug.cgi?id=76764
1319
1320         Reviewed by Dirk Pranke.
1321
1322         Move skipped_perf_tests from WebKit port to Base port so that Chromium port
1323         can also find skipped list. Chromium port only uses test_expectations.txt for
1324         layout tests but performacne tests don't use test_expectations.txt so Chromium port
1325         also needs to use Skipped list.
1326
1327         * Scripts/webkitpy/layout_tests/port/base.py:
1328         (Port._tests_from_skipped_file_contents):
1329         (Port):
1330         (Port._expectations_from_skipped_files):
1331         (Port.skipped_perf_tests):
1332         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1333         (PortTest.test_skipped_perf_tests):
1334         (PortTest.test_skipped_perf_tests.add_text_file):
1335         * Scripts/webkitpy/layout_tests/port/webkit.py:
1336         (WebKitPort._skipped_tests_for_unsupported_features):
1337         (WebKitPort._skipped_file_search_paths):
1338         (WebKitPort.skipped_layout_tests):
1339
1340 2012-01-23  Andy Estes  <aestes@apple.com>
1341
1342         Fix the build after r105635.
1343
1344         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp:
1345         (TestWebKitAPI::TEST):
1346
1347 2012-01-23  Xianzhu Wang  <wangxianzhu@chromium.org>
1348
1349         Basic enhancements to StringBuilder
1350         https://bugs.webkit.org/show_bug.cgi?id=67081
1351
1352         Reviewed by Darin Adler.
1353
1354         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp:
1355         (TestWebKitAPI::TEST):
1356
1357 2012-01-23  Mario Sanchez Prada  <msanchez@igalia.com>
1358
1359         [GTK] run-gtk-tests randomly fails while running the xprop comand
1360         https://bugs.webkit.org/show_bug.cgi?id=76817
1361
1362         Reviewed by Gustavo Noronha Silva.
1363
1364         No need to use xprop to remove the AT_SPI_BUS property since
1365         run-gtk-tests will always run new instances of Xvfb.
1366
1367         * Scripts/run-gtk-tests:
1368         (TestRunner): Unskip TestWebKitAccessibility.
1369         (TestRunner.run): Uncomment lines for launching the accessibility
1370         bus and registry daemon, and remove lines for running xprop.
1371
1372 2012-01-23  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1373
1374         [Qt] Don't warn about override and final being C++11 extensions
1375
1376         Clang will emit a warning when these extensions are used without passing
1377         --std=c++11, but we use feature checking to decide if we have the right
1378         extensions, so we can safely ignore these warnings. The XCode and Windows
1379         project files have the same workaround.
1380
1381         The reason for adding the flag to QMAKE_OBJECTIVE_CFLAGS as well is that
1382         we only have one extra compiler for Objective-C, which is also used for
1383         Objective-C++ sources, so we have to pass the flag, even if it doesn't
1384         make sense for Objective-C.
1385
1386         Reviewed by Simon Hausmann.
1387
1388         * qmake/mkspecs/features/unix/default_post.prf:
1389
1390 2012-01-20  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1391
1392         [Qt] Change how build-webkit decides when to do full incremental builds
1393
1394         Instead of relying on update-webkit (which isn't run on the bots) to
1395         decide when to do a full incremental build (make qmake), we let the
1396         build-webkit script itself check the current SVN revision against the
1397         previous build (by storing it in .webkit.config).
1398
1399         If the two differ we assume a full incremental build is needed, since
1400         the new revisions might have introduced problematic things like new
1401         Q_OBJECT macros. If not, we assume the developer is doing changes
1402         locally, and revert to doing a plain 'make'.
1403
1404         In addition, when the build fails in the latter case, we inform the
1405         developer of possible pitfalls and how to manually run 'make qmake'.
1406
1407         Reviewed by Ossy.
1408
1409         * Scripts/update-webkit:
1410         * Scripts/webkitdirs.pm:
1411         (buildQMakeProjects):
1412
1413 2012-01-23  Mario Sanchez Prada  <msanchez@igalia.com>
1414
1415         [GTK] run-gtk-tests randomly fails while running the xprop comand
1416         https://bugs.webkit.org/show_bug.cgi?id=76817
1417
1418         Reviewed by Philippe Normand.
1419
1420         Temporarily comment the lines related to launching the ATSPI bus
1421         and registry daemon, which are making the bots to fail randomly.
1422
1423         * Scripts/run-gtk-tests:
1424         (TestRunner): Skip TestWebKitAccessibility, as it won't run
1425         properly if the ATSPI infrastructure is not properly initialized.
1426         (TestRunner.run): Comment lines related to initialization of ATSPI.
1427
1428 2012-01-22  Ryosuke Niwa  <rniwa@webkit.org>
1429
1430         Windows python test build fix.
1431
1432         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1433         (PerfTestsRunner.__init__):
1434
1435 2012-01-22  Sergio Villar Senin  <svillar@igalia.com>
1436
1437         [GTK] DumpRenderTree converts "file:///" to a path differently
1438         https://bugs.webkit.org/show_bug.cgi?id=76631
1439
1440         Reviewed by Martin Robinson.
1441
1442         DumpRenderTree should print "/" as the last path component if the
1443         path is a lone slash instead of empty output.
1444
1445         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1446         (webViewConsoleMessage):
1447
1448 2012-01-20  David Levin  <levin@chromium.org>
1449
1450         The leaky DC check should use adoptPtr as a signal instead of OwnPtr to get less false positives.
1451         https://bugs.webkit.org/show_bug.cgi?id=76752
1452
1453         Reviewed by Eric Seidel.
1454
1455         * Scripts/webkitpy/style/checkers/cpp.py:
1456         (check_for_leaky_patterns):
1457         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1458         (LeakyPatternTest.test_create_dc):
1459         (LeakyPatternTest.test_create_compatible_dc):
1460
1461 2012-01-20  Shawn Singh  <shawnsingh@chromium.org>
1462
1463         Changed my status in commiters.py
1464
1465         Unreviewed.
1466
1467         * Scripts/webkitpy/common/config/committers.py:
1468
1469 2012-01-20  Ojan Vafai  <ojan@chromium.org>
1470
1471         run-webkit-tests --lint-test-files crawls the whole LayoutTests subtree
1472         https://bugs.webkit.org/show_bug.cgi?id=76748
1473
1474         Reviewed by Ryosuke Niwa.
1475
1476         It crawls the whole subtree and then doesn't use the data. Cutting this out
1477         saves 4 seconds warm and 17 seconds cold on my Mac Pro.
1478
1479         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1480         (run):
1481
1482 2012-01-20  Adam Barth  <abarth@webkit.org>
1483
1484         Another tiny tweak to the garden-o-matic CSS.  This makes things line
1485         up slightly nicer when there aren't any failures.
1486
1487         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
1488         (#summary):
1489
1490 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
1491
1492         run-perf-tests should support --test-results-server option
1493         https://bugs.webkit.org/show_bug.cgi?id=76680
1494
1495         Reviewed by Adam Barth.
1496
1497         Add --test-results-server, --builder-name, and --build-number options to run-perf-tests
1498         to be used by perf bots. Also refactor file_uploader as needed.
1499
1500         * Scripts/webkitpy/common/net/file_uploader.py:
1501         (FileUploader.__init__):
1502         (FileUploader.upload_single_file):
1503         (FileUploader.upload_as_multipart_form_data):
1504         (FileUploader):
1505         (FileUploader._upload_data):
1506         (FileUploader._upload_data.callback):
1507         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
1508         (JSONResultsGeneratorBase.upload_json_files):
1509         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1510         (PerfTestsRunner):
1511         (PerfTestsRunner._parse_args):
1512         (PerfTestsRunner.run):
1513         (PerfTestsRunner._generate_json):
1514         (PerfTestsRunner._upload_json):
1515         (PerfTestsRunner._run_tests_set):
1516         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1517         (create_runner):
1518         (test_run_with_upload_json):
1519         (test_run_with_upload_json.mock_upload_json):
1520         (test_upload_json):
1521         (test_upload_json.MockFileUploader):
1522         (test_upload_json.MockFileUploader.__init__):
1523         (test_upload_json.MockFileUploader.upload_single_file):
1524         (test_parse_args):
1525
1526 2012-01-20  Adam Barth  <abarth@webkit.org>
1527
1528         The party time image overlaps real content!  This patch makes the
1529         image centered vertically.
1530
1531         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
1532         (#onebar.partytime #summary):
1533
1534 2012-01-19  Ojan Vafai  <ojan@chromium.org>
1535
1536         Refactor TestExpectationsParser in preparation for caching the results
1537         https://bugs.webkit.org/show_bug.cgi?id=76669
1538
1539         Reviewed by Dimitri Glazkov.
1540
1541         Make everything private expect for the parse method. 
1542         Eventually, we'll need the expectations lines to not be modified
1543         outside of TestExpectationsParser so we can cache the results.
1544         This makes check-webkit-style of the chromium test_expectations.txt file
1545         go from ~17 seconds to ~12 seconds on my Mac Pro.
1546
1547         This patch is just a refactor in preparation, so no new tests.
1548
1549         * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor_unittest.py:
1550         (TestExpectationEditorTests.make_parsed_expectation_lines):
1551         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1552         (TestExpectationParser.parse):
1553         (TestExpectationParser):
1554         (TestExpectationParser._parse_line):
1555         (TestExpectationParser._tokenize):
1556         (TestExpectationParser._tokenize_list):
1557         (TestExpectationsModel._clear_expectations_for_test):
1558         (TestExpectations.__init__):
1559         (TestExpectations._add_expectations):
1560         (TestExpectations._add_skipped_tests):
1561         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1562         (TestExpectationParserTests.test_tokenize_blank):
1563         (TestExpectationParserTests.test_tokenize_missing_colon):
1564         (TestExpectationParserTests.test_tokenize_extra_colon):
1565         (TestExpectationParserTests.test_tokenize_empty_comment):
1566         (TestExpectationParserTests.test_tokenize_comment):
1567         (TestExpectationParserTests.test_tokenize_missing_equal):
1568         (TestExpectationParserTests.test_tokenize_extra_equal):
1569         (TestExpectationParserTests.test_tokenize_valid):
1570         (TestExpectationParserTests.test_tokenize_valid_with_comment):
1571         (TestExpectationParserTests.test_tokenize_valid_with_multiple_modifiers):
1572         (TestExpectationParserTests.test_parse_empty_string):
1573         (TestExpectationSerializerTests.assert_round_trip):
1574         (TestExpectationSerializerTests.assert_list_round_trip):
1575         * Scripts/webkitpy/tool/commands/expectations.py:
1576         (OptimizeExpectations.execute):
1577         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1578         (GardeningExpectationsUpdater.update_expectations):
1579
1580 2012-01-20  Adam Barth  <abarth@webkit.org>
1581
1582         Follow-up to previous patch: don't produce NaN when the revision number
1583         is missing.
1584
1585         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
1586         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model_unittests.js:
1587
1588 2012-01-20  Adam Barth  <abarth@webkit.org>
1589
1590         Garden-o-matic should tell me which revisions have been checked by all the bots
1591         https://bugs.webkit.org/show_bug.cgi?id=76722
1592
1593         Reviewed by Dimitri Glazkov.
1594
1595         This information helps me know when it's safe for me to roll a
1596         candidate revision.
1597
1598         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
1599         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
1600         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model_unittests.js:
1601
1602 2012-01-20  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1603
1604         build-webkit: Don't spit out congratulations message on Ctrl+C
1605
1606         Reviewed by Andreas Kling.
1607
1608         * Scripts/build-webkit:
1609
1610 2012-01-20  Philippe Normand  <pnormand@igalia.com>
1611
1612         [GTK] Allow extra jhbuild modulesets
1613         https://bugs.webkit.org/show_bug.cgi?id=76691
1614
1615         Reviewed by Gustavo Noronha Silva.
1616
1617         * gtk/jhbuildrc: Extra modulesets and modules can be added in
1618         JHBuild using the WEBKIT_EXTRA_MODULESETS and WEBKIT_EXTRA_MODULES
1619         env variables respectively. Use comma-separated values. Example:
1620         WEBKIT_EXTRA_MODULES=file:///path/to/module.set,file:///other/path.
1621
1622 2012-01-20  Mario Sanchez Prada  <msanchez@igalia.com>
1623
1624         [GTK] Expose accessibility hierarchy in WebKit2 to ATK/AT-SPI based ATs
1625         https://bugs.webkit.org/show_bug.cgi?id=72589
1626
1627         Reviewed by Martin Robinson.
1628
1629         Ensure the accessibility infrastructure is available when running
1630         the unit tests, otherwise WK2's accessibility tests won't work.
1631
1632         * Scripts/run-gtk-tests:
1633         (TestRunner):
1634         (TestRunner._lookup_atspi2_binary): New. Looks for the directory
1635         where at-spi2's stuff is located (like at-spi-bus-launcher and
1636         at-spi2-registryd).
1637         (TestRunner.run): Initialize the accessibility infrastructure.
1638
1639         * gtk/generate-gtkdoc:
1640         (get_webkit2_options): Add 'WebKitWebViewBaseAccessible.*'.
1641
1642         * gtk/jhbuild.modules: Added at-spi2-core and at-spi2-atk.
1643
1644 2012-01-20  Hayato Ito  <hayato@chromium.org>
1645
1646         Add ShadowRoot.idl which is enabled by newly introduced SHADOW_DOM flag.
1647         https://bugs.webkit.org/show_bug.cgi?id=76353
1648
1649         Reviewed by Hajime Morita.
1650
1651         Add ShadowRoot.idl, which is enabled only on chromium port since this is
1652         under development feature.
1653         ShadowRoot.idl contains minimum API so that we can test it.
1654         Other APIs should be added on other changes so that we can isolate issues.
1655
1656         * DumpRenderTree/chromium/TestShell.cpp:
1657         (TestShell::TestShell):
1658
1659 2012-01-20  Mario Sanchez Prada  <msanchez@igalia.com>
1660
1661         [GTK] GTK's DRT not to log events for already defunct objects
1662         https://bugs.webkit.org/show_bug.cgi?id=76620
1663
1664         Reviewed by Martin Robinson.
1665
1666         Do not log 'state-change:defunct' events.
1667
1668         * DumpRenderTree/gtk/AccessibilityCallbacks.cpp:
1669         (printAccessibilityEvent): Early return if a
1670         'state-change:defunct' signal is passed.
1671         (axObjectEventListener): Pass the signal name and value in
1672         separate parameters to printAccessibilityEvent.
1673
1674 2012-01-19  Gustavo Noronha Silva  <gns@gnome.org>
1675
1676         Unreviewed build fix. Wrong variable being used as the cwd for git
1677         remote update command.
1678
1679         * gtk/run-with-jhbuild:
1680         (update_jhbuild):
1681
1682 2012-01-19  Gustavo Noronha Silva  <gns@gnome.org>
1683
1684         [GTK] ensure the jhbuild used by webkit is as up-to-date as needed
1685         https://bugs.webkit.org/show_bug.cgi?id=76585
1686
1687         Reviewed by Martin Robinson.
1688
1689         This allows us to specify a minimum version of jhbuild required,
1690         while also providing stability against changes done to the master
1691         branch, by letting us choose what is the exact version that will
1692         be used.
1693
1694         * gtk/run-with-jhbuild: refactored for more hackability and easier
1695         reading
1696         (jhbuild_installed): checks whether jhbuild is installed
1697         (jhbuild_cloned): checks whether jhbuild has been cloned
1698         (jhbuild_at_expected_revision): checks whether jhbuild is at the
1699         expected revision
1700         (update_jhbuild): brings jhbuild to the expected revision
1701         (clone_jhbuild): clones jhbuild
1702         (install_jhbuild): build and installs jhbuild
1703         (update_webkitgtk_libs): runs update-webkitgtk-libs
1704         (ensure_jhbuild): high-level logic to decide whether and which of
1705         the steps above are executed
1706
1707 2012-01-19  Alexey Proskuryakov  <ap@apple.com>
1708
1709         [WK2] DumpRenderTree converts "file:///" to a path differently
1710         https://bugs.webkit.org/show_bug.cgi?id=76653
1711
1712         Reviewed by John Sullivan.
1713
1714         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp: (WTR::lastFileURLPathComponent):
1715         Do print a slash if there are no path components.
1716
1717 2012-01-19  Eric Seidel  <eric@webkit.org>
1718
1719         webkit-patch cannot rollout patches without changelog/bug number
1720         https://bugs.webkit.org/show_bug.cgi?id=75962
1721
1722         Reviewed by Adam Barth.
1723
1724         * Scripts/webkitpy/tool/commands/download.py:
1725         (AbstractRolloutPrepCommand._prepare_state):
1726         * Scripts/webkitpy/tool/commands/download_unittest.py:
1727         (AbstractRolloutPrepCommandTest.test_prepare_state):
1728
1729 2012-01-19  Alexey Proskuryakov  <ap@apple.com>
1730
1731         [WK2] fast/url/degenerate-file-base.html fails
1732         https://bugs.webkit.org/show_bug.cgi?id=76619
1733
1734         Reviewed by Adam Roben.
1735
1736         A URL generated form local path for loading had an extra slash, making document.URL
1737         incorrect in all WebKit2 tests (file://localhost//path/to/test.html).
1738
1739         * WebKitTestRunner/TestInvocation.cpp: (WTR::createWKURL):
1740
1741 2012-01-18  Ojan Vafai  <ojan@chromium.org>
1742
1743         check-webkit-style of the chromium test_expectations.txt file doesn't test all chromium ports
1744         https://bugs.webkit.org/show_bug.cgi?id=76510
1745
1746         Reviewed by Adam Barth.
1747
1748         Unlike other style rules, if you get the syntax of the test_expectations.txt
1749         file wrong, the layout tests won't run. Also, this check is simple and only
1750         slows down committing if you actually modify one of the test_expectations.txt files.
1751
1752         * Scripts/webkitpy/style/checkers/test_expectations.py:
1753         (TestExpectationsChecker.check_test_expectations):
1754         (TestExpectationsChecker.check):
1755         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
1756         (TestExpectationsTestCase.test_check_covers_all_configurations):
1757         (TestExpectationsTestCase.test_check_covers_all_configurations.mock_check_test_expectations):
1758         (TestExpectationsTestCase):
1759         (assert_lines_lint):
1760
1761 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
1762
1763         Build fix. Revert an unintentional change to WebKitDriver in r105443.
1764
1765         * Scripts/webkitpy/layout_tests/port/driver.py:
1766         (DriverProxy.__init__):
1767
1768 2012-01-19  Dirk Pranke  <dpranke@chromium.org>
1769
1770         remove the duplicated build rules from WebKit.gyp, Tools.gyp
1771         https://bugs.webkit.org/show_bug.cgi?id=73384
1772
1773         Reviewed by Tony Chang.
1774
1775         This patch removes the no-longer-necessary conditional logic for
1776         build_webkit_exes_from_webkit_gyp; we now always build the exes
1777         from their own dedicated gyp files.
1778
1779         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1780
1781 2012-01-19  Alexey Proskuryakov  <ap@apple.com>
1782
1783         make-new-script-test should work with HTTP tests
1784         https://bugs.webkit.org/show_bug.cgi?id=76603
1785
1786         Reviewed by Adam Roben.
1787
1788         * Scripts/make-new-script-test:
1789         (makePathToSharedSources): Renamed from makeTestRelativePathToSharedSources(), since it
1790         now also creates absolute paths for HTTP.
1791
1792 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
1793
1794         Some perf tests time out when ran by run-perf-tests
1795         https://bugs.webkit.org/show_bug.cgi?id=76612
1796
1797         Reviewed by Dirk Pranke and Eric Seidel.
1798
1799         Always pass --no-timeout to DumpRenderTree from run-perf-tests.
1800         Otherwise some tests such as Parser/xml-parser.html will timeout.
1801
1802         --no-timeout option is currently supported by Chromium and Mac ports.
1803
1804         * Scripts/webkitpy/layout_tests/port/base.py:
1805         (Port.to.create_driver):
1806         * Scripts/webkitpy/layout_tests/port/chromium.py:
1807         (ChromiumDriver.__init__):
1808         (ChromiumDriver._wrapper_options):
1809         * Scripts/webkitpy/layout_tests/port/driver.py:
1810         (Driver.__init__):
1811         (DriverProxy.__init__):
1812         * Scripts/webkitpy/layout_tests/port/webkit.py:
1813         (WebKitDriver.__init__):
1814         (WebKitDriver.cmd_line):
1815         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1816         (WebKitDriverTest.test_read_binary_block):
1817         (WebKitDriverTest):
1818         (WebKitDriverTest.test_no_timeout):
1819         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1820         (get_tests_run.RecordingTestDriver.__init__):
1821         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1822         (PerfTestsRunner._parse_args):
1823         (PerfTestsRunner._run_tests_set):
1824         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1825         (create_runner):
1826
1827 2012-01-19  Ben Wells  <benwells@chromium.org>
1828
1829         Add support for window.print to chromium DRT
1830         https://bugs.webkit.org/show_bug.cgi?id=76479
1831
1832         When used in a test window.print goes into print mode and then
1833         straight out again. This will allow the afterprint event to be
1834         tested by having the afterprint event happen before the dump
1835         happens.
1836
1837         Reviewed by Mihai Parparita.
1838
1839         * DumpRenderTree/chromium/WebViewHost.cpp:
1840         (WebViewHost::printPage):
1841         * DumpRenderTree/chromium/WebViewHost.h:
1842
1843 2012-01-18  Enrica Casucci  <enrica@apple.com>
1844
1845         editingAttributedStringFromRange in WebHTMLConverter does not handle NSUnderlineStyleAttributeName.
1846         https://bugs.webkit.org/show_bug.cgi?id=76588
1847         <rdar://problem/9325183>
1848
1849         Added test.
1850         
1851         Reviewed by Dan Bernstein.
1852
1853         * TestWebKitAPI/Tests/mac/InspectorBar.mm:
1854         (TestWebKitAPI::TEST):
1855
1856 2012-01-19  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1857
1858         [Qt] Do full incremental builds until bots use update-webkit
1859
1860         Rubber-stamped by Simon Hausmann.
1861
1862         * Scripts/webkitdirs.pm:
1863         (buildQMakeProjects):
1864
1865 2012-01-19  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1866
1867         Fix the Qt build
1868
1869         Unreviewed build fix.
1870
1871         * Scripts/webkitdirs.pm:
1872         (buildQMakeProjects):
1873
1874 2012-01-19  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1875
1876         [Qt] Make build-webkit always do safe incremental builds after update-webkit
1877
1878         When building against Qt5 with GCC we would just run 'make' in the build dir,
1879         and rely on the dependency files output by GCC (-MD) to handle dependency
1880         tracking, but that fails for special-cases like adding a Q_OBJECT macro
1881         to a header.
1882
1883         To guarantee that an incrmental build will work, we have to run 'make qmake',
1884         which we now do on every build-webkit that's followed by a successful run
1885         of update-webkit. The reasoning is that update-webkit can result in such
1886         potential corner-cases being applied, and since we can't know for sure
1887         unless we inspect the diff and account for all the corner cases we assume
1888         the worst and always run 'make qmake'.
1889
1890         After a succesful run of build-webkit we proceed to do just 'make' for any
1891         subsequent runs, since we assume that the developer knows what kind of
1892         changes he/she is doing, and when a 'make qmake' is needed.
1893
1894         Reviewed by Simon Hausmann.
1895
1896         * Scripts/update-webkit:
1897         * Scripts/webkitdirs.pm:
1898         (determineCurrentSVNRevision):
1899         (currentSVNRevision):
1900         (buildQMakeProjects):
1901         * qmake/mkspecs/features/default_post.prf:
1902
1903 2012-01-18  Ryosuke Niwa  <rniwa@webkit.org>
1904
1905         run-perf-tests should support Skipped list
1906         https://bugs.webkit.org/show_bug.cgi?id=76594
1907
1908         Reviewed by Adam Barth.
1909
1910         Add a support for Skipped list in run-perf-tests; also skip files in resources directories.
1911
1912         * Scripts/webkitpy/layout_tests/port/base.py:
1913         (Port.perf_tests_dir):
1914         (Port.skipped_perf_tests):
1915         (Port.skips_perf_test):
1916         (Port.relative_perf_test_filename):
1917         * Scripts/webkitpy/layout_tests/port/test.py:
1918         (TestPort.perf_tests_dir):
1919         * Scripts/webkitpy/layout_tests/port/webkit.py:
1920         (WebKitPort._expectations_from_skipped_files):
1921         (WebKitPort):
1922         (WebKitPort.skipped_layout_tests):
1923         (WebKitPort.skipped_perf_tests):
1924         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1925         (PerfTestsRunner._collect_tests):
1926         (PerfTestsRunner.run):
1927         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1928         (test_run_test_set_with_json_output):
1929         (test_run_test_set_with_json_source):
1930         (test_collect_tests_with_skipped_list):
1931         (test_collect_tests_with_skipped_list.add_file):
1932
1933 2012-01-18  Dirk Pranke  <dpranke@chromium.org>
1934
1935         Fix path to chromium_src_dir introduced in previous change.
1936
1937         Unreviewed, build fix.
1938
1939         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1940
1941 2012-01-18  Alexey Proskuryakov  <ap@apple.com>
1942
1943         Need infrastructure to test Content-Disposition filename encoding support
1944         https://bugs.webkit.org/show_bug.cgi?id=76572
1945         <rdar://problem/6421825>
1946
1947         Reviewed by Sam Weinig.
1948
1949         Made policy delegate log suggested file name when resource has Content-Disposition: attachment.
1950
1951         WebKitTestRunner does not have a policy delegate yet, bug 42546.
1952
1953         * DumpRenderTree/mac/PolicyDelegate.mm:
1954         (dispositionTypeFromContentDispositionHeader):
1955         (-[PolicyDelegate webView:decidePolicyForMIMEType:request:frame:decisionListener:]):
1956
1957 2012-01-18  Ojan Vafai  <ojan@chromium.org>
1958
1959         Fix tests from http://trac.webkit.org/changeset/105332.
1960
1961         The code assumed that if you passed an options object with a configuration
1962         attribute that it would have a valid value. A bunch of the testing mocks
1963         do things like having a configuration value of None.
1964
1965         * Scripts/webkitpy/layout_tests/port/base.py:
1966         (Port.__init__):
1967
1968 2012-01-18  Ojan Vafai  <ojan@chromium.org>
1969
1970         Shave 0.5 seconds off check-webkit-style runtime for test_expectations.txt
1971         https://bugs.webkit.org/show_bug.cgi?id=76576
1972
1973         Reviewed by Adam Barth.
1974
1975         Avoid calling default_configuration in base.py. We don't need it for
1976         checking test_expectations.txt style. It takes ~0.5 seconds to run on my
1977         Mac Pro. It's the call to "perl Tools/Scripts/webkit-build-directory --top-level"
1978         from common.executive.
1979
1980         At some point someone should probably look into why that call is so slow
1981         since it's on the critical path for run-webkit-tests and build-webkit startup.
1982
1983         * Scripts/webkitpy/layout_tests/port/base.py:
1984         (Port.__init__):
1985         * Scripts/webkitpy/style/checkers/test_expectations.py:
1986         (TestExpectationsChecker._determine_port_from_exepectations_path):
1987
1988 2012-01-18  Dirk Pranke  <dpranke@chromium.org>
1989
1990         [chromium] move Tools.gyp, switch build-webkit --chromium to All.gyp
1991         https://bugs.webkit.org/show_bug.cgi?id=76505
1992
1993         Reviewed by Tony Chang.
1994
1995         Update chromium build script to use All.gyp instead of WebKit.gyp
1996
1997         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp: Renamed from Source/WebKit/chromium/Tools.gyp.
1998         * Scripts/webkitdirs.pm:
1999
2000 2012-01-18  Sam Weinig  <sam@webkit.org>
2001
2002         Don’t hardcode the path to ‘make‘.
2003
2004         Reviewed by Dan Bernstein.
2005
2006         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
2007
2008 2012-01-18  Joshua Bell  <jsbell@chromium.org>
2009
2010         Adding myself to committers.py
2011         https://bugs.webkit.org/show_bug.cgi?id=76569
2012
2013         Unreviewed.
2014
2015         * Scripts/webkitpy/common/config/committers.py:
2016
2017 2012-01-18  Adam Barth  <abarth@webkit.org>
2018
2019         Up the cap on test failures tolerated by the EWS so the mac-ews can run to completion
2020         https://bugs.webkit.org/show_bug.cgi?id=76567
2021
2022         Reviewed by Eric Seidel.
2023
2024         Current the AppleMac port has slightly more failures than we allow on
2025         the EWS, causing the mac-ews to spin.  This patch ups the limit a bit,
2026         which should hopefully let the mac-ews run to completion.  Once
2027         AppleMac has fewer failures, we can lower the cap again.
2028
2029         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
2030         (LayoutTestResultsReaderTest.test_layout_test_results):
2031         * Scripts/webkitpy/tool/steps/runtests.py:
2032         (RunTests):
2033         * Scripts/webkitpy/tool/steps/steps_unittest.py:
2034
2035 2012-01-18  Eric Carlson  <eric.carlson@apple.com>
2036
2037         Provide access to user's list of preferred languages
2038         https://bugs.webkit.org/show_bug.cgi?id=76138
2039
2040         Reviewed by Alexey Proskuryakov.
2041
2042         * GNUmakefile.am: Add JSInternalsCustom.cpp.
2043
2044 2012-01-18  Adam Barth  <abarth@webkit.org>
2045
2046         Add Chromium Linux GPU 32 to the list of allowed Chromium Linux
2047         configurations.
2048
2049         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2050         (ChromiumLinuxPort.__init__):
2051
2052 2012-01-18  Dominic Mazzoni  <dmazzoni@google.com>
2053
2054         Accessibility: Chromium needs methods to scroll an object into view or to a specific location.
2055         https://bugs.webkit.org/show_bug.cgi?id=73460
2056
2057         Reviewed by Chris Fleizach.
2058
2059         * DumpRenderTree/AccessibilityUIElement.h:
2060         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
2061         (AccessibilityUIElement::AccessibilityUIElement):
2062         (AccessibilityUIElement::scrollToMakeVisibleCallback):
2063         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocusCallback):
2064         (AccessibilityUIElement::scrollToGlobalPointCallback):
2065         * DumpRenderTree/chromium/AccessibilityUIElement.h:
2066         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2067         (AccessibilityUIElement::scrollToMakeVisible):
2068         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2069         (AccessibilityUIElement::scrollToGlobalPoint):
2070         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2071         (AccessibilityUIElement::scrollToMakeVisible):
2072         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2073         (AccessibilityUIElement::scrollToGlobalPoint):
2074         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2075         (AccessibilityUIElement::scrollToMakeVisible):
2076         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2077         (AccessibilityUIElement::scrollToGlobalPoint):
2078
2079 2012-01-17  Enrica Casucci  <enrica@apple.com>
2080
2081         Missing NSForegroundColorAttributeName should be treated as black instead of transparent.
2082         https://bugs.webkit.org/show_bug.cgi?id=76490
2083         <rdar://problem/9460733>
2084
2085         Added TestWebKitAPI test.
2086         
2087         Reviewed by Dan Bernstein.
2088
2089         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added new test file.
2090         * TestWebKitAPI/Tests/mac/InspectorBar.mm: Added test that calls
2091         changeAttributes to the WebView.
2092         
2093 2012-01-18  Balazs Kelemen  <kbalazs@webkit.org>
2094
2095         [Qt] Consolidate layout test crash logging
2096         https://bugs.webkit.org/show_bug.cgi?id=75088
2097
2098         Reviewed by Simon Hausmann.
2099
2100         Move backtrace generating logic into WTFReportBacktrace
2101         and add a way to deinstall signal handlers if we know
2102         that we have already printed the backtrace.
2103
2104         * DumpRenderTree/qt/main.cpp:
2105         (crashHandler):
2106         (setupSignalHandlers):
2107         (WTFCrashHook):
2108         (main):
2109         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
2110         (WTR::crashHandler):
2111         (WTR::setupSignalHandlers):
2112         (WTR::crashHook):
2113         (WTR::InjectedBundle::platformInitialize):
2114
2115 2012-01-18  Ilya Tikhonovsky  <loislo@chromium.org>
2116
2117         Unreviewed build fix after r105256.
2118
2119         There was a cyclic dependency between self._port and self._host assigments.
2120
2121         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2122         (PerfTestsRunner.__init__):
2123
2124 2012-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2125
2126         Add new watchlist rule for the Qt build system
2127
2128         Reviewed by Simon Hausmann.
2129
2130         * Scripts/webkitpy/common/config/watchlist:
2131
2132 2012-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2133
2134         [Qt] Move OTHER_FILES from WebKit.pro to Tools.pro
2135
2136         Reviewed by Simon Hausmann.
2137
2138         * Tools.pro:
2139
2140 2012-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2141
2142         [Qt] Allow building only a subset of the WebKit sub-projects
2143
2144         Fixes build-jsc, broken in r104825.
2145
2146         We no longer need the intermediate buildQMakeQtProject(), as we're the
2147         only port using qmake, and buildQMakeProject is now highly Qt specific.
2148
2149         <http://webkit.org/b/76179>
2150
2151         Reviewed by Simon Hausmann.
2152
2153         * Scripts/build-jsc:
2154         * Scripts/build-webkit:
2155         * Scripts/webkitdirs.pm:
2156         (buildQMakeProjects):
2157
2158 2012-01-18  Ryosuke Niwa  <rniwa@webkit.org>
2159
2160         run-perf-tests should generate a json file that summaries the result
2161         https://bugs.webkit.org/show_bug.cgi?id=76504
2162
2163         Reviewed by Adam Barth.
2164
2165         Add an ability to generate a json file to run-perf-test in the preparation for perf bots.
2166         New option --outout-json-path specifies the json file's path, and --source-json-path specifies
2167         another json file to be merged into the generated json file.
2168
2169         Also fixed a bug that --build wasn't set by default.
2170
2171         * Scripts/webkitpy/layout_tests/port/test.py:
2172         (TestPort.webkit_base):
2173         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2174         (PerfTestsRunner.__init__):
2175         (PerfTestsRunner):
2176         (PerfTestsRunner._parse_args):
2177         (PerfTestsRunner.run):
2178         (PerfTestsRunner._generate_json_if_specified):
2179         (PerfTestsRunner._process_chromium_style_test_result):
2180         (PerfTestsRunner._process_parser_test_result):
2181         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2182         (MainTest):
2183         (create_runner):
2184         (test_run_test_set):
2185         (test_run_test_set_for_parser_tests):
2186         (test_run_test_set_with_summary_json):
2187         (test_collect_tests):
2188         (test_parse_args):
2189
2190 2012-01-17  Sheriff Bot  <webkit.review.bot@gmail.com>
2191
2192         Unreviewed, rolling out r105244.
2193         http://trac.webkit.org/changeset/105244
2194         https://bugs.webkit.org/show_bug.cgi?id=76518
2195
2196         broke Chromium Mac (Requested by rolandsteiner on #webkit).
2197
2198         * DumpRenderTree/AccessibilityUIElement.h:
2199         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
2200         (AccessibilityUIElement::AccessibilityUIElement):
2201         * DumpRenderTree/chromium/AccessibilityUIElement.h:
2202         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2203         (AccessibilityUIElement::removeSelection):
2204         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2205         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2206         (AccessibilityUIElement::removeSelection):
2207
2208 2012-01-17  Dominic Mazzoni  <dmazzoni@google.com>
2209
2210         Accessibility: Chromium needs methods to scroll an object into view or to a specific location.
2211         https://bugs.webkit.org/show_bug.cgi?id=73460
2212
2213         Reviewed by Chris Fleizach.
2214
2215         * DumpRenderTree/AccessibilityUIElement.h:
2216         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
2217         (AccessibilityUIElement::AccessibilityUIElement):
2218         (AccessibilityUIElement::scrollToMakeVisibleCallback):
2219         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocusCallback):
2220         (AccessibilityUIElement::scrollToGlobalPointCallback):
2221         * DumpRenderTree/chromium/AccessibilityUIElement.h:
2222         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2223         (AccessibilityUIElement::scrollToMakeVisible):
2224         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2225         (AccessibilityUIElement::scrollToGlobalPoint):
2226         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2227         (AccessibilityUIElement::scrollToMakeVisible):
2228         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2229         (AccessibilityUIElement::scrollToGlobalPoint):
2230         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2231         (AccessibilityUIElement::scrollToMakeVisible):
2232         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2233         (AccessibilityUIElement::scrollToGlobalPoint):
2234
2235 2012-01-17  Adam Barth  <abarth@webkit.org>
2236
2237         Enable tests on the mac-ews
2238         https://bugs.webkit.org/show_bug.cgi?id=76502
2239
2240         Reviewed by Simon Fraser.
2241
2242         Apple now has five machines running the mac-ews.  Lucas Forschler has
2243         asked that we try enabling tests to see if we have enough bandwidth.
2244
2245         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
2246         (MacEWS):
2247
2248 2012-01-17  Sanjoy Das  <sanjoy@playingwithpointers.com>
2249
2250         Add a check-webkit-style exception for GDBInterface.cpp
2251         https://bugs.webkit.org/show_bug.cgi?id=76187
2252
2253         Add an exception for Source/JavaScriptCore/jit/GDBInterface.cpp so
2254         that it does not fail the style check.
2255
2256         Reviewed by Martin Robinson.
2257
2258         * Scripts/webkitpy/style/checker.py: Tell the style checker to not run
2259         readability/naming tests on
2260         Source/JavaScriptCore/jit/GDBInterface.cpp.
2261         * Scripts/webkitpy/style/checker_unittest.py:
2262         (GlobalVariablesTest.test_path_rules_specifier): Add a test for
2263         the above exception.
2264
2265 2012-01-17  Hajime Morrita  <morrita@chromium.org>
2266
2267         https://bugs.webkit.org/show_bug.cgi?id=76245
2268         [Mac][DRT] should allow query string for the file URL.
2269
2270         Reviewed by Tony Chang.
2271
2272         Taught DRT about file:// as a url instead of a file path.
2273         We are now able to use a path file:///foo/bar.html?baz as a url
2274         with which DRT opens "/foo/bar.html" with "baz" as a query string.
2275
2276         * DumpRenderTree/mac/DumpRenderTree.mm:
2277         (runTest):
2278
2279 2012-01-17  Ojan Vafai  <ojan@chromium.org>
2280
2281         Check style on test_expectations.txt files before commit
2282         https://bugs.webkit.org/show_bug.cgi?id=76484
2283
2284         Reviewed by Adam Barth.
2285
2286         Unlike other style rules, if you get the syntax of the test_expectations.txt
2287         file wrong, the layout tests won't run. Also, this check is simple and only
2288         slows down committing if you actually modify one of the test_expectations.txt files.
2289
2290         * Scripts/webkitpy/tool/steps/commit.py:
2291         (Commit._check_test_expectations):
2292         (Commit.run):
2293         * Scripts/webkitpy/tool/steps/commit_unittest.py: Added.
2294         (CommitTest):
2295         (CommitTest.test_check_test_expectations):
2296
2297 2012-01-17  Viatcheslav Ostapenko  <ostapenko.viatcheslav@nokia.com>
2298
2299         [Qt] Debug build fails with debug qt5
2300         https://bugs.webkit.org/show_bug.cgi?id=76463
2301
2302         Reviewed by Simon Hausmann.
2303
2304         Add QNetworkCookie include in order to satisfy sizeof in QTypeInfo in debug builds.
2305
2306         * QtTestBrowser/cookiejar.h:
2307
2308 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2309
2310         webkitpy: fix json import on linux
2311         https://bugs.webkit.org/show_bug.cgi?id=76481
2312
2313         Reviewed by Adam Barth.
2314
2315         For some reason, the version of simplejson autoinstalled on my linux
2316         box seems to be broken. We should really only be using
2317         simplejson if json isn't available anyway, so this change looks
2318         for json first.
2319
2320         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2321         (JSONGeneratorTest.test_test_timings_trie):
2322
2323 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2324
2325         Fix failures in test-webkitpy caused by r105177.
2326
2327         Unreviewed, build fix.
2328
2329         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2330         (ChromiumLinuxPort._determine_driver_path_statically):
2331         (ChromiumLinuxPort._static_build_path):
2332
2333 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2334
2335         webkitpy: make PortFactory.get() be fully data-driven
2336         https://bugs.webkit.org/show_bug.cgi?id=76360
2337
2338         Reviewed by Eric Seidel.
2339
2340         Refactor PortFactory.get() to be fully data-driven.
2341
2342         * Scripts/webkitpy/layout_tests/port/factory.py:
2343         (PortFactory):
2344         (PortFactory._default_port):
2345         (PortFactory.get):
2346         (PortFactory.get.in):
2347         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
2348         (FactoryTest.assert_port):
2349
2350 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2351
2352         webkitpy: add determine_full_port_name(), clean up port.__init__()
2353         https://bugs.webkit.org/show_bug.cgi?id=76357
2354
2355         Reviewed by Adam Barth.
2356
2357         Each Port class is now required to implement a 'port_name' field
2358         that will match the start of all ports constructed by this class and
2359         a 'determine_full_port_name' field that will make the port
2360         fully-qualified.
2361
2362         * Scripts/webkitpy/layout_tests/port/apple.py:
2363         (ApplePort.__init__):
2364         * Scripts/webkitpy/layout_tests/port/base.py:
2365         (Port):
2366         (Port.determine_port_name):
2367         (Port.__init__):
2368         * Scripts/webkitpy/layout_tests/port/chromium.py:
2369         (ChromiumPort.for):
2370         (ChromiumPort.__init__):
2371         (ChromiumPort):
2372         (ChromiumPort._chromium_base_dir):
2373         (ChromiumPort.path_from_chromium_base):
2374         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2375         (ChromiumGpuLinuxPort):
2376         (ChromiumGpuLinuxPort.__init__):
2377         (ChromiumGpuMacPort):
2378         (ChromiumGpuMacPort.__init__):
2379         (ChromiumGpuWinPort):
2380         (ChromiumGpuWinPort.__init__):
2381         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2382         (ChromiumLinuxPort):
2383         (ChromiumLinuxPort.determine_port_name):
2384         (ChromiumLinuxPort.__init__):
2385         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2386         (ChromiumMacPort):
2387         (ChromiumMacPort.determine_port_name):
2388         (ChromiumMacPort.__init__):
2389         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2390         (ChromiumWinPort):
2391         (ChromiumWinPort.determine_port_name):
2392         (ChromiumWinPort.__init__):
2393         * Scripts/webkitpy/layout_tests/port/dryrun.py:
2394         (DryRunPort):
2395         (DryRunPort.determine_port_name):
2396         (DryRunPort.__init__):
2397         * Scripts/webkitpy/layout_tests/port/efl.py:
2398         (EflPort):
2399         * Scripts/webkitpy/layout_tests/port/factory.py:
2400         (PortFactory):
2401         (PortFactory._default_port):
2402         (PortFactory.get):
2403         (PortFactory.get.in):
2404         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
2405         (FactoryTest.assert_port):
2406         * Scripts/webkitpy/layout_tests/port/google_chrome.py:
2407         (GoogleChromeLinux32Port):
2408         (GoogleChromeLinux32Port.determine_port_name):
2409         (GoogleChromeLinux64Port):
2410         (GoogleChromeLinux64Port.determine_port_name):
2411         (GoogleChromeMacPort):
2412         (GoogleChromeMacPort.determine_port_name):
2413         (GoogleChromeWinPort):
2414         (GoogleChromeWinPort.determine_port_name):
2415         * Scripts/webkitpy/layout_tests/port/gtk.py:
2416         (GtkPort):
2417         * Scripts/webkitpy/layout_tests/port/mac.py:
2418         (MacPort):
2419         (MacPort.determine_port_name):
2420         (MacPort.__init__):
2421         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2422         (test_versions):
2423         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
2424         (MockDRTPort):
2425         (MockDRTPort.determine_port_name):
2426         (MockDRTPort.__init__):
2427         (MockDRT):
2428         (MockDRT.determine_port_name):
2429         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2430         (PortTestCase.make_port):
2431         * Scripts/webkitpy/layout_tests/port/qt.py:
2432         (QtPort):
2433         (QtPort.determine_port_name):
2434         (QtPort.__init__):
2435         * Scripts/webkitpy/layout_tests/port/test.py:
2436         (TestPort):
2437         (TestPort.determine_port_name):
2438         (TestPort.__init__):
2439         * Scripts/webkitpy/layout_tests/port/webkit.py:
2440         (WebKitPort.__init__):
2441         * Scripts/webkitpy/layout_tests/port/win.py:
2442         (WinPort):
2443         (WinPort.determine_port_name):
2444         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
2445         (get_test_baselines.AllPlatformsPort.__init__):
2446         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
2447         (get_test_config.TestMacPort):
2448         (get_test_config):
2449
2450 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2451
2452         webkitpy: clean up port code in preparation for static port names
2453         https://bugs.webkit.org/show_bug.cgi?id=76356
2454
2455         Reviewed by Adam Barth.
2456
2457         To fix bug 76215 and be able to determine appropriate port names
2458         in webkitpy without actually constructing Port objects, we need
2459         to shuffle some logic in the chromium ports to be able to figure
2460         out whether we built DRT in 32 or 64-bit mode.
2461
2462         * Scripts/webkitpy/layout_tests/port/chromium.py:
2463         (ChromiumPort):
2464         (ChromiumPort._chromium_base_dir):
2465         (ChromiumPort.__init__):
2466         (ChromiumPort.path_from_chromium_base):
2467         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2468         (ChromiumLinuxPort):
2469         (ChromiumLinuxPort._determine_driver_path_statically):
2470         (ChromiumLinuxPort._static_build_path):
2471         (ChromiumLinuxPort._determine_architecture):
2472         (ChromiumLinuxPort.__init__):
2473         (ChromiumLinuxPort._build_path):
2474
2475 2012-01-17  Hao Zheng  <zhenghao@chromium.org>
2476
2477         Do not remove pixel failures for ref tests.
2478         https://bugs.webkit.org/show_bug.cgi?id=76243
2479
2480         Reviewed by Tony Chang.
2481
2482         'NRWT --no-pixel-tests' complains when reftests are expected to
2483         be image mismatch.
2484
2485         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2486         (Manager._update_summary_with_result):
2487         * Scripts/webkitpy/layout_tests/models/test_failures.py:
2488         (is_reftest_failure):
2489         (determine_result_type):
2490
2491 2012-01-17  Balazs Ankes  <Ankes.Balazs@stud.u-szeged.hu>
2492
2493         REGRESSION(r81225): ORWT should ignore reftests
2494         https://bugs.webkit.org/show_bug.cgi?id=67936
2495
2496         Reviewed by Csaba Osztrogonác.
2497
2498         * Scripts/old-run-webkit-tests:
2499         (isUsedInReftest): $filename should be the first parameter
2500         (findTestsToRun): run test if it isn't reftest
2501
2502 2012-01-17  Gustavo Noronha Silva  <gns@gnome.org>
2503
2504         Unreviewed. Fix python unit test I broke with r105142
2505
2506         * Scripts/webkitpy/common/config/ports_unittest.py:
2507         (WebKitPortTest.test_gtk_port):
2508
2509 2012-01-17  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2510
2511         [Qt] Don't set the 'primary' flag when mocking touch-points in MiniBrowser
2512
2513         Qt 5 no longer has that flag.
2514
2515         Reviewed by Simon Hausmann.
2516
2517         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2518         (MiniBrowserApplication::notify):
2519
2520 2012-01-16  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
2521
2522         GTK+ EWS needs to run update-webkitgtk-libs after applying a patch
2523         https://bugs.webkit.org/show_bug.cgi?id=75857
2524
2525         Reviewed by Adam Barth.
2526
2527         * Scripts/webkitdirs.pm:
2528         (buildAutotoolsProject): move cleaning the build to happen early,
2529         and handle --update-gtk right after that
2530         * Scripts/webkitpy/common/config/ports.py:
2531         (GtkPort.build_webkit_command): pass --update-gtk as an option to
2532         build-webkit, so that it is used in the EWS
2533
2534 2012-01-16  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2535
2536         [Qt] Don't add _debug postfix to targets when building debug only
2537
2538         Reviewed by Simon Hausmann.
2539
2540         * qmake/mkspecs/features/default_post.prf:
2541         * qmake/mkspecs/features/functions.prf:
2542
2543 2012-01-16  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2544
2545         [Qt] Add more files to project-files for easier navigation in Qt Creator
2546
2547         Reviewed by Simon Hausmann.
2548
2549         * Tools.pro:
2550         * qmake/config.tests/README: Added.
2551
2552 2012-01-17  Philippe Normand  <pnormand@igalia.com>
2553
2554         [GTK] fast/dom/constructed-objects-prototypes.html fails
2555         https://bugs.webkit.org/show_bug.cgi?id=75432
2556
2557         Reviewed by Martin Robinson.
2558
2559         * Scripts/build-webkit: Enable video-track in GTK.
2560
2561 2012-01-16  Takashi Toyoshima  <toyoshim@chromium.org>
2562
2563         Add toyoshim as a committer.
2564         https://bugs.webkit.org/show_bug.cgi?id=76390
2565
2566         Reviewed by Hajime Morita.
2567
2568         * Scripts/webkitpy/common/config/committers.py:
2569
2570 2012-01-16  Brent Fulgham  <bfulgham@webkit.org>
2571
2572         [WinCairo] Build fix after r104155. Several debug targets were not
2573         properly including the '*Common' property sheets.
2574
2575         * DumpRenderTree/win/DumpRenderTreeLauncherDebugCairoCFLite.vsprops:
2576         * DumpRenderTree/win/ImageDiffLauncherDebugCairoCFLite.vsprops:
2577         * MiniBrowser/Configurations/MiniBrowserLauncherDebugCairoCFLite.vsprops:
2578         * WebKitTestRunner/win/WebKitTestRunnerLauncherDebugCairoCFLite.vsprops:
2579         * WinLauncher/WinLauncherLauncherDebugCairoCFLite.vsprops:
2580
2581 2012-01-16  Csaba Osztrogonác  <ossy@webkit.org>
2582
2583         REGRESSION(r105072): It broke committers_unittest.py
2584         https://bugs.webkit.org/show_bug.cgi?id=76396
2585
2586         Reviewed by Adam Barth.
2587
2588         * Scripts/webkitpy/common/config/committers_unittest.py:
2589         (CommittersTest._assert_fuzz_match): It should handle more names.
2590         (CommittersTest.test_contributors_by_fuzzy_match_with_legacy_names): Fix the testcase.
2591
2592 2012-01-16  Mario Sanchez Prada  <msanchez@igalia.com>
2593
2594         [GTK] Include gtk+3 in the jhbuild modules
2595         https://bugs.webkit.org/show_bug.cgi?id=76284
2596
2597         Reviewed by Philippe Normand.
2598
2599         * gtk/jhbuild.modules: Add some gtk+3 to the jhbuild moduleset.
2600
2601 2012-01-16  Dan Winship  <danw@gnome.org>
2602
2603         Add myself as a committer
2604
2605         Unreviewed
2606
2607         * Scripts/webkitpy/common/config/committers.py:
2608
2609 2012-01-16  Raphael Kubo da Costa  <kubo@profusion.mobi>
2610
2611         Unreviewed; fix the build when --as-needed and --no-copy-dt-needed-entries are passed to the linker.
2612
2613         Some recent versions of a few Linux distributions have started passing
2614         --as-needed and --no-copy-dt-needed-entries by default to the linker,
2615         which broke the build when SHARED_CORE was turned on, as libwtf_efl.so
2616         needed a few symbols from libjavascriptcore_efl.so.
2617
2618         * DumpRenderTree/efl/CMakeLists.txt: Explicitly pass WTF_LIBRARY_NAME
2619         before JavaScriptCore_LIBRARY_NAME in DumpRenderTree_LIBRARIES so the
2620         libraries are searched in the right order by the linker.
2621
2622 2012-01-16  Philippe Normand  <pnormand@igalia.com>
2623
2624         [GTK] set audio resources path in NRTW
2625         https://bugs.webkit.org/show_bug.cgi?id=76380
2626
2627         Reviewed by Martin Robinson.
2628
2629         * Scripts/webkitpy/layout_tests/port/gtk.py:
2630         (GtkPort.setup_environ_for_server): Set $AUDIO_RESOURCES_PATH so
2631         the uninstalled audio resources are used during webaudio tests.
2632
2633 2012-01-16  Carlos Garcia Campos  <cgarcia@igalia.com>
2634
2635         [GTK] ldflags and cflags should take precedence over the existing flags in gtkdoc.py
2636         https://bugs.webkit.org/show_bug.cgi?id=76369
2637
2638         Reviewed by Xan Lopez.
2639
2640         If there's an installed version of webkit and there are new
2641         symbols in the build tree, gtkdoc-scanobj fails due to undefined
2642         references, because the env ldflags are given before the local
2643         ones.
2644
2645         * gtk/gtkdoc.py:
2646         (GTKDoc._run_gtkdoc_scangobj):
2647
2648 2012-01-15  Pablo Flouret  <pablof@motorola.com>
2649
2650         Fix compilation errors on build-webkit --debug --no-video on mac.
2651         https://bugs.webkit.org/show_bug.cgi?id=75867
2652
2653         Reviewed by Philippe Normand.
2654
2655         Add --[no-]video-track to build-webkit, since video track feature
2656         depends on video.
2657
2658         * Scripts/build-webkit:
2659
2660 2012-01-14  James Robinson  <jamesr@chromium.org>
2661
2662         [chromium] Failing webkit_unit_tests does not make the cr-linux EWS bubble red
2663         https://bugs.webkit.org/show_bug.cgi?id=76313
2664
2665         Reviewed by Adam Barth.
2666
2667         The exit code of this script should be the exit code of webkit_unit_tests. Tested manually.
2668
2669         * Scripts/run-chromium-webkit-unit-tests:
2670
2671 2012-01-14  Sheriff Bot  <webkit.review.bot@gmail.com>
2672
2673         Unreviewed, rolling out r105017.
2674         http://trac.webkit.org/changeset/105017
2675         https://bugs.webkit.org/show_bug.cgi?id=76333
2676
2677         broke the build (Requested by philn-tp on #webkit).
2678
2679         * gtk/jhbuild.modules:
2680
2681 2012-01-14  Mario Sanchez Prada  <msanchez@igalia.com>
2682
2683         [GTK] Include gtk+3 in the jhbuild modules
2684         https://bugs.webkit.org/show_bug.cgi?id=76284
2685
2686         Reviewed by Philippe Normand.
2687
2688         * gtk/jhbuild.modules: Add some gtk+3 to the jhbuild moduleset.
2689
2690 2012-01-13  Adam Barth  <abarth@webkit.org>
2691
2692         Remove --dry-run support from webkit-patch
2693         https://bugs.webkit.org/show_bug.cgi?id=76300
2694
2695         Reviewed by Dirk Pranke.
2696
2697         This is the rest of the patch from Bug 76300.  I landed it separtely
2698         because main.py had some stray conflict markers that I needed to clean
2699         up first.
2700
2701         * Scripts/webkitpy/test/main.py:
2702         (Tester.parse_args):
2703         (Tester._run_tests):
2704
2705 2012-01-13  Adam Barth  <abarth@webkit.org>
2706
2707         test-webkitpy does not run
2708         https://bugs.webkit.org/show_bug.cgi?id=76318
2709
2710         Unreviewed.
2711
2712         Looks like some conflict markers got left in the file.
2713
2714         * Scripts/webkitpy/test/main.py:
2715         (Tester.run):
2716
2717 2012-01-13  Adam Barth  <abarth@webkit.org>
2718
2719         Remove --dry-run support from webkit-patch
2720         https://bugs.webkit.org/show_bug.cgi?id=76300
2721
2722         Reviewed by Dirk Pranke.
2723
2724         We don't use this functionality for anything, and it doesn't really
2725         work.  Someone in #webkit tried using it and and got sad/confused.  We
2726         should just remove it.
2727
2728         * Scripts/webkitpy/common/checkout/scm/git.py:
2729         (Git.push_local_commits_to_server):
2730         * Scripts/webkitpy/common/checkout/scm/scm.py:
2731         (SCM.__init__):
2732         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
2733         (_shared_test_commit_with_message):
2734         (GitSVNTest.test_commit_text_parsing):
2735         * Scripts/webkitpy/common/checkout/scm/svn.py:
2736         (SVN.commit_with_message):
2737         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
2738         (Bugzilla.__init__):
2739         (Bugzilla.quips):
2740         (Bugzilla.authenticate):
2741         (Bugzilla.add_attachment_to_bug):
2742         (Bugzilla.add_patch_to_bug):
2743         (Bugzilla.create_bug):
2744         (Bugzilla.clear_attachment_flags):
2745         (Bugzilla.set_flag_on_attachment):
2746         (Bugzilla.obsolete_attachment):
2747         (Bugzilla.add_cc_to_bug):
2748         (Bugzilla.post_comment_to_bug):
2749         (Bugzilla.close_bug_as_fixed):
2750         (Bugzilla.reassign_bug):
2751         (reopen_bug):
2752         * Scripts/webkitpy/test/main.py:
2753         (Tester.parse_args):
2754         (Tester.run_tests):
2755
2756 2012-01-13  David Levin  <levin@chromium.org>
2757
2758         check-webkit-style: should encourage the use of Own* classes for Windows DC.
2759         https://bugs.webkit.org/show_bug.cgi?id=76227
2760
2761         Reviewed by Dirk Pranke.
2762
2763         * Scripts/webkitpy/style/checkers/cpp.py:
2764         (check_for_leaky_patterns): The new check.
2765         (process_line): Added a call to the new check.
2766         (CppChecker): Added the new error type.
2767         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
2768         (CppStyleTestBase):
2769         (CppStyleTestBase.perform_leaky_pattern_check):
2770         The check for only leaky pattern errors.
2771         (LeakyPatternTest): Test cases.
2772
2773 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
2774
2775         test-webkitpy: clean up handling of test directories, QueueStatusServer
2776         https://bugs.webkit.org/show_bug.cgi?id=76238
2777
2778         Reviewed by Adam Barth.
2779
2780         This patch cleans up a bunch of hacks related to how we find all
2781         of the python unit tests and set up the sys.path variable in
2782         test-webkitpy.
2783
2784         This patch changes how test-webkitpy works slightly in that we
2785         will now look for test files under Tools/Scripts and Source/WebKit2/Scripts
2786         rather than just Tools/Scripts/webkitpy and Source/WebKit2/Scripts/webkit.
2787         This greatly simplifies the mapping of paths to module names and
2788         appears to have no significant performance impact or other drawbacks.
2789
2790         There are other ways that we could remove all of the hacks and
2791         still have a more flexible mapping of paths to module names, but
2792         the ones I've thought of so far add a decent amount of complexity to
2793         the implementation and don't seem to buy us much.
2794         
2795         * QueueStatusServer/__init__.py:
2796         * Scripts/test-webkitpy:
2797         * Scripts/webkitpy/test/main.py:
2798         (Tester._modules_from_paths):
2799         (Tester._is_module):
2800
2801 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
2802
2803         test-webkitpy: should support classes and individual test names as well as modules
2804         https://bugs.webkit.org/show_bug.cgi?id=76233
2805
2806         Reviewed by Adam Barth.
2807
2808         The default unittest.main() supports a list of individual
2809         TestCase classes or individual tests as well as modules. This
2810         change adds support for the same to test-webkitpy, so now you
2811         can say
2812
2813           % test-wekbitpy webkitpy.test.cat_unittest.CatTest
2814
2815         and
2816
2817           % test-wekbitpy webkitpy.test.cat_unittest.CatTest.test_basic
2818
2819         in addition to
2820
2821           % test-wekbitpy webkitpy.test.cat_unittest
2822
2823         * Scripts/test-webkitpy:
2824         * Scripts/webkitpy/test/cat_unittest.py:
2825         * Scripts/webkitpy/test/main.py:
2826         (Tester.parse_args):
2827         (Tester.run):
2828         (Tester):
2829         (Tester._find_modules):
2830         (Tester._exclude):
2831         (Tester._run_tests):
2832         (Tester._is_module):
2833         (Tester._log_exception):
2834
2835 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
2836
2837         test-webkitpy: add basic code coverage support
2838         https://bugs.webkit.org/show_bug.cgi?id=76201
2839
2840         Reviewed by Adam Barth.
2841
2842         This test adds basic code coverage support to test-webkitpy; if
2843         you pass the -c flag, we will generate a new coverage file. We
2844         can start simply with this and add more features as necessary.
2845
2846         This patch requires coverage to be already installed on the
2847         machine. I am not auto-installing this for now.
2848
2849         * Scripts/webkitpy/test/main.py:
2850         (Tester.parse_args):
2851         (Tester.run_tests):
2852
2853 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
2854
2855         StatusBubbleTest.test_build_bubble is failing
2856         https://bugs.webkit.org/show_bug.cgi?id=76222
2857
2858         Reviewed by Adam Barth.
2859
2860         * QueueStatusServer/handlers/statusbubble_unittest.py:
2861         (StatusBubbleTest.test_build_bubble):
2862
2863 2012-01-13  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
2864
2865         nrwt should use multiple http shards
2866         https://bugs.webkit.org/show_bug.cgi?id=75958
2867
2868         Add --max-locked-shards option to the nrwt for manual set
2869         the maximum number of the locked shards
2870
2871         Reviewed by Dirk Pranke.
2872
2873         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2874         (Manager._max_locked_shards):
2875         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2876         (ShardingTests.get_shards):
2877         (ShardingTests.test_multiple_locked_shards):
2878         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2879         (parse_args):
2880
2881 2012-01-13  Raphael Kubo da Costa  <kubo@profusion.mobi>
2882
2883         prepare-ChangeLog: Look for a text editor in a few more variables.
2884         https://bugs.webkit.org/show_bug.cgi?id=76289
2885
2886         Reviewed by Ryosuke Niwa.
2887
2888         A text editor was looked for only in the CHANGE_LOG_EDITOR and
2889         CHANGE_LOG_EDIT_APPLICATION environment variables before falling back
2890         to the OS X-specific open(1) command.
2891
2892         As there does not seem to be a common way of looking for a text editor
2893         among the scripts in Tools/Scripts right now, just look at $VISUAL and
2894         $EDITOR (set much more often than the other variables in Unix systems)
2895         before falling back to open(1).
2896
2897         * Scripts/prepare-ChangeLog:
2898         (openChangeLogs):
2899
2900 2012-01-13  Gabor Rapcsanyi  <rgabor@webkit.org>
2901
2902         [GTK] gtkdoc.py ignoring LDFLAGS and CFLAGS environment variables
2903         https://bugs.webkit.org/show_bug.cgi?id=76269
2904
2905         Reviewed by Martin Robinson.
2906
2907         * gtk/gtkdoc.py:
2908         (GTKDoc._run_gtkdoc_scangobj):
2909
2910 2012-01-13  Csaba Osztrogonác  <ossy@webkit.org>
2911
2912         [Qt][WK2]REGRESSION(r104881):It broke hundreds of tests
2913         https://bugs.webkit.org/show_bug.cgi?id=76247
2914
2915         Reviewed by Kenneth Rohde Christiansen.
2916
2917         * WebKitTestRunner/TestController.cpp:
2918         (WTR::TestController::resetStateToConsistentValues): Disable mock scrollbars on Qt by default. (workaround)
2919
2920 2012-01-13  Hajime Morrita  <morrita@chromium.org>
2921
2922         [Mac][DRT] Should support --no-timeout
2923         https://bugs.webkit.org/show_bug.cgi?id=76242
2924
2925         Reviewed by Darin Adler.
2926
2927         * DumpRenderTree/mac/DumpRenderTree.mm:
2928         (initializeGlobalsFromCommandLineOptions): Added "--no-timeout" to the options list.
2929         (setWaitToDumpWatchdog): Added to encapsulate the waitToDumpWatchdog global variable.
2930         (shouldSetWaitToDumpWatchdog): Added to check the useTimeoutWatchdog flag set by the option.
2931         * DumpRenderTree/mac/DumpRenderTreeMac.h:
2932         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
2933         (LayoutTestController::setWaitToDump): Now calls shouldSetWaitToDumpWatchdog() to check the flag.
2934
2935 2012-01-12  Simon Fraser  <simon.fraser@apple.com>
2936
2937         Mitigate scrollbar differences when running pixel tests
2938         https://bugs.webkit.org/show_bug.cgi?id=67217
2939
2940         Reviewed by Dan Bernstein.
2941         
2942         Enable mock scrollbars by default for tests.
2943
2944         * DumpRenderTree/mac/DumpRenderTree.mm:
2945         (resetDefaultsToConsistentValues):
2946         * WebKitTestRunner/TestController.cpp:
2947         (WTR::TestController::resetStateToConsistentValues):
2948
2949 2012-01-12  Ryosuke Niwa  <rniwa@webkit.org>
2950
2951         Need a script to run tests in PerformanceTests
2952         https://bugs.webkit.org/show_bug.cgi?id=76132
2953
2954         Reviewed by Adam Barth.
2955
2956         Add run-perf-tests to run performance tests using parser/resources/runner.js.
2957         Unfortunately, there isn't an easy way of telling which test uses which format
2958         so hard-code directory that uses Chromium perf-bot style (only inspector perf. tests for now).
2959
2960         All test outputs are re-formatted to match Chromium perf-bot style.
2961
2962         * Scripts/run-inspector-perf-tests.py:
2963         * Scripts/run-perf-tests: Added.
2964         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2965         (PerfTestsRunner):
2966         (PerfTestsRunner.__init__):
2967         (PerfTestsRunner._collect_tests):
2968         (PerfTestsRunner.run):
2969         (PerfTestsRunner._print_status):
2970         (PerfTestsRunner._run_tests_set):
2971         (PerfTestsRunner._process_chromium_style_test_result):
2972         (PerfTestsRunner._should_ignore_line_in_parser_test_result):
2973         (PerfTestsRunner._process_parser_test_result):
2974         (PerfTestsRunner._run_single_test):
2975         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2976         (MainTest.TestDriver.run_test):
2977         (create_runner):
2978         (run_test):
2979         (test_run_test_set):
2980         (test_run_test_set_for_parser_tests):
2981         (test_collect_tests):
2982
2983 2012-01-12  Dirk Pranke  <dpranke@chromium.org>
2984
2985         webkitpy: clean up port unit tests in preparation for making port_name mandatory
2986         https://bugs.webkit.org/show_bug.cgi?id=76123
2987
2988         Reviewed by Adam Barth.
2989
2990         This change updates the unit tests to be more consistent in how
2991         ports are created, in preparation for making port_name a
2992         mandatory parameter to the constructor and removing any version
2993         interpretation from it.
2994
2995         There should be no functional changes in this patch, but some
2996         tests that are no longer relevant have been removed or reworked.
2997
2998         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2999         (PortTest.test_get_option__default):
3000         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
3001         (ChromiumLinuxPortTest):
3002         (ChromiumLinuxPortTest.assert_architecture):
3003         (ChromiumLinuxPortTest.test_determine_architecture_fails):
3004         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
3005         (ChromiumMacPortTest):
3006         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3007         (ChromiumPortTest):
3008         (ChromiumPortTest.test_all_test_configurations):
3009         (ChromiumPortTest.TestMacPort.__init__):
3010         (ChromiumPortTest.TestLinuxPort.__init__):
3011         (ChromiumPortTest.TestWinPort.__init__):
3012         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
3013         (ChromiumWinTest):
3014         * Scripts/webkitpy/layout_tests/port/efl_unittest.py:
3015         (EflPortTest):
3016         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
3017         (GtkPortTest):
3018         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3019         (MacTest):
3020         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
3021         (MockDRTPortTest):
3022         (MockDRTPortTest.make_port):
3023         * Scripts/webkitpy/layout_tests/port/qt.py:
3024         (QtPort._port_flag_for_scripts):
3025         (QtPort.__init__):
3026         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
3027         (QtPortTest):
3028         (QtPortTest._assert_search_path):
3029         (QtPortTest.test_baseline_search_path):
3030         (QtPortTest.test_operating_system):
3031         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3032         (WebKitPortTest):
3033         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
3034         (WinPortTest):
3035         (WinPortTest._assert_search_path):
3036         (WinPortTest.test_baseline_search_path):
3037         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
3038         (get_test_config.TestMacPort.__init__):
3039
3040 2012-01-12  Dirk Pranke  <dpranke@chromium.org>
3041
3042         test-webkitpy: clean up logging and provide a real --help message
3043         https://bugs.webkit.org/show_bug.cgi?id=76142
3044
3045         Reviewed by Adam Barth.
3046
3047         This test is mostly more refactoring - it calls objects in
3048         unittest directly instead of routing through unittest.main() --
3049         in order to gain more control over the logging and to provide a
3050         correct --help message. This will also give us a better
3051         foundation for adding more feautures down the road.
3052
3053         This patch adds different levels of logging controlled by the
3054         '-s', '-q', and '-v' options (see --help for details), and
3055         removes the --verbose-logging option.
3056
3057         * Scripts/test-webkitpy:
3058         * Scripts/webkitpy/test/main.py:
3059         (Tester):
3060         (Tester.clean_packages):
3061         (Tester.__init__):
3062         (Tester.parse_args):
3063         (Tester.configure):
3064         (Tester._configure_logging):
3065         (Tester.find_modules):
3066         (Tester._exclude):
3067         (Tester._modules_from_paths):
3068         (Tester.run_tests):
3069
3070 2012-01-12  Simon Hausmann  <simon.hausmann@nokia.com>
3071
3072         Make the new WTF module build on Qt
3073         https://bugs.webkit.org/show_bug.cgi?id=76163
3074
3075         Reviewed by Tor Arne Vestbø.
3076
3077         * qmake/mkspecs/modules/wtf.prf: Pull in WTF from the new location.
3078
3079 2012-01-12  Vsevolod Vlasov  <vsevik@chromium.org>
3080
3081         Unreviewed, mac DumpRenderTree fix.
3082
3083         * DumpRenderTree/mac/UIDelegate.mm:
3084         (-[UIDelegate webView:addMessageToConsole:withSource:]):
3085
3086 2012-01-12  Vsevolod Vlasov  <vsevik@chromium.org>
3087
3088         WebKitTestRunner should not print console message line number when it is equal to zero.
3089         https://bugs.webkit.org/show_bug.cgi?id=76167
3090
3091         Reviewed by Csaba Osztrogonác.
3092
3093         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
3094         (WTR::InjectedBundlePage::willAddMessageToConsole):
3095
3096 2012-01-11  Vsevolod Vlasov  <vsevik@chromium.org>
3097
3098         Make default console messages line numbers consistent.
3099         https://bugs.webkit.org/show_bug.cgi?id=74075
3100
3101         Reviewed by Pavel Feldman.
3102
3103         Unset line numbers are not printed to console now.
3104         Added Source/WebCore/inspector to efl DumpRenderTree include path.
3105
3106         * DumpRenderTree/chromium/WebViewHost.cpp:
3107         (WebViewHost::didAddMessageToConsole):
3108         * DumpRenderTree/efl/CMakeLists.txt:
3109         * DumpRenderTree/efl/DumpRenderTreeView.cpp:
3110         (onConsoleMessage):
3111         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3112         (webViewConsoleMessage):
3113         * DumpRenderTree/mac/UIDelegate.mm:
3114         (-[UIDelegate webView:addMessageToConsole:withSource:]):
3115         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
3116         (WebCore::WebPage::javaScriptConsoleMessage):
3117         * DumpRenderTree/win/UIDelegate.cpp:
3118         (UIDelegate::webViewAddMessageToConsole):
3119         * DumpRenderTree/wx/DumpRenderTreeWx.cpp:
3120         (LayoutWebViewEventHandler::OnConsoleMessageEvent):
3121
3122 2012-01-10  Dirk Pranke  <dpranke@chromium.org>
3123
3124         test-webkitpy: push more logic into webkitpy.test.main, clean up code
3125         https://bugs.webkit.org/show_bug.cgi?id=76021
3126
3127         Reviewed by Eric Seidel.
3128
3129         I plan to add more functionality to test-webkitpy, but it is
3130         difficult to hack on now; this patch is the first of two that
3131         brings test-webkitpy more inline with current coding style by
3132         pushing more logic into the webkitpy.test.main.Tester class (so
3133         that it will be testable itself).
3134         
3135         There should be no functional changes in this patch.
3136
3137         * Scripts/test-webkitpy:
3138         (_path_from_webkit_root):
3139         * Scripts/webkitpy/test/main.py:
3140         (Tester):
3141         (Tester.init):
3142         (Tester.configure_logging):
3143         (Tester.configure_logging.filter):
3144         (Tester.clean_packages):
3145         (Tester.run_tests):
3146
3147 2012-01-11  Wei Jia  <wjia@chromium.org>
3148
3149         implement layout tests for <video> with media stream
3150         https://bugs.webkit.org/show_bug.cgi?id=74882
3151
3152         Reviewed by Kent Tamura.
3153
3154         * DumpRenderTree/DumpRenderTree.gypi:
3155         * DumpRenderTree/chromium/WebUserMediaClientMock.cpp: Added.
3156         (WebKit::WebUserMediaClientMock::create):
3157         (WebKit::WebUserMediaClientMock::IsMockStream):
3158         (WebKit::WebUserMediaClientMock::requestUserMedia):
3159         (WebKit::WebUserMediaClientMock::cancelUserMediaRequest):
3160         * DumpRenderTree/chromium/WebUserMediaClientMock.h: Added.
3161         (WebKit::WebUserMediaClientMock::~WebUserMediaClientMock):
3162         (WebKit::WebUserMediaClientMock::WebUserMediaClientMock):
3163         * DumpRenderTree/chromium/WebViewHost.cpp:
3164         (WebViewHost::userMediaClient):
3165         (WebViewHost::userMediaClientMock):
3166         (WebViewHost::createMediaPlayer):
3167         (WebViewHost::mediaStreamUtil):
3168         (WebViewHost::testMediaStreamClient):
3169         * DumpRenderTree/chromium/WebViewHost.h:
3170
3171 2012-01-11  Dirk Pranke  <dpranke@chromium.org>
3172
3173         Unreviewed, build fix.
3174
3175         Fix chromium-win's version remapping to work correctly on Win 7 SP1.
3176
3177         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
3178         (ChromiumWinPort.__init__):
3179
3180 2012-01-11  Ryosuke Niwa  <rniwa@webkit.org>
3181
3182         Sheriffbot build fix. Don't emit empty lines on IRC.
3183
3184         * Scripts/webkitpy/tool/bot/irc_command.py:
3185         (LastGreenRevision.execute):
3186
3187 2012-01-11  Ryosuke Niwa  <rniwa@webkit.org>
3188
3189         last-green-revision should report the revision that succeeded on all bots
3190         https://bugs.webkit.org/show_bug.cgi?id=76109
3191
3192         Reviewed by Adam Barth.
3193
3194         Add an algorithm to find the last known good revision among the specified bots. For a revision to be
3195         considered green by this algorithm, all matching builders must have a successful run at the revision
3196         or two consecutive successful runs before and after the revision.
3197
3198         Also fixed a bug in irc_command that the result wasn't posted on IRC properly and a bug in queries.py
3199         that resulted in an exception when the user didn't pass BUILDER_NAME argument.
3200
3201         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
3202         (BuildBot._revisions_for_builder):
3203         (BuildBot):
3204         (BuildBot._find_green_revision):
3205         (BuildBot.last_green_revision):
3206         * Scripts/webkitpy/common/net/buildbot/buildbot_mock.py:
3207         (MockBuildBot.last_green_revision):
3208         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
3209         (test_revisions_for_builder):
3210         (test_find_green_revision):
3211         (test_last_green_revision):
3212         * Scripts/webkitpy/tool/bot/irc_command.py:
3213         (LastGreenRevision.execute):
3214         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
3215         (SheriffIRCBotTest.test_lgr):
3216         * Scripts/webkitpy/tool/commands/queries.py:
3217         (LastGreenRevision):
3218
3219 2012-01-11  Dirk Pranke  <dpranke@chromium.org>
3220
3221         Unreviewed, fix build bustage on win32 introduced by r104725.
3222
3223         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
3224         (MockDRTPortTest.make_port):
3225         (MockChromiumDRTTest.test_pixeltest__fails):
3226
3227 2012-01-11  Jochen Eisinger  <jochen@chromium.org>
3228
3229         [Chromium] mimic the (old) behavior of the mac port for (re)setting the color profile
3230         https://bugs.webkit.org/show_bug.cgi?id=75618
3231
3232         Eventually, we should get away from setting the display's color profile, as the mac port is doing now.
3233
3234         Reviewed by Tony Chang.
3235
3236         * DumpRenderTree/chromium/LayoutTestHelper.mm:
3237         (installLayoutTestColorProfile):
3238         (restoreUserColorProfile):
3239         (main):
3240
3241 2012-01-11  Dirk Pranke  <dpranke@chromium.org>
3242
3243         test-webkitpy: fix -v and eliminate some spurious warnings
3244         https://bugs.webkit.org/show_bug.cgi?id=76099
3245
3246         Reviewed by Adam Barth.
3247
3248         when we added support for --xml output, we broke the -v flag.
3249
3250         Also, fix the code to swallow a harmless warning from
3251         VCSUtils.pm and fix a deprecated warning in jsonchecker.py.
3252
3253         test-webkitpy is now silent again except for the one known
3254         logging issue.
3255
3256         * Scripts/webkitpy/common/checkout/checkout.py:
3257         (Checkout.commit_message_for_this_commit):
3258         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
3259         (test_commit_message_for_this_commit):
3260         * Scripts/webkitpy/style/checkers/jsonchecker.py:
3261         (JSONChecker.check):
3262         (JSONChecker.line_number_from_json_exception):
3263         * Scripts/webkitpy/test/main.py:
3264         (Tester.run_tests):
3265
3266 2012-01-11  Dirk Pranke  <dpranke@chromium.org>
3267
3268         webkitpy: webkitpy.layout_tests.port tests should only use MockSystemHost, not MockHost
3269         https://bugs.webkit.org/show_bug.cgi?id=76084
3270
3271         Reviewed by Adam Barth.
3272
3273         This will help make it easier to catch layering violations. This
3274         change just refactors test code; there are no changes to
3275         production code.
3276
3277         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3278         (PortTest.make_port):
3279         (PortTest.test_find_no_paths_specified):
3280         (PortTest.test_find_one_test):
3281         (PortTest.test_find_glob):
3282         (PortTest.test_find_with_skipped_directories):
3283         (PortTest.test_find_with_skipped_directories_2):
3284         (PortTest.test_parse_reftest_list):
3285         (PortTest.test_httpd_returns_error_code):
3286         (PortTest.test_virtual_methods):
3287         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
3288         (ChromiumGpuTest.assert_port_works):
3289         (ChromiumGpuTest._assert_baseline_path):
3290         (ChromiumGpuTest.test_graphics_type):
3291         (ChromiumGpuTest.test_default_tests_paths.test_paths):
3292         (ChromiumGpuTest.test_test_files):
3293         (ChromiumGpuTest.test_test_files.test_paths):
3294         (ChromiumGpuTest):
3295         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
3296         (ChromiumLinuxPortTest.assert_architecture):
3297         (ChromiumLinuxPortTest.test_check_illegal_port_names):
3298         (ChromiumLinuxPortTest.test_determine_architecture_fails):
3299         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
3300         (DriverTest.make_port):
3301         (DriverTest):
3302         (DriverTest.assertVirtual):
3303         (DriverTest._assert_wrapper):
3304         (DriverTest.test_virtual_driver_methods):
3305         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
3306         (MockDRTPortTest.make_port):
3307         (MockDRTPortTest.test_port_name_in_constructor):
3308         (MockDRTTest.assertTest):
3309         (MockDRTTest.test_main):
3310         (MockChromiumDRTTest.test_pixeltest__fails):
3311         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
3312         (QtPortTest._assert_search_path):
3313         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3314         (TestWebKitPort.__init__):
3315         (WebKitPortUnitTests.test_default_options):
3316         (test_runtime_feature_list):
3317         (test_test_expectations):
3318
3319 2012-01-11  Ojan Vafai  <ojan@chromium.org>
3320
3321         Style skipped tests in the flakiness dashboard
3322         https://bugs.webkit.org/show_bug.cgi?id=76091
3323
3324         Reviewed by Tony Chang.
3325
3326         * TestResultServer/static-dashboards/dashboard_base.js:
3327         Removed "FLAKY" from this list since we don't ever mark individual runs
3328         as flaky. Also deleted an outdated line.
3329         * TestResultServer/static-dashboards/flakiness_dashboard.html:
3330
3331 2012-01-11  Ojan Vafai  <ojan@chromium.org>
3332
3333         Remove gtest normalization code from the test results server
3334         https://bugs.webkit.org/show_bug.cgi?id=76089
3335
3336         Reviewed by Tony Chang.
3337
3338         The normalization now happens on the gtest side.
3339
3340         * TestResultServer/model/jsonresults.py:
3341         (JsonResults.merge):
3342         * TestResultServer/model/jsonresults_unittest.py:
3343         (JsonResultsTest.test_gtest):
3344
3345 2012-01-11  Ojan Vafai  <ojan@chromium.org>
3346
3347         Be more aggressive about deleting unnecessary results from the test-results-server json files
3348         https://bugs.webkit.org/show_bug.cgi?id=76086
3349
3350         Reviewed by Tony Chang.
3351
3352         Remove any tests that only contain skip/pass/nodata values whose runtime
3353         is less than 5 seconds.
3354
3355         * TestResultServer/model/jsonresults.py:
3356         (JsonResults._normalize_results_json):
3357         * TestResultServer/model/jsonresults_unittest.py:
3358         (JsonResultsTest.test_merge_remove_test):
3359         (JsonResultsTest.test_merge_keep_test_with_all_pass_but_slow_time):
3360
3361 2012-01-11  Ryosuke Niwa  <rniwa@webkit.org>
3362
3363         last-green-revision should give us per-bot information
3364         https://bugs.webkit.org/show_bug.cgi?id=76011
3365
3366         Reviewed by Adam Barth.
3367
3368         Rewrote last-green-revision command. Instead of finding a revision for which all bots succeeded,
3369         we report the latest green run on each bot from the last 100 runs.
3370
3371         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
3372         (BuildBot._fetch_builder_page):
3373         (BuildBot):
3374         (BuildBot._green_revision_for_builder):
3375         (BuildBot.last_green_revision):
3376         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
3377         (test_green_revision_for_builder):
3378         (test_last_green_revision):
3379         * Scripts/webkitpy/tool/bot/irc_command.py:
3380         (LastGreenRevision.execute):
3381         * Scripts/webkitpy/tool/commands/queries.py:
3382         (LastGreenRevision.execute):
3383
3384 2012-01-11  Dirk Pranke  <dpranke@chromium.org>
3385
3386         webkitpy: clean up version detection in webkitpy.layout_tests.port
3387         https://bugs.webkit.org/show_bug.cgi?id=76016
3388
3389         Reviewed by Adam Barth.
3390
3391         This change updates the code in webkitpy.layout_tests.port to
3392         use the new os_name and os_version fields on host.platform. This
3393         gets layout_tests.port out of the version parsing business. Most
3394         of the diffs are just cleaning up test code to accomodate the
3395         changes.
3396
3397         * Scripts/webkitpy/common/system/executive_mock.py:
3398         (MockExecutive):
3399         (MockExecutive.ignore_error):
3400         (MockExecutive2):
3401         (MockExecutive2.ignore_error):
3402         * Scripts/webkitpy/common/system/platforminfo_mock.py:
3403         (MockPlatformInfo.is_mac):
3404         (MockPlatformInfo.is_linux):
3405         (MockPlatformInfo.is_win):
3406         * Scripts/webkitpy/common/system/systemhost_mock.py:
3407         (MockSystemHost.__init__):
3408         * Scripts/webkitpy/layout_tests/port/apple.py:
3409         (ApplePort.__init__):
3410         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
3411         (ChromiumMacPort.__init__):
3412         (ChromiumMacPort.check_wdiff):
3413         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
3414         (ChromiumMacPortTest):
3415         (ChromiumMacPortTest.assert_name):
3416         (ChromiumMacPortTest.test_versions):
3417         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3418         (ChromiumPortTest.test_all_test_configurations):
3419         (ChromiumPortTest.TestMacPort.__init__):
3420         (ChromiumPortTest.TestLinuxPort.__init__):
3421         (ChromiumPortTest.TestWinPort.__init__):
3422         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
3423         (ChromiumWinPort.__init__):
3424         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
3425         (ChromiumWinTest):
3426         (ChromiumWinTest.assert_name):
3427         (ChromiumWinTest.test_versions):
3428         (ChromiumWinTest.test_baseline_path):
3429         (ChromiumWinTest.test_build_path):
3430         * Scripts/webkitpy/layout_tests/port/factory.py:
3431         (PortFactory._port_name_from_arguments_and_options):
3432         (PortFactory.get):
3433         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
3434         (FactoryTest):
3435         (FactoryTest.setUp):
3436         (FactoryTest.assert_port):
3437         (FactoryTest.test_mac):
3438         (FactoryTest.test_win):
3439         (FactoryTest.test_google_chrome):
3440         (FactoryTest.test_gtk):
3441         (FactoryTest.test_qt):
3442         (FactoryTest.test_chromium_gpu):
3443         (FactoryTest.test_chromium_gpu_linux):
3444         (FactoryTest.test_chromium_gpu_mac):
3445         (FactoryTest.test_chromium_gpu_win):
3446         (FactoryTest.test_chromium_mac):
3447         (FactoryTest.test_chromium_linux):
3448         (FactoryTest.test_chromium_win):
3449         (FactoryTest.test_unknown_specified):
3450         (FactoryTest.test_unknown_default):
3451         * Scripts/webkitpy/layout_tests/port/google_chrome_unittest.py:
3452         (TestGoogleChromePort):
3453         (TestGoogleChromePort._verify_baseline_path):
3454         (TestGoogleChromePort._verify_expectations_overrides):
3455         (TestGoogleChromePort.test_get_google_chrome_port):
3456         * Scripts/webkitpy/layout_tests/port/mac.py:
3457         (MacPort):
3458         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3459         (MacTest):
3460         (assert_name):
3461         (test_tests_for_other_platforms):
3462         (test_version):
3463         (test_versions):
3464         (test_is_version_methods):
3465         (test_setup_environ_for_server):
3466         (_assert_search_path):
3467         (test_show_results_html_file):
3468         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
3469         (PortTestCase):
3470         (PortTestCase.make_port):
3471         * Scripts/webkitpy/layout_tests/port/win.py:
3472         (WinPort):
3473         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
3474         (WinPortTest):
3475         (WinPortTest.test_show_results_html_file):
3476         (WinPortTest._assert_search_path):
3477         (WinPortTest._assert_version):
3478
3479 2012-01-11  Hugo Parente Lima  <hugo.lima@openbossa.org>
3480
3481         [Qt][WK2] Remove workaround introduced into bug 75470
3482         https://bugs.webkit.org/show_bug.cgi?id=75889
3483
3484         Reviewed by Tor Arne Vestbø.
3485
3486         * MiniBrowser/qt/qml/BrowserWindow.qml:
3487
3488 2012-01-10  Dmitry Lomov  <dslomov@google.com>
3489
3490         [Chromium] Remove obsolete references to WebWorker class.
3491         https://bugs.webkit.org/show_bug.cgi?id=76020
3492
3493         Reviewed by David Levin.
3494
3495         * DumpRenderTree/chromium/TestWebWorker.h: Removed.
3496         * DumpRenderTree/chromium/WebViewHost.cpp:
3497         * DumpRenderTree/chromium/WebViewHost.h:
3498
3499 2012-01-11  Csaba Osztrogonác  <ossy@webkit.org>
3500
3501         [Qt] Some css3 filter tests are failing after r104698
3502         https://bugs.webkit.org/show_bug.cgi?id=76055
3503
3504         Reviewed by Simon Hausmann.
3505
3506         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
3507         (LayoutTestController::overridePreference): Add WebKitAcceleratedCompositingEnabled preference.
3508
3509 2012-01-11  Tony Chang  <tony@chromium.org>
3510
3511         Don't remove 0 byte files when running build-webkit --chromium
3512         https://bugs.webkit.org/show_bug.cgi?id=75997
3513
3514         Reviewed by Csaba Osztrogonác.
3515
3516         If QTDIR is set, isQt() is true even when running build-webkit --chromium.
3517
3518         * Scripts/webkitdirs.pm:
3519         (determineIsQt):
3520
3521 2012-01-10  David Kilzer  <ddkilzer@apple.com>
3522
3523         Add TestWebKitAPI tests for KURL
3524         <http://webkit.org/b/75774>
3525
3526         Reviewed by Adam Barth.
3527
3528         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Add a
3529         WebCore folder and KURL.cpp source file to the project.
3530         * TestWebKitAPI/Tests/WebCore/KURL.cpp: Added.
3531         (TestWebKitAPI::WebCore::SetUp): Initialize the main thread
3532         during setup since this test doesn't go through WebKit or
3533         WebKit2.
3534         (TestWebKitAPI::TEST_F): Add two basic tests.
3535
3536 2012-01-10  David Kilzer  <ddkilzer@apple.com>
3537
3538         Extract common code for comparing WTF::String objects into WTFStringUtilities.h
3539
3540         Reviewed by Adam Barth.
3541
3542         Needed for: <http://webkit.org/b/75774> Add TestWebKitAPI tests for KURL
3543
3544         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Add
3545         WTFStringUtilities.h to the project.
3546         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp: Update to use
3547         WTFStringUtilities.h.
3548         * TestWebKitAPI/WTFStringUtilities.h: Added.
3549         (WTF::operator<<):
3550
3551 2012-01-10  Vincent Scheib  <scheib@chromium.org>
3552
3553         Clarify run-webkit-tests options reset-results and new-baseline
3554         https://bugs.webkit.org/show_bug.cgi?id=75993
3555
3556         Reviewed by Dirk Pranke.
3557
3558         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3559         (parse_args):
3560
3561 2012-01-10  Simon Fraser  <simon.fraser@apple.com>
3562
3563         Disabled mock scrollbars should draw differently
3564         https://bugs.webkit.org/show_bug.cgi?id=75995
3565
3566         Reviewed by James Robinson.
3567
3568         When the scrollbar is disabled, paint the entire track of
3569         mock scrollbars with a lighter gray, and hide the thumb.
3570
3571         * DumpRenderTree/mac/DumpRenderTree.mm:
3572         (-[DRTMockScroller drawKnob]):
3573         (-[DRTMockScroller drawRect:]):
3574
3575 2012-01-10  James Robinson  <jamesr@chromium.org>
3576
3577         [chromium] Run webkit_unit_tests as part of webkit-patch build-and-test
3578         https://bugs.webkit.org/show_bug.cgi?id=75988
3579
3580         Reviewed by Adam Barth.
3581
3582         * Scripts/webkitpy/common/config/ports.py:
3583         (WebKitPort.run_webkit_unit_tests_command):
3584         (WebKitPort):
3585         (ChromiumPort.run_webkit_unit_tests_command):
3586         (ChromiumPort):
3587         * Scripts/webkitpy/common/config/ports_mock.py:
3588         (MockPort.run_webkit_unit_tests_command):
3589         * Scripts/webkitpy/tool/commands/download_unittest.py:
3590         (DownloadCommandsTest.test_build_and_test):
3591         (test_land):
3592         (test_land_red_builders):
3593         * Scripts/webkitpy/tool/steps/runtests.py:
3594         (RunTests.run):
3595         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
3596
3597 2012-01-09  Dirk Pranke  <dpranke@chromium.org>
3598
3599         webkitpy: add os_name, os_version to platforminfo
3600         https://bugs.webkit.org/show_bug.cgi?id=75931
3601
3602         Reviewed by Eric Siedel.
3603
3604         As a first step in cleaning up the version parsing logic in
3605         webkitpy.layout_tests.port, this adds common routines for
3606         getting the os_name and os_version fields to the PlatformInfo
3607         class. Nothing uses them yet but I've added FIXME's to some of
3608         the code that needs to be deleted in the port files.
3609
3610         * Scripts/webkitpy/common/system/platforminfo.py:
3611         (PlatformInfo.__init__):
3612         (PlatformInfo):
3613         (PlatformInfo._determine_os_name):
3614         (PlatformInfo._determine_mac_version):
3615         (PlatformInfo._determine_linux_version):
3616         (PlatformInfo._determine_win_version):
3617         (PlatformInfo.display_name):
3618         (PlatformInfo.total_bytes_memory):
3619         (PlatformInfo._compute_bytes_from_vm_stat_output):
3620         (PlatformInfo.free_bytes_memory):
3621         * Scripts/webkitpy/common/system/platforminfo_unittest.py: Added.
3622         (TestPlatformInfo):
3623         (TestPlatformInfo.make_info):
3624         (TestPlatformInfo.make_info.mock_run_command):
3625         (TestPlatformInfo.setUp):
3626         (TestPlatformInfo.tearDown):
3627         (TestPlatformInfo.test_basic):
3628         (TestPlatformInfo.integration_test_basic):
3629         (TestPlatformInfo.test_display_name_mac):
3630         (TestPlatformInfo.test_display_name_win32):
3631         (TestPlatformInfo.test_memory_mac):
3632         (TestPlatformInfo.test_memory_win32):
3633         (TestPlatformInfo.test_determine_os_name):
3634         (TestPlatformInfo.test_determine_mac_version):
3635         (TestPlatformInfo.test_determine_linux_version):
3636         (TestPlatformInfo.test_determine_win_version_from_tuple):
3637         (TestPlatformInfo.test_determine_win_version_from_cmd):
3638         * Scripts/webkitpy/layout_tests/port/apple.py:
3639         (ApplePort.__init__):
3640         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
3641         (ChromiumMacPort.__init__):
3642         * Scripts/webkitpy/layout_tests/port/mac.py:
3643         (MacPort):
3644         * Scripts/webkitpy/layout_tests/port/win.py:
3645         (WinPort):
3646
3647 2012-01-10  Adam Roben  <aroben@apple.com>
3648
3649         Make it possible to type data: URLs into MiniBrowser on Windows
3650
3651         Fixes <http://webkit.org/b/75084> Crash when trying to navigate to a data: URL in
3652         MiniBrowser on Windows
3653
3654         Reviewed by Darin Adler.
3655
3656         * MiniBrowser/win/BrowserWindow.cpp:
3657         (BrowserWindow::handleMessage): Instead of checking for an "http://" prefix when deciding
3658         whether to prepend "http://", just look for whether the typed URL contains a colon. If it
3659         does, we assume it already has a scheme and don't modify it.
3660
3661 2012-01-10  Adam Roben  <aroben@apple.com>
3662
3663         Make MiniBrowser automatically escape invalid URL characters typed in its URL field
3664
3665         Fixes <http://webkit.org/b/75086> Crash when typing a data: URL containing double-quotes in
3666         MiniBrowser on Windows
3667
3668         Reviewed by Darin Adler.
3669
3670         * MiniBrowser/win/BrowserView.cpp:
3671         (BrowserView::goToURL): Pass the typed string through
3672         CFURLCreateStringByAddingPercentEscapes before trying to create a URL from it.
3673
3674 2012-01-10  Ben Wells  <benwells@chromium.org>
3675
3676         Adding myself (benwells) to committers.py
3677         https://bugs.webkit.org/show_bug.cgi?id=75971
3678
3679         Reviewed by Ojan Vafai.
3680
3681         * Scripts/webkitpy/common/config/committers.py:
3682
3683 2012-01-10  Mario Sanchez Prada  <msanchez@igalia.com>
3684
3685         AX: support helpText() in DumpRenderTree
3686         https://bugs.webkit.org/show_bug.cgi?id=40193
3687
3688         Reviewed by Chris Fleizach.