Source/WebKit2: Add API to get the parent frame in WKBundleFrameRef
[WebKit-https.git] / Tools / ChangeLog
1 2012-01-27  Ada Chan  <adachan@apple.com>
2
3         Add test for WKBundleFrameGetParentFrame().
4         https://bugs.webkit.org/show_bug.cgi?id=77161
5
6         Reviewed by Anders Carlsson.
7
8         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Add ParentFrame.cpp and ParentFrame_Bundle.cpp.
9         * TestWebKitAPI/Tests/WebKit2/ParentFrame.cpp: Added.
10         (TestWebKitAPI):
11         (TestWebKitAPI::didReceiveMessageFromInjectedBundle): Get the check result from the injected bundle.
12         (TestWebKitAPI::setInjectedBundleClient):
13         (TestWebKitAPI::TEST): Load simple-iframe.html and then wait for the injected bundle to post result.
14         Make sure the check is successful.
15         * TestWebKitAPI/Tests/WebKit2/ParentFrame_Bundle.cpp: Added.
16         (TestWebKitAPI):
17         (ParentFrameTest):
18         (TestWebKitAPI::ParentFrameTest::ParentFrameTest):
19         (TestWebKitAPI::didFinishLoadForFrame): If the frame is a subframe, store it off for checking later.
20         If the frame is the main frame, check whether it's indeed the subframe's parent frame and post the result
21         to the test controller.
22         (TestWebKitAPI::ParentFrameTest::didCreatePage): Set the page loader client on this page.
23
24 2012-01-27  Zan Dobersek  <zandobersek@gmail.com>
25
26         [Gtk] DumpRenderTree lacks --no-timeout command line option
27         https://bugs.webkit.org/show_bug.cgi?id=77196
28
29         Reviewed by Martin Robinson.
30
31         Add the --no-timeout command line option and reorganize
32         watchdog code to make running DumpRenderTree without
33         timeout possible.
34
35         * DumpRenderTree/gtk/DumpRenderTree.cpp:
36         (setWaitToDumpWatchdog):
37         (shouldSetWaitToDumpWatchdog):
38         (initializeGlobalsFromCommandLineOptions):
39         * DumpRenderTree/gtk/DumpRenderTreeGtk.h:
40         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
41         (waitToDumpWatchdogFired):
42         (LayoutTestController::setWaitToDump):
43
44 2012-01-27  Andras Becsi  <andras.becsi@nokia.com>
45
46         [Qt][WK2] Enable the automatic touch->mouse mocking on the new Qt5 hashes
47         https://bugs.webkit.org/show_bug.cgi?id=77122
48
49         Reviewed by Kenneth Rohde Christiansen.
50
51         * MiniBrowser/qt/MiniBrowserApplication.cpp:
52         (MiniBrowserApplication::MiniBrowserApplication):
53
54 2012-01-27  Zeno Albisser  <zeno@webkit.org>
55
56         [Qt][Mac] Build fails after adding ICU support (r105997).
57         https://bugs.webkit.org/show_bug.cgi?id=77118
58
59         Define WTF_USE_ICU_UNICODE if we are on mac.
60
61         Reviewed by Tor Arne Vestbø.
62
63         * qmake/mkspecs/features/features.prf:
64
65 2012-01-27  Roland Takacs  <takacs.roland@stud.u-szeged.hu>
66
67         Update committers.py with new contributors.
68         https://bugs.webkit.org/show_bug.cgi?id=77207
69
70         I added six new contributors into committers.py:
71
72         - Adam Kallai
73         - Balazs Ankes
74         - Janos Badics
75         - Nandor Huszka
76         - Roland Takacs
77         - Szilard Ledan-Muntean
78
79         * Scripts/webkitpy/common/config/committers.py:
80
81 2012-01-27  Vsevolod Vlasov  <vsevik@chromium.org>
82
83         Add Vsevolod Vlasov to reviewers list.
84
85         Reviewed by Pavel Feldman.
86
87         * Scripts/webkitpy/common/config/committers.py:
88
89 2012-01-26  No'am Rosenthal  <noam.rosenthal@nokia.com> and Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
90
91         [Qt] WKTR: Use a software rendering pipiline when running tests.
92         https://bugs.webkit.org/show_bug.cgi?id=76708
93
94         Reviewed by Kenneth Rohde Christiansen.
95
96         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
97         (WTR::WrapperWindow::WrapperWindow):
98         (WTR::PlatformWebView::PlatformWebView): Use software rendering of layers since the wrapping QQuickView isn't shown.
99
100 2012-01-26  Kevin Ollivier  <kevino@theolliviers.com>
101
102         [wx] Unreviewed. Build fix, add WebCore/testing/js
103         to the list of source directories we build.
104
105         * waf/build/settings.py:
106
107 2012-01-26  Filip Pizlo  <fpizlo@apple.com>
108
109         Tools/Scripts/commit-log-editor is broken due to $_ getting clobbered
110         https://bugs.webkit.org/show_bug.cgi?id=77177
111
112         Reviewed by Jon Honeycutt.
113
114         * Scripts/commit-log-editor:
115
116 2012-01-26  Ojan Vafai  <ojan@chromium.org>
117
118         Decrease sleep time when killing server_process on Mac.
119         https://bugs.webkit.org/show_bug.cgi?id=77169
120
121         Reviewed by Eric Seidel.
122
123         This shaves another ~70ms off single test runtime.
124
125         * Scripts/webkitpy/layout_tests/port/server_process.py:
126         (ServerProcess.stop):
127
128 2012-01-26  Ojan Vafai  <ojan@chromium.org>
129
130         Remove builder name so http://trac.webkit.org/changeset/106035 actually works.
131         https://bugs.webkit.org/show_bug.cgi?id=77164
132
133         Reviewed by Dirk Pranke.
134
135         http://trac.webkit.org/changeset/106035 depends on builder_name being
136         falsey by default. There's other code that requires it to be iterable
137         by default. So, make it be the empty string.
138
139         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
140         (parse_args):
141
142 2012-01-26  Ryosuke Niwa  <rniwa@webkit.org>
143
144         run-perf-test should include relative file paths in test names
145         https://bugs.webkit.org/show_bug.cgi?id=77144
146
147         Reviewed by Adam Birth.
148
149         Include all directory names in the test name separated by "/".
150         e.g. Bindings/event-target-wrapper.html translates to Bindings/event-target-wrapper
151
152         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
153         (PerfTestsRunner):
154         (PerfTestsRunner._process_parser_test_result):
155         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
156         (test_run_test_set_with_json_output):
157         (test_run_test_set_with_json_source):
158
159 2012-01-26  Dirk Pranke  <dpranke@chromium.org>
160
161         webkitpy: re-land cleanup of test scaffolding
162         https://bugs.webkit.org/show_bug.cgi?id=77153
163
164         Reviewed by Eric Seidel.
165
166         This is an attempt to reland the change in bug 76973 / 
167         r105935. My earlier attempt to manipulate PYTHONPATH
168         seems to only work some of the time, for reasons that
169         elude me. Directly adding Tools/Scripts to sys.path from
170         inside the test script appears to work (more?) reliably.
171
172         Also, it seemed like the files didn't actually get deleted
173         the last time; I don't know why.
174
175         * Scripts/test-webkitpy:
176         * Scripts/webkitpy/common/system/executive_unittest.py:
177         (command_line):
178         (ExecutiveTest):
179         (ExecutiveTest.test_run_command_args_type):
180         (ExecutiveTest.test_run_command_with_unicode):
181         (ExecutiveTest.test_running_pids):
182         (main):
183         * Scripts/webkitpy/common/system/fileutils.py: Removed.
184         * Scripts/webkitpy/test/cat.py: Removed.
185         * Scripts/webkitpy/test/cat_unittest.py: Removed.
186         * Scripts/webkitpy/test/echo.py: Removed.
187         * Scripts/webkitpy/test/echo_unittest.py: Removed.
188
189 2012-01-26  Ojan Vafai  <ojan@chromium.org>
190
191         run-webkit-tests calls "nm" when it doesn't need to
192         https://bugs.webkit.org/show_bug.cgi?id=77143
193
194         Reviewed by Dirk Pranke.
195
196         If we are not running any test that can be skipped by checking
197         the DRT feature list or looking at the list of symbols, avoid calling
198         out to DRT or calling "nm".
199
200         This saves ~700ms on my Mac Pro. In cases where you still have to call out
201         to "nm" it's only a couple ms slower.
202
203         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
204         (TestExpectations.__init__):
205         * Scripts/webkitpy/layout_tests/port/base.py:
206         (Port.skipped_tests):
207         * Scripts/webkitpy/layout_tests/port/webkit.py:
208         (WebKitPort._should_check_for_missing_features):
209         (WebKitPort):
210         (WebKitPort._should_check_for_missing_symbols):
211         (WebKitPort._skipped_tests_for_unsupported_features):
212         (WebKitPort.skipped_layout_tests):
213         (WebKitPort.skipped_tests):
214         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
215         (WebKitPortTest.test_skipped_directories_for_symbols):
216         (test_skipped_directories_for_features):
217         (test_skipped_directories_for_features_no_matching_tests_in_test_list):
218         (test_skipped_layout_tests):
219
220 2012-01-26  Benjamin Poulain  <bpoulain@apple.com>
221
222         Support a suffix on ChangeLog filenames based on a configuration file
223         https://bugs.webkit.org/show_bug.cgi?id=76956
224
225         Reviewed by Kentaro Hara.
226
227         Modify the tools used to generate, edit and resolve ChangeLogs to support
228         the addition of a suffix to the ChangeLog filename.
229
230         This patch makes it easier to support alternate ChangeLog files in branches.
231
232         * Scripts/VCSUtils.pm:
233         (changeLogSuffix):
234         (changeLogFileName):
235         * Scripts/commit-log-editor:
236         * Scripts/prepare-ChangeLog:
237         (getLatestChangeLogs):
238         (generateNewChangeLogs):
239         (generateFileList):
240         * Scripts/resolve-ChangeLogs:
241         (findChangeLog):
242
243 2012-01-26  Sheriff Bot  <webkit.review.bot@gmail.com>
244
245         Unreviewed, rolling out r106038.
246         http://trac.webkit.org/changeset/106038
247         https://bugs.webkit.org/show_bug.cgi?id=77142
248
249         Caused a bunch of skipped tests to not be skipped. (Requested
250         by ojan on #webkit).
251
252         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
253         (TestExpectationParser):
254         (TestExpectationParser.parse):
255         (TestExpectationParser._parse_line):
256         (TestExpectationParser._collect_matching_tests):
257         (TestExpectations.__init__):
258         (TestExpectations._add_skipped_tests):
259         * Scripts/webkitpy/layout_tests/port/base.py:
260         (Port.test_dirs):
261         (Port.normalize_test_name):
262         (Port.update_baseline):
263         (Port.layout_tests_dir):
264         (Port.relative_perf_test_filename):
265         (Port.abspath_for_test):
266
267 2012-01-25  Ojan Vafai  <ojan@chromium.org>
268
269         Parsing test_expecations.txt + Skipped lists takes too long
270         https://bugs.webkit.org/show_bug.cgi?id=77059
271
272         Reviewed by Dirk Pranke.
273
274         This saves ~100ms on the Apple Mac port.
275         -memoize a bunch of path methods.
276         -Avoid doing multiple disk accesses per line.
277         -Parse the skipped list directly instead of turning it into a test_expecations.txt
278         formatting string and parsing that.
279
280         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
281         (TestExpectationParser):
282         (TestExpectationParser.parse_skipped_tests):
283         (TestExpectationParser._parse_line):
284         (TestExpectationParser._collect_matching_tests):
285         (TestExpectations.__init__):
286         (TestExpectations._add_skipped_tests):
287         * Scripts/webkitpy/layout_tests/port/base.py:
288         (Port):
289         (Port.test_isfile):
290         (Port.normalize_test_name):
291         (Port.layout_tests_dir):
292         (Port.abspath_for_test):
293
294 2012-01-25  Ojan Vafai  <ojan@chromium.org>
295
296         Only store the SVN revision in the summarized results if we're on a builder
297         https://bugs.webkit.org/show_bug.cgi?id=76976
298
299         Reviewed by Dirk Pranke.
300
301         This shaves another ~130ms off the single test run-webkit-tests runtime.
302
303         * Scripts/webkitpy/layout_tests/controllers/manager.py:
304         (summarize_results):
305         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
306         (ResultSummaryTest.get_result):
307         (ResultSummaryTest):
308         (ResultSummaryTest.get_result_summary):
309         (ResultSummaryTest.get_unexpected_results):
310         (ResultSummaryTest.test_no_svn_revision):
311         (ResultSummaryTest.test_svn_revision):
312         * Scripts/webkitpy/layout_tests/port/base.py:
313         (Port.use_trac_links_in_results_html):
314
315 2012-01-26  Luciano Wolf  <luciano.wolf@openbossa.org>
316
317         [Qt][WK2] fast/loader tests failed after r94178
318         https://bugs.webkit.org/show_bug.cgi?id=67388
319
320         Reviewed by Simon Hausmann.
321
322         Applying the same fix used for Qt-WebKit1 from bug #67254 -
323         [Qt][DRT] Normalize file:///tmp/LayoutTests in LayoutTestController::pathToLocalResource()
324         https://bugs.webkit.org/show_bug.cgi?id=67254
325         Translate file:///tmp/LayoutTests/* urls into the repository LayoutTests directory, which is
326         derived from the running location of DumpRenderTree binary.
327
328         * WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
329         (WTR::LayoutTestController::pathToLocalResource):
330
331 2012-01-26  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
332
333         Implement overridePreference for boolean preferences in WebKitTestRunner
334         https://bugs.webkit.org/show_bug.cgi?id=77033
335
336         Reviewed by Alexey Proskuryakov.
337
338         For now overridePreference supports boolean values.
339
340         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
341         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
342         (WTR::LayoutTestController::setXSSAuditorEnabled):
343         (WTR::LayoutTestController::overridePreference):
344         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
345         * WebKitTestRunner/TestController.cpp:
346         (WTR::TestController::resetStateToConsistentValues): Reset the overrides
347         explicitly and properly disable the use of PageCache when resetting the
348         state. Tests that use PageCache can now enable it via overridePreference.
349
350 2012-01-26  Alexander Færøy  <alexander.faeroy@nokia.com>
351
352         [Qt] Support opening multiple URL's from the command line in the Qt MiniBrowser
353         https://bugs.webkit.org/show_bug.cgi?id=77100
354
355         Reviewed by Simon Hausmann.
356
357         This patch makes the MiniBrowser read multiple URL's from the command
358         line such that there will be one browser window opened for each URL.
359
360         * MiniBrowser/qt/MiniBrowserApplication.cpp:
361         (MiniBrowserApplication::handleUserOptions):
362
363 2012-01-26  Alexander Færøy  <alexander.faeroy@nokia.com>
364
365         [Qt] The Qt MiniBrowser crashes when given multiple URL's
366         https://bugs.webkit.org/show_bug.cgi?id=77009
367
368         Reviewed by Simon Hausmann.
369
370         This patch fixes a crash in BrowserWindow::newWindow() where a
371         BrowserWindow-object is constructed with a 0 pointer as the
372         WindowOptions parameter which leads to a crash.
373
374         * MiniBrowser/qt/BrowserWindow.cpp:
375         (BrowserWindow::BrowserWindow):
376         (BrowserWindow::newWindow):
377         * MiniBrowser/qt/BrowserWindow.h:
378
379 2012-01-26  Jesus Sanchez-Palencia  <jesus.palencia@openbossa.org>
380
381         [Qt] Use ICU if available
382         https://bugs.webkit.org/show_bug.cgi?id=76821
383
384         Reviewed by Simon Hausmann.
385
386         Adding unicode defines for a Qt 5 based build.
387
388         * qmake/mkspecs/features/features.prf:
389
390 2012-01-26  Nándor Huszka  <huszka.nandor@stud.u-szeged.hu>
391
392         [WK2] WebKitTestRunner needs layoutTestController.setGeolocationPermission
393         https://bugs.webkit.org/show_bug.cgi?id=42545
394
395         Reviewed by Zoltan Herczeg.
396
397         Implement the setGeolocationPermission method.
398
399         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
400         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
401         (WTR::LayoutTestController::setGeolocationPermission):
402         (WTR):
403         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
404         (LayoutTestController):
405
406 2012-01-26  Balazs Kelemen  <kbalazs@webkit.org>
407
408         [NRWT] Support --ignore-metrics
409         https://bugs.webkit.org/show_bug.cgi?id=76278
410
411         Reviewed by Tony Chang.
412
413         Port the feature from ORWT.
414
415         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
416         (SingleTestRunner._run_compare_test):
417         * Scripts/webkitpy/layout_tests/port/driver.py:
418         (DriverOutput):
419         (DriverOutput.strip_metrics):
420         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
421         (DriverOutputTest):
422         (DriverOutputTest.test_strip_metrics):
423         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
424         (_set_up_derived_options):
425         (parse_args):
426
427 2012-01-26  Sheriff Bot  <webkit.review.bot@gmail.com>
428
429         Unreviewed, rolling out r105935 and r105954.
430         http://trac.webkit.org/changeset/105935
431         http://trac.webkit.org/changeset/105954
432         https://bugs.webkit.org/show_bug.cgi?id=77080
433
434         test-webkitpy is still broken (Requested by abarth on
435         #webkit).
436
437         * Scripts/test-webkitpy:
438         * Scripts/webkitpy/common/system/executive_unittest.py:
439         (never_ending_command):
440         (ExecutiveTest.test_run_command_args_type):
441         (ExecutiveTest.test_run_command_with_unicode):
442         (ExecutiveTest.test_running_pids):
443         * Scripts/webkitpy/common/system/fileutils.py:
444         (make_stdout_binary):
445         * Scripts/webkitpy/test/cat.py:
446         (command_arguments):
447         (main):
448         * Scripts/webkitpy/test/cat_unittest.py:
449         (CatTest):
450         (CatTest.assert_cat):
451         (CatTest.test_basic):
452         (CatTest.test_no_newline):
453         (CatTest.test_unicode):
454         (CatTest.test_as_command):
455         * Scripts/webkitpy/test/echo.py:
456         (command_arguments):
457         (main):
458         * Scripts/webkitpy/test/echo_unittest.py:
459         (EchoTest):
460         (EchoTest.test_basic):
461         (EchoTest.test_no_newline):
462         (EchoTest.test_unicode):
463         (EchoTest.test_argument_order):
464         (EchoTest.test_empty_arguments):
465         (EchoTest.test_no_arguments):
466         (EchoTest.test_as_command):
467
468 2012-01-26  Gabor Rapcsanyi  <rgabor@webkit.org>
469
470         [GTK] Detect cross-compilation in webkitdirs.pm to not generate gtkdoc
471         https://bugs.webkit.org/show_bug.cgi?id=76138
472
473         Reviewed by Martin Robinson.
474
475         * Scripts/webkitdirs.pm:
476         (isCrossCompilation):
477         (buildAutotoolsProject):
478
479 2012-01-25  Adam Barth  <abarth@webkit.org>
480
481         Kentaro is a reviewer now.  Yay!
482
483         * Scripts/webkitpy/common/config/committers.py:
484
485 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
486
487         nrwt: running w/ no arguments fails
488         https://bugs.webkit.org/show_bug.cgi?id=77064
489
490         Unreviewed, build fix.
491
492         This change fixes a regression introduced in r105944
493         where running without --platform <foo> or one of the other
494         platform switches would cause a null deref :(.
495
496         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
497         (main):
498
499 2012-01-25  Eric Seidel  <eric@webkit.org>
500
501         Automatically CC someone from Apple on any changes to Obj-C public API
502         https://bugs.webkit.org/show_bug.cgi?id=77053
503
504         Reviewed by Timothy Hatcher.
505
506         This has long been de-facto practice, seems now that we have
507         fancy watchlist support, we should just make the tools
508         do this automatically.
509
510         * Scripts/webkitpy/common/config/watchlist:
511
512 2012-01-25  Antoine Labour  <piman@chromium.org>
513
514         Updating status in committers.py.
515
516         Unreviewed.
517
518         * Scripts/webkitpy/common/config/committers.py:
519
520 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
521
522         test-webkitpy: executive_unittest fails if PYTHONPATH isn't set
523         https://bugs.webkit.org/show_bug.cgi?id=77058
524
525         Reviewed by Eric Seidel.
526
527         r105935 broke executive_unittest by changing which script we
528         were running and not ensuring that the PYTHONPATH was set up
529         properly.
530
531         This change also fixes a regression introduced recently where
532         test-webkitpy wasn't figuring out the path to Tools/Scripts
533         right, either (so running ./test-webkitpy would fail).
534
535         * Scripts/test-webkitpy:
536         * Scripts/webkitpy/common/system/executive_unittest.py:
537         (script_dir):
538         (ExecutiveTest.setUp):
539         (ExecutiveTest):
540         (ExecutiveTest.tearDown):
541
542 2012-01-25  Adam Barth  <abarth@webkit.org>
543
544         Disable tests on the mac-ews while we sort out hte stability issues.
545
546         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
547         (MacEWS):
548
549 2012-01-23  Dirk Pranke  <dpranke@chromium.org>
550
551         investigate use of 'mac' and 'win' as fully-specified port names for the apple ports
552         https://bugs.webkit.org/show_bug.cgi?id=76475
553
554         Reviewed by Eric Seidel.
555
556         This patch cleans up the internal usage of port names for the
557         Apple mac and Apple win ports so that 'mac' and 'win are never
558         considered "fully specified" port names: they are still legal
559         input to the PortFactory.get() routine, but only if run on the given
560         platform (i.e. trying to PortFactory.get('mac') on windows or
561         linux will return an error, since we don't know which version of
562         the port is desired.
563
564         This also cleans up handling of the webkit2 variants, so that
565         the full port names consistently follow the
566         <operating_system>-<version>[-<wk2>] convention.
567
568         Lastly this patch adds some assertions and removes some
569         try/catch logic to catch more programming and usage errors;
570         previously mac-tiger might've been allowed -- and it would've
571         translated into using just the baselines in platform/mac -- but
572         now it should fail.
573
574         * Scripts/webkitpy/layout_tests/port/apple.py:
575         (ApplePort.determine_full_port_name):
576         (ApplePort.__init__):
577         (ApplePort):
578         (ApplePort._port_name_with_version):
579         * Scripts/webkitpy/layout_tests/port/builders.py:
580         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
581         (FactoryTest.test_mac):
582         (FactoryTest.test_win):
583         * Scripts/webkitpy/layout_tests/port/mac.py:
584         (MacPort.baseline_search_path):
585         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
586         (MacTest.test_versions):
587         * Scripts/webkitpy/layout_tests/port/win.py:
588         (WinPort.baseline_search_path):
589         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
590         (WinTest.test_versions):
591
592 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
593
594         webkitpy: print nicer errors while linting expectations files, remove redundant tests
595         https://bugs.webkit.org/show_bug.cgi?id=76955
596
597         Reviewed by Adam Barth.
598
599         This change cleans up the errors that are printed so that the
600         are more "quickfix" compatible (path:lineno).
601
602         Also, there were a bunch of redundant tests between
603         layout_tests.models.test_expectations_unittest and
604         style.checkers.test_expectations_unittest. The latter should
605         just expect the former to do most of the testing.
606
607         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
608         (TestExpectationParser._parse_modifiers):
609         (TestExpectations._report_errors):
610         (TestExpectations._add_skipped_tests):
611         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
612         (test_parse_error_nonfatal):
613         (SemanticTests.test_bad_bugid):
614         * Scripts/webkitpy/style/checkers/test_expectations.py:
615         (TestExpectationsChecker.__init__):
616         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
617         (TestExpectationsTestCase.assert_lines_lint):
618         (TestExpectationsTestCase.test_valid_expectations):
619         (TestExpectationsTestCase.test_invalid_expectations):
620         (TestExpectationsTestCase.test_tab):
621
622 2012-01-25  Eric Seidel  <eric@webkit.org>
623
624         webkit-patch apply-* should always continue after failures
625         https://bugs.webkit.org/show_bug.cgi?id=77057
626
627         Reviewed by Adam Barth.
628
629         As far as I can tell there is only one potential drawback to always
630         forcing: that if you're somehow in the wrong directory it will create new
631         directories for new files.  Since webkit-patch always cd's to the root
632         it seems that's not a drawback.  The drawback of not using --force for
633         svn-apply, is that it will stop after the first failure, which is rarely
634         (if ever) the desired behavior.  This just removes the force option
635         (which was strangely hidden behind --non-interactive).  This should
636         make for a better user experiance.
637
638         * Scripts/webkitpy/common/checkout/checkout.py:
639         (Checkout.apply_patch):
640         * Scripts/webkitpy/common/checkout/checkout_mock.py:
641         (MockCheckout.apply_patch):
642         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
643         (CheckoutTest.test_chromium_deps):
644         (CheckoutTest):
645         (CheckoutTest.test_apply_patch):
646         * Scripts/webkitpy/tool/commands/download_unittest.py:
647         (DownloadCommandsTest._default_options):
648         * Scripts/webkitpy/tool/steps/applypatch.py:
649         (ApplyPatch.options):
650         (ApplyPatch.run):
651         * Scripts/webkitpy/tool/steps/options.py:
652         (Options):
653
654 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
655
656         nrwt: should be able to run --platform test interactively
657         https://bugs.webkit.org/show_bug.cgi?id=76959
658
659         Reviewed by Adam Barth.
660
661         As part of refactoring the Port interfaces, in r103254 I made
662         passing a Host object to the Port mandatory; previously TestPort
663         objects would create their own (mock) hosts. However,
664         new-run-webkit-tests always passed a real Host, not a MockHost,
665         and so you couldn't run new-run-webkit-tests --platform test
666         interactively to debug test failures.
667
668         This change fixes that by creating a MockHost instead of a real
669         host if you say --platform test*.
670
671         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
672         (main):
673
674 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
675
676         run-webkit-tests --lint-test-files should lint all the ports by default
677         https://bugs.webkit.org/show_bug.cgi?id=76749
678
679         Reviewed by Ojan Vafai.
680
681         Currently run-webkit-tests --lint-test-files will only lint the
682         test_expectations for the single port determined by the command
683         line options. This is not obvious and can produce unintended results
684         if you want to lint the chromium port (probably the normal case
685         for using --lint-test-files) but don't specify a port name or
686         --chromium.
687
688         I think we should probably lint *all* of the ports if no port
689         name is supplied. This change implements that, and also moves
690         the linting logic out of the Manager class, since this really
691         has nothing to do with Managers. It has little to do with
692         run-webkit-tests, but splitting it into a different script would
693         introduce compatibility issues. It might make sense to do that
694         anyway, in a separate patch and combining that with cleaning up
695         the style checker to share more code.
696
697         * Scripts/webkitpy/layout_tests/controllers/manager.py:
698         (Manager._strip_test_dir_prefix):
699         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
700         (lint):
701         (run):
702         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
703         (LintTest.test_all_configurations): added.
704         (LintTest.test_lint_test_files): added.
705         (LintTest.test_lint_test_files__errors): added.
706         (MainTest.test_lint_test_files): removed.
707         (MainTest.test_lint_test_files__errors): removed.
708
709 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
710
711         webkitpy: clean up a bunch of test scaffolding
712         https://bugs.webkit.org/show_bug.cgi?id=76973
713
714         Reviewed by Eric Seidel.
715
716         There were several helper modules in the webkitpy.test package
717         that appear to only be used by executive_unittest.py. I am
718         rolling them all into that file to make that clearer.
719
720         * Scripts/webkitpy/common/system/executive_unittest.py:
721         (ExecutiveTest.test_run_command_args_type):
722         (ExecutiveTest.test_run_command_with_unicode):
723         (ExecutiveTest.test_running_pids):
724         (command_line):
725         (main):
726         * Scripts/webkitpy/common/system/fileutils.py: Removed.
727         * Scripts/webkitpy/test/cat.py: Removed.
728         * Scripts/webkitpy/test/cat_unittest.py: Removed.
729         * Scripts/webkitpy/test/echo.py: Removed.
730         * Scripts/webkitpy/test/echo_unittest.py: Removed.
731
732 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
733
734         webkitpy: clean up a bunch of test scaffolding
735         https://bugs.webkit.org/show_bug.cgi?id=76973
736
737         Reviewed by Eric Seidel.
738
739         There were several helper modules in the webkitpy.test package
740         that appear to only be used by executive_unittest.py. I am
741         rolling them all into that file to make that clearer.
742
743         * Scripts/webkitpy/common/system/executive_unittest.py:
744         (ExecutiveTest.test_run_command_args_type):
745         (ExecutiveTest.test_run_command_with_unicode):
746         (ExecutiveTest.test_running_pids):
747         (command_line):
748         (main):
749         * Scripts/webkitpy/common/system/fileutils.py: Removed.
750         * Scripts/webkitpy/test/cat.py: Removed.
751         * Scripts/webkitpy/test/cat_unittest.py: Removed.
752         * Scripts/webkitpy/test/echo.py: Removed.
753         * Scripts/webkitpy/test/echo_unittest.py: Removed.
754
755 2012-01-25  Eric Seidel  <eric@webkit.org>
756
757         Remove Python 2.5 support from WebKit
758         https://bugs.webkit.org/show_bug.cgi?id=71593
759
760         Reviewed by Adam Barth.
761
762         Just removing support from this one core file.
763         I'll commit the rest of the (large) change once we've
764         gone 24 hours and seen that no bots have exploded.
765         We can roll out this one change if we break bots
766         and then roll in the rest once those bots are fixed.
767
768         * Scripts/webkitpy/common/system/executive.py:
769         (Executive.cpu_count):
770
771 2012-01-25  Hajime Morita  <morrita@google.com>
772
773         ENABLE_SHADOW_DOM should be available via build-webkit --shadow-dom
774         https://bugs.webkit.org/show_bug.cgi?id=76863
775
776         Reviewed by Dimitri Glazkov.
777
778         Added a feature flag.
779
780         * Scripts/build-webkit:
781         * qmake/mkspecs/features/features.prf:
782
783 2012-01-25  Martin Robinson  <mrobinson@igalia.com>
784
785         [WK2] Navigation events not triggered by a mouse event claim to be triggered by the left mouse button during policy decisions
786         https://bugs.webkit.org/show_bug.cgi?id=76344
787
788         Reviewed by Sam Weinig.
789
790         Add a test verifying that the mouse button during policy decisions for
791         events not triggered by mouse events is NoButton.
792
793         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
794         (TestWebKitAPI::decidePolicyForNavigationAction): Added test.
795         (TestWebKitAPI::decidePolicyForNewWindowAction): Added test.
796
797 2012-01-25  Ryosuke Niwa  <rniwa@webkit.org>
798
799         run-perf-tests should timeout sooner
800         https://bugs.webkit.org/show_bug.cgi?id=77026
801
802         Reviewed by Adam Barth.
803
804         Bump off the timeout from 10 minutes to 4 minutes.
805
806         Also use float instead of str in test result values to compress the size of json files bots upload.
807
808         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
809         (PerfTestsRunner._parse_args):
810         (PerfTestsRunner._process_chromium_style_test_result):
811         (PerfTestsRunner._process_parser_test_result):
812         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
813         (test_run_test_set_for_parser_tests):
814         (test_run_test_set_with_json_output):
815         (test_run_test_set_with_json_source):
816
817 2012-01-25  Enrica Casucci  <enrica@apple.com>
818
819         WebView should implement typingAttributes methods to work correctly with the Inspector bar in Mail.
820         https://bugs.webkit.org/show_bug.cgi?id=76951
821         <rdar://problem/9325158>
822
823         Reviewed by Alexey Proskuryakov.
824
825         * TestWebKitAPI/Tests/mac/InspectorBar.mm: Modified to reflect the changes to WebView
826         and WebHTMLView.
827         (TestWebKitAPI::TEST):
828
829 2012-01-23  MORITA Hajime  <morrita@google.com>
830
831         [Refactoring][Internals] Should have InternalSettings
832         https://bugs.webkit.org/show_bug.cgi?id=76424
833
834         Reviewed by Kent Tamura.
835
836         * GNUmakefile.am: Added so new files.
837
838 2012-01-25  Takashi Toyoshima  <toyoshim@chromium.org>
839
840         webkitpy: new-run-webkit-websocket is broken
841         https://bugs.webkit.org/show_bug.cgi?id=76999
842
843         Reviewed by Dirk Pranke.
844
845         r105674 broke new-run-webkit-websocket, since options.chromium
846         doesn't exist any more. new-run-webkit-httpd was also broken
847         and already fixed at r105835.
848         * Scripts/new-run-webkit-websocketserver:
849         (main):
850
851 2012-01-25  Nándor Huszka  <huszka.nandor@stud.u-szeged.hu>
852
853         [WK2] WebKitTestRunner needs layoutTestController.setPopupBlockingEnabled
854         https://bugs.webkit.org/show_bug.cgi?id=63458
855
856         Reviewed by Adam Roben.
857
858         Implement the setPopupBlockingEnabled method.
859
860         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
861         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
862         (WTR::LayoutTestController::setPopupBlockingEnabled):
863         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
864
865 2012-01-24  Balazs Kelemen  <kbalazs@webkit.org>
866
867         [Qt] Cleanup testfonts handling
868         https://bugs.webkit.org/show_bug.cgi?id=76068
869
870         Reviewed by Simon Hausmann.
871
872         Factor test font initialization into it's own file  so test
873         font users can share the implementation by adding the file
874         to their build.
875         Use fonts.conf from the test font directory to not require the
876         current directory to be the WebKit root.
877         Fix the -use-test-fonts option of QtTestBrowser by initializing
878         the fonts earlier, before it is too late (namely before calling
879         show on the main window).
880
881         * DumpRenderTree/qt/DumpRenderTree.pro:
882         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
883         (WebCore::DumpRenderTree::open):
884         * DumpRenderTree/qt/DumpRenderTreeQt.h:
885         * DumpRenderTree/qt/QtInitializeTestFonts.cpp: Copied from Tools/WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp.
886         (WebKit::initializeTestFonts):
887         * DumpRenderTree/qt/QtInitializeTestFonts.h: Added.
888         * DumpRenderTree/qt/fonts.conf: Removed. No need for this since
889         we will use the one from the test fonts directory.
890         * DumpRenderTree/qt/main.cpp:
891         (main): Fix the ifdefs to make it clear that the code to
892         set up the Dpi has not been ported to Qt5 yet.
893         * QtTestBrowser/QtTestBrowser.pro:
894         * QtTestBrowser/launcherwindow.h:
895         (WindowOptions::WindowOptions):
896         * QtTestBrowser/main.cpp:
897         (launcherMain):
898         (LauncherApplication::handleUserOptions):
899         * WebKitTestRunner/InjectedBundle/Target.pri:
900         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
901         (WTR::activateFonts): Remove the code to set up the Dpi
902         because it does not work with Qt5.
903
904 2012-01-25  Simon Hausmann  <simon.hausmann@nokia.com>
905
906         [Qt] build-webkit --qt requires a git-svn or SVN checkout.
907
908         Reviewed by Tor Arne Vestø.
909
910         Don't require it by making svnRevisionForDirectory return "unknown"
911         instead of bailing out with an error.
912
913         * Scripts/VCSUtils.pm:
914         (svnRevisionForDirectory):
915
916 2012-01-25  Adam Barth  <abarth@webkit.org>
917
918         Temporarily disable running the unit tests on the Chromium EWS.  These
919         tests don't work properly (see
920         https://bugs.webkit.org/show_bug.cgi?id=76979) and are currently
921         blocking the EWS.
922
923         * Scripts/webkitpy/common/config/ports.py:
924         (ChromiumPort):
925
926 2012-01-25  Adam Barth  <abarth@webkit.org>
927
928         Update results.html version of PixelZoomer
929         https://bugs.webkit.org/show_bug.cgi?id=76965
930
931         Reviewed by Ojan Vafai.
932
933         Improve readability as suggested by Ojan.
934
935         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/pixelzoomer.js:
936
937 2012-01-25  Carlos Garcia Campos  <cgarcia@igalia.com>
938
939         [GTK] Page clients don't need to be GObjects anymore in WebKit2 GTK+ API
940         https://bugs.webkit.org/show_bug.cgi?id=76899
941
942         Reviewed by Martin Robinson.
943
944         * gtk/generate-gtkdoc:
945         (get_webkit2_options): WebKitWebLoaderClient has been renamed to
946         WebKitLoaderClient.
947
948 2012-01-24  Ryosuke Niwa  <rniwa@webkit.org>
949
950         Build fix. Trigger Chromium Mac Release (Perf) on new commits.
951
952         * BuildSlaveSupport/build.webkit.org-config/config.json:
953
954 2012-01-23  Ryosuke Niwa  <rniwa@webkit.org>
955
956         WebKit needs public performance tests running on public bots
957         https://bugs.webkit.org/show_bug.cgi?id=10266
958
959         Reviewed by Adam Barth.
960
961         Completes the initial implementation of WebKit Perf bots. Add a build step for run-perf-test
962         and make google-mac-4 run it, and upload the results to webkit-perf.webkit.org.
963
964         * BuildSlaveSupport/build.webkit.org-config/config.json:
965         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
966         (RunAndUploadPerfTests):
967         (RunAndUploadPerfTests.start):
968         (UploadPerfTestResults):
969         (BuildAndPerfTestFactory):
970         (BuildAndPerfTestFactory.__init__):
971
972 2012-01-24  Ojan Vafai  <ojan@chromium.org>
973
974         Initializing the browser property of the Bugzilla class takes too long
975         https://bugs.webkit.org/show_bug.cgi?id=76960
976
977         Reviewed by Adam Barth.
978
979         Lazily initialize it. This saves ~150ms on a no-op run-webkit-tests call
980         on my Mac Pro.
981
982         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
983         (Bugzilla.__init__):
984         (Bugzilla._get_browser):
985         (Bugzilla):
986         (Bugzilla._set_browser):
987
988 2012-01-24  Mark Rowe  <mrowe@apple.com>
989
990         Death to JavaScriptGlue.
991
992         Rubber-stamped by Geoffrey Garen.
993
994         * BuildSlaveSupport/build-launcher-app:
995         (copyNightlyLauncher):
996         * Scripts/build-webkit:
997         * Scripts/do-file-rename:
998         * Scripts/do-webcore-rename:
999         * Scripts/webkitpy/common/config/build.py:
1000         (_should_file_trigger_build):
1001         * Scripts/webkitpy/common/config/build_unittest.py: Update the test to use another Mac-specific
1002         path in conjunction with a cross-platform path.
1003         (ShouldBuildTest):
1004
1005 2012-01-24  Kentaro Hara  <haraken@chromium.org>
1006
1007         webkitperl/prepare-ChangeLog_unittest/parser_unittests.pl should test stdout and stderr
1008         https://bugs.webkit.org/show_bug.cgi?id=75836
1009
1010         Reviewed by David Kilzer.
1011
1012         Previously parser_unittests.pl can just test the returned value of
1013         get_function_line_ranges() in prepare-ChangeLog. This patch fixes
1014         parser_unittests.pl so that it also tests the stdout and stderr of
1015         get_function_line_ranges().
1016
1017         This patch adds css_unittests_warning.css, which contains test cases
1018         for warning messages of the CSS parser of prepare-ChangeLog.
1019
1020         * Scripts/webkitperl/prepare-ChangeLog_unittest/parser_unittests.pl:
1021         (captureOutput): Captures the stdout and stderr of a given method.
1022         (convertAbsolutepathToWebKitPath): Converts absolute paths in a given
1023         string to relative paths.
1024
1025         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/css_unittests_warning-expected.txt: Added.
1026         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/css_unittests_warning.css: Added.
1027         (element1):
1028         (element2):
1029         (element3):
1030         (element4):
1031         ():
1032         (element5):
1033         (element6):
1034
1035         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests-expected.txt:
1036         Just updated the test results, i.e. added stdout and stderr results.
1037         No essential changes in the test results.
1038         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/css_unittests-expected.txt: Ditto.
1039         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/java_unittests-expected.txt: Ditto.
1040         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/javascript_unittests-expected.txt: Ditto.
1041         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/perl_unittests-expected.txt: Ditto.
1042         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/python_unittests-expected.txt: Ditto.
1043
1044 2012-01-24  Dirk Pranke  <dpranke@chromium.org>
1045
1046         webkitpy: new-run-webkit-httpd is broken
1047         https://bugs.webkit.org/show_bug.cgi?id=76963
1048
1049         Reviewed by Adam Barth.
1050
1051         r105674 broke new-run-webkit-httpd, since options.chromium
1052         doesn't exist any more. We didn't catch this because we don't
1053         run the integration tests by default anywhere :(.
1054
1055         * Scripts/new-run-webkit-httpd:
1056         (run):
1057         (main):
1058
1059 2012-01-24  Adam Barth  <abarth@webkit.org>
1060
1061         garden-o-matic pixel zoomer should let you look at every pixel of the image
1062         https://bugs.webkit.org/show_bug.cgi?id=76962
1063
1064         Reviewed by Ojan Vafai.
1065
1066         Previously, you couldn't get up to the right and bottom pixels.
1067
1068         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/pixelzoomer.js:
1069
1070 2012-01-24  Enrica Casucci  <enrica@apple.com>
1071
1072         WebHTMLView should implement typingAttributes methods to work correctly with the Inspector bar in Mail.
1073         https://bugs.webkit.org/show_bug.cgi?id=76951
1074         <rdar://problem/9325158>
1075
1076         Added test to verify that WebHTMLView supports the typingAttributes method and that it
1077         works as expected.
1078         
1079         Reviewed by Alexey Proskuryakov.
1080
1081         * TestWebKitAPI/Tests/mac/InspectorBar.mm:
1082         (TestWebKitAPI::TEST):
1083
1084 2012-01-24  James Robinson  <jamesr@chromium.org>
1085
1086         Put a space between the trac link and ellipsis in sheriffbot rollout message so it linkifies better in some IRC clients
1087         https://bugs.webkit.org/show_bug.cgi?id=76884
1088
1089         Reviewed by Adam Barth.
1090
1091         * Scripts/webkitpy/tool/bot/irc_command.py:
1092         (Rollout.execute):
1093         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
1094         (SheriffIRCBotTest.test_rollout):
1095         (SheriffIRCBotTest.test_revert):
1096         (SheriffIRCBotTest.test_multi_rollout):
1097         (SheriffIRCBotTest.test_rollout_with_r_in_svn_revision):
1098         (SheriffIRCBotTest.test_multi_rollout_with_r_in_svn_revision):
1099         (SheriffIRCBotTest.test_rollout_invalidate_reason):
1100         (test_multi_rollout_invalidate_reason):
1101
1102 2012-01-24  Kentaro Hara  <haraken@chromium.org>
1103
1104         The cpp parser of prepare-ChangeLog cannot detect a change in classes and namespaces
1105         https://bugs.webkit.org/show_bug.cgi?id=75531
1106
1107         Reviewed by David Kilzer.
1108
1109         Previously, the cpp parser of prepare-ChangeLog could not detect a change
1110         outside methods. Consider the following cpp file.
1111
1112             namespace N {
1113             int a;     // this change does not appear on ChangeLog.
1114             class C {
1115                 int b;     // this change does not appear on ChangeLog.
1116                 void f()
1117                 {
1118                     int c;     // this change appears on ChangeLog.
1119                 }
1120                 int d;     // this change does not appear on ChangeLog.
1121             };
1122             int e;     // this change does not appear on ChangeLog.
1123             };
1124
1125         The previous prepare-ChangeLog outputs just methods in which a change is found:
1126             (N::C::f):
1127
1128         This patch fixes prepare-ChangeLog so that it outputs namespaces, classes
1129         and methods in which a change is found:
1130             (N):
1131             (N::C):
1132             (N::C::f):
1133
1134         * Scripts/prepare-ChangeLog:
1135         (get_function_line_ranges_for_cpp): Modified as described above.
1136
1137         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests.cpp: Added test cases.
1138         (Class104):
1139         (Class105):
1140         (Class106):
1141         (Class106::func32):
1142         (Class106::func33):
1143         (NameSpace3):
1144         (NameSpace4):
1145         (NameSpace5):
1146         (NameSpace6):
1147         (Class107):
1148         (NameSpace5::NameSpace6::Class107::func34):
1149         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests-expected.txt:
1150
1151 2012-01-24  Kentaro Hara  <haraken@chromium.org>
1152
1153         prepare-ChangeLog outputs warnings for top-level { ... } blocks
1154         https://bugs.webkit.org/show_bug.cgi?id=75943
1155
1156         Reviewed by David Kilzer.
1157
1158         Currently, prepare-ChangeLog outputs warnings for top-level { ... } blocks,
1159         because it does not distinguish '}' at the end of a subroutine
1160         with '}' at the end of a if or for statement.
1161
1162         foo.pl:
1163             if (1) {
1164             }
1165             for (@v) {
1166             }
1167
1168         output:
1169             nested functions found at top-level at foo.pl:2
1170             nested functions found at top-level at foo.pl:4
1171
1172         This patch fixes the bug.
1173
1174         * Scripts/prepare-ChangeLog:
1175         (get_function_line_ranges_for_perl):
1176         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/perl_unittests.pl:
1177         Added test cases. No change in the test results.
1178
1179 2012-01-24  Dirk Pranke  <dpranke@chromium.org>
1180
1181         r105674 broke check-webkit-style for chromium's test_expectations.
1182
1183         Unreviewed, build fix.
1184         
1185         We now need to pass in a platform name of 'chromium' if we want
1186         the chromium port; options.chromium is no longer recognized. I
1187         will follow this patch up with a separate patch that has a test,
1188         but this at least fixes the build.
1189
1190         * Scripts/webkitpy/style/checkers/test_expectations.py:
1191         (TestExpectationsChecker._determine_port_from_exepectations_path):
1192
1193 2012-01-20  Ojan Vafai  <ojan@chromium.org>
1194
1195         check-webkit-style of the chromium test_expectations.txt file takes too long
1196         https://bugs.webkit.org/show_bug.cgi?id=76745
1197
1198         Reviewed by Dimitri Glazkov.
1199
1200         When in lint mode, have TestExpectations test all configurations instead
1201         of looping over each configuration. This also has the benefit of making
1202         the error output considerably more concise.
1203
1204         Also, got rid of the double-printing of errors when linting through check-webkit-style.
1205
1206         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1207         (Manager.lint):
1208         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1209         (TestExpectations._report_errors):
1210         (TestExpectations._add_expectations):
1211         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1212         (test_parse_error_nonfatal):
1213         (test_error_on_different_platform):
1214         * Scripts/webkitpy/style/checkers/test_expectations.py:
1215         (TestExpectationsChecker.check_test_expectations):
1216         (TestExpectationsChecker.check):
1217         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
1218         (TestExpectationsTestCase.test_determine_port_from_exepectations_path):
1219         (TestExpectationsTestCase.assert_lines_lint):
1220
1221 2012-01-24  Ryosuke Niwa  <rniwa@webkit.org>
1222
1223         Another build fix attempt after r105543.
1224
1225         * Scripts/webkitpy/common/net/file_uploader.py:
1226         (FileUploader.upload_as_multipart_form_data):
1227
1228 2012-01-24  Ryosuke Niwa  <rniwa@webkit.org>
1229
1230         Build fix attempt after r105543.
1231
1232         * Scripts/webkitpy/common/net/file_uploader.py:
1233         (FileUploader.upload_as_multipart_form_data):
1234
1235 2012-01-24  Vincent Scheib  <scheib@chromium.org>
1236
1237         [Chromium] Add Pointer Lock test hooks and mock implementation to DumpRenderTree
1238         https://bugs.webkit.org/show_bug.cgi?id=76411
1239
1240         Reviewed by Adam Barth.
1241
1242         Breaking up https://bugs.webkit.org/show_bug.cgi?id=75762
1243         into a series of patches, this change adds a trivial
1244         pointer lock implementation to the Chromium port's DumpRenderTree
1245         and exposes LayoutTestController hooks to force failure conditions.
1246
1247         * DumpRenderTree/chromium/LayoutTestController.cpp:
1248         (LayoutTestController::LayoutTestController):
1249         (LayoutTestController::didLosePointerLock):
1250         (LayoutTestController::setPointerLockWillFailAsynchronously):
1251         (LayoutTestController::setPointerLockWillFailSynchronously):
1252         * DumpRenderTree/chromium/LayoutTestController.h:
1253         * DumpRenderTree/chromium/WebViewHost.cpp:
1254         (WebViewHost::requestPointerLock):
1255         (WebViewHost::requestPointerUnlock):
1256         (WebViewHost::isPointerLocked):
1257         (WebViewHost::didCompletePointerLock):
1258         (WebViewHost::didNotCompletePointerLock):
1259         (WebViewHost::didLosePointerLock):
1260         (WebViewHost::reset):
1261         * DumpRenderTree/chromium/WebViewHost.h:
1262         (WebViewHost::setPointerLockWillFailAsynchronously):
1263         (WebViewHost::setPointerLockWillFailSynchronously):
1264
1265 2012-01-24  Mario Sanchez Prada  <msanchez@igalia.com>
1266
1267         Unreviewed, skip WK2 API test timing out on some GTK bots.
1268
1269         WebKit2APITests/TestDownloads is consistently timing out on the
1270         32bit release and 64bit debug bots.
1271
1272         * Scripts/run-gtk-tests:
1273         (TestRunner): Skip WebKit2APITests/TestDownloads.
1274
1275 2012-01-24  Andras Becsi  <andras.becsi@nokia.com>
1276
1277         [Qt] Fix the build with the newes Qt5 hashes
1278         https://bugs.webkit.org/show_bug.cgi?id=76657
1279
1280         Reviewed by Simon Hausmann.
1281
1282         * MiniBrowser/qt/MiniBrowserApplication.cpp:
1283         (MiniBrowserApplication::MiniBrowserApplication): Disable the automatic touch->mouse event synthesis
1284         for now to prevent an infinite loop where Qt always creates mouse events out of our touch mock events.
1285
1286 2012-01-24  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1287
1288         [Qt] Distinguish between Cygwin and cmd.exe in the way we set env variables
1289
1290         We can't use win32-msvc* to decide whether or not to use (set FOO=bar), as
1291         building with MinGW inside cmd.exe will require the same construct. Instead
1292         we assume that when the first character in PWD is a '/' we're running under
1293         Cygwin, and use the normal unix way of setting environment variables.
1294
1295         Reviewed by Simon Hausmann.
1296
1297         * qmake/mkspecs/features/default_pre.prf:
1298
1299 2012-01-24  Simon Hausmann  <simon.hausmann@nokia.com>
1300
1301         Some qmake build files must include Qt5's "quick" module.
1302         https://bugs.webkit.org/show_bug.cgi?id=76671
1303
1304         Unreviewed build fix.
1305
1306         * MiniBrowser/qt/MiniBrowser.pro:
1307         * WebKitTestRunner/Target.pri:
1308
1309 2012-01-24  Andras Becsi  <andras.becsi@nokia.com>
1310
1311         [Qt] Fix the event timestamps for touch mocking
1312
1313         Reviewed by Simon Hausmann.
1314
1315         * MiniBrowser/qt/MiniBrowserApplication.cpp:
1316         (MiniBrowserApplication::sendTouchEvent): We should not set QEvent::None as the
1317         timestamp argument in the QWindowSystemInterface::handleTouchEvent call.
1318
1319 2012-01-24  Carlos Garcia Campos  <cgarcia@igalia.com>
1320
1321         [GTK] Implement DownloadClient in WebKit2 GTK+ API
1322         https://bugs.webkit.org/show_bug.cgi?id=72952
1323
1324         Reviewed by Martin Robinson.
1325
1326         * gtk/generate-gtkdoc:
1327         (get_webkit2_options): Ignore WebKitDownloadClient.
1328
1329 2012-01-23  Zan Dobersek  <zandobersek@gmail.com>
1330
1331         [GTK] editing/deleting/5408255.html results are incorrect
1332         https://bugs.webkit.org/show_bug.cgi?id=53644
1333
1334         Reviewed by Martin Robinson.
1335
1336         WEBKIT_TOP_LEVEL environment variable is now set directly in either
1337         WebKitTestRunner or DumpRenderTree through usage of a compilation-time
1338         macro. This way both tools can be run outside the test harness without
1339         the need to manually set the environment variable.
1340
1341         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1342         (getTopLevelPath):
1343         * GNUmakefile.am:
1344         * Scripts/webkitpy/layout_tests/port/gtk.py:
1345         (GtkPort.setup_environ_for_server):
1346         * WebKitTestRunner/GNUmakefile.am:
1347         * WebKitTestRunner/InjectedBundle/gtk/InjectedBundleGtk.cpp:
1348         (WTR::InjectedBundle::platformInitialize):
1349
1350 2012-01-23  Dmitry Lomov  <dslomov@google.com>
1351
1352         [Chromium] Implement layoutTestController.workerThreadCount in DRT
1353         https://bugs.webkit.org/show_bug.cgi?id=74653.
1354
1355         Reviewed by Darin Fisher.
1356
1357         * DumpRenderTree/chromium/LayoutTestController.cpp:
1358         (LayoutTestController::LayoutTestController):
1359         (LayoutTestController::workerThreadCount):
1360         * DumpRenderTree/chromium/LayoutTestController.h:
1361
1362 2012-01-23  Dirk Pranke  <dpranke@chromium.org>
1363
1364         nrwt: make --chromium work like --qt
1365         https://bugs.webkit.org/show_bug.cgi?id=76875
1366
1367         Reviewed by Adam Barth.
1368
1369         --chromium used to have to be handled differently from --qt
1370         due to the way the PortFactory was implemented; there's not
1371         really a good reason for that any more so this patch makes
1372         things slightly more consistent and eliminates the
1373         options.chromium flag (--chromium is now truly a synonym for
1374         --platform chromium).
1375
1376         * Scripts/webkitpy/layout_tests/port/factory.py:
1377         (PortFactory._default_port):
1378         (PortFactory.get):
1379         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
1380         (FactoryTest.setUp):
1381         (FactoryTest.test_chromium_mac):
1382         (FactoryTest.test_chromium_linux):
1383         (FactoryTest.test_chromium_win):
1384         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1385         (parse_args):
1386
1387 2012-01-23  Dirk Pranke  <dpranke@chromium.org>
1388
1389         run-webkit-tests needs to propagate --chromium
1390         https://bugs.webkit.org/show_bug.cgi?id=76870
1391
1392         Reviewed by Eric Seidel.
1393
1394         run-webkit-tests removes '--chromium' argument from @ARGV when
1395         determining which port to run, which means that that doesn't
1396         propagate to new-run-webkit-tests. That's bad (and is handled
1397         for the other ports by re-adding the flag, but apparently we're
1398         just now noticing for Chromium).
1399
1400         * Scripts/run-webkit-tests:
1401
1402 2012-01-23  Brian Weinstein  <bweinstein@apple.com>
1403
1404         More build fixing after r105646.
1405
1406         * TestWebKitAPI/Tests/WTF/RedBlackTree.cpp:
1407         (TestWebKitAPI::TestNode::key):
1408
1409 2012-01-23  Gavin Barraclough  <barraclough@apple.com>
1410
1411         Unreviewed build fix - r105646 broke this.
1412
1413         * TestWebKitAPI/Tests/WTF/RedBlackTree.cpp:
1414         (TestWebKitAPI::TestNode::TestNode):
1415         (TestWebKitAPI::TestNode::key):
1416         (TestWebKitAPI::RedBlackTreeTest::assertEqual):
1417         (TestWebKitAPI::RedBlackTreeTest::assertSameValuesForKey):
1418         (TestWebKitAPI::RedBlackTreeTest::testDriver):
1419
1420 2012-01-23  Sheriff Bot  <webkit.review.bot@gmail.com>
1421
1422         Unreviewed, rolling out r105658.
1423         http://trac.webkit.org/changeset/105658
1424         https://bugs.webkit.org/show_bug.cgi?id=76883
1425
1426         We want this eventually, but not right at this moment
1427         (Requested by abarth on #webkit).
1428
1429         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1430
1431 2012-01-23  Adam Barth  <abarth@webkit.org>
1432
1433         garden-o-matic should support Chromium Mac Lion
1434         https://bugs.webkit.org/show_bug.cgi?id=76880
1435
1436         Reviewed by Eric Seidel.
1437
1438         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1439
1440 2012-01-23  Emil A Eklund  <eae@chromium.org>
1441
1442         check-webkit-style whitespace/operators triggers on overloaded division operator
1443         https://bugs.webkit.org/show_bug.cgi?id=76650
1444
1445         Reviewed by Darin Adler.
1446
1447         * Scripts/webkitpy/style/checkers/cpp.py:
1448         (check_spacing):
1449         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1450         (CppStyleTest.test_operator_methods):
1451
1452 2012-01-23  Ryosuke Niwa  <rniwa@webkit.org>
1453
1454         run-perf-tests should report server-side errors
1455         https://bugs.webkit.org/show_bug.cgi?id=76802
1456
1457         Reviewed by Tony Chang.
1458
1459         Report errors whenever server's response doesn't match "OK".
1460
1461         * Scripts/webkitpy/common/net/file_uploader.py:
1462         (FileUploader.upload_single_text_file):
1463         (FileUploader.upload_as_multipart_form_data):
1464         (FileUploader._upload_data.callback):
1465         (FileUploader):
1466         (FileUploader._upload_data):
1467         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1468         (PerfTestsRunner._upload_json):
1469         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1470         (test_upload_json):
1471         (test_upload_json.MockFileUploader.upload_single_text_file):
1472
1473 2012-01-22  Ryosuke Niwa  <rniwa@webkit.org>
1474
1475         run-perf-tests ignore Skipped list on chromium
1476         https://bugs.webkit.org/show_bug.cgi?id=76764
1477
1478         Reviewed by Dirk Pranke.
1479
1480         Move skipped_perf_tests from WebKit port to Base port so that Chromium port
1481         can also find skipped list. Chromium port only uses test_expectations.txt for
1482         layout tests but performacne tests don't use test_expectations.txt so Chromium port
1483         also needs to use Skipped list.
1484
1485         * Scripts/webkitpy/layout_tests/port/base.py:
1486         (Port._tests_from_skipped_file_contents):
1487         (Port):
1488         (Port._expectations_from_skipped_files):
1489         (Port.skipped_perf_tests):
1490         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1491         (PortTest.test_skipped_perf_tests):
1492         (PortTest.test_skipped_perf_tests.add_text_file):
1493         * Scripts/webkitpy/layout_tests/port/webkit.py:
1494         (WebKitPort._skipped_tests_for_unsupported_features):
1495         (WebKitPort._skipped_file_search_paths):
1496         (WebKitPort.skipped_layout_tests):
1497
1498 2012-01-23  Andy Estes  <aestes@apple.com>
1499
1500         Fix the build after r105635.
1501
1502         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp:
1503         (TestWebKitAPI::TEST):
1504
1505 2012-01-23  Xianzhu Wang  <wangxianzhu@chromium.org>
1506
1507         Basic enhancements to StringBuilder
1508         https://bugs.webkit.org/show_bug.cgi?id=67081
1509
1510         Reviewed by Darin Adler.
1511
1512         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp:
1513         (TestWebKitAPI::TEST):
1514
1515 2012-01-23  Mario Sanchez Prada  <msanchez@igalia.com>
1516
1517         [GTK] run-gtk-tests randomly fails while running the xprop comand
1518         https://bugs.webkit.org/show_bug.cgi?id=76817
1519
1520         Reviewed by Gustavo Noronha Silva.
1521
1522         No need to use xprop to remove the AT_SPI_BUS property since
1523         run-gtk-tests will always run new instances of Xvfb.
1524
1525         * Scripts/run-gtk-tests:
1526         (TestRunner): Unskip TestWebKitAccessibility.
1527         (TestRunner.run): Uncomment lines for launching the accessibility
1528         bus and registry daemon, and remove lines for running xprop.
1529
1530 2012-01-23  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1531
1532         [Qt] Don't warn about override and final being C++11 extensions
1533
1534         Clang will emit a warning when these extensions are used without passing
1535         --std=c++11, but we use feature checking to decide if we have the right
1536         extensions, so we can safely ignore these warnings. The XCode and Windows
1537         project files have the same workaround.
1538
1539         The reason for adding the flag to QMAKE_OBJECTIVE_CFLAGS as well is that
1540         we only have one extra compiler for Objective-C, which is also used for
1541         Objective-C++ sources, so we have to pass the flag, even if it doesn't
1542         make sense for Objective-C.
1543
1544         Reviewed by Simon Hausmann.
1545
1546         * qmake/mkspecs/features/unix/default_post.prf:
1547
1548 2012-01-20  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1549
1550         [Qt] Change how build-webkit decides when to do full incremental builds
1551
1552         Instead of relying on update-webkit (which isn't run on the bots) to
1553         decide when to do a full incremental build (make qmake), we let the
1554         build-webkit script itself check the current SVN revision against the
1555         previous build (by storing it in .webkit.config).
1556
1557         If the two differ we assume a full incremental build is needed, since
1558         the new revisions might have introduced problematic things like new
1559         Q_OBJECT macros. If not, we assume the developer is doing changes
1560         locally, and revert to doing a plain 'make'.
1561
1562         In addition, when the build fails in the latter case, we inform the
1563         developer of possible pitfalls and how to manually run 'make qmake'.
1564
1565         Reviewed by Ossy.
1566
1567         * Scripts/update-webkit:
1568         * Scripts/webkitdirs.pm:
1569         (buildQMakeProjects):
1570
1571 2012-01-23  Mario Sanchez Prada  <msanchez@igalia.com>
1572
1573         [GTK] run-gtk-tests randomly fails while running the xprop comand
1574         https://bugs.webkit.org/show_bug.cgi?id=76817
1575
1576         Reviewed by Philippe Normand.
1577
1578         Temporarily comment the lines related to launching the ATSPI bus
1579         and registry daemon, which are making the bots to fail randomly.
1580
1581         * Scripts/run-gtk-tests:
1582         (TestRunner): Skip TestWebKitAccessibility, as it won't run
1583         properly if the ATSPI infrastructure is not properly initialized.
1584         (TestRunner.run): Comment lines related to initialization of ATSPI.
1585
1586 2012-01-22  Ryosuke Niwa  <rniwa@webkit.org>
1587
1588         Windows python test build fix.
1589
1590         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1591         (PerfTestsRunner.__init__):
1592
1593 2012-01-22  Sergio Villar Senin  <svillar@igalia.com>
1594
1595         [GTK] DumpRenderTree converts "file:///" to a path differently
1596         https://bugs.webkit.org/show_bug.cgi?id=76631
1597
1598         Reviewed by Martin Robinson.
1599
1600         DumpRenderTree should print "/" as the last path component if the
1601         path is a lone slash instead of empty output.
1602
1603         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1604         (webViewConsoleMessage):
1605
1606 2012-01-20  David Levin  <levin@chromium.org>
1607
1608         The leaky DC check should use adoptPtr as a signal instead of OwnPtr to get less false positives.
1609         https://bugs.webkit.org/show_bug.cgi?id=76752
1610
1611         Reviewed by Eric Seidel.
1612
1613         * Scripts/webkitpy/style/checkers/cpp.py:
1614         (check_for_leaky_patterns):
1615         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1616         (LeakyPatternTest.test_create_dc):
1617         (LeakyPatternTest.test_create_compatible_dc):
1618
1619 2012-01-20  Shawn Singh  <shawnsingh@chromium.org>
1620
1621         Changed my status in commiters.py
1622
1623         Unreviewed.
1624
1625         * Scripts/webkitpy/common/config/committers.py:
1626
1627 2012-01-20  Ojan Vafai  <ojan@chromium.org>
1628
1629         run-webkit-tests --lint-test-files crawls the whole LayoutTests subtree
1630         https://bugs.webkit.org/show_bug.cgi?id=76748
1631
1632         Reviewed by Ryosuke Niwa.
1633
1634         It crawls the whole subtree and then doesn't use the data. Cutting this out
1635         saves 4 seconds warm and 17 seconds cold on my Mac Pro.
1636
1637         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1638         (run):
1639
1640 2012-01-20  Adam Barth  <abarth@webkit.org>
1641
1642         Another tiny tweak to the garden-o-matic CSS.  This makes things line
1643         up slightly nicer when there aren't any failures.
1644
1645         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
1646         (#summary):
1647
1648 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
1649
1650         run-perf-tests should support --test-results-server option
1651         https://bugs.webkit.org/show_bug.cgi?id=76680
1652
1653         Reviewed by Adam Barth.
1654
1655         Add --test-results-server, --builder-name, and --build-number options to run-perf-tests
1656         to be used by perf bots. Also refactor file_uploader as needed.
1657
1658         * Scripts/webkitpy/common/net/file_uploader.py:
1659         (FileUploader.__init__):
1660         (FileUploader.upload_single_file):
1661         (FileUploader.upload_as_multipart_form_data):
1662         (FileUploader):
1663         (FileUploader._upload_data):
1664         (FileUploader._upload_data.callback):
1665         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
1666         (JSONResultsGeneratorBase.upload_json_files):
1667         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1668         (PerfTestsRunner):
1669         (PerfTestsRunner._parse_args):
1670         (PerfTestsRunner.run):
1671         (PerfTestsRunner._generate_json):
1672         (PerfTestsRunner._upload_json):
1673         (PerfTestsRunner._run_tests_set):
1674         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1675         (create_runner):
1676         (test_run_with_upload_json):
1677         (test_run_with_upload_json.mock_upload_json):
1678         (test_upload_json):
1679         (test_upload_json.MockFileUploader):
1680         (test_upload_json.MockFileUploader.__init__):
1681         (test_upload_json.MockFileUploader.upload_single_file):
1682         (test_parse_args):
1683
1684 2012-01-20  Adam Barth  <abarth@webkit.org>
1685
1686         The party time image overlaps real content!  This patch makes the
1687         image centered vertically.
1688
1689         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
1690         (#onebar.partytime #summary):
1691
1692 2012-01-19  Ojan Vafai  <ojan@chromium.org>
1693
1694         Refactor TestExpectationsParser in preparation for caching the results
1695         https://bugs.webkit.org/show_bug.cgi?id=76669
1696
1697         Reviewed by Dimitri Glazkov.
1698
1699         Make everything private expect for the parse method. 
1700         Eventually, we'll need the expectations lines to not be modified
1701         outside of TestExpectationsParser so we can cache the results.
1702         This makes check-webkit-style of the chromium test_expectations.txt file
1703         go from ~17 seconds to ~12 seconds on my Mac Pro.
1704
1705         This patch is just a refactor in preparation, so no new tests.
1706
1707         * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor_unittest.py:
1708         (TestExpectationEditorTests.make_parsed_expectation_lines):
1709         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1710         (TestExpectationParser.parse):
1711         (TestExpectationParser):
1712         (TestExpectationParser._parse_line):
1713         (TestExpectationParser._tokenize):
1714         (TestExpectationParser._tokenize_list):
1715         (TestExpectationsModel._clear_expectations_for_test):
1716         (TestExpectations.__init__):
1717         (TestExpectations._add_expectations):
1718         (TestExpectations._add_skipped_tests):
1719         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1720         (TestExpectationParserTests.test_tokenize_blank):
1721         (TestExpectationParserTests.test_tokenize_missing_colon):
1722         (TestExpectationParserTests.test_tokenize_extra_colon):
1723         (TestExpectationParserTests.test_tokenize_empty_comment):
1724         (TestExpectationParserTests.test_tokenize_comment):
1725         (TestExpectationParserTests.test_tokenize_missing_equal):
1726         (TestExpectationParserTests.test_tokenize_extra_equal):
1727         (TestExpectationParserTests.test_tokenize_valid):
1728         (TestExpectationParserTests.test_tokenize_valid_with_comment):
1729         (TestExpectationParserTests.test_tokenize_valid_with_multiple_modifiers):
1730         (TestExpectationParserTests.test_parse_empty_string):
1731         (TestExpectationSerializerTests.assert_round_trip):
1732         (TestExpectationSerializerTests.assert_list_round_trip):
1733         * Scripts/webkitpy/tool/commands/expectations.py:
1734         (OptimizeExpectations.execute):
1735         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1736         (GardeningExpectationsUpdater.update_expectations):
1737
1738 2012-01-20  Adam Barth  <abarth@webkit.org>
1739
1740         Follow-up to previous patch: don't produce NaN when the revision number
1741         is missing.
1742
1743         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
1744         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model_unittests.js:
1745
1746 2012-01-20  Adam Barth  <abarth@webkit.org>
1747
1748         Garden-o-matic should tell me which revisions have been checked by all the bots
1749         https://bugs.webkit.org/show_bug.cgi?id=76722
1750
1751         Reviewed by Dimitri Glazkov.
1752
1753         This information helps me know when it's safe for me to roll a
1754         candidate revision.
1755
1756         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
1757         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
1758         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model_unittests.js:
1759
1760 2012-01-20  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1761
1762         build-webkit: Don't spit out congratulations message on Ctrl+C
1763
1764         Reviewed by Andreas Kling.
1765
1766         * Scripts/build-webkit:
1767
1768 2012-01-20  Philippe Normand  <pnormand@igalia.com>
1769
1770         [GTK] Allow extra jhbuild modulesets
1771         https://bugs.webkit.org/show_bug.cgi?id=76691
1772
1773         Reviewed by Gustavo Noronha Silva.
1774
1775         * gtk/jhbuildrc: Extra modulesets and modules can be added in
1776         JHBuild using the WEBKIT_EXTRA_MODULESETS and WEBKIT_EXTRA_MODULES
1777         env variables respectively. Use comma-separated values. Example:
1778         WEBKIT_EXTRA_MODULES=file:///path/to/module.set,file:///other/path.
1779
1780 2012-01-20  Mario Sanchez Prada  <msanchez@igalia.com>
1781
1782         [GTK] Expose accessibility hierarchy in WebKit2 to ATK/AT-SPI based ATs
1783         https://bugs.webkit.org/show_bug.cgi?id=72589
1784
1785         Reviewed by Martin Robinson.
1786
1787         Ensure the accessibility infrastructure is available when running
1788         the unit tests, otherwise WK2's accessibility tests won't work.
1789
1790         * Scripts/run-gtk-tests:
1791         (TestRunner):
1792         (TestRunner._lookup_atspi2_binary): New. Looks for the directory
1793         where at-spi2's stuff is located (like at-spi-bus-launcher and
1794         at-spi2-registryd).
1795         (TestRunner.run): Initialize the accessibility infrastructure.
1796
1797         * gtk/generate-gtkdoc:
1798         (get_webkit2_options): Add 'WebKitWebViewBaseAccessible.*'.
1799
1800         * gtk/jhbuild.modules: Added at-spi2-core and at-spi2-atk.
1801
1802 2012-01-20  Hayato Ito  <hayato@chromium.org>
1803
1804         Add ShadowRoot.idl which is enabled by newly introduced SHADOW_DOM flag.
1805         https://bugs.webkit.org/show_bug.cgi?id=76353
1806
1807         Reviewed by Hajime Morita.
1808
1809         Add ShadowRoot.idl, which is enabled only on chromium port since this is
1810         under development feature.
1811         ShadowRoot.idl contains minimum API so that we can test it.
1812         Other APIs should be added on other changes so that we can isolate issues.
1813
1814         * DumpRenderTree/chromium/TestShell.cpp:
1815         (TestShell::TestShell):
1816
1817 2012-01-20  Mario Sanchez Prada  <msanchez@igalia.com>
1818
1819         [GTK] GTK's DRT not to log events for already defunct objects
1820         https://bugs.webkit.org/show_bug.cgi?id=76620
1821
1822         Reviewed by Martin Robinson.
1823
1824         Do not log 'state-change:defunct' events.
1825
1826         * DumpRenderTree/gtk/AccessibilityCallbacks.cpp:
1827         (printAccessibilityEvent): Early return if a
1828         'state-change:defunct' signal is passed.
1829         (axObjectEventListener): Pass the signal name and value in
1830         separate parameters to printAccessibilityEvent.
1831
1832 2012-01-19  Gustavo Noronha Silva  <gns@gnome.org>
1833
1834         Unreviewed build fix. Wrong variable being used as the cwd for git
1835         remote update command.
1836
1837         * gtk/run-with-jhbuild:
1838         (update_jhbuild):
1839
1840 2012-01-19  Gustavo Noronha Silva  <gns@gnome.org>
1841
1842         [GTK] ensure the jhbuild used by webkit is as up-to-date as needed
1843         https://bugs.webkit.org/show_bug.cgi?id=76585
1844
1845         Reviewed by Martin Robinson.
1846
1847         This allows us to specify a minimum version of jhbuild required,
1848         while also providing stability against changes done to the master
1849         branch, by letting us choose what is the exact version that will
1850         be used.
1851
1852         * gtk/run-with-jhbuild: refactored for more hackability and easier
1853         reading
1854         (jhbuild_installed): checks whether jhbuild is installed
1855         (jhbuild_cloned): checks whether jhbuild has been cloned
1856         (jhbuild_at_expected_revision): checks whether jhbuild is at the
1857         expected revision
1858         (update_jhbuild): brings jhbuild to the expected revision
1859         (clone_jhbuild): clones jhbuild
1860         (install_jhbuild): build and installs jhbuild
1861         (update_webkitgtk_libs): runs update-webkitgtk-libs
1862         (ensure_jhbuild): high-level logic to decide whether and which of
1863         the steps above are executed
1864
1865 2012-01-19  Alexey Proskuryakov  <ap@apple.com>
1866
1867         [WK2] DumpRenderTree converts "file:///" to a path differently
1868         https://bugs.webkit.org/show_bug.cgi?id=76653
1869
1870         Reviewed by John Sullivan.
1871
1872         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp: (WTR::lastFileURLPathComponent):
1873         Do print a slash if there are no path components.
1874
1875 2012-01-19  Eric Seidel  <eric@webkit.org>
1876
1877         webkit-patch cannot rollout patches without changelog/bug number
1878         https://bugs.webkit.org/show_bug.cgi?id=75962
1879
1880         Reviewed by Adam Barth.
1881
1882         * Scripts/webkitpy/tool/commands/download.py:
1883         (AbstractRolloutPrepCommand._prepare_state):
1884         * Scripts/webkitpy/tool/commands/download_unittest.py:
1885         (AbstractRolloutPrepCommandTest.test_prepare_state):
1886
1887 2012-01-19  Alexey Proskuryakov  <ap@apple.com>
1888
1889         [WK2] fast/url/degenerate-file-base.html fails
1890         https://bugs.webkit.org/show_bug.cgi?id=76619
1891
1892         Reviewed by Adam Roben.
1893
1894         A URL generated form local path for loading had an extra slash, making document.URL
1895         incorrect in all WebKit2 tests (file://localhost//path/to/test.html).
1896
1897         * WebKitTestRunner/TestInvocation.cpp: (WTR::createWKURL):
1898
1899 2012-01-18  Ojan Vafai  <ojan@chromium.org>
1900
1901         check-webkit-style of the chromium test_expectations.txt file doesn't test all chromium ports
1902         https://bugs.webkit.org/show_bug.cgi?id=76510
1903
1904         Reviewed by Adam Barth.
1905
1906         Unlike other style rules, if you get the syntax of the test_expectations.txt
1907         file wrong, the layout tests won't run. Also, this check is simple and only
1908         slows down committing if you actually modify one of the test_expectations.txt files.
1909
1910         * Scripts/webkitpy/style/checkers/test_expectations.py:
1911         (TestExpectationsChecker.check_test_expectations):
1912         (TestExpectationsChecker.check):
1913         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
1914         (TestExpectationsTestCase.test_check_covers_all_configurations):
1915         (TestExpectationsTestCase.test_check_covers_all_configurations.mock_check_test_expectations):
1916         (TestExpectationsTestCase):
1917         (assert_lines_lint):
1918
1919 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
1920
1921         Build fix. Revert an unintentional change to WebKitDriver in r105443.
1922
1923         * Scripts/webkitpy/layout_tests/port/driver.py:
1924         (DriverProxy.__init__):
1925
1926 2012-01-19  Dirk Pranke  <dpranke@chromium.org>
1927
1928         remove the duplicated build rules from WebKit.gyp, Tools.gyp
1929         https://bugs.webkit.org/show_bug.cgi?id=73384
1930
1931         Reviewed by Tony Chang.
1932
1933         This patch removes the no-longer-necessary conditional logic for
1934         build_webkit_exes_from_webkit_gyp; we now always build the exes
1935         from their own dedicated gyp files.
1936
1937         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1938
1939 2012-01-19  Alexey Proskuryakov  <ap@apple.com>
1940
1941         make-new-script-test should work with HTTP tests
1942         https://bugs.webkit.org/show_bug.cgi?id=76603
1943
1944         Reviewed by Adam Roben.
1945
1946         * Scripts/make-new-script-test:
1947         (makePathToSharedSources): Renamed from makeTestRelativePathToSharedSources(), since it
1948         now also creates absolute paths for HTTP.
1949
1950 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
1951
1952         Some perf tests time out when ran by run-perf-tests
1953         https://bugs.webkit.org/show_bug.cgi?id=76612
1954
1955         Reviewed by Dirk Pranke and Eric Seidel.
1956
1957         Always pass --no-timeout to DumpRenderTree from run-perf-tests.
1958         Otherwise some tests such as Parser/xml-parser.html will timeout.
1959
1960         --no-timeout option is currently supported by Chromium and Mac ports.
1961
1962         * Scripts/webkitpy/layout_tests/port/base.py:
1963         (Port.to.create_driver):
1964         * Scripts/webkitpy/layout_tests/port/chromium.py:
1965         (ChromiumDriver.__init__):
1966         (ChromiumDriver._wrapper_options):
1967         * Scripts/webkitpy/layout_tests/port/driver.py:
1968         (Driver.__init__):
1969         (DriverProxy.__init__):
1970         * Scripts/webkitpy/layout_tests/port/webkit.py:
1971         (WebKitDriver.__init__):
1972         (WebKitDriver.cmd_line):
1973         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1974         (WebKitDriverTest.test_read_binary_block):
1975         (WebKitDriverTest):
1976         (WebKitDriverTest.test_no_timeout):
1977         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1978         (get_tests_run.RecordingTestDriver.__init__):
1979         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1980         (PerfTestsRunner._parse_args):
1981         (PerfTestsRunner._run_tests_set):
1982         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1983         (create_runner):
1984
1985 2012-01-19  Ben Wells  <benwells@chromium.org>
1986
1987         Add support for window.print to chromium DRT
1988         https://bugs.webkit.org/show_bug.cgi?id=76479
1989
1990         When used in a test window.print goes into print mode and then
1991         straight out again. This will allow the afterprint event to be
1992         tested by having the afterprint event happen before the dump
1993         happens.
1994
1995         Reviewed by Mihai Parparita.
1996
1997         * DumpRenderTree/chromium/WebViewHost.cpp:
1998         (WebViewHost::printPage):
1999         * DumpRenderTree/chromium/WebViewHost.h:
2000
2001 2012-01-18  Enrica Casucci  <enrica@apple.com>
2002
2003         editingAttributedStringFromRange in WebHTMLConverter does not handle NSUnderlineStyleAttributeName.
2004         https://bugs.webkit.org/show_bug.cgi?id=76588
2005         <rdar://problem/9325183>
2006
2007         Added test.
2008         
2009         Reviewed by Dan Bernstein.
2010
2011         * TestWebKitAPI/Tests/mac/InspectorBar.mm:
2012         (TestWebKitAPI::TEST):
2013
2014 2012-01-19  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2015
2016         [Qt] Do full incremental builds until bots use update-webkit
2017
2018         Rubber-stamped by Simon Hausmann.
2019
2020         * Scripts/webkitdirs.pm:
2021         (buildQMakeProjects):
2022
2023 2012-01-19  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2024
2025         Fix the Qt build
2026
2027         Unreviewed build fix.
2028
2029         * Scripts/webkitdirs.pm:
2030         (buildQMakeProjects):
2031
2032 2012-01-19  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2033
2034         [Qt] Make build-webkit always do safe incremental builds after update-webkit
2035
2036         When building against Qt5 with GCC we would just run 'make' in the build dir,
2037         and rely on the dependency files output by GCC (-MD) to handle dependency
2038         tracking, but that fails for special-cases like adding a Q_OBJECT macro
2039         to a header.
2040
2041         To guarantee that an incrmental build will work, we have to run 'make qmake',
2042         which we now do on every build-webkit that's followed by a successful run
2043         of update-webkit. The reasoning is that update-webkit can result in such
2044         potential corner-cases being applied, and since we can't know for sure
2045         unless we inspect the diff and account for all the corner cases we assume
2046         the worst and always run 'make qmake'.
2047
2048         After a succesful run of build-webkit we proceed to do just 'make' for any
2049         subsequent runs, since we assume that the developer knows what kind of
2050         changes he/she is doing, and when a 'make qmake' is needed.
2051
2052         Reviewed by Simon Hausmann.
2053
2054         * Scripts/update-webkit:
2055         * Scripts/webkitdirs.pm:
2056         (determineCurrentSVNRevision):
2057         (currentSVNRevision):
2058         (buildQMakeProjects):
2059         * qmake/mkspecs/features/default_post.prf:
2060
2061 2012-01-18  Ryosuke Niwa  <rniwa@webkit.org>
2062
2063         run-perf-tests should support Skipped list
2064         https://bugs.webkit.org/show_bug.cgi?id=76594
2065
2066         Reviewed by Adam Barth.
2067
2068         Add a support for Skipped list in run-perf-tests; also skip files in resources directories.
2069
2070         * Scripts/webkitpy/layout_tests/port/base.py:
2071         (Port.perf_tests_dir):
2072         (Port.skipped_perf_tests):
2073         (Port.skips_perf_test):
2074         (Port.relative_perf_test_filename):
2075         * Scripts/webkitpy/layout_tests/port/test.py:
2076         (TestPort.perf_tests_dir):
2077         * Scripts/webkitpy/layout_tests/port/webkit.py:
2078         (WebKitPort._expectations_from_skipped_files):
2079         (WebKitPort):
2080         (WebKitPort.skipped_layout_tests):
2081         (WebKitPort.skipped_perf_tests):
2082         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2083         (PerfTestsRunner._collect_tests):
2084         (PerfTestsRunner.run):
2085         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2086         (test_run_test_set_with_json_output):
2087         (test_run_test_set_with_json_source):
2088         (test_collect_tests_with_skipped_list):
2089         (test_collect_tests_with_skipped_list.add_file):
2090
2091 2012-01-18  Dirk Pranke  <dpranke@chromium.org>
2092
2093         Fix path to chromium_src_dir introduced in previous change.
2094
2095         Unreviewed, build fix.
2096
2097         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2098
2099 2012-01-18  Alexey Proskuryakov  <ap@apple.com>
2100
2101         Need infrastructure to test Content-Disposition filename encoding support
2102         https://bugs.webkit.org/show_bug.cgi?id=76572
2103         <rdar://problem/6421825>
2104
2105         Reviewed by Sam Weinig.
2106
2107         Made policy delegate log suggested file name when resource has Content-Disposition: attachment.
2108
2109         WebKitTestRunner does not have a policy delegate yet, bug 42546.
2110
2111         * DumpRenderTree/mac/PolicyDelegate.mm:
2112         (dispositionTypeFromContentDispositionHeader):
2113         (-[PolicyDelegate webView:decidePolicyForMIMEType:request:frame:decisionListener:]):
2114
2115 2012-01-18  Ojan Vafai  <ojan@chromium.org>
2116
2117         Fix tests from http://trac.webkit.org/changeset/105332.
2118
2119         The code assumed that if you passed an options object with a configuration
2120         attribute that it would have a valid value. A bunch of the testing mocks
2121         do things like having a configuration value of None.
2122
2123         * Scripts/webkitpy/layout_tests/port/base.py:
2124         (Port.__init__):
2125
2126 2012-01-18  Ojan Vafai  <ojan@chromium.org>
2127
2128         Shave 0.5 seconds off check-webkit-style runtime for test_expectations.txt
2129         https://bugs.webkit.org/show_bug.cgi?id=76576
2130
2131         Reviewed by Adam Barth.
2132
2133         Avoid calling default_configuration in base.py. We don't need it for
2134         checking test_expectations.txt style. It takes ~0.5 seconds to run on my
2135         Mac Pro. It's the call to "perl Tools/Scripts/webkit-build-directory --top-level"
2136         from common.executive.
2137
2138         At some point someone should probably look into why that call is so slow
2139         since it's on the critical path for run-webkit-tests and build-webkit startup.
2140
2141         * Scripts/webkitpy/layout_tests/port/base.py:
2142         (Port.__init__):
2143         * Scripts/webkitpy/style/checkers/test_expectations.py:
2144         (TestExpectationsChecker._determine_port_from_exepectations_path):
2145
2146 2012-01-18  Dirk Pranke  <dpranke@chromium.org>
2147
2148         [chromium] move Tools.gyp, switch build-webkit --chromium to All.gyp
2149         https://bugs.webkit.org/show_bug.cgi?id=76505
2150
2151         Reviewed by Tony Chang.
2152
2153         Update chromium build script to use All.gyp instead of WebKit.gyp
2154
2155         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp: Renamed from Source/WebKit/chromium/Tools.gyp.
2156         * Scripts/webkitdirs.pm:
2157
2158 2012-01-18  Sam Weinig  <sam@webkit.org>
2159
2160         Don’t hardcode the path to ‘make‘.
2161
2162         Reviewed by Dan Bernstein.
2163
2164         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
2165
2166 2012-01-18  Joshua Bell  <jsbell@chromium.org>
2167
2168         Adding myself to committers.py
2169         https://bugs.webkit.org/show_bug.cgi?id=76569
2170
2171         Unreviewed.
2172
2173         * Scripts/webkitpy/common/config/committers.py:
2174
2175 2012-01-18  Adam Barth  <abarth@webkit.org>
2176
2177         Up the cap on test failures tolerated by the EWS so the mac-ews can run to completion
2178         https://bugs.webkit.org/show_bug.cgi?id=76567
2179
2180         Reviewed by Eric Seidel.
2181
2182         Current the AppleMac port has slightly more failures than we allow on
2183         the EWS, causing the mac-ews to spin.  This patch ups the limit a bit,
2184         which should hopefully let the mac-ews run to completion.  Once
2185         AppleMac has fewer failures, we can lower the cap again.
2186
2187         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
2188         (LayoutTestResultsReaderTest.test_layout_test_results):
2189         * Scripts/webkitpy/tool/steps/runtests.py:
2190         (RunTests):
2191         * Scripts/webkitpy/tool/steps/steps_unittest.py:
2192
2193 2012-01-18  Eric Carlson  <eric.carlson@apple.com>
2194
2195         Provide access to user's list of preferred languages
2196         https://bugs.webkit.org/show_bug.cgi?id=76138
2197
2198         Reviewed by Alexey Proskuryakov.
2199
2200         * GNUmakefile.am: Add JSInternalsCustom.cpp.
2201
2202 2012-01-18  Adam Barth  <abarth@webkit.org>
2203
2204         Add Chromium Linux GPU 32 to the list of allowed Chromium Linux
2205         configurations.
2206
2207         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2208         (ChromiumLinuxPort.__init__):
2209
2210 2012-01-18  Dominic Mazzoni  <dmazzoni@google.com>
2211
2212         Accessibility: Chromium needs methods to scroll an object into view or to a specific location.
2213         https://bugs.webkit.org/show_bug.cgi?id=73460
2214
2215         Reviewed by Chris Fleizach.
2216
2217         * DumpRenderTree/AccessibilityUIElement.h:
2218         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
2219         (AccessibilityUIElement::AccessibilityUIElement):
2220         (AccessibilityUIElement::scrollToMakeVisibleCallback):
2221         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocusCallback):
2222         (AccessibilityUIElement::scrollToGlobalPointCallback):
2223         * DumpRenderTree/chromium/AccessibilityUIElement.h:
2224         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2225         (AccessibilityUIElement::scrollToMakeVisible):
2226         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2227         (AccessibilityUIElement::scrollToGlobalPoint):
2228         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2229         (AccessibilityUIElement::scrollToMakeVisible):
2230         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2231         (AccessibilityUIElement::scrollToGlobalPoint):
2232         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2233         (AccessibilityUIElement::scrollToMakeVisible):
2234         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2235         (AccessibilityUIElement::scrollToGlobalPoint):
2236
2237 2012-01-17  Enrica Casucci  <enrica@apple.com>
2238
2239         Missing NSForegroundColorAttributeName should be treated as black instead of transparent.
2240         https://bugs.webkit.org/show_bug.cgi?id=76490
2241         <rdar://problem/9460733>
2242
2243         Added TestWebKitAPI test.
2244         
2245         Reviewed by Dan Bernstein.
2246
2247         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added new test file.
2248         * TestWebKitAPI/Tests/mac/InspectorBar.mm: Added test that calls
2249         changeAttributes to the WebView.
2250         
2251 2012-01-18  Balazs Kelemen  <kbalazs@webkit.org>
2252
2253         [Qt] Consolidate layout test crash logging
2254         https://bugs.webkit.org/show_bug.cgi?id=75088
2255
2256         Reviewed by Simon Hausmann.
2257
2258         Move backtrace generating logic into WTFReportBacktrace
2259         and add a way to deinstall signal handlers if we know
2260         that we have already printed the backtrace.
2261
2262         * DumpRenderTree/qt/main.cpp:
2263         (crashHandler):
2264         (setupSignalHandlers):
2265         (WTFCrashHook):
2266         (main):
2267         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
2268         (WTR::crashHandler):
2269         (WTR::setupSignalHandlers):
2270         (WTR::crashHook):
2271         (WTR::InjectedBundle::platformInitialize):
2272
2273 2012-01-18  Ilya Tikhonovsky  <loislo@chromium.org>
2274
2275         Unreviewed build fix after r105256.
2276
2277         There was a cyclic dependency between self._port and self._host assigments.
2278
2279         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2280         (PerfTestsRunner.__init__):
2281
2282 2012-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2283
2284         Add new watchlist rule for the Qt build system
2285
2286         Reviewed by Simon Hausmann.
2287
2288         * Scripts/webkitpy/common/config/watchlist:
2289
2290 2012-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2291
2292         [Qt] Move OTHER_FILES from WebKit.pro to Tools.pro
2293
2294         Reviewed by Simon Hausmann.
2295
2296         * Tools.pro:
2297
2298 2012-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2299
2300         [Qt] Allow building only a subset of the WebKit sub-projects
2301
2302         Fixes build-jsc, broken in r104825.
2303
2304         We no longer need the intermediate buildQMakeQtProject(), as we're the
2305         only port using qmake, and buildQMakeProject is now highly Qt specific.
2306
2307         <http://webkit.org/b/76179>
2308
2309         Reviewed by Simon Hausmann.
2310
2311         * Scripts/build-jsc:
2312         * Scripts/build-webkit:
2313         * Scripts/webkitdirs.pm:
2314         (buildQMakeProjects):
2315
2316 2012-01-18  Ryosuke Niwa  <rniwa@webkit.org>
2317
2318         run-perf-tests should generate a json file that summaries the result
2319         https://bugs.webkit.org/show_bug.cgi?id=76504
2320
2321         Reviewed by Adam Barth.
2322
2323         Add an ability to generate a json file to run-perf-test in the preparation for perf bots.
2324         New option --outout-json-path specifies the json file's path, and --source-json-path specifies
2325         another json file to be merged into the generated json file.
2326
2327         Also fixed a bug that --build wasn't set by default.
2328
2329         * Scripts/webkitpy/layout_tests/port/test.py:
2330         (TestPort.webkit_base):
2331         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2332         (PerfTestsRunner.__init__):
2333         (PerfTestsRunner):
2334         (PerfTestsRunner._parse_args):
2335         (PerfTestsRunner.run):
2336         (PerfTestsRunner._generate_json_if_specified):
2337         (PerfTestsRunner._process_chromium_style_test_result):
2338         (PerfTestsRunner._process_parser_test_result):
2339         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2340         (MainTest):
2341         (create_runner):
2342         (test_run_test_set):
2343         (test_run_test_set_for_parser_tests):
2344         (test_run_test_set_with_summary_json):
2345         (test_collect_tests):
2346         (test_parse_args):
2347
2348 2012-01-17  Sheriff Bot  <webkit.review.bot@gmail.com>
2349
2350         Unreviewed, rolling out r105244.
2351         http://trac.webkit.org/changeset/105244
2352         https://bugs.webkit.org/show_bug.cgi?id=76518
2353
2354         broke Chromium Mac (Requested by rolandsteiner on #webkit).
2355
2356         * DumpRenderTree/AccessibilityUIElement.h:
2357         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
2358         (AccessibilityUIElement::AccessibilityUIElement):
2359         * DumpRenderTree/chromium/AccessibilityUIElement.h:
2360         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2361         (AccessibilityUIElement::removeSelection):
2362         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2363         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2364         (AccessibilityUIElement::removeSelection):
2365
2366 2012-01-17  Dominic Mazzoni  <dmazzoni@google.com>
2367
2368         Accessibility: Chromium needs methods to scroll an object into view or to a specific location.
2369         https://bugs.webkit.org/show_bug.cgi?id=73460
2370
2371         Reviewed by Chris Fleizach.
2372
2373         * DumpRenderTree/AccessibilityUIElement.h:
2374         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
2375         (AccessibilityUIElement::AccessibilityUIElement):
2376         (AccessibilityUIElement::scrollToMakeVisibleCallback):
2377         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocusCallback):
2378         (AccessibilityUIElement::scrollToGlobalPointCallback):
2379         * DumpRenderTree/chromium/AccessibilityUIElement.h:
2380         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2381         (AccessibilityUIElement::scrollToMakeVisible):
2382         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2383         (AccessibilityUIElement::scrollToGlobalPoint):
2384         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2385         (AccessibilityUIElement::scrollToMakeVisible):
2386         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2387         (AccessibilityUIElement::scrollToGlobalPoint):
2388         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2389         (AccessibilityUIElement::scrollToMakeVisible):
2390         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2391         (AccessibilityUIElement::scrollToGlobalPoint):
2392
2393 2012-01-17  Adam Barth  <abarth@webkit.org>
2394
2395         Enable tests on the mac-ews
2396         https://bugs.webkit.org/show_bug.cgi?id=76502
2397
2398         Reviewed by Simon Fraser.
2399
2400         Apple now has five machines running the mac-ews.  Lucas Forschler has
2401         asked that we try enabling tests to see if we have enough bandwidth.
2402
2403         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
2404         (MacEWS):
2405
2406 2012-01-17  Sanjoy Das  <sanjoy@playingwithpointers.com>
2407
2408         Add a check-webkit-style exception for GDBInterface.cpp
2409         https://bugs.webkit.org/show_bug.cgi?id=76187
2410
2411         Add an exception for Source/JavaScriptCore/jit/GDBInterface.cpp so
2412         that it does not fail the style check.
2413
2414         Reviewed by Martin Robinson.
2415
2416         * Scripts/webkitpy/style/checker.py: Tell the style checker to not run
2417         readability/naming tests on
2418         Source/JavaScriptCore/jit/GDBInterface.cpp.
2419         * Scripts/webkitpy/style/checker_unittest.py:
2420         (GlobalVariablesTest.test_path_rules_specifier): Add a test for
2421         the above exception.
2422
2423 2012-01-17  Hajime Morrita  <morrita@chromium.org>
2424
2425         https://bugs.webkit.org/show_bug.cgi?id=76245
2426         [Mac][DRT] should allow query string for the file URL.
2427
2428         Reviewed by Tony Chang.
2429
2430         Taught DRT about file:// as a url instead of a file path.
2431         We are now able to use a path file:///foo/bar.html?baz as a url
2432         with which DRT opens "/foo/bar.html" with "baz" as a query string.
2433
2434         * DumpRenderTree/mac/DumpRenderTree.mm:
2435         (runTest):
2436
2437 2012-01-17  Ojan Vafai  <ojan@chromium.org>
2438
2439         Check style on test_expectations.txt files before commit
2440         https://bugs.webkit.org/show_bug.cgi?id=76484
2441
2442         Reviewed by Adam Barth.
2443
2444         Unlike other style rules, if you get the syntax of the test_expectations.txt
2445         file wrong, the layout tests won't run. Also, this check is simple and only
2446         slows down committing if you actually modify one of the test_expectations.txt files.
2447
2448         * Scripts/webkitpy/tool/steps/commit.py:
2449         (Commit._check_test_expectations):
2450         (Commit.run):
2451         * Scripts/webkitpy/tool/steps/commit_unittest.py: Added.
2452         (CommitTest):
2453         (CommitTest.test_check_test_expectations):
2454
2455 2012-01-17  Viatcheslav Ostapenko  <ostapenko.viatcheslav@nokia.com>
2456
2457         [Qt] Debug build fails with debug qt5
2458         https://bugs.webkit.org/show_bug.cgi?id=76463
2459
2460         Reviewed by Simon Hausmann.
2461
2462         Add QNetworkCookie include in order to satisfy sizeof in QTypeInfo in debug builds.
2463
2464         * QtTestBrowser/cookiejar.h:
2465
2466 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2467
2468         webkitpy: fix json import on linux
2469         https://bugs.webkit.org/show_bug.cgi?id=76481
2470
2471         Reviewed by Adam Barth.
2472
2473         For some reason, the version of simplejson autoinstalled on my linux
2474         box seems to be broken. We should really only be using
2475         simplejson if json isn't available anyway, so this change looks
2476         for json first.
2477
2478         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2479         (JSONGeneratorTest.test_test_timings_trie):
2480
2481 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2482
2483         Fix failures in test-webkitpy caused by r105177.
2484
2485         Unreviewed, build fix.
2486
2487         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2488         (ChromiumLinuxPort._determine_driver_path_statically):
2489         (ChromiumLinuxPort._static_build_path):
2490
2491 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2492
2493         webkitpy: make PortFactory.get() be fully data-driven
2494         https://bugs.webkit.org/show_bug.cgi?id=76360
2495
2496         Reviewed by Eric Seidel.
2497
2498         Refactor PortFactory.get() to be fully data-driven.
2499
2500         * Scripts/webkitpy/layout_tests/port/factory.py:
2501         (PortFactory):
2502         (PortFactory._default_port):
2503         (PortFactory.get):
2504         (PortFactory.get.in):
2505         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
2506         (FactoryTest.assert_port):
2507
2508 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2509
2510         webkitpy: add determine_full_port_name(), clean up port.__init__()
2511         https://bugs.webkit.org/show_bug.cgi?id=76357
2512
2513         Reviewed by Adam Barth.
2514
2515         Each Port class is now required to implement a 'port_name' field
2516         that will match the start of all ports constructed by this class and
2517         a 'determine_full_port_name' field that will make the port
2518         fully-qualified.
2519
2520         * Scripts/webkitpy/layout_tests/port/apple.py:
2521         (ApplePort.__init__):
2522         * Scripts/webkitpy/layout_tests/port/base.py:
2523         (Port):
2524         (Port.determine_port_name):
2525         (Port.__init__):
2526         * Scripts/webkitpy/layout_tests/port/chromium.py:
2527         (ChromiumPort.for):
2528         (ChromiumPort.__init__):
2529         (ChromiumPort):
2530         (ChromiumPort._chromium_base_dir):
2531         (ChromiumPort.path_from_chromium_base):
2532         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2533         (ChromiumGpuLinuxPort):
2534         (ChromiumGpuLinuxPort.__init__):
2535         (ChromiumGpuMacPort):
2536         (ChromiumGpuMacPort.__init__):
2537         (ChromiumGpuWinPort):
2538         (ChromiumGpuWinPort.__init__):
2539         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2540         (ChromiumLinuxPort):
2541         (ChromiumLinuxPort.determine_port_name):
2542         (ChromiumLinuxPort.__init__):
2543         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2544         (ChromiumMacPort):
2545         (ChromiumMacPort.determine_port_name):
2546         (ChromiumMacPort.__init__):
2547         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2548         (ChromiumWinPort):
2549         (ChromiumWinPort.determine_port_name):
2550         (ChromiumWinPort.__init__):
2551         * Scripts/webkitpy/layout_tests/port/dryrun.py:
2552         (DryRunPort):
2553         (DryRunPort.determine_port_name):
2554         (DryRunPort.__init__):
2555         * Scripts/webkitpy/layout_tests/port/efl.py:
2556         (EflPort):
2557         * Scripts/webkitpy/layout_tests/port/factory.py:
2558         (PortFactory):
2559         (PortFactory._default_port):
2560         (PortFactory.get):
2561         (PortFactory.get.in):
2562         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
2563         (FactoryTest.assert_port):
2564         * Scripts/webkitpy/layout_tests/port/google_chrome.py:
2565         (GoogleChromeLinux32Port):
2566         (GoogleChromeLinux32Port.determine_port_name):
2567         (GoogleChromeLinux64Port):
2568         (GoogleChromeLinux64Port.determine_port_name):
2569         (GoogleChromeMacPort):
2570         (GoogleChromeMacPort.determine_port_name):
2571         (GoogleChromeWinPort):
2572         (GoogleChromeWinPort.determine_port_name):
2573         * Scripts/webkitpy/layout_tests/port/gtk.py:
2574         (GtkPort):
2575         * Scripts/webkitpy/layout_tests/port/mac.py:
2576         (MacPort):
2577         (MacPort.determine_port_name):
2578         (MacPort.__init__):
2579         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2580         (test_versions):
2581         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
2582         (MockDRTPort):
2583         (MockDRTPort.determine_port_name):
2584         (MockDRTPort.__init__):
2585         (MockDRT):
2586         (MockDRT.determine_port_name):
2587         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2588         (PortTestCase.make_port):
2589         * Scripts/webkitpy/layout_tests/port/qt.py:
2590         (QtPort):
2591         (QtPort.determine_port_name):
2592         (QtPort.__init__):
2593         * Scripts/webkitpy/layout_tests/port/test.py:
2594         (TestPort):
2595         (TestPort.determine_port_name):
2596         (TestPort.__init__):
2597         * Scripts/webkitpy/layout_tests/port/webkit.py:
2598         (WebKitPort.__init__):
2599         * Scripts/webkitpy/layout_tests/port/win.py:
2600         (WinPort):
2601         (WinPort.determine_port_name):
2602         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
2603         (get_test_baselines.AllPlatformsPort.__init__):
2604         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
2605         (get_test_config.TestMacPort):
2606         (get_test_config):
2607
2608 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2609
2610         webkitpy: clean up port code in preparation for static port names
2611         https://bugs.webkit.org/show_bug.cgi?id=76356
2612
2613         Reviewed by Adam Barth.
2614
2615         To fix bug 76215 and be able to determine appropriate port names
2616         in webkitpy without actually constructing Port objects, we need
2617         to shuffle some logic in the chromium ports to be able to figure
2618         out whether we built DRT in 32 or 64-bit mode.
2619
2620         * Scripts/webkitpy/layout_tests/port/chromium.py:
2621         (ChromiumPort):
2622         (ChromiumPort._chromium_base_dir):
2623         (ChromiumPort.__init__):
2624         (ChromiumPort.path_from_chromium_base):
2625         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2626         (ChromiumLinuxPort):
2627         (ChromiumLinuxPort._determine_driver_path_statically):
2628         (ChromiumLinuxPort._static_build_path):
2629         (ChromiumLinuxPort._determine_architecture):
2630         (ChromiumLinuxPort.__init__):
2631         (ChromiumLinuxPort._build_path):
2632
2633 2012-01-17  Hao Zheng  <zhenghao@chromium.org>
2634
2635         Do not remove pixel failures for ref tests.
2636         https://bugs.webkit.org/show_bug.cgi?id=76243
2637
2638         Reviewed by Tony Chang.
2639
2640         'NRWT --no-pixel-tests' complains when reftests are expected to
2641         be image mismatch.
2642
2643         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2644         (Manager._update_summary_with_result):
2645         * Scripts/webkitpy/layout_tests/models/test_failures.py:
2646         (is_reftest_failure):
2647         (determine_result_type):
2648
2649 2012-01-17  Balazs Ankes  <Ankes.Balazs@stud.u-szeged.hu>
2650
2651         REGRESSION(r81225): ORWT should ignore reftests
2652         https://bugs.webkit.org/show_bug.cgi?id=67936
2653
2654         Reviewed by Csaba Osztrogonác.
2655
2656         * Scripts/old-run-webkit-tests:
2657         (isUsedInReftest): $filename should be the first parameter
2658         (findTestsToRun): run test if it isn't reftest
2659
2660 2012-01-17  Gustavo Noronha Silva  <gns@gnome.org>
2661
2662         Unreviewed. Fix python unit test I broke with r105142
2663
2664         * Scripts/webkitpy/common/config/ports_unittest.py:
2665         (WebKitPortTest.test_gtk_port):
2666
2667 2012-01-17  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2668
2669         [Qt] Don't set the 'primary' flag when mocking touch-points in MiniBrowser
2670
2671         Qt 5 no longer has that flag.
2672
2673         Reviewed by Simon Hausmann.
2674
2675         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2676         (MiniBrowserApplication::notify):
2677
2678 2012-01-16  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
2679
2680         GTK+ EWS needs to run update-webkitgtk-libs after applying a patch
2681         https://bugs.webkit.org/show_bug.cgi?id=75857
2682
2683         Reviewed by Adam Barth.
2684
2685         * Scripts/webkitdirs.pm:
2686         (buildAutotoolsProject): move cleaning the build to happen early,
2687         and handle --update-gtk right after that
2688         * Scripts/webkitpy/common/config/ports.py:
2689         (GtkPort.build_webkit_command): pass --update-gtk as an option to
2690         build-webkit, so that it is used in the EWS
2691
2692 2012-01-16  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2693
2694         [Qt] Don't add _debug postfix to targets when building debug only
2695
2696         Reviewed by Simon Hausmann.
2697
2698         * qmake/mkspecs/features/default_post.prf:
2699         * qmake/mkspecs/features/functions.prf:
2700
2701 2012-01-16  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2702
2703         [Qt] Add more files to project-files for easier navigation in Qt Creator
2704
2705         Reviewed by Simon Hausmann.
2706
2707         * Tools.pro:
2708         * qmake/config.tests/README: Added.
2709
2710 2012-01-17  Philippe Normand  <pnormand@igalia.com>
2711
2712         [GTK] fast/dom/constructed-objects-prototypes.html fails
2713         https://bugs.webkit.org/show_bug.cgi?id=75432
2714
2715         Reviewed by Martin Robinson.
2716
2717         * Scripts/build-webkit: Enable video-track in GTK.
2718
2719 2012-01-16  Takashi Toyoshima  <toyoshim@chromium.org>
2720
2721         Add toyoshim as a committer.
2722         https://bugs.webkit.org/show_bug.cgi?id=76390
2723
2724         Reviewed by Hajime Morita.
2725
2726         * Scripts/webkitpy/common/config/committers.py:
2727
2728 2012-01-16  Brent Fulgham  <bfulgham@webkit.org>
2729
2730         [WinCairo] Build fix after r104155. Several debug targets were not
2731         properly including the '*Common' property sheets.
2732
2733         * DumpRenderTree/win/DumpRenderTreeLauncherDebugCairoCFLite.vsprops:
2734         * DumpRenderTree/win/ImageDiffLauncherDebugCairoCFLite.vsprops:
2735         * MiniBrowser/Configurations/MiniBrowserLauncherDebugCairoCFLite.vsprops:
2736         * WebKitTestRunner/win/WebKitTestRunnerLauncherDebugCairoCFLite.vsprops:
2737         * WinLauncher/WinLauncherLauncherDebugCairoCFLite.vsprops:
2738
2739 2012-01-16  Csaba Osztrogonác  <ossy@webkit.org>
2740
2741         REGRESSION(r105072): It broke committers_unittest.py
2742         https://bugs.webkit.org/show_bug.cgi?id=76396
2743
2744         Reviewed by Adam Barth.
2745
2746         * Scripts/webkitpy/common/config/committers_unittest.py:
2747         (CommittersTest._assert_fuzz_match): It should handle more names.
2748         (CommittersTest.test_contributors_by_fuzzy_match_with_legacy_names): Fix the testcase.
2749
2750 2012-01-16  Mario Sanchez Prada  <msanchez@igalia.com>
2751
2752         [GTK] Include gtk+3 in the jhbuild modules
2753         https://bugs.webkit.org/show_bug.cgi?id=76284
2754
2755         Reviewed by Philippe Normand.
2756
2757         * gtk/jhbuild.modules: Add some gtk+3 to the jhbuild moduleset.
2758
2759 2012-01-16  Dan Winship  <danw@gnome.org>
2760
2761         Add myself as a committer
2762
2763         Unreviewed
2764
2765         * Scripts/webkitpy/common/config/committers.py:
2766
2767 2012-01-16  Raphael Kubo da Costa  <kubo@profusion.mobi>
2768
2769         Unreviewed; fix the build when --as-needed and --no-copy-dt-needed-entries are passed to the linker.
2770
2771         Some recent versions of a few Linux distributions have started passing
2772         --as-needed and --no-copy-dt-needed-entries by default to the linker,
2773         which broke the build when SHARED_CORE was turned on, as libwtf_efl.so
2774         needed a few symbols from libjavascriptcore_efl.so.
2775
2776         * DumpRenderTree/efl/CMakeLists.txt: Explicitly pass WTF_LIBRARY_NAME
2777         before JavaScriptCore_LIBRARY_NAME in DumpRenderTree_LIBRARIES so the
2778         libraries are searched in the right order by the linker.
2779
2780 2012-01-16  Philippe Normand  <pnormand@igalia.com>
2781
2782         [GTK] set audio resources path in NRTW
2783         https://bugs.webkit.org/show_bug.cgi?id=76380
2784
2785         Reviewed by Martin Robinson.
2786
2787         * Scripts/webkitpy/layout_tests/port/gtk.py:
2788         (GtkPort.setup_environ_for_server): Set $AUDIO_RESOURCES_PATH so
2789         the uninstalled audio resources are used during webaudio tests.
2790
2791 2012-01-16  Carlos Garcia Campos  <cgarcia@igalia.com>
2792
2793         [GTK] ldflags and cflags should take precedence over the existing flags in gtkdoc.py
2794         https://bugs.webkit.org/show_bug.cgi?id=76369
2795
2796         Reviewed by Xan Lopez.
2797
2798         If there's an installed version of webkit and there are new
2799         symbols in the build tree, gtkdoc-scanobj fails due to undefined
2800         references, because the env ldflags are given before the local
2801         ones.
2802
2803         * gtk/gtkdoc.py:
2804         (GTKDoc._run_gtkdoc_scangobj):
2805
2806 2012-01-15  Pablo Flouret  <pablof@motorola.com>
2807
2808         Fix compilation errors on build-webkit --debug --no-video on mac.
2809         https://bugs.webkit.org/show_bug.cgi?id=75867
2810
2811         Reviewed by Philippe Normand.
2812
2813         Add --[no-]video-track to build-webkit, since video track feature
2814         depends on video.
2815
2816         * Scripts/build-webkit:
2817
2818 2012-01-14  James Robinson  <jamesr@chromium.org>
2819
2820         [chromium] Failing webkit_unit_tests does not make the cr-linux EWS bubble red
2821         https://bugs.webkit.org/show_bug.cgi?id=76313
2822
2823         Reviewed by Adam Barth.
2824
2825         The exit code of this script should be the exit code of webkit_unit_tests. Tested manually.
2826
2827         * Scripts/run-chromium-webkit-unit-tests:
2828
2829 2012-01-14  Sheriff Bot  <webkit.review.bot@gmail.com>
2830
2831         Unreviewed, rolling out r105017.
2832         http://trac.webkit.org/changeset/105017
2833         https://bugs.webkit.org/show_bug.cgi?id=76333
2834
2835         broke the build (Requested by philn-tp on #webkit).
2836
2837         * gtk/jhbuild.modules:
2838
2839 2012-01-14  Mario Sanchez Prada  <msanchez@igalia.com>
2840
2841         [GTK] Include gtk+3 in the jhbuild modules
2842         https://bugs.webkit.org/show_bug.cgi?id=76284
2843
2844         Reviewed by Philippe Normand.
2845
2846         * gtk/jhbuild.modules: Add some gtk+3 to the jhbuild moduleset.
2847
2848 2012-01-13  Adam Barth  <abarth@webkit.org>
2849
2850         Remove --dry-run support from webkit-patch
2851         https://bugs.webkit.org/show_bug.cgi?id=76300
2852
2853         Reviewed by Dirk Pranke.
2854
2855         This is the rest of the patch from Bug 76300.  I landed it separtely
2856         because main.py had some stray conflict markers that I needed to clean
2857         up first.
2858
2859         * Scripts/webkitpy/test/main.py:
2860         (Tester.parse_args):
2861         (Tester._run_tests):
2862
2863 2012-01-13  Adam Barth  <abarth@webkit.org>
2864
2865         test-webkitpy does not run
2866         https://bugs.webkit.org/show_bug.cgi?id=76318
2867
2868         Unreviewed.
2869
2870         Looks like some conflict markers got left in the file.
2871
2872         * Scripts/webkitpy/test/main.py:
2873         (Tester.run):
2874
2875 2012-01-13  Adam Barth  <abarth@webkit.org>
2876
2877         Remove --dry-run support from webkit-patch
2878         https://bugs.webkit.org/show_bug.cgi?id=76300
2879
2880         Reviewed by Dirk Pranke.
2881
2882         We don't use this functionality for anything, and it doesn't really
2883         work.  Someone in #webkit tried using it and and got sad/confused.  We
2884         should just remove it.
2885
2886         * Scripts/webkitpy/common/checkout/scm/git.py:
2887         (Git.push_local_commits_to_server):
2888         * Scripts/webkitpy/common/checkout/scm/scm.py:
2889         (SCM.__init__):
2890         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
2891         (_shared_test_commit_with_message):
2892         (GitSVNTest.test_commit_text_parsing):
2893         * Scripts/webkitpy/common/checkout/scm/svn.py:
2894         (SVN.commit_with_message):
2895         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
2896         (Bugzilla.__init__):
2897         (Bugzilla.quips):
2898         (Bugzilla.authenticate):
2899         (Bugzilla.add_attachment_to_bug):
2900         (Bugzilla.add_patch_to_bug):
2901         (Bugzilla.create_bug):
2902         (Bugzilla.clear_attachment_flags):
2903         (Bugzilla.set_flag_on_attachment):
2904         (Bugzilla.obsolete_attachment):
2905         (Bugzilla.add_cc_to_bug):
2906         (Bugzilla.post_comment_to_bug):
2907         (Bugzilla.close_bug_as_fixed):
2908         (Bugzilla.reassign_bug):
2909         (reopen_bug):
2910         * Scripts/webkitpy/test/main.py:
2911         (Tester.parse_args):
2912         (Tester.run_tests):
2913
2914 2012-01-13  David Levin  <levin@chromium.org>
2915
2916         check-webkit-style: should encourage the use of Own* classes for Windows DC.
2917         https://bugs.webkit.org/show_bug.cgi?id=76227
2918
2919         Reviewed by Dirk Pranke.
2920
2921         * Scripts/webkitpy/style/checkers/cpp.py:
2922         (check_for_leaky_patterns): The new check.
2923         (process_line): Added a call to the new check.
2924         (CppChecker): Added the new error type.
2925         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
2926         (CppStyleTestBase):
2927         (CppStyleTestBase.perform_leaky_pattern_check):
2928         The check for only leaky pattern errors.
2929         (LeakyPatternTest): Test cases.
2930
2931 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
2932
2933         test-webkitpy: clean up handling of test directories, QueueStatusServer
2934         https://bugs.webkit.org/show_bug.cgi?id=76238
2935
2936         Reviewed by Adam Barth.
2937
2938         This patch cleans up a bunch of hacks related to how we find all
2939         of the python unit tests and set up the sys.path variable in
2940         test-webkitpy.
2941
2942         This patch changes how test-webkitpy works slightly in that we
2943         will now look for test files under Tools/Scripts and Source/WebKit2/Scripts
2944         rather than just Tools/Scripts/webkitpy and Source/WebKit2/Scripts/webkit.
2945         This greatly simplifies the mapping of paths to module names and
2946         appears to have no significant performance impact or other drawbacks.
2947
2948         There are other ways that we could remove all of the hacks and
2949         still have a more flexible mapping of paths to module names, but
2950         the ones I've thought of so far add a decent amount of complexity to
2951         the implementation and don't seem to buy us much.
2952         
2953         * QueueStatusServer/__init__.py:
2954         * Scripts/test-webkitpy:
2955         * Scripts/webkitpy/test/main.py:
2956         (Tester._modules_from_paths):
2957         (Tester._is_module):
2958
2959 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
2960
2961         test-webkitpy: should support classes and individual test names as well as modules
2962         https://bugs.webkit.org/show_bug.cgi?id=76233
2963
2964         Reviewed by Adam Barth.
2965
2966         The default unittest.main() supports a list of individual
2967         TestCase classes or individual tests as well as modules. This
2968         change adds support for the same to test-webkitpy, so now you
2969         can say
2970
2971           % test-wekbitpy webkitpy.test.cat_unittest.CatTest
2972
2973         and
2974
2975           % test-wekbitpy webkitpy.test.cat_unittest.CatTest.test_basic
2976
2977         in addition to
2978
2979           % test-wekbitpy webkitpy.test.cat_unittest
2980
2981         * Scripts/test-webkitpy:
2982         * Scripts/webkitpy/test/cat_unittest.py:
2983         * Scripts/webkitpy/test/main.py:
2984         (Tester.parse_args):
2985         (Tester.run):
2986         (Tester):
2987         (Tester._find_modules):
2988         (Tester._exclude):
2989         (Tester._run_tests):
2990         (Tester._is_module):
2991         (Tester._log_exception):
2992
2993 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
2994
2995         test-webkitpy: add basic code coverage support
2996         https://bugs.webkit.org/show_bug.cgi?id=76201
2997
2998         Reviewed by Adam Barth.
2999
3000         This test adds basic code coverage support to test-webkitpy; if
3001         you pass the -c flag, we will generate a new coverage file. We
3002         can start simply with this and add more features as necessary.
3003
3004         This patch requires coverage to be already installed on the
3005         machine. I am not auto-installing this for now.
3006
3007         * Scripts/webkitpy/test/main.py:
3008         (Tester.parse_args):
3009         (Tester.run_tests):
3010
3011 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
3012
3013         StatusBubbleTest.test_build_bubble is failing
3014         https://bugs.webkit.org/show_bug.cgi?id=76222
3015
3016         Reviewed by Adam Barth.
3017
3018         * QueueStatusServer/handlers/statusbubble_unittest.py:
3019         (StatusBubbleTest.test_build_bubble):
3020
3021 2012-01-13  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
3022
3023         nrwt should use multiple http shards
3024         https://bugs.webkit.org/show_bug.cgi?id=75958
3025
3026         Add --max-locked-shards option to the nrwt for manual set
3027         the maximum number of the locked shards
3028
3029         Reviewed by Dirk Pranke.
3030
3031         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3032         (Manager._max_locked_shards):
3033         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3034         (ShardingTests.get_shards):
3035         (ShardingTests.test_multiple_locked_shards):
3036         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3037         (parse_args):
3038
3039 2012-01-13  Raphael Kubo da Costa  <kubo@profusion.mobi>
3040
3041         prepare-ChangeLog: Look for a text editor in a few more variables.
3042         https://bugs.webkit.org/show_bug.cgi?id=76289
3043
3044         Reviewed by Ryosuke Niwa.
3045
3046         A text editor was looked for only in the CHANGE_LOG_EDITOR and
3047         CHANGE_LOG_EDIT_APPLICATION environment variables before falling back
3048         to the OS X-specific open(1) command.
3049
3050         As there does not seem to be a common way of looking for a text editor
3051         among the scripts in Tools/Scripts right now, just look at $VISUAL and
3052         $EDITOR (set much more often than the other variables in Unix systems)
3053         before falling back to open(1).
3054
3055         * Scripts/prepare-ChangeLog:
3056         (openChangeLogs):
3057
3058 2012-01-13  Gabor Rapcsanyi  <rgabor@webkit.org>
3059
3060         [GTK] gtkdoc.py ignoring LDFLAGS and CFLAGS environment variables
3061         https://bugs.webkit.org/show_bug.cgi?id=76269
3062
3063         Reviewed by Martin Robinson.
3064
3065         * gtk/gtkdoc.py:
3066         (GTKDoc._run_gtkdoc_scangobj):
3067
3068 2012-01-13  Csaba Osztrogonác  <ossy@webkit.org>
3069
3070         [Qt][WK2]REGRESSION(r104881):It broke hundreds of tests
3071         https://bugs.webkit.org/show_bug.cgi?id=76247
3072
3073         Reviewed by Kenneth Rohde Christiansen.
3074
3075         * WebKitTestRunner/TestController.cpp:
3076         (WTR::TestController::resetStateToConsistentValues): Disable mock scrollbars on Qt by default. (workaround)
3077
3078 2012-01-13  Hajime Morrita  <morrita@chromium.org>
3079
3080         [Mac][DRT] Should support --no-timeout
3081         https://bugs.webkit.org/show_bug.cgi?id=76242
3082
3083         Reviewed by Darin Adler.
3084
3085         * DumpRenderTree/mac/DumpRenderTree.mm:
3086         (initializeGlobalsFromCommandLineOptions): Added "--no-timeout" to the options list.
3087         (setWaitToDumpWatchdog): Added to encapsulate the waitToDumpWatchdog global variable.
3088         (shouldSetWaitToDumpWatchdog): Added to check the useTimeoutWatchdog flag set by the option.
3089         * DumpRenderTree/mac/DumpRenderTreeMac.h:
3090         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
3091         (LayoutTestController::setWaitToDump): Now calls shouldSetWaitToDumpWatchdog() to check the flag.
3092
3093 2012-01-12  Simon Fraser  <simon.fraser@apple.com>
3094
3095         Mitigate scrollbar differences when running pixel tests
3096         https://bugs.webkit.org/show_bug.cgi?id=67217
3097
3098         Reviewed by Dan Bernstein.
3099         
3100         Enable mock scrollbars by default for tests.
3101
3102         * DumpRenderTree/mac/DumpRenderTree.mm:
3103         (resetDefaultsToConsistentValues):
3104         * WebKitTestRunner/TestController.cpp:
3105         (WTR::TestController::resetStateToConsistentValues):
3106
3107 2012-01-12  Ryosuke Niwa  <rniwa@webkit.org>
3108
3109         Need a script to run tests in PerformanceTests
3110         https://bugs.webkit.org/show_bug.cgi?id=76132
3111
3112         Reviewed by Adam Barth.
3113
3114         Add run-perf-tests to run performance tests using parser/resources/runner.js.
3115         Unfortunately, there isn't an easy way of telling which test uses which format
3116         so hard-code directory that uses Chromium perf-bot style (only inspector perf. tests for now).
3117
3118         All test outputs are re-formatted to match Chromium perf-bot style.
3119
3120         * Scripts/run-inspector-perf-tests.py:
3121         * Scripts/run-perf-tests: Added.
3122         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3123         (PerfTestsRunner):
3124         (PerfTestsRunner.__init__):
3125         (PerfTestsRunner._collect_tests):
3126         (PerfTestsRunner.run):
3127         (PerfTestsRunner._print_status):
3128         (PerfTestsRunner._run_tests_set):
3129         (PerfTestsRunner._process_chromium_style_test_result):
3130         (PerfTestsRunner._should_ignore_line_in_parser_test_result):
3131         (PerfTestsRunner._process_parser_test_result):
3132         (PerfTestsRunner._run_single_test):
3133         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
3134         (MainTest.TestDriver.run_test):
3135         (create_runner):
3136         (run_test):
3137         (test_run_test_set):
3138         (test_run_test_set_for_parser_tests):
3139         (test_collect_tests):
3140
3141 2012-01-12  Dirk Pranke  <dpranke@chromium.org>
3142
3143         webkitpy: clean up port unit tests in preparation for making port_name mandatory
3144         https://bugs.webkit.org/show_bug.cgi?id=76123
3145
3146         Reviewed by Adam Barth.
3147
3148         This change updates the unit tests to be more consistent in how
3149         ports are created, in preparation for making port_name a
3150         mandatory parameter to the constructor and removing any version
3151         interpretation from it.
3152
3153         There should be no functional changes in this patch, but some
3154         tests that are no longer relevant have been removed or reworked.
3155
3156         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3157         (PortTest.test_get_option__default):
3158         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
3159         (ChromiumLinuxPortTest):
3160         (ChromiumLinuxPortTest.assert_architecture):
3161         (ChromiumLinuxPortTest.test_determine_architecture_fails):
3162         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
3163         (ChromiumMacPortTest):
3164         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3165         (ChromiumPortTest):
3166         (ChromiumPortTest.test_all_test_configurations):
3167         (ChromiumPortTest.TestMacPort.__init__):
3168         (ChromiumPortTest.TestLinuxPort.__init__):
3169         (ChromiumPortTest.TestWinPort.__init__):
3170         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
3171         (ChromiumWinTest):
3172         * Scripts/webkitpy/layout_tests/port/efl_unittest.py:
3173         (EflPortTest):
3174         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
3175         (GtkPortTest):
3176         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3177         (MacTest):
3178         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
3179         (MockDRTPortTest):
3180         (MockDRTPortTest.make_port):
3181         * Scripts/webkitpy/layout_tests/port/qt.py:
3182         (QtPort._port_flag_for_scripts):
3183         (QtPort.__init__):
3184         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
3185         (QtPortTest):
3186         (QtPortTest._assert_search_path):
3187         (QtPortTest.test_baseline_search_path):
3188         (QtPortTest.test_operating_system):
3189         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3190         (WebKitPortTest):
3191         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
3192         (WinPortTest):
3193         (WinPortTest._assert_search_path):
3194         (WinPortTest.test_baseline_search_path):
3195         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
3196         (get_test_config.TestMacPort.__init__):
3197
3198 2012-01-12  Dirk Pranke  <dpranke@chromium.org>
3199
3200         test-webkitpy: clean up logging and provide a real --help message
3201         https://bugs.webkit.org/show_bug.cgi?id=76142
3202
3203         Reviewed by Adam Barth.
3204
3205         This test is mostly more refactoring - it calls objects in
3206         unittest directly instead of routing through unittest.main() --
3207         in order to gain more control over the logging and to provide a
3208         correct --help message. This will also give us a better
3209         foundation for adding more feautures down the road.
3210
3211         This patch adds different levels of logging controlled by the
3212         '-s', '-q', and '-v' options (see --help for details), and
3213         removes the --verbose-logging option.
3214
3215         * Scripts/test-webkitpy:
3216         * Scripts/webkitpy/test/main.py:
3217         (Tester):
3218         (Tester.clean_packages):
3219         (Tester.__init__):
3220         (Tester.parse_args):
3221         (Tester.configure):
3222         (Tester._configure_logging):
3223         (Tester.find_modules):
3224         (Tester._exclude):
3225         (Tester._modules_from_paths):
3226         (Tester.run_tests):
3227
3228 2012-01-12  Simon Hausmann  <simon.hausmann@nokia.com>
3229
3230         Make the new WTF module build on Qt
3231         https://bugs.webkit.org/show_bug.cgi?id=76163
3232
3233         Reviewed by Tor Arne Vestbø.
3234
3235         * qmake/mkspecs/modules/wtf.prf: Pull in WTF from the new location.
3236
3237 2012-01-12  Vsevolod Vlasov  <vsevik@chromium.org>
3238
3239         Unreviewed, mac DumpRenderTree fix.
3240
3241         * DumpRenderTree/mac/UIDelegate.mm:
3242         (-[UIDelegate webView:addMessageToConsole:withSource:]):
3243
3244 2012-01-12  Vsevolod Vlasov  <vsevik@chromium.org>
3245
3246         WebKitTestRunner should not print console message line number when it is equal to zero.
3247         https://bugs.webkit.org/show_bug.cgi?id=76167
3248
3249         Reviewed by Csaba Osztrogonác.
3250
3251         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
3252         (WTR::InjectedBundlePage::willAddMessageToConsole):
3253
3254 2012-01-11  Vsevolod Vlasov  <vsevik@chromium.org>
3255
3256         Make default console messages line numbers consistent.
3257         https://bugs.webkit.org/show_bug.cgi?id=74075
3258
3259         Reviewed by Pavel Feldman.
3260
3261         Unset line numbers are not printed to console now.
3262         Added Source/WebCore/inspector to efl DumpRenderTree include path.
3263
3264         * DumpRenderTree/chromium/WebViewHost.cpp:
3265         (WebViewHost::didAddMessageToConsole):
3266         * DumpRenderTree/efl/CMakeLists.txt:
3267         * DumpRenderTree/efl/DumpRenderTreeView.cpp:
3268         (onConsoleMessage):
3269         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3270         (webViewConsoleMessage):
3271         * DumpRenderTree/mac/UIDelegate.mm:
3272         (-[UIDelegate webView:addMessageToConsole:withSource:]):
3273         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
3274         (WebCore::WebPage::javaScriptConsoleMessage):
3275         * DumpRenderTree/win/UIDelegate.cpp:
3276         (UIDelegate::webViewAddMessageToConsole):
3277         * DumpRenderTree/wx/DumpRenderTreeWx.cpp:
3278         (LayoutWebViewEventHandler::OnConsoleMessageEvent):
3279
3280 2012-01-10  Dirk Pranke  <dpranke@chromium.org>
3281
3282         test-webkitpy: push more logic into webkitpy.test.main, clean up code
3283         https://bugs.webkit.org/show_bug.cgi?id=76021
3284
3285         Reviewed by Eric Seidel.
3286
3287         I plan to add more functionality to test-webkitpy, but it is
3288         difficult to hack on now; this patch is the first of two that
3289         brings test-webkitpy more inline with current coding style by
3290         pushing more logic into the webkitpy.test.main.Tester class (so
3291         that it will be testable itself).
3292         
3293         There should be no functional changes in this patch.
3294
3295         * Scripts/test-webkitpy:
3296         (_path_from_webkit_root):
3297         * Scripts/webkitpy/test/main.py:
3298         (Tester):
3299         (Tester.init):
3300         (Tester.configure_logging):
3301         (Tester.configure_logging.filter):
3302         (Tester.clean_packages):
3303         (Tester.run_tests):
3304
3305 2012-01-11  Wei Jia  <wjia@chromium.org>
3306
3307         implement layout tests for <video> with media stream
3308         https://bugs.webkit.org/show_bug.cgi?id=74882
3309
3310         Reviewed by Kent Tamura.
3311
3312         * DumpRenderTree/DumpRenderTree.gypi:
3313         * DumpRenderTree/chromium/WebUserMediaClientMock.cpp: Added.
3314         (WebKit::WebUserMediaClientMock::create):
3315         (WebKit::WebUserMediaClientMock::IsMockStream):
3316         (WebKit::WebUserMediaClientMock::requestUserMedia):
3317         (WebKit::WebUserMediaClientMock::cancelUserMediaRequest):
3318         * DumpRenderTree/chromium/WebUserMediaClientMock.h: Added.
3319         (WebKit::WebUserMediaClientMock::~WebUserMediaClientMock):
3320         (WebKit::WebUserMediaClientMock::WebUserMediaClientMock):
3321         * DumpRenderTree/chromium/WebViewHost.cpp:
3322         (WebViewHost::userMediaClient):
3323         (WebViewHost::userMediaClientMock):
3324         (WebViewHost::createMediaPlayer):
3325         (WebViewHost::mediaStreamUtil):
3326         (WebViewHost::testMediaStreamClient):
3327         * DumpRenderTree/chromium/WebViewHost.h:
3328
3329 2012-01-11  Dirk Pranke  <dpranke@chromium.org>
3330
3331         Unreviewed, build fix.
3332
3333         Fix chromium-win's version remapping to work correctly on Win 7 SP1.
3334
3335         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
3336         (ChromiumWinPort.__init__):
3337
3338 2012-01-11  Ryosuke Niwa  <rniwa@webkit.org>
3339
3340         Sheriffbot build fix. Don't emit empty lines on IRC.
3341
3342         * Scripts/webkitpy/tool/bot/irc_command.py:
3343         (LastGreenRevision.execute):
3344
3345 2012-01-11  Ryosuke Niwa  <rniwa@webkit.org>
3346
3347         last-green-revision should report the revision that succeeded on all bots
3348         https://bugs.webkit.org/show_bug.cgi?id=76109
3349
3350         Reviewed by Adam Barth.
3351
3352         Add an algorithm to find the last known good revision among the specified bots. For a revision to be
3353         considered green by this algorithm, all matching builders must have a successful run at the revision
3354         or two consecutive successful runs before and after the revision.
3355
3356         Also fixed a bug in irc_command that the result wasn't posted on IRC properly and a bug in queries.py
3357         that resulted in an exception when the user didn't pass BUILDER_NAME argument.
3358
3359         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
3360         (BuildBot._revisions_for_builder):
3361         (BuildBot):
3362         (BuildBot._find_green_revision):
3363         (BuildBot.last_green_revision):
3364         * Scripts/webkitpy/common/net/buildbot/buildbot_mock.py:
3365         (MockBuildBot.last_green_revision):
3366         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
3367         (test_revisions_for_builder):
3368         (test_find_green_revision):
3369         (test_last_green_revision):
3370         * Scripts/webkitpy/tool/bot/irc_command.py:
3371         (LastGreenRevision.execute):
3372         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
3373         (SheriffIRCBotTest.test_lgr):
3374         * Scripts/webkitpy/tool/commands/queries.py:
3375         (LastGreenRevision):
3376
3377 2012-01-11  Dirk Pranke  <dpranke@chromium.org>
3378
3379         Unreviewed, fix build bustage on win32 introduced by r104725.
3380
3381         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
3382         (MockDRTPortTest.make_port):
3383         (MockChromiumDRTTest.test_pixeltest__fails):
3384
3385 2012-01-11  Jochen Eisinger  <jochen@chromium.org>
3386
3387         [Chromium] mimic the (old) behavior of the mac port for (re)setting the color profile
3388         https://bugs.webkit.org/show_bug.cgi?id=75618
3389
3390         Eventually, we should get away from setting the display's color profile, as the mac port is doing now.
3391
3392         Reviewed by Tony Chang.
3393
3394         * DumpRenderTree/chromium/LayoutTestHelper.mm:
3395         (installLayoutTestColorProfile):
3396         (restoreUserColorProfile):
3397         (main):
3398
3399 2012-01-11  Dirk Pranke  <dpranke@chromium.org>
3400
3401         test-webkitpy: fix -v and eliminate some spurious warnings
3402         https://bugs.webkit.org/show_bug.cgi?id=76099
3403
3404         Reviewed by Adam Barth.
3405
3406         when we added support for --xml output, we broke the -v flag.
3407
3408         Also, fix the code to swallow a harmless warning from
3409         VCSUtils.pm and fix a deprecated warning in jsonchecker.py.
3410
3411         test-webkitpy is now silent again except for the one known
3412         logging issue.
3413
3414         * Scripts/webkitpy/common/checkout/checkout.py:
3415         (Checkout.commit_message_for_this_commit):
3416         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
3417         (test_commit_message_for_this_commit):
3418         * Scripts/webkitpy/style/checkers/jsonchecker.py:
3419         (JSONChecker.check):
3420         (JSONChecker.line_number_from_json_exception):
3421         * Scripts/webkitpy/test/main.py:
3422         (Tester.run_tests):
3423
3424 2012-01-11  Dirk Pranke  <dpranke@chromium.org>
3425
3426         webkitpy: webkitpy.layout_tests.port tests should only use MockSystemHost, not MockHost
3427         https://bugs.webkit.org/show_bug.cgi?id=76084
3428
3429         Reviewed by Adam Barth.
3430
3431         This will help make it easier to catch layering violations. This
3432         change just refactors test code; there are no changes to
3433         production code.
3434
3435         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3436         (PortTest.make_port):
3437         (PortTest.test_find_no_paths_specified):
3438         (PortTest.test_find_one_test):
3439         (PortTest.test_find_glob):
3440         (PortTest.test_find_with_skipped_directories):
3441         (PortTest.test_find_with_skipped_directories_2):
3442         (PortTest.test_parse_reftest_list):
3443         (PortTest.test_httpd_returns_error_code):
3444         (PortTest.test_virtual_methods):
3445         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
3446         (ChromiumGpuTest.assert_port_works):
3447         (ChromiumGpuTest._assert_baseline_path):
3448         (ChromiumGpuTest.test_graphics_type):
3449         (ChromiumGpuTest.test_default_tests_paths.test_paths):
3450         (ChromiumGpuTest.test_test_files):
3451         (ChromiumGpuTest.test_test_files.test_paths):
3452         (ChromiumGpuTest):
3453         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
3454         (ChromiumLinuxPortTest.assert_architecture):
3455         (ChromiumLinuxPortTest.test_check_illegal_port_names):
3456         (ChromiumLinuxPortTest.test_determine_architecture_fails):
3457         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
3458         (DriverTest.make_port):
3459         (DriverTest):
3460         (DriverTest.assertVirtual):
3461         (DriverTest._assert_wrapper):
3462         (DriverTest.test_virtual_driver_methods):
3463         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
3464         (MockDRTPortTest.make_port):
3465         (MockDRTPortTest.test_port_name_in_constructor):
3466         (MockDRTTest.assertTest):
3467         (MockDRTTest.test_main):
3468         (MockChromiumDRTTest.test_pixeltest__fails):
3469         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
3470         (QtPortTest._assert_search_path):
3471         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3472         (TestWebKitPort.__init__):
3473         (WebKitPortUnitTests.test_default_options):
3474         (test_runtime_feature_list):
3475         (test_test_expectations):
3476
3477 2012-01-11  Ojan Vafai  <ojan@chromium.org>
3478
3479         Style skipped tests in the flakiness dashboard
3480         https://bugs.webkit.org/show_bug.cgi?id=76091
3481
3482         Reviewed by Tony Chang.
3483
3484         * TestResultServer/static-dashboards/dashboard_base.js:
3485         Removed "FLAKY" from this list since we don't ever mark individual runs
3486         as flaky. Also deleted an outdated line.
3487         * TestResultServer/static-dashboards/flakiness_dashboard.html:
3488
3489 2012-01-11  Ojan Vafai  <ojan@chromium.org>
3490
3491         Remove gtest normalization code from the test results server
3492         https://bugs.webkit.org/show_bug.cgi?id=76089
3493
3494         Reviewed by Tony Chang.
3495
3496         The normalization now happens on the gtest side.
3497
3498         * TestResultServer/model/jsonresults.py:
3499         (JsonResults.merge):
3500         * TestResultServer/model/jsonresults_unittest.py:
3501         (JsonResultsTest.test_gtest):
3502
3503 2012-01-11  Ojan Vafai  <ojan@chromium.org>
3504
3505         Be more aggressive about deleting unnecessary results from the test-results-server json files
3506         https://bugs.webkit.org/show_bug.cgi?id=76086
3507
3508         Reviewed by Tony Chang.
3509
3510         Remove any tests that only contain skip/pass/nodata values whose runtime
3511         is less than 5 seconds.
3512
3513         * TestResultServer/model/jsonresults.py:
3514         (JsonResults._normalize_results_json):
3515         * TestResultServer/model/jsonresults_unittest.py:
3516         (JsonResultsTest.test_merge_remove_test):
3517         (JsonResultsTest.test_merge_keep_test_with_all_pass_but_slow_time):
3518
3519 2012-01-11  Ryosuke Niwa  <rniwa@webkit.org>
3520
3521         last-green-revision should give us per-bot information
3522         https://bugs.webkit.org/show_bug.cgi?id=76011
3523
3524         Reviewed by Adam Barth.
3525
3526         Rewrote last-green-revision command. Instead of finding a revision for which all bots succeeded,
3527         we report the latest green run on each bot from the last 100 runs.
3528
3529         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
3530         (BuildBot._fetch_builder_page):
3531         (BuildBot):
3532         (BuildBot._green_revision_for_builder):
3533         (BuildBot.last_green_revision):
3534         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
3535         (test_green_revision_for_builder):
3536         (test_last_green_revision):
3537         * Scripts/webkitpy/tool/bot/irc_command.py:
3538         (LastGreenRevision.execute):
3539         * Scripts/webkitpy/tool/commands/queries.py:
3540         (LastGreenRevision.execute):
3541
3542 2012-01-11  Dirk Pranke  <dpranke@chromium.org>
3543
3544         webkitpy: clean up version detection in webkitpy.layout_tests.port
3545         https://bugs.webkit.org/show_bug.cgi?id=76016
3546
3547         Reviewed by Adam Barth.
3548
3549         This change updates the code in webkitpy.layout_tests.port to
3550         use the new os_name and os_version fields on host.platform. This
3551         gets layout_tests.port out of the version parsing business. Most
3552         of the diffs are just cleaning up test code to accomodate the
3553         changes.
3554
3555         * Scripts/webkitpy/common/system/executive_mock.py:
3556         (MockExecutive):
3557         (MockExecutive.ignore_error):
3558         (MockExecutive2):
3559         (MockExecutive2.ignore_error):
3560         * Scripts/webkitpy/common/system/platforminfo_mock.py:
3561         (MockPlatformInfo.is_mac):
3562         (MockPlatformInfo.is_linux):
3563         (MockPlatformInfo.is_win):
3564         * Scripts/webkitpy/common/system/systemhost_mock.py:
3565         (MockSystemHost.__init__):
3566         * Scripts/webkitpy/layout_tests/port/apple.py:
3567         (ApplePort.__init__):
3568         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
3569         (ChromiumMacPort.__init__):
3570         (ChromiumMacPort.check_wdiff):
3571         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
3572         (ChromiumMacPortTest):
3573         (ChromiumMacPortTest.assert_name):
3574         (ChromiumMacPortTest.test_versions):
3575         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3576         (ChromiumPortTest.test_all_test_configurations):
3577         (ChromiumPortTest.TestMacPort.__init__):
3578         (ChromiumPortTest.TestLinuxPort.__init__):
3579         (ChromiumPortTest.TestWinPort.__init__):
3580         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
3581         (ChromiumWinPort.__init__):
3582         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
3583         (ChromiumWinTest):
3584         (ChromiumWinTest.assert_name):
3585         (ChromiumWinTest.test_versions):
3586         (ChromiumWinTest.test_baseline_path):
3587         (ChromiumWinTest.test_build_path):
3588         * Scripts/webkitpy/layout_tests/port/factory.py:
3589         (PortFactory._port_name_from_arguments_and_options):
3590         (PortFactory.get):
3591         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
3592         (FactoryTest):
3593         (FactoryTest.setUp):
3594         (FactoryTest.assert_port):
3595         (FactoryTest.test_mac):
3596         (FactoryTest.test_win):
3597         (FactoryTest.test_google_chrome):
3598         (FactoryTest.test_gtk):
3599         (FactoryTest.test_qt):
3600         (FactoryTest.test_chromium_gpu):
3601         (FactoryTest.test_chromium_gpu_linux):
3602         (FactoryTest.test_chromium_gpu_mac):
3603         (FactoryTest.test_chromium_gpu_win):
3604         (FactoryTest.test_chromium_mac):
3605         (FactoryTest.test_chromium_linux):
3606         (FactoryTest.test_chromium_win):
3607         (FactoryTest.test_unknown_specified):
3608         (FactoryTest.test_unknown_default):
3609         * Scripts/webkitpy/layout_tests/port/google_chrome_unittest.py:
3610         (TestGoogleChromePort):
3611         (TestGoogleChromePort._verify_baseline_path):
3612         (TestGoogleChromePort._verify_expectations_overrides):
3613         (TestGoogleChromePort.test_get_google_chrome_port):
3614         * Scripts/webkitpy/layout_tests/port/mac.py:
3615         (MacPort):
3616         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3617         (MacTest):
3618         (assert_name):
3619         (test_tests_for_other_platforms):
3620         (test_version):
3621         (test_versions):
3622         (test_is_version_methods):
3623         (test_setup_environ_for_server):
3624         (_assert_search_path):
3625         (test_show_results_html_file):
3626         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
3627         (PortTestCase):
3628         (PortTestCase.make_port):
3629         * Scripts/webkitpy/layout_tests/port/win.py:
3630         (WinPort):
3631         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
3632         (WinPortTest):
3633         (WinPortTest.test_show_results_html_file):
3634         (WinPortTest._assert_search_path):
3635         (WinPortTest._assert_version):
3636
3637 2012-01-11  Hugo Parente Lima  <hugo.lima@openbossa.org>
3638
3639         [Qt][WK2] Remove workaround introduced into bug 75470
3640         https://bugs.webkit.org/show_bug.cgi?id=75889
3641
3642         Reviewed by Tor Arne Vestbø.
3643
3644         * MiniBrowser/qt/qml/BrowserWindow.qml:
3645
3646 2012-01-10  Dmitry Lomov  <dslomov@google.com>
3647
3648         [Chromium] Remove obsolete references to WebWorker class.
3649         https://bugs.webkit.org/show_bug.cgi?id=76020
3650
3651         Reviewed by David Levin.
3652
3653         * DumpRenderTree/chromium/TestWebWorker.h: Removed.
3654         * DumpRenderTree/chromium/WebViewHost.cpp:
3655         * DumpRenderTree/chromium/WebViewHost.h:
3656
3657 2012-01-11  Csaba Osztrogonác  <ossy@webkit.org>
3658
3659         [Qt] Some css3 filter tests are failing after r104698
3660         https://bugs.webkit.org/show_bug.cgi?id=76055
3661
3662         Reviewed by Simon Hausmann.
3663
3664         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
3665         (LayoutTestController::overridePreference): Add WebKitAcceleratedCompositingEnabled preference.
3666
3667 2012-01-11  Tony Chang  <tony@chromium.org>
3668
3669         Don't remove 0 byte files when running build-webkit --chromium
3670         https://bugs.webkit.org/show_bug.cgi?id=75997
3671
3672         Reviewed by Csaba Osztrogonác.
3673
3674         If QTDIR is set, isQt() is true even when running build-webkit --chromium.
3675
3676         * Scripts/webkitdirs.pm:
3677         (determineIsQt):
3678
3679 2012-01-10  David Kilzer  <ddkilzer@apple.com>
3680
3681         Add TestWebKitAPI tests for KURL
3682         <http://webkit.org/b/75774>
3683
3684         Reviewed by Adam Barth.
3685
3686         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Add a
3687         WebCore folder and KURL.cpp source file to the project.