971fefd930c3f59658d0388c8a93963d58c393be
[WebKit-https.git] / Tools / ChangeLog
1 2012-01-26  Ryosuke Niwa  <rniwa@webkit.org>
2
3         run-perf-test should include relative file paths in test names
4         https://bugs.webkit.org/show_bug.cgi?id=77144
5
6         Reviewed by Adam Birth.
7
8         Include all directory names in the test name separated by "/".
9         e.g. Bindings/event-target-wrapper.html translates to Bindings/event-target-wrapper
10
11         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
12         (PerfTestsRunner):
13         (PerfTestsRunner._process_parser_test_result):
14         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
15         (test_run_test_set_with_json_output):
16         (test_run_test_set_with_json_source):
17
18 2012-01-26  Dirk Pranke  <dpranke@chromium.org>
19
20         webkitpy: re-land cleanup of test scaffolding
21         https://bugs.webkit.org/show_bug.cgi?id=77153
22
23         Reviewed by Eric Seidel.
24
25         This is an attempt to reland the change in bug 76973 / 
26         r105935. My earlier attempt to manipulate PYTHONPATH
27         seems to only work some of the time, for reasons that
28         elude me. Directly adding Tools/Scripts to sys.path from
29         inside the test script appears to work (more?) reliably.
30
31         Also, it seemed like the files didn't actually get deleted
32         the last time; I don't know why.
33
34         * Scripts/test-webkitpy:
35         * Scripts/webkitpy/common/system/executive_unittest.py:
36         (command_line):
37         (ExecutiveTest):
38         (ExecutiveTest.test_run_command_args_type):
39         (ExecutiveTest.test_run_command_with_unicode):
40         (ExecutiveTest.test_running_pids):
41         (main):
42         * Scripts/webkitpy/common/system/fileutils.py: Removed.
43         * Scripts/webkitpy/test/cat.py: Removed.
44         * Scripts/webkitpy/test/cat_unittest.py: Removed.
45         * Scripts/webkitpy/test/echo.py: Removed.
46         * Scripts/webkitpy/test/echo_unittest.py: Removed.
47
48 2012-01-26  Ojan Vafai  <ojan@chromium.org>
49
50         run-webkit-tests calls "nm" when it doesn't need to
51         https://bugs.webkit.org/show_bug.cgi?id=77143
52
53         Reviewed by Dirk Pranke.
54
55         If we are not running any test that can be skipped by checking
56         the DRT feature list or looking at the list of symbols, avoid calling
57         out to DRT or calling "nm".
58
59         This saves ~700ms on my Mac Pro. In cases where you still have to call out
60         to "nm" it's only a couple ms slower.
61
62         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
63         (TestExpectations.__init__):
64         * Scripts/webkitpy/layout_tests/port/base.py:
65         (Port.skipped_tests):
66         * Scripts/webkitpy/layout_tests/port/webkit.py:
67         (WebKitPort._should_check_for_missing_features):
68         (WebKitPort):
69         (WebKitPort._should_check_for_missing_symbols):
70         (WebKitPort._skipped_tests_for_unsupported_features):
71         (WebKitPort.skipped_layout_tests):
72         (WebKitPort.skipped_tests):
73         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
74         (WebKitPortTest.test_skipped_directories_for_symbols):
75         (test_skipped_directories_for_features):
76         (test_skipped_directories_for_features_no_matching_tests_in_test_list):
77         (test_skipped_layout_tests):
78
79 2012-01-26  Benjamin Poulain  <bpoulain@apple.com>
80
81         Support a suffix on ChangeLog filenames based on a configuration file
82         https://bugs.webkit.org/show_bug.cgi?id=76956
83
84         Reviewed by Kentaro Hara.
85
86         Modify the tools used to generate, edit and resolve ChangeLogs to support
87         the addition of a suffix to the ChangeLog filename.
88
89         This patch makes it easier to support alternate ChangeLog files in branches.
90
91         * Scripts/VCSUtils.pm:
92         (changeLogSuffix):
93         (changeLogFileName):
94         * Scripts/commit-log-editor:
95         * Scripts/prepare-ChangeLog:
96         (getLatestChangeLogs):
97         (generateNewChangeLogs):
98         (generateFileList):
99         * Scripts/resolve-ChangeLogs:
100         (findChangeLog):
101
102 2012-01-26  Sheriff Bot  <webkit.review.bot@gmail.com>
103
104         Unreviewed, rolling out r106038.
105         http://trac.webkit.org/changeset/106038
106         https://bugs.webkit.org/show_bug.cgi?id=77142
107
108         Caused a bunch of skipped tests to not be skipped. (Requested
109         by ojan on #webkit).
110
111         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
112         (TestExpectationParser):
113         (TestExpectationParser.parse):
114         (TestExpectationParser._parse_line):
115         (TestExpectationParser._collect_matching_tests):
116         (TestExpectations.__init__):
117         (TestExpectations._add_skipped_tests):
118         * Scripts/webkitpy/layout_tests/port/base.py:
119         (Port.test_dirs):
120         (Port.normalize_test_name):
121         (Port.update_baseline):
122         (Port.layout_tests_dir):
123         (Port.relative_perf_test_filename):
124         (Port.abspath_for_test):
125
126 2012-01-25  Ojan Vafai  <ojan@chromium.org>
127
128         Parsing test_expecations.txt + Skipped lists takes too long
129         https://bugs.webkit.org/show_bug.cgi?id=77059
130
131         Reviewed by Dirk Pranke.
132
133         This saves ~100ms on the Apple Mac port.
134         -memoize a bunch of path methods.
135         -Avoid doing multiple disk accesses per line.
136         -Parse the skipped list directly instead of turning it into a test_expecations.txt
137         formatting string and parsing that.
138
139         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
140         (TestExpectationParser):
141         (TestExpectationParser.parse_skipped_tests):
142         (TestExpectationParser._parse_line):
143         (TestExpectationParser._collect_matching_tests):
144         (TestExpectations.__init__):
145         (TestExpectations._add_skipped_tests):
146         * Scripts/webkitpy/layout_tests/port/base.py:
147         (Port):
148         (Port.test_isfile):
149         (Port.normalize_test_name):
150         (Port.layout_tests_dir):
151         (Port.abspath_for_test):
152
153 2012-01-25  Ojan Vafai  <ojan@chromium.org>
154
155         Only store the SVN revision in the summarized results if we're on a builder
156         https://bugs.webkit.org/show_bug.cgi?id=76976
157
158         Reviewed by Dirk Pranke.
159
160         This shaves another ~130ms off the single test run-webkit-tests runtime.
161
162         * Scripts/webkitpy/layout_tests/controllers/manager.py:
163         (summarize_results):
164         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
165         (ResultSummaryTest.get_result):
166         (ResultSummaryTest):
167         (ResultSummaryTest.get_result_summary):
168         (ResultSummaryTest.get_unexpected_results):
169         (ResultSummaryTest.test_no_svn_revision):
170         (ResultSummaryTest.test_svn_revision):
171         * Scripts/webkitpy/layout_tests/port/base.py:
172         (Port.use_trac_links_in_results_html):
173
174 2012-01-26  Luciano Wolf  <luciano.wolf@openbossa.org>
175
176         [Qt][WK2] fast/loader tests failed after r94178
177         https://bugs.webkit.org/show_bug.cgi?id=67388
178
179         Reviewed by Simon Hausmann.
180
181         Applying the same fix used for Qt-WebKit1 from bug #67254 -
182         [Qt][DRT] Normalize file:///tmp/LayoutTests in LayoutTestController::pathToLocalResource()
183         https://bugs.webkit.org/show_bug.cgi?id=67254
184         Translate file:///tmp/LayoutTests/* urls into the repository LayoutTests directory, which is
185         derived from the running location of DumpRenderTree binary.
186
187         * WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
188         (WTR::LayoutTestController::pathToLocalResource):
189
190 2012-01-26  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
191
192         Implement overridePreference for boolean preferences in WebKitTestRunner
193         https://bugs.webkit.org/show_bug.cgi?id=77033
194
195         Reviewed by Alexey Proskuryakov.
196
197         For now overridePreference supports boolean values.
198
199         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
200         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
201         (WTR::LayoutTestController::setXSSAuditorEnabled):
202         (WTR::LayoutTestController::overridePreference):
203         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
204         * WebKitTestRunner/TestController.cpp:
205         (WTR::TestController::resetStateToConsistentValues): Reset the overrides
206         explicitly and properly disable the use of PageCache when resetting the
207         state. Tests that use PageCache can now enable it via overridePreference.
208
209 2012-01-26  Alexander Færøy  <alexander.faeroy@nokia.com>
210
211         [Qt] Support opening multiple URL's from the command line in the Qt MiniBrowser
212         https://bugs.webkit.org/show_bug.cgi?id=77100
213
214         Reviewed by Simon Hausmann.
215
216         This patch makes the MiniBrowser read multiple URL's from the command
217         line such that there will be one browser window opened for each URL.
218
219         * MiniBrowser/qt/MiniBrowserApplication.cpp:
220         (MiniBrowserApplication::handleUserOptions):
221
222 2012-01-26  Alexander Færøy  <alexander.faeroy@nokia.com>
223
224         [Qt] The Qt MiniBrowser crashes when given multiple URL's
225         https://bugs.webkit.org/show_bug.cgi?id=77009
226
227         Reviewed by Simon Hausmann.
228
229         This patch fixes a crash in BrowserWindow::newWindow() where a
230         BrowserWindow-object is constructed with a 0 pointer as the
231         WindowOptions parameter which leads to a crash.
232
233         * MiniBrowser/qt/BrowserWindow.cpp:
234         (BrowserWindow::BrowserWindow):
235         (BrowserWindow::newWindow):
236         * MiniBrowser/qt/BrowserWindow.h:
237
238 2012-01-26  Jesus Sanchez-Palencia  <jesus.palencia@openbossa.org>
239
240         [Qt] Use ICU if available
241         https://bugs.webkit.org/show_bug.cgi?id=76821
242
243         Reviewed by Simon Hausmann.
244
245         Adding unicode defines for a Qt 5 based build.
246
247         * qmake/mkspecs/features/features.prf:
248
249 2012-01-26  Nándor Huszka  <huszka.nandor@stud.u-szeged.hu>
250
251         [WK2] WebKitTestRunner needs layoutTestController.setGeolocationPermission
252         https://bugs.webkit.org/show_bug.cgi?id=42545
253
254         Reviewed by Zoltan Herczeg.
255
256         Implement the setGeolocationPermission method.
257
258         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
259         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
260         (WTR::LayoutTestController::setGeolocationPermission):
261         (WTR):
262         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
263         (LayoutTestController):
264
265 2012-01-26  Balazs Kelemen  <kbalazs@webkit.org>
266
267         [NRWT] Support --ignore-metrics
268         https://bugs.webkit.org/show_bug.cgi?id=76278
269
270         Reviewed by Tony Chang.
271
272         Port the feature from ORWT.
273
274         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
275         (SingleTestRunner._run_compare_test):
276         * Scripts/webkitpy/layout_tests/port/driver.py:
277         (DriverOutput):
278         (DriverOutput.strip_metrics):
279         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
280         (DriverOutputTest):
281         (DriverOutputTest.test_strip_metrics):
282         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
283         (_set_up_derived_options):
284         (parse_args):
285
286 2012-01-26  Sheriff Bot  <webkit.review.bot@gmail.com>
287
288         Unreviewed, rolling out r105935 and r105954.
289         http://trac.webkit.org/changeset/105935
290         http://trac.webkit.org/changeset/105954
291         https://bugs.webkit.org/show_bug.cgi?id=77080
292
293         test-webkitpy is still broken (Requested by abarth on
294         #webkit).
295
296         * Scripts/test-webkitpy:
297         * Scripts/webkitpy/common/system/executive_unittest.py:
298         (never_ending_command):
299         (ExecutiveTest.test_run_command_args_type):
300         (ExecutiveTest.test_run_command_with_unicode):
301         (ExecutiveTest.test_running_pids):
302         * Scripts/webkitpy/common/system/fileutils.py:
303         (make_stdout_binary):
304         * Scripts/webkitpy/test/cat.py:
305         (command_arguments):
306         (main):
307         * Scripts/webkitpy/test/cat_unittest.py:
308         (CatTest):
309         (CatTest.assert_cat):
310         (CatTest.test_basic):
311         (CatTest.test_no_newline):
312         (CatTest.test_unicode):
313         (CatTest.test_as_command):
314         * Scripts/webkitpy/test/echo.py:
315         (command_arguments):
316         (main):
317         * Scripts/webkitpy/test/echo_unittest.py:
318         (EchoTest):
319         (EchoTest.test_basic):
320         (EchoTest.test_no_newline):
321         (EchoTest.test_unicode):
322         (EchoTest.test_argument_order):
323         (EchoTest.test_empty_arguments):
324         (EchoTest.test_no_arguments):
325         (EchoTest.test_as_command):
326
327 2012-01-26  Gabor Rapcsanyi  <rgabor@webkit.org>
328
329         [GTK] Detect cross-compilation in webkitdirs.pm to not generate gtkdoc
330         https://bugs.webkit.org/show_bug.cgi?id=76138
331
332         Reviewed by Martin Robinson.
333
334         * Scripts/webkitdirs.pm:
335         (isCrossCompilation):
336         (buildAutotoolsProject):
337
338 2012-01-25  Adam Barth  <abarth@webkit.org>
339
340         Kentaro is a reviewer now.  Yay!
341
342         * Scripts/webkitpy/common/config/committers.py:
343
344 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
345
346         nrwt: running w/ no arguments fails
347         https://bugs.webkit.org/show_bug.cgi?id=77064
348
349         Unreviewed, build fix.
350
351         This change fixes a regression introduced in r105944
352         where running without --platform <foo> or one of the other
353         platform switches would cause a null deref :(.
354
355         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
356         (main):
357
358 2012-01-25  Eric Seidel  <eric@webkit.org>
359
360         Automatically CC someone from Apple on any changes to Obj-C public API
361         https://bugs.webkit.org/show_bug.cgi?id=77053
362
363         Reviewed by Timothy Hatcher.
364
365         This has long been de-facto practice, seems now that we have
366         fancy watchlist support, we should just make the tools
367         do this automatically.
368
369         * Scripts/webkitpy/common/config/watchlist:
370
371 2012-01-25  Antoine Labour  <piman@chromium.org>
372
373         Updating status in committers.py.
374
375         Unreviewed.
376
377         * Scripts/webkitpy/common/config/committers.py:
378
379 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
380
381         test-webkitpy: executive_unittest fails if PYTHONPATH isn't set
382         https://bugs.webkit.org/show_bug.cgi?id=77058
383
384         Reviewed by Eric Seidel.
385
386         r105935 broke executive_unittest by changing which script we
387         were running and not ensuring that the PYTHONPATH was set up
388         properly.
389
390         This change also fixes a regression introduced recently where
391         test-webkitpy wasn't figuring out the path to Tools/Scripts
392         right, either (so running ./test-webkitpy would fail).
393
394         * Scripts/test-webkitpy:
395         * Scripts/webkitpy/common/system/executive_unittest.py:
396         (script_dir):
397         (ExecutiveTest.setUp):
398         (ExecutiveTest):
399         (ExecutiveTest.tearDown):
400
401 2012-01-25  Adam Barth  <abarth@webkit.org>
402
403         Disable tests on the mac-ews while we sort out hte stability issues.
404
405         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
406         (MacEWS):
407
408 2012-01-23  Dirk Pranke  <dpranke@chromium.org>
409
410         investigate use of 'mac' and 'win' as fully-specified port names for the apple ports
411         https://bugs.webkit.org/show_bug.cgi?id=76475
412
413         Reviewed by Eric Seidel.
414
415         This patch cleans up the internal usage of port names for the
416         Apple mac and Apple win ports so that 'mac' and 'win are never
417         considered "fully specified" port names: they are still legal
418         input to the PortFactory.get() routine, but only if run on the given
419         platform (i.e. trying to PortFactory.get('mac') on windows or
420         linux will return an error, since we don't know which version of
421         the port is desired.
422
423         This also cleans up handling of the webkit2 variants, so that
424         the full port names consistently follow the
425         <operating_system>-<version>[-<wk2>] convention.
426
427         Lastly this patch adds some assertions and removes some
428         try/catch logic to catch more programming and usage errors;
429         previously mac-tiger might've been allowed -- and it would've
430         translated into using just the baselines in platform/mac -- but
431         now it should fail.
432
433         * Scripts/webkitpy/layout_tests/port/apple.py:
434         (ApplePort.determine_full_port_name):
435         (ApplePort.__init__):
436         (ApplePort):
437         (ApplePort._port_name_with_version):
438         * Scripts/webkitpy/layout_tests/port/builders.py:
439         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
440         (FactoryTest.test_mac):
441         (FactoryTest.test_win):
442         * Scripts/webkitpy/layout_tests/port/mac.py:
443         (MacPort.baseline_search_path):
444         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
445         (MacTest.test_versions):
446         * Scripts/webkitpy/layout_tests/port/win.py:
447         (WinPort.baseline_search_path):
448         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
449         (WinTest.test_versions):
450
451 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
452
453         webkitpy: print nicer errors while linting expectations files, remove redundant tests
454         https://bugs.webkit.org/show_bug.cgi?id=76955
455
456         Reviewed by Adam Barth.
457
458         This change cleans up the errors that are printed so that the
459         are more "quickfix" compatible (path:lineno).
460
461         Also, there were a bunch of redundant tests between
462         layout_tests.models.test_expectations_unittest and
463         style.checkers.test_expectations_unittest. The latter should
464         just expect the former to do most of the testing.
465
466         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
467         (TestExpectationParser._parse_modifiers):
468         (TestExpectations._report_errors):
469         (TestExpectations._add_skipped_tests):
470         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
471         (test_parse_error_nonfatal):
472         (SemanticTests.test_bad_bugid):
473         * Scripts/webkitpy/style/checkers/test_expectations.py:
474         (TestExpectationsChecker.__init__):
475         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
476         (TestExpectationsTestCase.assert_lines_lint):
477         (TestExpectationsTestCase.test_valid_expectations):
478         (TestExpectationsTestCase.test_invalid_expectations):
479         (TestExpectationsTestCase.test_tab):
480
481 2012-01-25  Eric Seidel  <eric@webkit.org>
482
483         webkit-patch apply-* should always continue after failures
484         https://bugs.webkit.org/show_bug.cgi?id=77057
485
486         Reviewed by Adam Barth.
487
488         As far as I can tell there is only one potential drawback to always
489         forcing: that if you're somehow in the wrong directory it will create new
490         directories for new files.  Since webkit-patch always cd's to the root
491         it seems that's not a drawback.  The drawback of not using --force for
492         svn-apply, is that it will stop after the first failure, which is rarely
493         (if ever) the desired behavior.  This just removes the force option
494         (which was strangely hidden behind --non-interactive).  This should
495         make for a better user experiance.
496
497         * Scripts/webkitpy/common/checkout/checkout.py:
498         (Checkout.apply_patch):
499         * Scripts/webkitpy/common/checkout/checkout_mock.py:
500         (MockCheckout.apply_patch):
501         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
502         (CheckoutTest.test_chromium_deps):
503         (CheckoutTest):
504         (CheckoutTest.test_apply_patch):
505         * Scripts/webkitpy/tool/commands/download_unittest.py:
506         (DownloadCommandsTest._default_options):
507         * Scripts/webkitpy/tool/steps/applypatch.py:
508         (ApplyPatch.options):
509         (ApplyPatch.run):
510         * Scripts/webkitpy/tool/steps/options.py:
511         (Options):
512
513 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
514
515         nrwt: should be able to run --platform test interactively
516         https://bugs.webkit.org/show_bug.cgi?id=76959
517
518         Reviewed by Adam Barth.
519
520         As part of refactoring the Port interfaces, in r103254 I made
521         passing a Host object to the Port mandatory; previously TestPort
522         objects would create their own (mock) hosts. However,
523         new-run-webkit-tests always passed a real Host, not a MockHost,
524         and so you couldn't run new-run-webkit-tests --platform test
525         interactively to debug test failures.
526
527         This change fixes that by creating a MockHost instead of a real
528         host if you say --platform test*.
529
530         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
531         (main):
532
533 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
534
535         run-webkit-tests --lint-test-files should lint all the ports by default
536         https://bugs.webkit.org/show_bug.cgi?id=76749
537
538         Reviewed by Ojan Vafai.
539
540         Currently run-webkit-tests --lint-test-files will only lint the
541         test_expectations for the single port determined by the command
542         line options. This is not obvious and can produce unintended results
543         if you want to lint the chromium port (probably the normal case
544         for using --lint-test-files) but don't specify a port name or
545         --chromium.
546
547         I think we should probably lint *all* of the ports if no port
548         name is supplied. This change implements that, and also moves
549         the linting logic out of the Manager class, since this really
550         has nothing to do with Managers. It has little to do with
551         run-webkit-tests, but splitting it into a different script would
552         introduce compatibility issues. It might make sense to do that
553         anyway, in a separate patch and combining that with cleaning up
554         the style checker to share more code.
555
556         * Scripts/webkitpy/layout_tests/controllers/manager.py:
557         (Manager._strip_test_dir_prefix):
558         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
559         (lint):
560         (run):
561         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
562         (LintTest.test_all_configurations): added.
563         (LintTest.test_lint_test_files): added.
564         (LintTest.test_lint_test_files__errors): added.
565         (MainTest.test_lint_test_files): removed.
566         (MainTest.test_lint_test_files__errors): removed.
567
568 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
569
570         webkitpy: clean up a bunch of test scaffolding
571         https://bugs.webkit.org/show_bug.cgi?id=76973
572
573         Reviewed by Eric Seidel.
574
575         There were several helper modules in the webkitpy.test package
576         that appear to only be used by executive_unittest.py. I am
577         rolling them all into that file to make that clearer.
578
579         * Scripts/webkitpy/common/system/executive_unittest.py:
580         (ExecutiveTest.test_run_command_args_type):
581         (ExecutiveTest.test_run_command_with_unicode):
582         (ExecutiveTest.test_running_pids):
583         (command_line):
584         (main):
585         * Scripts/webkitpy/common/system/fileutils.py: Removed.
586         * Scripts/webkitpy/test/cat.py: Removed.
587         * Scripts/webkitpy/test/cat_unittest.py: Removed.
588         * Scripts/webkitpy/test/echo.py: Removed.
589         * Scripts/webkitpy/test/echo_unittest.py: Removed.
590
591 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
592
593         webkitpy: clean up a bunch of test scaffolding
594         https://bugs.webkit.org/show_bug.cgi?id=76973
595
596         Reviewed by Eric Seidel.
597
598         There were several helper modules in the webkitpy.test package
599         that appear to only be used by executive_unittest.py. I am
600         rolling them all into that file to make that clearer.
601
602         * Scripts/webkitpy/common/system/executive_unittest.py:
603         (ExecutiveTest.test_run_command_args_type):
604         (ExecutiveTest.test_run_command_with_unicode):
605         (ExecutiveTest.test_running_pids):
606         (command_line):
607         (main):
608         * Scripts/webkitpy/common/system/fileutils.py: Removed.
609         * Scripts/webkitpy/test/cat.py: Removed.
610         * Scripts/webkitpy/test/cat_unittest.py: Removed.
611         * Scripts/webkitpy/test/echo.py: Removed.
612         * Scripts/webkitpy/test/echo_unittest.py: Removed.
613
614 2012-01-25  Eric Seidel  <eric@webkit.org>
615
616         Remove Python 2.5 support from WebKit
617         https://bugs.webkit.org/show_bug.cgi?id=71593
618
619         Reviewed by Adam Barth.
620
621         Just removing support from this one core file.
622         I'll commit the rest of the (large) change once we've
623         gone 24 hours and seen that no bots have exploded.
624         We can roll out this one change if we break bots
625         and then roll in the rest once those bots are fixed.
626
627         * Scripts/webkitpy/common/system/executive.py:
628         (Executive.cpu_count):
629
630 2012-01-25  Hajime Morita  <morrita@google.com>
631
632         ENABLE_SHADOW_DOM should be available via build-webkit --shadow-dom
633         https://bugs.webkit.org/show_bug.cgi?id=76863
634
635         Reviewed by Dimitri Glazkov.
636
637         Added a feature flag.
638
639         * Scripts/build-webkit:
640         * qmake/mkspecs/features/features.prf:
641
642 2012-01-25  Martin Robinson  <mrobinson@igalia.com>
643
644         [WK2] Navigation events not triggered by a mouse event claim to be triggered by the left mouse button during policy decisions
645         https://bugs.webkit.org/show_bug.cgi?id=76344
646
647         Reviewed by Sam Weinig.
648
649         Add a test verifying that the mouse button during policy decisions for
650         events not triggered by mouse events is NoButton.
651
652         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
653         (TestWebKitAPI::decidePolicyForNavigationAction): Added test.
654         (TestWebKitAPI::decidePolicyForNewWindowAction): Added test.
655
656 2012-01-25  Ryosuke Niwa  <rniwa@webkit.org>
657
658         run-perf-tests should timeout sooner
659         https://bugs.webkit.org/show_bug.cgi?id=77026
660
661         Reviewed by Adam Barth.
662
663         Bump off the timeout from 10 minutes to 4 minutes.
664
665         Also use float instead of str in test result values to compress the size of json files bots upload.
666
667         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
668         (PerfTestsRunner._parse_args):
669         (PerfTestsRunner._process_chromium_style_test_result):
670         (PerfTestsRunner._process_parser_test_result):
671         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
672         (test_run_test_set_for_parser_tests):
673         (test_run_test_set_with_json_output):
674         (test_run_test_set_with_json_source):
675
676 2012-01-25  Enrica Casucci  <enrica@apple.com>
677
678         WebView should implement typingAttributes methods to work correctly with the Inspector bar in Mail.
679         https://bugs.webkit.org/show_bug.cgi?id=76951
680         <rdar://problem/9325158>
681
682         Reviewed by Alexey Proskuryakov.
683
684         * TestWebKitAPI/Tests/mac/InspectorBar.mm: Modified to reflect the changes to WebView
685         and WebHTMLView.
686         (TestWebKitAPI::TEST):
687
688 2012-01-23  MORITA Hajime  <morrita@google.com>
689
690         [Refactoring][Internals] Should have InternalSettings
691         https://bugs.webkit.org/show_bug.cgi?id=76424
692
693         Reviewed by Kent Tamura.
694
695         * GNUmakefile.am: Added so new files.
696
697 2012-01-25  Takashi Toyoshima  <toyoshim@chromium.org>
698
699         webkitpy: new-run-webkit-websocket is broken
700         https://bugs.webkit.org/show_bug.cgi?id=76999
701
702         Reviewed by Dirk Pranke.
703
704         r105674 broke new-run-webkit-websocket, since options.chromium
705         doesn't exist any more. new-run-webkit-httpd was also broken
706         and already fixed at r105835.
707         * Scripts/new-run-webkit-websocketserver:
708         (main):
709
710 2012-01-25  Nándor Huszka  <huszka.nandor@stud.u-szeged.hu>
711
712         [WK2] WebKitTestRunner needs layoutTestController.setPopupBlockingEnabled
713         https://bugs.webkit.org/show_bug.cgi?id=63458
714
715         Reviewed by Adam Roben.
716
717         Implement the setPopupBlockingEnabled method.
718
719         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
720         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
721         (WTR::LayoutTestController::setPopupBlockingEnabled):
722         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
723
724 2012-01-24  Balazs Kelemen  <kbalazs@webkit.org>
725
726         [Qt] Cleanup testfonts handling
727         https://bugs.webkit.org/show_bug.cgi?id=76068
728
729         Reviewed by Simon Hausmann.
730
731         Factor test font initialization into it's own file  so test
732         font users can share the implementation by adding the file
733         to their build.
734         Use fonts.conf from the test font directory to not require the
735         current directory to be the WebKit root.
736         Fix the -use-test-fonts option of QtTestBrowser by initializing
737         the fonts earlier, before it is too late (namely before calling
738         show on the main window).
739
740         * DumpRenderTree/qt/DumpRenderTree.pro:
741         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
742         (WebCore::DumpRenderTree::open):
743         * DumpRenderTree/qt/DumpRenderTreeQt.h:
744         * DumpRenderTree/qt/QtInitializeTestFonts.cpp: Copied from Tools/WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp.
745         (WebKit::initializeTestFonts):
746         * DumpRenderTree/qt/QtInitializeTestFonts.h: Added.
747         * DumpRenderTree/qt/fonts.conf: Removed. No need for this since
748         we will use the one from the test fonts directory.
749         * DumpRenderTree/qt/main.cpp:
750         (main): Fix the ifdefs to make it clear that the code to
751         set up the Dpi has not been ported to Qt5 yet.
752         * QtTestBrowser/QtTestBrowser.pro:
753         * QtTestBrowser/launcherwindow.h:
754         (WindowOptions::WindowOptions):
755         * QtTestBrowser/main.cpp:
756         (launcherMain):
757         (LauncherApplication::handleUserOptions):
758         * WebKitTestRunner/InjectedBundle/Target.pri:
759         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
760         (WTR::activateFonts): Remove the code to set up the Dpi
761         because it does not work with Qt5.
762
763 2012-01-25  Simon Hausmann  <simon.hausmann@nokia.com>
764
765         [Qt] build-webkit --qt requires a git-svn or SVN checkout.
766
767         Reviewed by Tor Arne Vestø.
768
769         Don't require it by making svnRevisionForDirectory return "unknown"
770         instead of bailing out with an error.
771
772         * Scripts/VCSUtils.pm:
773         (svnRevisionForDirectory):
774
775 2012-01-25  Adam Barth  <abarth@webkit.org>
776
777         Temporarily disable running the unit tests on the Chromium EWS.  These
778         tests don't work properly (see
779         https://bugs.webkit.org/show_bug.cgi?id=76979) and are currently
780         blocking the EWS.
781
782         * Scripts/webkitpy/common/config/ports.py:
783         (ChromiumPort):
784
785 2012-01-25  Adam Barth  <abarth@webkit.org>
786
787         Update results.html version of PixelZoomer
788         https://bugs.webkit.org/show_bug.cgi?id=76965
789
790         Reviewed by Ojan Vafai.
791
792         Improve readability as suggested by Ojan.
793
794         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/pixelzoomer.js:
795
796 2012-01-25  Carlos Garcia Campos  <cgarcia@igalia.com>
797
798         [GTK] Page clients don't need to be GObjects anymore in WebKit2 GTK+ API
799         https://bugs.webkit.org/show_bug.cgi?id=76899
800
801         Reviewed by Martin Robinson.
802
803         * gtk/generate-gtkdoc:
804         (get_webkit2_options): WebKitWebLoaderClient has been renamed to
805         WebKitLoaderClient.
806
807 2012-01-24  Ryosuke Niwa  <rniwa@webkit.org>
808
809         Build fix. Trigger Chromium Mac Release (Perf) on new commits.
810
811         * BuildSlaveSupport/build.webkit.org-config/config.json:
812
813 2012-01-23  Ryosuke Niwa  <rniwa@webkit.org>
814
815         WebKit needs public performance tests running on public bots
816         https://bugs.webkit.org/show_bug.cgi?id=10266
817
818         Reviewed by Adam Barth.
819
820         Completes the initial implementation of WebKit Perf bots. Add a build step for run-perf-test
821         and make google-mac-4 run it, and upload the results to webkit-perf.webkit.org.
822
823         * BuildSlaveSupport/build.webkit.org-config/config.json:
824         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
825         (RunAndUploadPerfTests):
826         (RunAndUploadPerfTests.start):
827         (UploadPerfTestResults):
828         (BuildAndPerfTestFactory):
829         (BuildAndPerfTestFactory.__init__):
830
831 2012-01-24  Ojan Vafai  <ojan@chromium.org>
832
833         Initializing the browser property of the Bugzilla class takes too long
834         https://bugs.webkit.org/show_bug.cgi?id=76960
835
836         Reviewed by Adam Barth.
837
838         Lazily initialize it. This saves ~150ms on a no-op run-webkit-tests call
839         on my Mac Pro.
840
841         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
842         (Bugzilla.__init__):
843         (Bugzilla._get_browser):
844         (Bugzilla):
845         (Bugzilla._set_browser):
846
847 2012-01-24  Mark Rowe  <mrowe@apple.com>
848
849         Death to JavaScriptGlue.
850
851         Rubber-stamped by Geoffrey Garen.
852
853         * BuildSlaveSupport/build-launcher-app:
854         (copyNightlyLauncher):
855         * Scripts/build-webkit:
856         * Scripts/do-file-rename:
857         * Scripts/do-webcore-rename:
858         * Scripts/webkitpy/common/config/build.py:
859         (_should_file_trigger_build):
860         * Scripts/webkitpy/common/config/build_unittest.py: Update the test to use another Mac-specific
861         path in conjunction with a cross-platform path.
862         (ShouldBuildTest):
863
864 2012-01-24  Kentaro Hara  <haraken@chromium.org>
865
866         webkitperl/prepare-ChangeLog_unittest/parser_unittests.pl should test stdout and stderr
867         https://bugs.webkit.org/show_bug.cgi?id=75836
868
869         Reviewed by David Kilzer.
870
871         Previously parser_unittests.pl can just test the returned value of
872         get_function_line_ranges() in prepare-ChangeLog. This patch fixes
873         parser_unittests.pl so that it also tests the stdout and stderr of
874         get_function_line_ranges().
875
876         This patch adds css_unittests_warning.css, which contains test cases
877         for warning messages of the CSS parser of prepare-ChangeLog.
878
879         * Scripts/webkitperl/prepare-ChangeLog_unittest/parser_unittests.pl:
880         (captureOutput): Captures the stdout and stderr of a given method.
881         (convertAbsolutepathToWebKitPath): Converts absolute paths in a given
882         string to relative paths.
883
884         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/css_unittests_warning-expected.txt: Added.
885         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/css_unittests_warning.css: Added.
886         (element1):
887         (element2):
888         (element3):
889         (element4):
890         ():
891         (element5):
892         (element6):
893
894         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests-expected.txt:
895         Just updated the test results, i.e. added stdout and stderr results.
896         No essential changes in the test results.
897         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/css_unittests-expected.txt: Ditto.
898         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/java_unittests-expected.txt: Ditto.
899         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/javascript_unittests-expected.txt: Ditto.
900         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/perl_unittests-expected.txt: Ditto.
901         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/python_unittests-expected.txt: Ditto.
902
903 2012-01-24  Dirk Pranke  <dpranke@chromium.org>
904
905         webkitpy: new-run-webkit-httpd is broken
906         https://bugs.webkit.org/show_bug.cgi?id=76963
907
908         Reviewed by Adam Barth.
909
910         r105674 broke new-run-webkit-httpd, since options.chromium
911         doesn't exist any more. We didn't catch this because we don't
912         run the integration tests by default anywhere :(.
913
914         * Scripts/new-run-webkit-httpd:
915         (run):
916         (main):
917
918 2012-01-24  Adam Barth  <abarth@webkit.org>
919
920         garden-o-matic pixel zoomer should let you look at every pixel of the image
921         https://bugs.webkit.org/show_bug.cgi?id=76962
922
923         Reviewed by Ojan Vafai.
924
925         Previously, you couldn't get up to the right and bottom pixels.
926
927         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/pixelzoomer.js:
928
929 2012-01-24  Enrica Casucci  <enrica@apple.com>
930
931         WebHTMLView should implement typingAttributes methods to work correctly with the Inspector bar in Mail.
932         https://bugs.webkit.org/show_bug.cgi?id=76951
933         <rdar://problem/9325158>
934
935         Added test to verify that WebHTMLView supports the typingAttributes method and that it
936         works as expected.
937         
938         Reviewed by Alexey Proskuryakov.
939
940         * TestWebKitAPI/Tests/mac/InspectorBar.mm:
941         (TestWebKitAPI::TEST):
942
943 2012-01-24  James Robinson  <jamesr@chromium.org>
944
945         Put a space between the trac link and ellipsis in sheriffbot rollout message so it linkifies better in some IRC clients
946         https://bugs.webkit.org/show_bug.cgi?id=76884
947
948         Reviewed by Adam Barth.
949
950         * Scripts/webkitpy/tool/bot/irc_command.py:
951         (Rollout.execute):
952         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
953         (SheriffIRCBotTest.test_rollout):
954         (SheriffIRCBotTest.test_revert):
955         (SheriffIRCBotTest.test_multi_rollout):
956         (SheriffIRCBotTest.test_rollout_with_r_in_svn_revision):
957         (SheriffIRCBotTest.test_multi_rollout_with_r_in_svn_revision):
958         (SheriffIRCBotTest.test_rollout_invalidate_reason):
959         (test_multi_rollout_invalidate_reason):
960
961 2012-01-24  Kentaro Hara  <haraken@chromium.org>
962
963         The cpp parser of prepare-ChangeLog cannot detect a change in classes and namespaces
964         https://bugs.webkit.org/show_bug.cgi?id=75531
965
966         Reviewed by David Kilzer.
967
968         Previously, the cpp parser of prepare-ChangeLog could not detect a change
969         outside methods. Consider the following cpp file.
970
971             namespace N {
972             int a;     // this change does not appear on ChangeLog.
973             class C {
974                 int b;     // this change does not appear on ChangeLog.
975                 void f()
976                 {
977                     int c;     // this change appears on ChangeLog.
978                 }
979                 int d;     // this change does not appear on ChangeLog.
980             };
981             int e;     // this change does not appear on ChangeLog.
982             };
983
984         The previous prepare-ChangeLog outputs just methods in which a change is found:
985             (N::C::f):
986
987         This patch fixes prepare-ChangeLog so that it outputs namespaces, classes
988         and methods in which a change is found:
989             (N):
990             (N::C):
991             (N::C::f):
992
993         * Scripts/prepare-ChangeLog:
994         (get_function_line_ranges_for_cpp): Modified as described above.
995
996         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests.cpp: Added test cases.
997         (Class104):
998         (Class105):
999         (Class106):
1000         (Class106::func32):
1001         (Class106::func33):
1002         (NameSpace3):
1003         (NameSpace4):
1004         (NameSpace5):
1005         (NameSpace6):
1006         (Class107):
1007         (NameSpace5::NameSpace6::Class107::func34):
1008         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests-expected.txt:
1009
1010 2012-01-24  Kentaro Hara  <haraken@chromium.org>
1011
1012         prepare-ChangeLog outputs warnings for top-level { ... } blocks
1013         https://bugs.webkit.org/show_bug.cgi?id=75943
1014
1015         Reviewed by David Kilzer.
1016
1017         Currently, prepare-ChangeLog outputs warnings for top-level { ... } blocks,
1018         because it does not distinguish '}' at the end of a subroutine
1019         with '}' at the end of a if or for statement.
1020
1021         foo.pl:
1022             if (1) {
1023             }
1024             for (@v) {
1025             }
1026
1027         output:
1028             nested functions found at top-level at foo.pl:2
1029             nested functions found at top-level at foo.pl:4
1030
1031         This patch fixes the bug.
1032
1033         * Scripts/prepare-ChangeLog:
1034         (get_function_line_ranges_for_perl):
1035         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/perl_unittests.pl:
1036         Added test cases. No change in the test results.
1037
1038 2012-01-24  Dirk Pranke  <dpranke@chromium.org>
1039
1040         r105674 broke check-webkit-style for chromium's test_expectations.
1041
1042         Unreviewed, build fix.
1043         
1044         We now need to pass in a platform name of 'chromium' if we want
1045         the chromium port; options.chromium is no longer recognized. I
1046         will follow this patch up with a separate patch that has a test,
1047         but this at least fixes the build.
1048
1049         * Scripts/webkitpy/style/checkers/test_expectations.py:
1050         (TestExpectationsChecker._determine_port_from_exepectations_path):
1051
1052 2012-01-20  Ojan Vafai  <ojan@chromium.org>
1053
1054         check-webkit-style of the chromium test_expectations.txt file takes too long
1055         https://bugs.webkit.org/show_bug.cgi?id=76745
1056
1057         Reviewed by Dimitri Glazkov.
1058
1059         When in lint mode, have TestExpectations test all configurations instead
1060         of looping over each configuration. This also has the benefit of making
1061         the error output considerably more concise.
1062
1063         Also, got rid of the double-printing of errors when linting through check-webkit-style.
1064
1065         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1066         (Manager.lint):
1067         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1068         (TestExpectations._report_errors):
1069         (TestExpectations._add_expectations):
1070         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1071         (test_parse_error_nonfatal):
1072         (test_error_on_different_platform):
1073         * Scripts/webkitpy/style/checkers/test_expectations.py:
1074         (TestExpectationsChecker.check_test_expectations):
1075         (TestExpectationsChecker.check):
1076         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
1077         (TestExpectationsTestCase.test_determine_port_from_exepectations_path):
1078         (TestExpectationsTestCase.assert_lines_lint):
1079
1080 2012-01-24  Ryosuke Niwa  <rniwa@webkit.org>
1081
1082         Another build fix attempt after r105543.
1083
1084         * Scripts/webkitpy/common/net/file_uploader.py:
1085         (FileUploader.upload_as_multipart_form_data):
1086
1087 2012-01-24  Ryosuke Niwa  <rniwa@webkit.org>
1088
1089         Build fix attempt after r105543.
1090
1091         * Scripts/webkitpy/common/net/file_uploader.py:
1092         (FileUploader.upload_as_multipart_form_data):
1093
1094 2012-01-24  Vincent Scheib  <scheib@chromium.org>
1095
1096         [Chromium] Add Pointer Lock test hooks and mock implementation to DumpRenderTree
1097         https://bugs.webkit.org/show_bug.cgi?id=76411
1098
1099         Reviewed by Adam Barth.
1100
1101         Breaking up https://bugs.webkit.org/show_bug.cgi?id=75762
1102         into a series of patches, this change adds a trivial
1103         pointer lock implementation to the Chromium port's DumpRenderTree
1104         and exposes LayoutTestController hooks to force failure conditions.
1105
1106         * DumpRenderTree/chromium/LayoutTestController.cpp:
1107         (LayoutTestController::LayoutTestController):
1108         (LayoutTestController::didLosePointerLock):
1109         (LayoutTestController::setPointerLockWillFailAsynchronously):
1110         (LayoutTestController::setPointerLockWillFailSynchronously):
1111         * DumpRenderTree/chromium/LayoutTestController.h:
1112         * DumpRenderTree/chromium/WebViewHost.cpp:
1113         (WebViewHost::requestPointerLock):
1114         (WebViewHost::requestPointerUnlock):
1115         (WebViewHost::isPointerLocked):
1116         (WebViewHost::didCompletePointerLock):
1117         (WebViewHost::didNotCompletePointerLock):
1118         (WebViewHost::didLosePointerLock):
1119         (WebViewHost::reset):
1120         * DumpRenderTree/chromium/WebViewHost.h:
1121         (WebViewHost::setPointerLockWillFailAsynchronously):
1122         (WebViewHost::setPointerLockWillFailSynchronously):
1123
1124 2012-01-24  Mario Sanchez Prada  <msanchez@igalia.com>
1125
1126         Unreviewed, skip WK2 API test timing out on some GTK bots.
1127
1128         WebKit2APITests/TestDownloads is consistently timing out on the
1129         32bit release and 64bit debug bots.
1130
1131         * Scripts/run-gtk-tests:
1132         (TestRunner): Skip WebKit2APITests/TestDownloads.
1133
1134 2012-01-24  Andras Becsi  <andras.becsi@nokia.com>
1135
1136         [Qt] Fix the build with the newes Qt5 hashes
1137         https://bugs.webkit.org/show_bug.cgi?id=76657
1138
1139         Reviewed by Simon Hausmann.
1140
1141         * MiniBrowser/qt/MiniBrowserApplication.cpp:
1142         (MiniBrowserApplication::MiniBrowserApplication): Disable the automatic touch->mouse event synthesis
1143         for now to prevent an infinite loop where Qt always creates mouse events out of our touch mock events.
1144
1145 2012-01-24  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1146
1147         [Qt] Distinguish between Cygwin and cmd.exe in the way we set env variables
1148
1149         We can't use win32-msvc* to decide whether or not to use (set FOO=bar), as
1150         building with MinGW inside cmd.exe will require the same construct. Instead
1151         we assume that when the first character in PWD is a '/' we're running under
1152         Cygwin, and use the normal unix way of setting environment variables.
1153
1154         Reviewed by Simon Hausmann.
1155
1156         * qmake/mkspecs/features/default_pre.prf:
1157
1158 2012-01-24  Simon Hausmann  <simon.hausmann@nokia.com>
1159
1160         Some qmake build files must include Qt5's "quick" module.
1161         https://bugs.webkit.org/show_bug.cgi?id=76671
1162
1163         Unreviewed build fix.
1164
1165         * MiniBrowser/qt/MiniBrowser.pro:
1166         * WebKitTestRunner/Target.pri:
1167
1168 2012-01-24  Andras Becsi  <andras.becsi@nokia.com>
1169
1170         [Qt] Fix the event timestamps for touch mocking
1171
1172         Reviewed by Simon Hausmann.
1173
1174         * MiniBrowser/qt/MiniBrowserApplication.cpp:
1175         (MiniBrowserApplication::sendTouchEvent): We should not set QEvent::None as the
1176         timestamp argument in the QWindowSystemInterface::handleTouchEvent call.
1177
1178 2012-01-24  Carlos Garcia Campos  <cgarcia@igalia.com>
1179
1180         [GTK] Implement DownloadClient in WebKit2 GTK+ API
1181         https://bugs.webkit.org/show_bug.cgi?id=72952
1182
1183         Reviewed by Martin Robinson.
1184
1185         * gtk/generate-gtkdoc:
1186         (get_webkit2_options): Ignore WebKitDownloadClient.
1187
1188 2012-01-23  Zan Dobersek  <zandobersek@gmail.com>
1189
1190         [GTK] editing/deleting/5408255.html results are incorrect
1191         https://bugs.webkit.org/show_bug.cgi?id=53644
1192
1193         Reviewed by Martin Robinson.
1194
1195         WEBKIT_TOP_LEVEL environment variable is now set directly in either
1196         WebKitTestRunner or DumpRenderTree through usage of a compilation-time
1197         macro. This way both tools can be run outside the test harness without
1198         the need to manually set the environment variable.
1199
1200         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1201         (getTopLevelPath):
1202         * GNUmakefile.am:
1203         * Scripts/webkitpy/layout_tests/port/gtk.py:
1204         (GtkPort.setup_environ_for_server):
1205         * WebKitTestRunner/GNUmakefile.am:
1206         * WebKitTestRunner/InjectedBundle/gtk/InjectedBundleGtk.cpp:
1207         (WTR::InjectedBundle::platformInitialize):
1208
1209 2012-01-23  Dmitry Lomov  <dslomov@google.com>
1210
1211         [Chromium] Implement layoutTestController.workerThreadCount in DRT
1212         https://bugs.webkit.org/show_bug.cgi?id=74653.
1213
1214         Reviewed by Darin Fisher.
1215
1216         * DumpRenderTree/chromium/LayoutTestController.cpp:
1217         (LayoutTestController::LayoutTestController):
1218         (LayoutTestController::workerThreadCount):
1219         * DumpRenderTree/chromium/LayoutTestController.h:
1220
1221 2012-01-23  Dirk Pranke  <dpranke@chromium.org>
1222
1223         nrwt: make --chromium work like --qt
1224         https://bugs.webkit.org/show_bug.cgi?id=76875
1225
1226         Reviewed by Adam Barth.
1227
1228         --chromium used to have to be handled differently from --qt
1229         due to the way the PortFactory was implemented; there's not
1230         really a good reason for that any more so this patch makes
1231         things slightly more consistent and eliminates the
1232         options.chromium flag (--chromium is now truly a synonym for
1233         --platform chromium).
1234
1235         * Scripts/webkitpy/layout_tests/port/factory.py:
1236         (PortFactory._default_port):
1237         (PortFactory.get):
1238         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
1239         (FactoryTest.setUp):
1240         (FactoryTest.test_chromium_mac):
1241         (FactoryTest.test_chromium_linux):
1242         (FactoryTest.test_chromium_win):
1243         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1244         (parse_args):
1245
1246 2012-01-23  Dirk Pranke  <dpranke@chromium.org>
1247
1248         run-webkit-tests needs to propagate --chromium
1249         https://bugs.webkit.org/show_bug.cgi?id=76870
1250
1251         Reviewed by Eric Seidel.
1252
1253         run-webkit-tests removes '--chromium' argument from @ARGV when
1254         determining which port to run, which means that that doesn't
1255         propagate to new-run-webkit-tests. That's bad (and is handled
1256         for the other ports by re-adding the flag, but apparently we're
1257         just now noticing for Chromium).
1258
1259         * Scripts/run-webkit-tests:
1260
1261 2012-01-23  Brian Weinstein  <bweinstein@apple.com>
1262
1263         More build fixing after r105646.
1264
1265         * TestWebKitAPI/Tests/WTF/RedBlackTree.cpp:
1266         (TestWebKitAPI::TestNode::key):
1267
1268 2012-01-23  Gavin Barraclough  <barraclough@apple.com>
1269
1270         Unreviewed build fix - r105646 broke this.
1271
1272         * TestWebKitAPI/Tests/WTF/RedBlackTree.cpp:
1273         (TestWebKitAPI::TestNode::TestNode):
1274         (TestWebKitAPI::TestNode::key):
1275         (TestWebKitAPI::RedBlackTreeTest::assertEqual):
1276         (TestWebKitAPI::RedBlackTreeTest::assertSameValuesForKey):
1277         (TestWebKitAPI::RedBlackTreeTest::testDriver):
1278
1279 2012-01-23  Sheriff Bot  <webkit.review.bot@gmail.com>
1280
1281         Unreviewed, rolling out r105658.
1282         http://trac.webkit.org/changeset/105658
1283         https://bugs.webkit.org/show_bug.cgi?id=76883
1284
1285         We want this eventually, but not right at this moment
1286         (Requested by abarth on #webkit).
1287
1288         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1289
1290 2012-01-23  Adam Barth  <abarth@webkit.org>
1291
1292         garden-o-matic should support Chromium Mac Lion
1293         https://bugs.webkit.org/show_bug.cgi?id=76880
1294
1295         Reviewed by Eric Seidel.
1296
1297         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1298
1299 2012-01-23  Emil A Eklund  <eae@chromium.org>
1300
1301         check-webkit-style whitespace/operators triggers on overloaded division operator
1302         https://bugs.webkit.org/show_bug.cgi?id=76650
1303
1304         Reviewed by Darin Adler.
1305
1306         * Scripts/webkitpy/style/checkers/cpp.py:
1307         (check_spacing):
1308         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1309         (CppStyleTest.test_operator_methods):
1310
1311 2012-01-23  Ryosuke Niwa  <rniwa@webkit.org>
1312
1313         run-perf-tests should report server-side errors
1314         https://bugs.webkit.org/show_bug.cgi?id=76802
1315
1316         Reviewed by Tony Chang.
1317
1318         Report errors whenever server's response doesn't match "OK".
1319
1320         * Scripts/webkitpy/common/net/file_uploader.py:
1321         (FileUploader.upload_single_text_file):
1322         (FileUploader.upload_as_multipart_form_data):
1323         (FileUploader._upload_data.callback):
1324         (FileUploader):
1325         (FileUploader._upload_data):
1326         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1327         (PerfTestsRunner._upload_json):
1328         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1329         (test_upload_json):
1330         (test_upload_json.MockFileUploader.upload_single_text_file):
1331
1332 2012-01-22  Ryosuke Niwa  <rniwa@webkit.org>
1333
1334         run-perf-tests ignore Skipped list on chromium
1335         https://bugs.webkit.org/show_bug.cgi?id=76764
1336
1337         Reviewed by Dirk Pranke.
1338
1339         Move skipped_perf_tests from WebKit port to Base port so that Chromium port
1340         can also find skipped list. Chromium port only uses test_expectations.txt for
1341         layout tests but performacne tests don't use test_expectations.txt so Chromium port
1342         also needs to use Skipped list.
1343
1344         * Scripts/webkitpy/layout_tests/port/base.py:
1345         (Port._tests_from_skipped_file_contents):
1346         (Port):
1347         (Port._expectations_from_skipped_files):
1348         (Port.skipped_perf_tests):
1349         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1350         (PortTest.test_skipped_perf_tests):
1351         (PortTest.test_skipped_perf_tests.add_text_file):
1352         * Scripts/webkitpy/layout_tests/port/webkit.py:
1353         (WebKitPort._skipped_tests_for_unsupported_features):
1354         (WebKitPort._skipped_file_search_paths):
1355         (WebKitPort.skipped_layout_tests):
1356
1357 2012-01-23  Andy Estes  <aestes@apple.com>
1358
1359         Fix the build after r105635.
1360
1361         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp:
1362         (TestWebKitAPI::TEST):
1363
1364 2012-01-23  Xianzhu Wang  <wangxianzhu@chromium.org>
1365
1366         Basic enhancements to StringBuilder
1367         https://bugs.webkit.org/show_bug.cgi?id=67081
1368
1369         Reviewed by Darin Adler.
1370
1371         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp:
1372         (TestWebKitAPI::TEST):
1373
1374 2012-01-23  Mario Sanchez Prada  <msanchez@igalia.com>
1375
1376         [GTK] run-gtk-tests randomly fails while running the xprop comand
1377         https://bugs.webkit.org/show_bug.cgi?id=76817
1378
1379         Reviewed by Gustavo Noronha Silva.
1380
1381         No need to use xprop to remove the AT_SPI_BUS property since
1382         run-gtk-tests will always run new instances of Xvfb.
1383
1384         * Scripts/run-gtk-tests:
1385         (TestRunner): Unskip TestWebKitAccessibility.
1386         (TestRunner.run): Uncomment lines for launching the accessibility
1387         bus and registry daemon, and remove lines for running xprop.
1388
1389 2012-01-23  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1390
1391         [Qt] Don't warn about override and final being C++11 extensions
1392
1393         Clang will emit a warning when these extensions are used without passing
1394         --std=c++11, but we use feature checking to decide if we have the right
1395         extensions, so we can safely ignore these warnings. The XCode and Windows
1396         project files have the same workaround.
1397
1398         The reason for adding the flag to QMAKE_OBJECTIVE_CFLAGS as well is that
1399         we only have one extra compiler for Objective-C, which is also used for
1400         Objective-C++ sources, so we have to pass the flag, even if it doesn't
1401         make sense for Objective-C.
1402
1403         Reviewed by Simon Hausmann.
1404
1405         * qmake/mkspecs/features/unix/default_post.prf:
1406
1407 2012-01-20  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1408
1409         [Qt] Change how build-webkit decides when to do full incremental builds
1410
1411         Instead of relying on update-webkit (which isn't run on the bots) to
1412         decide when to do a full incremental build (make qmake), we let the
1413         build-webkit script itself check the current SVN revision against the
1414         previous build (by storing it in .webkit.config).
1415
1416         If the two differ we assume a full incremental build is needed, since
1417         the new revisions might have introduced problematic things like new
1418         Q_OBJECT macros. If not, we assume the developer is doing changes
1419         locally, and revert to doing a plain 'make'.
1420
1421         In addition, when the build fails in the latter case, we inform the
1422         developer of possible pitfalls and how to manually run 'make qmake'.
1423
1424         Reviewed by Ossy.
1425
1426         * Scripts/update-webkit:
1427         * Scripts/webkitdirs.pm:
1428         (buildQMakeProjects):
1429
1430 2012-01-23  Mario Sanchez Prada  <msanchez@igalia.com>
1431
1432         [GTK] run-gtk-tests randomly fails while running the xprop comand
1433         https://bugs.webkit.org/show_bug.cgi?id=76817
1434
1435         Reviewed by Philippe Normand.
1436
1437         Temporarily comment the lines related to launching the ATSPI bus
1438         and registry daemon, which are making the bots to fail randomly.
1439
1440         * Scripts/run-gtk-tests:
1441         (TestRunner): Skip TestWebKitAccessibility, as it won't run
1442         properly if the ATSPI infrastructure is not properly initialized.
1443         (TestRunner.run): Comment lines related to initialization of ATSPI.
1444
1445 2012-01-22  Ryosuke Niwa  <rniwa@webkit.org>
1446
1447         Windows python test build fix.
1448
1449         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1450         (PerfTestsRunner.__init__):
1451
1452 2012-01-22  Sergio Villar Senin  <svillar@igalia.com>
1453
1454         [GTK] DumpRenderTree converts "file:///" to a path differently
1455         https://bugs.webkit.org/show_bug.cgi?id=76631
1456
1457         Reviewed by Martin Robinson.
1458
1459         DumpRenderTree should print "/" as the last path component if the
1460         path is a lone slash instead of empty output.
1461
1462         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1463         (webViewConsoleMessage):
1464
1465 2012-01-20  David Levin  <levin@chromium.org>
1466
1467         The leaky DC check should use adoptPtr as a signal instead of OwnPtr to get less false positives.
1468         https://bugs.webkit.org/show_bug.cgi?id=76752
1469
1470         Reviewed by Eric Seidel.
1471
1472         * Scripts/webkitpy/style/checkers/cpp.py:
1473         (check_for_leaky_patterns):
1474         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1475         (LeakyPatternTest.test_create_dc):
1476         (LeakyPatternTest.test_create_compatible_dc):
1477
1478 2012-01-20  Shawn Singh  <shawnsingh@chromium.org>
1479
1480         Changed my status in commiters.py
1481
1482         Unreviewed.
1483
1484         * Scripts/webkitpy/common/config/committers.py:
1485
1486 2012-01-20  Ojan Vafai  <ojan@chromium.org>
1487
1488         run-webkit-tests --lint-test-files crawls the whole LayoutTests subtree
1489         https://bugs.webkit.org/show_bug.cgi?id=76748
1490
1491         Reviewed by Ryosuke Niwa.
1492
1493         It crawls the whole subtree and then doesn't use the data. Cutting this out
1494         saves 4 seconds warm and 17 seconds cold on my Mac Pro.
1495
1496         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1497         (run):
1498
1499 2012-01-20  Adam Barth  <abarth@webkit.org>
1500
1501         Another tiny tweak to the garden-o-matic CSS.  This makes things line
1502         up slightly nicer when there aren't any failures.
1503
1504         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
1505         (#summary):
1506
1507 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
1508
1509         run-perf-tests should support --test-results-server option
1510         https://bugs.webkit.org/show_bug.cgi?id=76680
1511
1512         Reviewed by Adam Barth.
1513
1514         Add --test-results-server, --builder-name, and --build-number options to run-perf-tests
1515         to be used by perf bots. Also refactor file_uploader as needed.
1516
1517         * Scripts/webkitpy/common/net/file_uploader.py:
1518         (FileUploader.__init__):
1519         (FileUploader.upload_single_file):
1520         (FileUploader.upload_as_multipart_form_data):
1521         (FileUploader):
1522         (FileUploader._upload_data):
1523         (FileUploader._upload_data.callback):
1524         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
1525         (JSONResultsGeneratorBase.upload_json_files):
1526         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1527         (PerfTestsRunner):
1528         (PerfTestsRunner._parse_args):
1529         (PerfTestsRunner.run):
1530         (PerfTestsRunner._generate_json):
1531         (PerfTestsRunner._upload_json):
1532         (PerfTestsRunner._run_tests_set):
1533         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1534         (create_runner):
1535         (test_run_with_upload_json):
1536         (test_run_with_upload_json.mock_upload_json):
1537         (test_upload_json):
1538         (test_upload_json.MockFileUploader):
1539         (test_upload_json.MockFileUploader.__init__):
1540         (test_upload_json.MockFileUploader.upload_single_file):
1541         (test_parse_args):
1542
1543 2012-01-20  Adam Barth  <abarth@webkit.org>
1544
1545         The party time image overlaps real content!  This patch makes the
1546         image centered vertically.
1547
1548         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
1549         (#onebar.partytime #summary):
1550
1551 2012-01-19  Ojan Vafai  <ojan@chromium.org>
1552
1553         Refactor TestExpectationsParser in preparation for caching the results
1554         https://bugs.webkit.org/show_bug.cgi?id=76669
1555
1556         Reviewed by Dimitri Glazkov.
1557
1558         Make everything private expect for the parse method. 
1559         Eventually, we'll need the expectations lines to not be modified
1560         outside of TestExpectationsParser so we can cache the results.
1561         This makes check-webkit-style of the chromium test_expectations.txt file
1562         go from ~17 seconds to ~12 seconds on my Mac Pro.
1563
1564         This patch is just a refactor in preparation, so no new tests.
1565
1566         * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor_unittest.py:
1567         (TestExpectationEditorTests.make_parsed_expectation_lines):
1568         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1569         (TestExpectationParser.parse):
1570         (TestExpectationParser):
1571         (TestExpectationParser._parse_line):
1572         (TestExpectationParser._tokenize):
1573         (TestExpectationParser._tokenize_list):
1574         (TestExpectationsModel._clear_expectations_for_test):
1575         (TestExpectations.__init__):
1576         (TestExpectations._add_expectations):
1577         (TestExpectations._add_skipped_tests):
1578         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1579         (TestExpectationParserTests.test_tokenize_blank):
1580         (TestExpectationParserTests.test_tokenize_missing_colon):
1581         (TestExpectationParserTests.test_tokenize_extra_colon):
1582         (TestExpectationParserTests.test_tokenize_empty_comment):
1583         (TestExpectationParserTests.test_tokenize_comment):
1584         (TestExpectationParserTests.test_tokenize_missing_equal):
1585         (TestExpectationParserTests.test_tokenize_extra_equal):
1586         (TestExpectationParserTests.test_tokenize_valid):
1587         (TestExpectationParserTests.test_tokenize_valid_with_comment):
1588         (TestExpectationParserTests.test_tokenize_valid_with_multiple_modifiers):
1589         (TestExpectationParserTests.test_parse_empty_string):
1590         (TestExpectationSerializerTests.assert_round_trip):
1591         (TestExpectationSerializerTests.assert_list_round_trip):
1592         * Scripts/webkitpy/tool/commands/expectations.py:
1593         (OptimizeExpectations.execute):
1594         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1595         (GardeningExpectationsUpdater.update_expectations):
1596
1597 2012-01-20  Adam Barth  <abarth@webkit.org>
1598
1599         Follow-up to previous patch: don't produce NaN when the revision number
1600         is missing.
1601
1602         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
1603         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model_unittests.js:
1604
1605 2012-01-20  Adam Barth  <abarth@webkit.org>
1606
1607         Garden-o-matic should tell me which revisions have been checked by all the bots
1608         https://bugs.webkit.org/show_bug.cgi?id=76722
1609
1610         Reviewed by Dimitri Glazkov.
1611
1612         This information helps me know when it's safe for me to roll a
1613         candidate revision.
1614
1615         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
1616         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
1617         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model_unittests.js:
1618
1619 2012-01-20  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1620
1621         build-webkit: Don't spit out congratulations message on Ctrl+C
1622
1623         Reviewed by Andreas Kling.
1624
1625         * Scripts/build-webkit:
1626
1627 2012-01-20  Philippe Normand  <pnormand@igalia.com>
1628
1629         [GTK] Allow extra jhbuild modulesets
1630         https://bugs.webkit.org/show_bug.cgi?id=76691
1631
1632         Reviewed by Gustavo Noronha Silva.
1633
1634         * gtk/jhbuildrc: Extra modulesets and modules can be added in
1635         JHBuild using the WEBKIT_EXTRA_MODULESETS and WEBKIT_EXTRA_MODULES
1636         env variables respectively. Use comma-separated values. Example:
1637         WEBKIT_EXTRA_MODULES=file:///path/to/module.set,file:///other/path.
1638
1639 2012-01-20  Mario Sanchez Prada  <msanchez@igalia.com>
1640
1641         [GTK] Expose accessibility hierarchy in WebKit2 to ATK/AT-SPI based ATs
1642         https://bugs.webkit.org/show_bug.cgi?id=72589
1643
1644         Reviewed by Martin Robinson.
1645
1646         Ensure the accessibility infrastructure is available when running
1647         the unit tests, otherwise WK2's accessibility tests won't work.
1648
1649         * Scripts/run-gtk-tests:
1650         (TestRunner):
1651         (TestRunner._lookup_atspi2_binary): New. Looks for the directory
1652         where at-spi2's stuff is located (like at-spi-bus-launcher and
1653         at-spi2-registryd).
1654         (TestRunner.run): Initialize the accessibility infrastructure.
1655
1656         * gtk/generate-gtkdoc:
1657         (get_webkit2_options): Add 'WebKitWebViewBaseAccessible.*'.
1658
1659         * gtk/jhbuild.modules: Added at-spi2-core and at-spi2-atk.
1660
1661 2012-01-20  Hayato Ito  <hayato@chromium.org>
1662
1663         Add ShadowRoot.idl which is enabled by newly introduced SHADOW_DOM flag.
1664         https://bugs.webkit.org/show_bug.cgi?id=76353
1665
1666         Reviewed by Hajime Morita.
1667
1668         Add ShadowRoot.idl, which is enabled only on chromium port since this is
1669         under development feature.
1670         ShadowRoot.idl contains minimum API so that we can test it.
1671         Other APIs should be added on other changes so that we can isolate issues.
1672
1673         * DumpRenderTree/chromium/TestShell.cpp:
1674         (TestShell::TestShell):
1675
1676 2012-01-20  Mario Sanchez Prada  <msanchez@igalia.com>
1677
1678         [GTK] GTK's DRT not to log events for already defunct objects
1679         https://bugs.webkit.org/show_bug.cgi?id=76620
1680
1681         Reviewed by Martin Robinson.
1682
1683         Do not log 'state-change:defunct' events.
1684
1685         * DumpRenderTree/gtk/AccessibilityCallbacks.cpp:
1686         (printAccessibilityEvent): Early return if a
1687         'state-change:defunct' signal is passed.
1688         (axObjectEventListener): Pass the signal name and value in
1689         separate parameters to printAccessibilityEvent.
1690
1691 2012-01-19  Gustavo Noronha Silva  <gns@gnome.org>
1692
1693         Unreviewed build fix. Wrong variable being used as the cwd for git
1694         remote update command.
1695
1696         * gtk/run-with-jhbuild:
1697         (update_jhbuild):
1698
1699 2012-01-19  Gustavo Noronha Silva  <gns@gnome.org>
1700
1701         [GTK] ensure the jhbuild used by webkit is as up-to-date as needed
1702         https://bugs.webkit.org/show_bug.cgi?id=76585
1703
1704         Reviewed by Martin Robinson.
1705
1706         This allows us to specify a minimum version of jhbuild required,
1707         while also providing stability against changes done to the master
1708         branch, by letting us choose what is the exact version that will
1709         be used.
1710
1711         * gtk/run-with-jhbuild: refactored for more hackability and easier
1712         reading
1713         (jhbuild_installed): checks whether jhbuild is installed
1714         (jhbuild_cloned): checks whether jhbuild has been cloned
1715         (jhbuild_at_expected_revision): checks whether jhbuild is at the
1716         expected revision
1717         (update_jhbuild): brings jhbuild to the expected revision
1718         (clone_jhbuild): clones jhbuild
1719         (install_jhbuild): build and installs jhbuild
1720         (update_webkitgtk_libs): runs update-webkitgtk-libs
1721         (ensure_jhbuild): high-level logic to decide whether and which of
1722         the steps above are executed
1723
1724 2012-01-19  Alexey Proskuryakov  <ap@apple.com>
1725
1726         [WK2] DumpRenderTree converts "file:///" to a path differently
1727         https://bugs.webkit.org/show_bug.cgi?id=76653
1728
1729         Reviewed by John Sullivan.
1730
1731         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp: (WTR::lastFileURLPathComponent):
1732         Do print a slash if there are no path components.
1733
1734 2012-01-19  Eric Seidel  <eric@webkit.org>
1735
1736         webkit-patch cannot rollout patches without changelog/bug number
1737         https://bugs.webkit.org/show_bug.cgi?id=75962
1738
1739         Reviewed by Adam Barth.
1740
1741         * Scripts/webkitpy/tool/commands/download.py:
1742         (AbstractRolloutPrepCommand._prepare_state):
1743         * Scripts/webkitpy/tool/commands/download_unittest.py:
1744         (AbstractRolloutPrepCommandTest.test_prepare_state):
1745
1746 2012-01-19  Alexey Proskuryakov  <ap@apple.com>
1747
1748         [WK2] fast/url/degenerate-file-base.html fails
1749         https://bugs.webkit.org/show_bug.cgi?id=76619
1750
1751         Reviewed by Adam Roben.
1752
1753         A URL generated form local path for loading had an extra slash, making document.URL
1754         incorrect in all WebKit2 tests (file://localhost//path/to/test.html).
1755
1756         * WebKitTestRunner/TestInvocation.cpp: (WTR::createWKURL):
1757
1758 2012-01-18  Ojan Vafai  <ojan@chromium.org>
1759
1760         check-webkit-style of the chromium test_expectations.txt file doesn't test all chromium ports
1761         https://bugs.webkit.org/show_bug.cgi?id=76510
1762
1763         Reviewed by Adam Barth.
1764
1765         Unlike other style rules, if you get the syntax of the test_expectations.txt
1766         file wrong, the layout tests won't run. Also, this check is simple and only
1767         slows down committing if you actually modify one of the test_expectations.txt files.
1768
1769         * Scripts/webkitpy/style/checkers/test_expectations.py:
1770         (TestExpectationsChecker.check_test_expectations):
1771         (TestExpectationsChecker.check):
1772         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
1773         (TestExpectationsTestCase.test_check_covers_all_configurations):
1774         (TestExpectationsTestCase.test_check_covers_all_configurations.mock_check_test_expectations):
1775         (TestExpectationsTestCase):
1776         (assert_lines_lint):
1777
1778 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
1779
1780         Build fix. Revert an unintentional change to WebKitDriver in r105443.
1781
1782         * Scripts/webkitpy/layout_tests/port/driver.py:
1783         (DriverProxy.__init__):
1784
1785 2012-01-19  Dirk Pranke  <dpranke@chromium.org>
1786
1787         remove the duplicated build rules from WebKit.gyp, Tools.gyp
1788         https://bugs.webkit.org/show_bug.cgi?id=73384
1789
1790         Reviewed by Tony Chang.
1791
1792         This patch removes the no-longer-necessary conditional logic for
1793         build_webkit_exes_from_webkit_gyp; we now always build the exes
1794         from their own dedicated gyp files.
1795
1796         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1797
1798 2012-01-19  Alexey Proskuryakov  <ap@apple.com>
1799
1800         make-new-script-test should work with HTTP tests
1801         https://bugs.webkit.org/show_bug.cgi?id=76603
1802
1803         Reviewed by Adam Roben.
1804
1805         * Scripts/make-new-script-test:
1806         (makePathToSharedSources): Renamed from makeTestRelativePathToSharedSources(), since it
1807         now also creates absolute paths for HTTP.
1808
1809 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
1810
1811         Some perf tests time out when ran by run-perf-tests
1812         https://bugs.webkit.org/show_bug.cgi?id=76612
1813
1814         Reviewed by Dirk Pranke and Eric Seidel.
1815
1816         Always pass --no-timeout to DumpRenderTree from run-perf-tests.
1817         Otherwise some tests such as Parser/xml-parser.html will timeout.
1818
1819         --no-timeout option is currently supported by Chromium and Mac ports.
1820
1821         * Scripts/webkitpy/layout_tests/port/base.py:
1822         (Port.to.create_driver):
1823         * Scripts/webkitpy/layout_tests/port/chromium.py:
1824         (ChromiumDriver.__init__):
1825         (ChromiumDriver._wrapper_options):
1826         * Scripts/webkitpy/layout_tests/port/driver.py:
1827         (Driver.__init__):
1828         (DriverProxy.__init__):
1829         * Scripts/webkitpy/layout_tests/port/webkit.py:
1830         (WebKitDriver.__init__):
1831         (WebKitDriver.cmd_line):
1832         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1833         (WebKitDriverTest.test_read_binary_block):
1834         (WebKitDriverTest):
1835         (WebKitDriverTest.test_no_timeout):
1836         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1837         (get_tests_run.RecordingTestDriver.__init__):
1838         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1839         (PerfTestsRunner._parse_args):
1840         (PerfTestsRunner._run_tests_set):
1841         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1842         (create_runner):
1843
1844 2012-01-19  Ben Wells  <benwells@chromium.org>
1845
1846         Add support for window.print to chromium DRT
1847         https://bugs.webkit.org/show_bug.cgi?id=76479
1848
1849         When used in a test window.print goes into print mode and then
1850         straight out again. This will allow the afterprint event to be
1851         tested by having the afterprint event happen before the dump
1852         happens.
1853
1854         Reviewed by Mihai Parparita.
1855
1856         * DumpRenderTree/chromium/WebViewHost.cpp:
1857         (WebViewHost::printPage):
1858         * DumpRenderTree/chromium/WebViewHost.h:
1859
1860 2012-01-18  Enrica Casucci  <enrica@apple.com>
1861
1862         editingAttributedStringFromRange in WebHTMLConverter does not handle NSUnderlineStyleAttributeName.
1863         https://bugs.webkit.org/show_bug.cgi?id=76588
1864         <rdar://problem/9325183>
1865
1866         Added test.
1867         
1868         Reviewed by Dan Bernstein.
1869
1870         * TestWebKitAPI/Tests/mac/InspectorBar.mm:
1871         (TestWebKitAPI::TEST):
1872
1873 2012-01-19  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1874
1875         [Qt] Do full incremental builds until bots use update-webkit
1876
1877         Rubber-stamped by Simon Hausmann.
1878
1879         * Scripts/webkitdirs.pm:
1880         (buildQMakeProjects):
1881
1882 2012-01-19  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1883
1884         Fix the Qt build
1885
1886         Unreviewed build fix.
1887
1888         * Scripts/webkitdirs.pm:
1889         (buildQMakeProjects):
1890
1891 2012-01-19  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1892
1893         [Qt] Make build-webkit always do safe incremental builds after update-webkit
1894
1895         When building against Qt5 with GCC we would just run 'make' in the build dir,
1896         and rely on the dependency files output by GCC (-MD) to handle dependency
1897         tracking, but that fails for special-cases like adding a Q_OBJECT macro
1898         to a header.
1899
1900         To guarantee that an incrmental build will work, we have to run 'make qmake',
1901         which we now do on every build-webkit that's followed by a successful run
1902         of update-webkit. The reasoning is that update-webkit can result in such
1903         potential corner-cases being applied, and since we can't know for sure
1904         unless we inspect the diff and account for all the corner cases we assume
1905         the worst and always run 'make qmake'.
1906
1907         After a succesful run of build-webkit we proceed to do just 'make' for any
1908         subsequent runs, since we assume that the developer knows what kind of
1909         changes he/she is doing, and when a 'make qmake' is needed.
1910
1911         Reviewed by Simon Hausmann.
1912
1913         * Scripts/update-webkit:
1914         * Scripts/webkitdirs.pm:
1915         (determineCurrentSVNRevision):
1916         (currentSVNRevision):
1917         (buildQMakeProjects):
1918         * qmake/mkspecs/features/default_post.prf:
1919
1920 2012-01-18  Ryosuke Niwa  <rniwa@webkit.org>
1921
1922         run-perf-tests should support Skipped list
1923         https://bugs.webkit.org/show_bug.cgi?id=76594
1924
1925         Reviewed by Adam Barth.
1926
1927         Add a support for Skipped list in run-perf-tests; also skip files in resources directories.
1928
1929         * Scripts/webkitpy/layout_tests/port/base.py:
1930         (Port.perf_tests_dir):
1931         (Port.skipped_perf_tests):
1932         (Port.skips_perf_test):
1933         (Port.relative_perf_test_filename):
1934         * Scripts/webkitpy/layout_tests/port/test.py:
1935         (TestPort.perf_tests_dir):
1936         * Scripts/webkitpy/layout_tests/port/webkit.py:
1937         (WebKitPort._expectations_from_skipped_files):
1938         (WebKitPort):
1939         (WebKitPort.skipped_layout_tests):
1940         (WebKitPort.skipped_perf_tests):
1941         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1942         (PerfTestsRunner._collect_tests):
1943         (PerfTestsRunner.run):
1944         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1945         (test_run_test_set_with_json_output):
1946         (test_run_test_set_with_json_source):
1947         (test_collect_tests_with_skipped_list):
1948         (test_collect_tests_with_skipped_list.add_file):
1949
1950 2012-01-18  Dirk Pranke  <dpranke@chromium.org>
1951
1952         Fix path to chromium_src_dir introduced in previous change.
1953
1954         Unreviewed, build fix.
1955
1956         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1957
1958 2012-01-18  Alexey Proskuryakov  <ap@apple.com>
1959
1960         Need infrastructure to test Content-Disposition filename encoding support
1961         https://bugs.webkit.org/show_bug.cgi?id=76572
1962         <rdar://problem/6421825>
1963
1964         Reviewed by Sam Weinig.
1965
1966         Made policy delegate log suggested file name when resource has Content-Disposition: attachment.
1967
1968         WebKitTestRunner does not have a policy delegate yet, bug 42546.
1969
1970         * DumpRenderTree/mac/PolicyDelegate.mm:
1971         (dispositionTypeFromContentDispositionHeader):
1972         (-[PolicyDelegate webView:decidePolicyForMIMEType:request:frame:decisionListener:]):
1973
1974 2012-01-18  Ojan Vafai  <ojan@chromium.org>
1975
1976         Fix tests from http://trac.webkit.org/changeset/105332.
1977
1978         The code assumed that if you passed an options object with a configuration
1979         attribute that it would have a valid value. A bunch of the testing mocks
1980         do things like having a configuration value of None.
1981
1982         * Scripts/webkitpy/layout_tests/port/base.py:
1983         (Port.__init__):
1984
1985 2012-01-18  Ojan Vafai  <ojan@chromium.org>
1986
1987         Shave 0.5 seconds off check-webkit-style runtime for test_expectations.txt
1988         https://bugs.webkit.org/show_bug.cgi?id=76576
1989
1990         Reviewed by Adam Barth.
1991
1992         Avoid calling default_configuration in base.py. We don't need it for
1993         checking test_expectations.txt style. It takes ~0.5 seconds to run on my
1994         Mac Pro. It's the call to "perl Tools/Scripts/webkit-build-directory --top-level"
1995         from common.executive.
1996
1997         At some point someone should probably look into why that call is so slow
1998         since it's on the critical path for run-webkit-tests and build-webkit startup.
1999
2000         * Scripts/webkitpy/layout_tests/port/base.py:
2001         (Port.__init__):
2002         * Scripts/webkitpy/style/checkers/test_expectations.py:
2003         (TestExpectationsChecker._determine_port_from_exepectations_path):
2004
2005 2012-01-18  Dirk Pranke  <dpranke@chromium.org>
2006
2007         [chromium] move Tools.gyp, switch build-webkit --chromium to All.gyp
2008         https://bugs.webkit.org/show_bug.cgi?id=76505
2009
2010         Reviewed by Tony Chang.
2011
2012         Update chromium build script to use All.gyp instead of WebKit.gyp
2013
2014         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp: Renamed from Source/WebKit/chromium/Tools.gyp.
2015         * Scripts/webkitdirs.pm:
2016
2017 2012-01-18  Sam Weinig  <sam@webkit.org>
2018
2019         Don’t hardcode the path to ‘make‘.
2020
2021         Reviewed by Dan Bernstein.
2022
2023         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
2024
2025 2012-01-18  Joshua Bell  <jsbell@chromium.org>
2026
2027         Adding myself to committers.py
2028         https://bugs.webkit.org/show_bug.cgi?id=76569
2029
2030         Unreviewed.
2031
2032         * Scripts/webkitpy/common/config/committers.py:
2033
2034 2012-01-18  Adam Barth  <abarth@webkit.org>
2035
2036         Up the cap on test failures tolerated by the EWS so the mac-ews can run to completion
2037         https://bugs.webkit.org/show_bug.cgi?id=76567
2038
2039         Reviewed by Eric Seidel.
2040
2041         Current the AppleMac port has slightly more failures than we allow on
2042         the EWS, causing the mac-ews to spin.  This patch ups the limit a bit,
2043         which should hopefully let the mac-ews run to completion.  Once
2044         AppleMac has fewer failures, we can lower the cap again.
2045
2046         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
2047         (LayoutTestResultsReaderTest.test_layout_test_results):
2048         * Scripts/webkitpy/tool/steps/runtests.py:
2049         (RunTests):
2050         * Scripts/webkitpy/tool/steps/steps_unittest.py:
2051
2052 2012-01-18  Eric Carlson  <eric.carlson@apple.com>
2053
2054         Provide access to user's list of preferred languages
2055         https://bugs.webkit.org/show_bug.cgi?id=76138
2056
2057         Reviewed by Alexey Proskuryakov.
2058
2059         * GNUmakefile.am: Add JSInternalsCustom.cpp.
2060
2061 2012-01-18  Adam Barth  <abarth@webkit.org>
2062
2063         Add Chromium Linux GPU 32 to the list of allowed Chromium Linux
2064         configurations.
2065
2066         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2067         (ChromiumLinuxPort.__init__):
2068
2069 2012-01-18  Dominic Mazzoni  <dmazzoni@google.com>
2070
2071         Accessibility: Chromium needs methods to scroll an object into view or to a specific location.
2072         https://bugs.webkit.org/show_bug.cgi?id=73460
2073
2074         Reviewed by Chris Fleizach.
2075
2076         * DumpRenderTree/AccessibilityUIElement.h:
2077         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
2078         (AccessibilityUIElement::AccessibilityUIElement):
2079         (AccessibilityUIElement::scrollToMakeVisibleCallback):
2080         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocusCallback):
2081         (AccessibilityUIElement::scrollToGlobalPointCallback):
2082         * DumpRenderTree/chromium/AccessibilityUIElement.h:
2083         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2084         (AccessibilityUIElement::scrollToMakeVisible):
2085         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2086         (AccessibilityUIElement::scrollToGlobalPoint):
2087         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2088         (AccessibilityUIElement::scrollToMakeVisible):
2089         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2090         (AccessibilityUIElement::scrollToGlobalPoint):
2091         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2092         (AccessibilityUIElement::scrollToMakeVisible):
2093         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2094         (AccessibilityUIElement::scrollToGlobalPoint):
2095
2096 2012-01-17  Enrica Casucci  <enrica@apple.com>
2097
2098         Missing NSForegroundColorAttributeName should be treated as black instead of transparent.
2099         https://bugs.webkit.org/show_bug.cgi?id=76490
2100         <rdar://problem/9460733>
2101
2102         Added TestWebKitAPI test.
2103         
2104         Reviewed by Dan Bernstein.
2105
2106         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added new test file.
2107         * TestWebKitAPI/Tests/mac/InspectorBar.mm: Added test that calls
2108         changeAttributes to the WebView.
2109         
2110 2012-01-18  Balazs Kelemen  <kbalazs@webkit.org>
2111
2112         [Qt] Consolidate layout test crash logging
2113         https://bugs.webkit.org/show_bug.cgi?id=75088
2114
2115         Reviewed by Simon Hausmann.
2116
2117         Move backtrace generating logic into WTFReportBacktrace
2118         and add a way to deinstall signal handlers if we know
2119         that we have already printed the backtrace.
2120
2121         * DumpRenderTree/qt/main.cpp:
2122         (crashHandler):
2123         (setupSignalHandlers):
2124         (WTFCrashHook):
2125         (main):
2126         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
2127         (WTR::crashHandler):
2128         (WTR::setupSignalHandlers):
2129         (WTR::crashHook):
2130         (WTR::InjectedBundle::platformInitialize):
2131
2132 2012-01-18  Ilya Tikhonovsky  <loislo@chromium.org>
2133
2134         Unreviewed build fix after r105256.
2135
2136         There was a cyclic dependency between self._port and self._host assigments.
2137
2138         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2139         (PerfTestsRunner.__init__):
2140
2141 2012-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2142
2143         Add new watchlist rule for the Qt build system
2144
2145         Reviewed by Simon Hausmann.
2146
2147         * Scripts/webkitpy/common/config/watchlist:
2148
2149 2012-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2150
2151         [Qt] Move OTHER_FILES from WebKit.pro to Tools.pro
2152
2153         Reviewed by Simon Hausmann.
2154
2155         * Tools.pro:
2156
2157 2012-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2158
2159         [Qt] Allow building only a subset of the WebKit sub-projects
2160
2161         Fixes build-jsc, broken in r104825.
2162
2163         We no longer need the intermediate buildQMakeQtProject(), as we're the
2164         only port using qmake, and buildQMakeProject is now highly Qt specific.
2165
2166         <http://webkit.org/b/76179>
2167
2168         Reviewed by Simon Hausmann.
2169
2170         * Scripts/build-jsc:
2171         * Scripts/build-webkit:
2172         * Scripts/webkitdirs.pm:
2173         (buildQMakeProjects):
2174
2175 2012-01-18  Ryosuke Niwa  <rniwa@webkit.org>
2176
2177         run-perf-tests should generate a json file that summaries the result
2178         https://bugs.webkit.org/show_bug.cgi?id=76504
2179
2180         Reviewed by Adam Barth.
2181
2182         Add an ability to generate a json file to run-perf-test in the preparation for perf bots.
2183         New option --outout-json-path specifies the json file's path, and --source-json-path specifies
2184         another json file to be merged into the generated json file.
2185
2186         Also fixed a bug that --build wasn't set by default.
2187
2188         * Scripts/webkitpy/layout_tests/port/test.py:
2189         (TestPort.webkit_base):
2190         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2191         (PerfTestsRunner.__init__):
2192         (PerfTestsRunner):
2193         (PerfTestsRunner._parse_args):
2194         (PerfTestsRunner.run):
2195         (PerfTestsRunner._generate_json_if_specified):
2196         (PerfTestsRunner._process_chromium_style_test_result):
2197         (PerfTestsRunner._process_parser_test_result):
2198         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2199         (MainTest):
2200         (create_runner):
2201         (test_run_test_set):
2202         (test_run_test_set_for_parser_tests):
2203         (test_run_test_set_with_summary_json):
2204         (test_collect_tests):
2205         (test_parse_args):
2206
2207 2012-01-17  Sheriff Bot  <webkit.review.bot@gmail.com>
2208
2209         Unreviewed, rolling out r105244.
2210         http://trac.webkit.org/changeset/105244
2211         https://bugs.webkit.org/show_bug.cgi?id=76518
2212
2213         broke Chromium Mac (Requested by rolandsteiner on #webkit).
2214
2215         * DumpRenderTree/AccessibilityUIElement.h:
2216         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
2217         (AccessibilityUIElement::AccessibilityUIElement):
2218         * DumpRenderTree/chromium/AccessibilityUIElement.h:
2219         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2220         (AccessibilityUIElement::removeSelection):
2221         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2222         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2223         (AccessibilityUIElement::removeSelection):
2224
2225 2012-01-17  Dominic Mazzoni  <dmazzoni@google.com>
2226
2227         Accessibility: Chromium needs methods to scroll an object into view or to a specific location.
2228         https://bugs.webkit.org/show_bug.cgi?id=73460
2229
2230         Reviewed by Chris Fleizach.
2231
2232         * DumpRenderTree/AccessibilityUIElement.h:
2233         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
2234         (AccessibilityUIElement::AccessibilityUIElement):
2235         (AccessibilityUIElement::scrollToMakeVisibleCallback):
2236         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocusCallback):
2237         (AccessibilityUIElement::scrollToGlobalPointCallback):
2238         * DumpRenderTree/chromium/AccessibilityUIElement.h:
2239         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2240         (AccessibilityUIElement::scrollToMakeVisible):
2241         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2242         (AccessibilityUIElement::scrollToGlobalPoint):
2243         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2244         (AccessibilityUIElement::scrollToMakeVisible):
2245         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2246         (AccessibilityUIElement::scrollToGlobalPoint):
2247         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2248         (AccessibilityUIElement::scrollToMakeVisible):
2249         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2250         (AccessibilityUIElement::scrollToGlobalPoint):
2251
2252 2012-01-17  Adam Barth  <abarth@webkit.org>
2253
2254         Enable tests on the mac-ews
2255         https://bugs.webkit.org/show_bug.cgi?id=76502
2256
2257         Reviewed by Simon Fraser.
2258
2259         Apple now has five machines running the mac-ews.  Lucas Forschler has
2260         asked that we try enabling tests to see if we have enough bandwidth.
2261
2262         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
2263         (MacEWS):
2264
2265 2012-01-17  Sanjoy Das  <sanjoy@playingwithpointers.com>
2266
2267         Add a check-webkit-style exception for GDBInterface.cpp
2268         https://bugs.webkit.org/show_bug.cgi?id=76187
2269
2270         Add an exception for Source/JavaScriptCore/jit/GDBInterface.cpp so
2271         that it does not fail the style check.
2272
2273         Reviewed by Martin Robinson.
2274
2275         * Scripts/webkitpy/style/checker.py: Tell the style checker to not run
2276         readability/naming tests on
2277         Source/JavaScriptCore/jit/GDBInterface.cpp.
2278         * Scripts/webkitpy/style/checker_unittest.py:
2279         (GlobalVariablesTest.test_path_rules_specifier): Add a test for
2280         the above exception.
2281
2282 2012-01-17  Hajime Morrita  <morrita@chromium.org>
2283
2284         https://bugs.webkit.org/show_bug.cgi?id=76245
2285         [Mac][DRT] should allow query string for the file URL.
2286
2287         Reviewed by Tony Chang.
2288
2289         Taught DRT about file:// as a url instead of a file path.
2290         We are now able to use a path file:///foo/bar.html?baz as a url
2291         with which DRT opens "/foo/bar.html" with "baz" as a query string.
2292
2293         * DumpRenderTree/mac/DumpRenderTree.mm:
2294         (runTest):
2295
2296 2012-01-17  Ojan Vafai  <ojan@chromium.org>
2297
2298         Check style on test_expectations.txt files before commit
2299         https://bugs.webkit.org/show_bug.cgi?id=76484
2300
2301         Reviewed by Adam Barth.
2302
2303         Unlike other style rules, if you get the syntax of the test_expectations.txt
2304         file wrong, the layout tests won't run. Also, this check is simple and only
2305         slows down committing if you actually modify one of the test_expectations.txt files.
2306
2307         * Scripts/webkitpy/tool/steps/commit.py:
2308         (Commit._check_test_expectations):
2309         (Commit.run):
2310         * Scripts/webkitpy/tool/steps/commit_unittest.py: Added.
2311         (CommitTest):
2312         (CommitTest.test_check_test_expectations):
2313
2314 2012-01-17  Viatcheslav Ostapenko  <ostapenko.viatcheslav@nokia.com>
2315
2316         [Qt] Debug build fails with debug qt5
2317         https://bugs.webkit.org/show_bug.cgi?id=76463
2318
2319         Reviewed by Simon Hausmann.
2320
2321         Add QNetworkCookie include in order to satisfy sizeof in QTypeInfo in debug builds.
2322
2323         * QtTestBrowser/cookiejar.h:
2324
2325 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2326
2327         webkitpy: fix json import on linux
2328         https://bugs.webkit.org/show_bug.cgi?id=76481
2329
2330         Reviewed by Adam Barth.
2331
2332         For some reason, the version of simplejson autoinstalled on my linux
2333         box seems to be broken. We should really only be using
2334         simplejson if json isn't available anyway, so this change looks
2335         for json first.
2336
2337         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2338         (JSONGeneratorTest.test_test_timings_trie):
2339
2340 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2341
2342         Fix failures in test-webkitpy caused by r105177.
2343
2344         Unreviewed, build fix.
2345
2346         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2347         (ChromiumLinuxPort._determine_driver_path_statically):
2348         (ChromiumLinuxPort._static_build_path):
2349
2350 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2351
2352         webkitpy: make PortFactory.get() be fully data-driven
2353         https://bugs.webkit.org/show_bug.cgi?id=76360
2354
2355         Reviewed by Eric Seidel.
2356
2357         Refactor PortFactory.get() to be fully data-driven.
2358
2359         * Scripts/webkitpy/layout_tests/port/factory.py:
2360         (PortFactory):
2361         (PortFactory._default_port):
2362         (PortFactory.get):
2363         (PortFactory.get.in):
2364         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
2365         (FactoryTest.assert_port):
2366
2367 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2368
2369         webkitpy: add determine_full_port_name(), clean up port.__init__()
2370         https://bugs.webkit.org/show_bug.cgi?id=76357
2371
2372         Reviewed by Adam Barth.
2373
2374         Each Port class is now required to implement a 'port_name' field
2375         that will match the start of all ports constructed by this class and
2376         a 'determine_full_port_name' field that will make the port
2377         fully-qualified.
2378
2379         * Scripts/webkitpy/layout_tests/port/apple.py:
2380         (ApplePort.__init__):
2381         * Scripts/webkitpy/layout_tests/port/base.py:
2382         (Port):
2383         (Port.determine_port_name):
2384         (Port.__init__):
2385         * Scripts/webkitpy/layout_tests/port/chromium.py:
2386         (ChromiumPort.for):
2387         (ChromiumPort.__init__):
2388         (ChromiumPort):
2389         (ChromiumPort._chromium_base_dir):
2390         (ChromiumPort.path_from_chromium_base):
2391         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2392         (ChromiumGpuLinuxPort):
2393         (ChromiumGpuLinuxPort.__init__):
2394         (ChromiumGpuMacPort):
2395         (ChromiumGpuMacPort.__init__):
2396         (ChromiumGpuWinPort):
2397         (ChromiumGpuWinPort.__init__):
2398         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2399         (ChromiumLinuxPort):
2400         (ChromiumLinuxPort.determine_port_name):
2401         (ChromiumLinuxPort.__init__):
2402         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2403         (ChromiumMacPort):
2404         (ChromiumMacPort.determine_port_name):
2405         (ChromiumMacPort.__init__):
2406         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2407         (ChromiumWinPort):
2408         (ChromiumWinPort.determine_port_name):
2409         (ChromiumWinPort.__init__):
2410         * Scripts/webkitpy/layout_tests/port/dryrun.py:
2411         (DryRunPort):
2412         (DryRunPort.determine_port_name):
2413         (DryRunPort.__init__):
2414         * Scripts/webkitpy/layout_tests/port/efl.py:
2415         (EflPort):
2416         * Scripts/webkitpy/layout_tests/port/factory.py:
2417         (PortFactory):
2418         (PortFactory._default_port):
2419         (PortFactory.get):
2420         (PortFactory.get.in):
2421         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
2422         (FactoryTest.assert_port):
2423         * Scripts/webkitpy/layout_tests/port/google_chrome.py:
2424         (GoogleChromeLinux32Port):
2425         (GoogleChromeLinux32Port.determine_port_name):
2426         (GoogleChromeLinux64Port):
2427         (GoogleChromeLinux64Port.determine_port_name):
2428         (GoogleChromeMacPort):
2429         (GoogleChromeMacPort.determine_port_name):
2430         (GoogleChromeWinPort):
2431         (GoogleChromeWinPort.determine_port_name):
2432         * Scripts/webkitpy/layout_tests/port/gtk.py:
2433         (GtkPort):
2434         * Scripts/webkitpy/layout_tests/port/mac.py:
2435         (MacPort):
2436         (MacPort.determine_port_name):
2437         (MacPort.__init__):
2438         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2439         (test_versions):
2440         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
2441         (MockDRTPort):
2442         (MockDRTPort.determine_port_name):
2443         (MockDRTPort.__init__):
2444         (MockDRT):
2445         (MockDRT.determine_port_name):
2446         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2447         (PortTestCase.make_port):
2448         * Scripts/webkitpy/layout_tests/port/qt.py:
2449         (QtPort):
2450         (QtPort.determine_port_name):
2451         (QtPort.__init__):
2452         * Scripts/webkitpy/layout_tests/port/test.py:
2453         (TestPort):
2454         (TestPort.determine_port_name):
2455         (TestPort.__init__):
2456         * Scripts/webkitpy/layout_tests/port/webkit.py:
2457         (WebKitPort.__init__):
2458         * Scripts/webkitpy/layout_tests/port/win.py:
2459         (WinPort):
2460         (WinPort.determine_port_name):
2461         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
2462         (get_test_baselines.AllPlatformsPort.__init__):
2463         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
2464         (get_test_config.TestMacPort):
2465         (get_test_config):
2466
2467 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2468
2469         webkitpy: clean up port code in preparation for static port names
2470         https://bugs.webkit.org/show_bug.cgi?id=76356
2471
2472         Reviewed by Adam Barth.
2473
2474         To fix bug 76215 and be able to determine appropriate port names
2475         in webkitpy without actually constructing Port objects, we need
2476         to shuffle some logic in the chromium ports to be able to figure
2477         out whether we built DRT in 32 or 64-bit mode.
2478
2479         * Scripts/webkitpy/layout_tests/port/chromium.py:
2480         (ChromiumPort):
2481         (ChromiumPort._chromium_base_dir):
2482         (ChromiumPort.__init__):
2483         (ChromiumPort.path_from_chromium_base):
2484         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2485         (ChromiumLinuxPort):
2486         (ChromiumLinuxPort._determine_driver_path_statically):
2487         (ChromiumLinuxPort._static_build_path):
2488         (ChromiumLinuxPort._determine_architecture):
2489         (ChromiumLinuxPort.__init__):
2490         (ChromiumLinuxPort._build_path):
2491
2492 2012-01-17  Hao Zheng  <zhenghao@chromium.org>
2493
2494         Do not remove pixel failures for ref tests.
2495         https://bugs.webkit.org/show_bug.cgi?id=76243
2496
2497         Reviewed by Tony Chang.
2498
2499         'NRWT --no-pixel-tests' complains when reftests are expected to
2500         be image mismatch.
2501
2502         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2503         (Manager._update_summary_with_result):
2504         * Scripts/webkitpy/layout_tests/models/test_failures.py:
2505         (is_reftest_failure):
2506         (determine_result_type):
2507
2508 2012-01-17  Balazs Ankes  <Ankes.Balazs@stud.u-szeged.hu>
2509
2510         REGRESSION(r81225): ORWT should ignore reftests
2511         https://bugs.webkit.org/show_bug.cgi?id=67936
2512
2513         Reviewed by Csaba Osztrogonác.
2514
2515         * Scripts/old-run-webkit-tests:
2516         (isUsedInReftest): $filename should be the first parameter
2517         (findTestsToRun): run test if it isn't reftest
2518
2519 2012-01-17  Gustavo Noronha Silva  <gns@gnome.org>
2520
2521         Unreviewed. Fix python unit test I broke with r105142
2522
2523         * Scripts/webkitpy/common/config/ports_unittest.py:
2524         (WebKitPortTest.test_gtk_port):
2525
2526 2012-01-17  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2527
2528         [Qt] Don't set the 'primary' flag when mocking touch-points in MiniBrowser
2529
2530         Qt 5 no longer has that flag.
2531
2532         Reviewed by Simon Hausmann.
2533
2534         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2535         (MiniBrowserApplication::notify):
2536
2537 2012-01-16  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
2538
2539         GTK+ EWS needs to run update-webkitgtk-libs after applying a patch
2540         https://bugs.webkit.org/show_bug.cgi?id=75857
2541
2542         Reviewed by Adam Barth.
2543
2544         * Scripts/webkitdirs.pm:
2545         (buildAutotoolsProject): move cleaning the build to happen early,
2546         and handle --update-gtk right after that
2547         * Scripts/webkitpy/common/config/ports.py:
2548         (GtkPort.build_webkit_command): pass --update-gtk as an option to
2549         build-webkit, so that it is used in the EWS
2550
2551 2012-01-16  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2552
2553         [Qt] Don't add _debug postfix to targets when building debug only
2554
2555         Reviewed by Simon Hausmann.
2556
2557         * qmake/mkspecs/features/default_post.prf:
2558         * qmake/mkspecs/features/functions.prf:
2559
2560 2012-01-16  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2561
2562         [Qt] Add more files to project-files for easier navigation in Qt Creator
2563
2564         Reviewed by Simon Hausmann.
2565
2566         * Tools.pro:
2567         * qmake/config.tests/README: Added.
2568
2569 2012-01-17  Philippe Normand  <pnormand@igalia.com>
2570
2571         [GTK] fast/dom/constructed-objects-prototypes.html fails
2572         https://bugs.webkit.org/show_bug.cgi?id=75432
2573
2574         Reviewed by Martin Robinson.
2575
2576         * Scripts/build-webkit: Enable video-track in GTK.
2577
2578 2012-01-16  Takashi Toyoshima  <toyoshim@chromium.org>
2579
2580         Add toyoshim as a committer.
2581         https://bugs.webkit.org/show_bug.cgi?id=76390
2582
2583         Reviewed by Hajime Morita.
2584
2585         * Scripts/webkitpy/common/config/committers.py:
2586
2587 2012-01-16  Brent Fulgham  <bfulgham@webkit.org>
2588
2589         [WinCairo] Build fix after r104155. Several debug targets were not
2590         properly including the '*Common' property sheets.
2591
2592         * DumpRenderTree/win/DumpRenderTreeLauncherDebugCairoCFLite.vsprops:
2593         * DumpRenderTree/win/ImageDiffLauncherDebugCairoCFLite.vsprops:
2594         * MiniBrowser/Configurations/MiniBrowserLauncherDebugCairoCFLite.vsprops:
2595         * WebKitTestRunner/win/WebKitTestRunnerLauncherDebugCairoCFLite.vsprops:
2596         * WinLauncher/WinLauncherLauncherDebugCairoCFLite.vsprops:
2597
2598 2012-01-16  Csaba Osztrogonác  <ossy@webkit.org>
2599
2600         REGRESSION(r105072): It broke committers_unittest.py
2601         https://bugs.webkit.org/show_bug.cgi?id=76396
2602
2603         Reviewed by Adam Barth.
2604
2605         * Scripts/webkitpy/common/config/committers_unittest.py:
2606         (CommittersTest._assert_fuzz_match): It should handle more names.
2607         (CommittersTest.test_contributors_by_fuzzy_match_with_legacy_names): Fix the testcase.
2608
2609 2012-01-16  Mario Sanchez Prada  <msanchez@igalia.com>
2610
2611         [GTK] Include gtk+3 in the jhbuild modules
2612         https://bugs.webkit.org/show_bug.cgi?id=76284
2613
2614         Reviewed by Philippe Normand.
2615
2616         * gtk/jhbuild.modules: Add some gtk+3 to the jhbuild moduleset.
2617
2618 2012-01-16  Dan Winship  <danw@gnome.org>
2619
2620         Add myself as a committer
2621
2622         Unreviewed
2623
2624         * Scripts/webkitpy/common/config/committers.py:
2625
2626 2012-01-16  Raphael Kubo da Costa  <kubo@profusion.mobi>
2627
2628         Unreviewed; fix the build when --as-needed and --no-copy-dt-needed-entries are passed to the linker.
2629
2630         Some recent versions of a few Linux distributions have started passing
2631         --as-needed and --no-copy-dt-needed-entries by default to the linker,
2632         which broke the build when SHARED_CORE was turned on, as libwtf_efl.so
2633         needed a few symbols from libjavascriptcore_efl.so.
2634
2635         * DumpRenderTree/efl/CMakeLists.txt: Explicitly pass WTF_LIBRARY_NAME
2636         before JavaScriptCore_LIBRARY_NAME in DumpRenderTree_LIBRARIES so the
2637         libraries are searched in the right order by the linker.
2638
2639 2012-01-16  Philippe Normand  <pnormand@igalia.com>
2640
2641         [GTK] set audio resources path in NRTW
2642         https://bugs.webkit.org/show_bug.cgi?id=76380
2643
2644         Reviewed by Martin Robinson.
2645
2646         * Scripts/webkitpy/layout_tests/port/gtk.py:
2647         (GtkPort.setup_environ_for_server): Set $AUDIO_RESOURCES_PATH so
2648         the uninstalled audio resources are used during webaudio tests.
2649
2650 2012-01-16  Carlos Garcia Campos  <cgarcia@igalia.com>
2651
2652         [GTK] ldflags and cflags should take precedence over the existing flags in gtkdoc.py
2653         https://bugs.webkit.org/show_bug.cgi?id=76369
2654
2655         Reviewed by Xan Lopez.
2656
2657         If there's an installed version of webkit and there are new
2658         symbols in the build tree, gtkdoc-scanobj fails due to undefined
2659         references, because the env ldflags are given before the local
2660         ones.
2661
2662         * gtk/gtkdoc.py:
2663         (GTKDoc._run_gtkdoc_scangobj):
2664
2665 2012-01-15  Pablo Flouret  <pablof@motorola.com>
2666
2667         Fix compilation errors on build-webkit --debug --no-video on mac.
2668         https://bugs.webkit.org/show_bug.cgi?id=75867
2669
2670         Reviewed by Philippe Normand.
2671
2672         Add --[no-]video-track to build-webkit, since video track feature
2673         depends on video.
2674
2675         * Scripts/build-webkit:
2676
2677 2012-01-14  James Robinson  <jamesr@chromium.org>
2678
2679         [chromium] Failing webkit_unit_tests does not make the cr-linux EWS bubble red
2680         https://bugs.webkit.org/show_bug.cgi?id=76313
2681
2682         Reviewed by Adam Barth.
2683
2684         The exit code of this script should be the exit code of webkit_unit_tests. Tested manually.
2685
2686         * Scripts/run-chromium-webkit-unit-tests:
2687
2688 2012-01-14  Sheriff Bot  <webkit.review.bot@gmail.com>
2689
2690         Unreviewed, rolling out r105017.
2691         http://trac.webkit.org/changeset/105017
2692         https://bugs.webkit.org/show_bug.cgi?id=76333
2693
2694         broke the build (Requested by philn-tp on #webkit).
2695
2696         * gtk/jhbuild.modules:
2697
2698 2012-01-14  Mario Sanchez Prada  <msanchez@igalia.com>
2699
2700         [GTK] Include gtk+3 in the jhbuild modules
2701         https://bugs.webkit.org/show_bug.cgi?id=76284
2702
2703         Reviewed by Philippe Normand.
2704
2705         * gtk/jhbuild.modules: Add some gtk+3 to the jhbuild moduleset.
2706
2707 2012-01-13  Adam Barth  <abarth@webkit.org>
2708
2709         Remove --dry-run support from webkit-patch
2710         https://bugs.webkit.org/show_bug.cgi?id=76300
2711
2712         Reviewed by Dirk Pranke.
2713
2714         This is the rest of the patch from Bug 76300.  I landed it separtely
2715         because main.py had some stray conflict markers that I needed to clean
2716         up first.
2717
2718         * Scripts/webkitpy/test/main.py:
2719         (Tester.parse_args):
2720         (Tester._run_tests):
2721
2722 2012-01-13  Adam Barth  <abarth@webkit.org>
2723
2724         test-webkitpy does not run
2725         https://bugs.webkit.org/show_bug.cgi?id=76318
2726
2727         Unreviewed.
2728
2729         Looks like some conflict markers got left in the file.
2730
2731         * Scripts/webkitpy/test/main.py:
2732         (Tester.run):
2733
2734 2012-01-13  Adam Barth  <abarth@webkit.org>
2735
2736         Remove --dry-run support from webkit-patch
2737         https://bugs.webkit.org/show_bug.cgi?id=76300
2738
2739         Reviewed by Dirk Pranke.
2740
2741         We don't use this functionality for anything, and it doesn't really
2742         work.  Someone in #webkit tried using it and and got sad/confused.  We
2743         should just remove it.
2744
2745         * Scripts/webkitpy/common/checkout/scm/git.py:
2746         (Git.push_local_commits_to_server):
2747         * Scripts/webkitpy/common/checkout/scm/scm.py:
2748         (SCM.__init__):
2749         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
2750         (_shared_test_commit_with_message):
2751         (GitSVNTest.test_commit_text_parsing):
2752         * Scripts/webkitpy/common/checkout/scm/svn.py:
2753         (SVN.commit_with_message):
2754         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
2755         (Bugzilla.__init__):
2756         (Bugzilla.quips):
2757         (Bugzilla.authenticate):
2758         (Bugzilla.add_attachment_to_bug):
2759         (Bugzilla.add_patch_to_bug):
2760         (Bugzilla.create_bug):
2761         (Bugzilla.clear_attachment_flags):
2762         (Bugzilla.set_flag_on_attachment):
2763         (Bugzilla.obsolete_attachment):
2764         (Bugzilla.add_cc_to_bug):
2765         (Bugzilla.post_comment_to_bug):
2766         (Bugzilla.close_bug_as_fixed):
2767         (Bugzilla.reassign_bug):
2768         (reopen_bug):
2769         * Scripts/webkitpy/test/main.py:
2770         (Tester.parse_args):
2771         (Tester.run_tests):
2772
2773 2012-01-13  David Levin  <levin@chromium.org>
2774
2775         check-webkit-style: should encourage the use of Own* classes for Windows DC.
2776         https://bugs.webkit.org/show_bug.cgi?id=76227
2777
2778         Reviewed by Dirk Pranke.
2779
2780         * Scripts/webkitpy/style/checkers/cpp.py:
2781         (check_for_leaky_patterns): The new check.
2782         (process_line): Added a call to the new check.
2783         (CppChecker): Added the new error type.
2784         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
2785         (CppStyleTestBase):
2786         (CppStyleTestBase.perform_leaky_pattern_check):
2787         The check for only leaky pattern errors.
2788         (LeakyPatternTest): Test cases.
2789
2790 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
2791
2792         test-webkitpy: clean up handling of test directories, QueueStatusServer
2793         https://bugs.webkit.org/show_bug.cgi?id=76238
2794
2795         Reviewed by Adam Barth.
2796
2797         This patch cleans up a bunch of hacks related to how we find all
2798         of the python unit tests and set up the sys.path variable in
2799         test-webkitpy.
2800
2801         This patch changes how test-webkitpy works slightly in that we
2802         will now look for test files under Tools/Scripts and Source/WebKit2/Scripts
2803         rather than just Tools/Scripts/webkitpy and Source/WebKit2/Scripts/webkit.
2804         This greatly simplifies the mapping of paths to module names and
2805         appears to have no significant performance impact or other drawbacks.
2806
2807         There are other ways that we could remove all of the hacks and
2808         still have a more flexible mapping of paths to module names, but
2809         the ones I've thought of so far add a decent amount of complexity to
2810         the implementation and don't seem to buy us much.
2811         
2812         * QueueStatusServer/__init__.py:
2813         * Scripts/test-webkitpy:
2814         * Scripts/webkitpy/test/main.py:
2815         (Tester._modules_from_paths):
2816         (Tester._is_module):
2817
2818 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
2819
2820         test-webkitpy: should support classes and individual test names as well as modules
2821         https://bugs.webkit.org/show_bug.cgi?id=76233
2822
2823         Reviewed by Adam Barth.
2824
2825         The default unittest.main() supports a list of individual
2826         TestCase classes or individual tests as well as modules. This
2827         change adds support for the same to test-webkitpy, so now you
2828         can say
2829
2830           % test-wekbitpy webkitpy.test.cat_unittest.CatTest
2831
2832         and
2833
2834           % test-wekbitpy webkitpy.test.cat_unittest.CatTest.test_basic
2835
2836         in addition to
2837
2838           % test-wekbitpy webkitpy.test.cat_unittest
2839
2840         * Scripts/test-webkitpy:
2841         * Scripts/webkitpy/test/cat_unittest.py:
2842         * Scripts/webkitpy/test/main.py:
2843         (Tester.parse_args):
2844         (Tester.run):
2845         (Tester):
2846         (Tester._find_modules):
2847         (Tester._exclude):
2848         (Tester._run_tests):
2849         (Tester._is_module):
2850         (Tester._log_exception):
2851
2852 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
2853
2854         test-webkitpy: add basic code coverage support
2855         https://bugs.webkit.org/show_bug.cgi?id=76201
2856
2857         Reviewed by Adam Barth.
2858
2859         This test adds basic code coverage support to test-webkitpy; if
2860         you pass the -c flag, we will generate a new coverage file. We
2861         can start simply with this and add more features as necessary.
2862
2863         This patch requires coverage to be already installed on the
2864         machine. I am not auto-installing this for now.
2865
2866         * Scripts/webkitpy/test/main.py:
2867         (Tester.parse_args):
2868         (Tester.run_tests):
2869
2870 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
2871
2872         StatusBubbleTest.test_build_bubble is failing
2873         https://bugs.webkit.org/show_bug.cgi?id=76222
2874
2875         Reviewed by Adam Barth.
2876
2877         * QueueStatusServer/handlers/statusbubble_unittest.py:
2878         (StatusBubbleTest.test_build_bubble):
2879
2880 2012-01-13  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
2881
2882         nrwt should use multiple http shards
2883         https://bugs.webkit.org/show_bug.cgi?id=75958
2884
2885         Add --max-locked-shards option to the nrwt for manual set
2886         the maximum number of the locked shards
2887
2888         Reviewed by Dirk Pranke.
2889
2890         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2891         (Manager._max_locked_shards):
2892         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2893         (ShardingTests.get_shards):
2894         (ShardingTests.test_multiple_locked_shards):
2895         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2896         (parse_args):
2897
2898 2012-01-13  Raphael Kubo da Costa  <kubo@profusion.mobi>
2899
2900         prepare-ChangeLog: Look for a text editor in a few more variables.
2901         https://bugs.webkit.org/show_bug.cgi?id=76289
2902
2903         Reviewed by Ryosuke Niwa.
2904
2905         A text editor was looked for only in the CHANGE_LOG_EDITOR and
2906         CHANGE_LOG_EDIT_APPLICATION environment variables before falling back
2907         to the OS X-specific open(1) command.
2908
2909         As there does not seem to be a common way of looking for a text editor
2910         among the scripts in Tools/Scripts right now, just look at $VISUAL and
2911         $EDITOR (set much more often than the other variables in Unix systems)
2912         before falling back to open(1).
2913
2914         * Scripts/prepare-ChangeLog:
2915         (openChangeLogs):
2916
2917 2012-01-13  Gabor Rapcsanyi  <rgabor@webkit.org>
2918
2919         [GTK] gtkdoc.py ignoring LDFLAGS and CFLAGS environment variables
2920         https://bugs.webkit.org/show_bug.cgi?id=76269
2921
2922         Reviewed by Martin Robinson.
2923
2924         * gtk/gtkdoc.py:
2925         (GTKDoc._run_gtkdoc_scangobj):
2926
2927 2012-01-13  Csaba Osztrogonác  <ossy@webkit.org>
2928
2929         [Qt][WK2]REGRESSION(r104881):It broke hundreds of tests
2930         https://bugs.webkit.org/show_bug.cgi?id=76247
2931
2932         Reviewed by Kenneth Rohde Christiansen.
2933
2934         * WebKitTestRunner/TestController.cpp:
2935         (WTR::TestController::resetStateToConsistentValues): Disable mock scrollbars on Qt by default. (workaround)
2936
2937 2012-01-13  Hajime Morrita  <morrita@chromium.org>
2938
2939         [Mac][DRT] Should support --no-timeout
2940         https://bugs.webkit.org/show_bug.cgi?id=76242
2941
2942         Reviewed by Darin Adler.
2943
2944         * DumpRenderTree/mac/DumpRenderTree.mm:
2945         (initializeGlobalsFromCommandLineOptions): Added "--no-timeout" to the options list.
2946         (setWaitToDumpWatchdog): Added to encapsulate the waitToDumpWatchdog global variable.
2947         (shouldSetWaitToDumpWatchdog): Added to check the useTimeoutWatchdog flag set by the option.
2948         * DumpRenderTree/mac/DumpRenderTreeMac.h:
2949         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
2950         (LayoutTestController::setWaitToDump): Now calls shouldSetWaitToDumpWatchdog() to check the flag.
2951
2952 2012-01-12  Simon Fraser  <simon.fraser@apple.com>
2953
2954         Mitigate scrollbar differences when running pixel tests
2955         https://bugs.webkit.org/show_bug.cgi?id=67217
2956
2957         Reviewed by Dan Bernstein.
2958         
2959         Enable mock scrollbars by default for tests.
2960
2961         * DumpRenderTree/mac/DumpRenderTree.mm:
2962         (resetDefaultsToConsistentValues):
2963         * WebKitTestRunner/TestController.cpp:
2964         (WTR::TestController::resetStateToConsistentValues):
2965
2966 2012-01-12  Ryosuke Niwa  <rniwa@webkit.org>
2967
2968         Need a script to run tests in PerformanceTests
2969         https://bugs.webkit.org/show_bug.cgi?id=76132
2970
2971         Reviewed by Adam Barth.
2972
2973         Add run-perf-tests to run performance tests using parser/resources/runner.js.
2974         Unfortunately, there isn't an easy way of telling which test uses which format
2975         so hard-code directory that uses Chromium perf-bot style (only inspector perf. tests for now).
2976
2977         All test outputs are re-formatted to match Chromium perf-bot style.
2978
2979         * Scripts/run-inspector-perf-tests.py:
2980         * Scripts/run-perf-tests: Added.
2981         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2982         (PerfTestsRunner):
2983         (PerfTestsRunner.__init__):
2984         (PerfTestsRunner._collect_tests):
2985         (PerfTestsRunner.run):
2986         (PerfTestsRunner._print_status):
2987         (PerfTestsRunner._run_tests_set):
2988         (PerfTestsRunner._process_chromium_style_test_result):
2989         (PerfTestsRunner._should_ignore_line_in_parser_test_result):
2990         (PerfTestsRunner._process_parser_test_result):
2991         (PerfTestsRunner._run_single_test):
2992         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2993         (MainTest.TestDriver.run_test):
2994         (create_runner):
2995         (run_test):
2996         (test_run_test_set):
2997         (test_run_test_set_for_parser_tests):
2998         (test_collect_tests):
2999
3000 2012-01-12  Dirk Pranke  <dpranke@chromium.org>
3001
3002         webkitpy: clean up port unit tests in preparation for making port_name mandatory
3003         https://bugs.webkit.org/show_bug.cgi?id=76123
3004
3005         Reviewed by Adam Barth.
3006
3007         This change updates the unit tests to be more consistent in how
3008         ports are created, in preparation for making port_name a
3009         mandatory parameter to the constructor and removing any version
3010         interpretation from it.
3011
3012         There should be no functional changes in this patch, but some
3013         tests that are no longer relevant have been removed or reworked.
3014
3015         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3016         (PortTest.test_get_option__default):
3017         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
3018         (ChromiumLinuxPortTest):
3019         (ChromiumLinuxPortTest.assert_architecture):
3020         (ChromiumLinuxPortTest.test_determine_architecture_fails):
3021         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
3022         (ChromiumMacPortTest):
3023         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3024         (ChromiumPortTest):
3025         (ChromiumPortTest.test_all_test_configurations):
3026         (ChromiumPortTest.TestMacPort.__init__):
3027         (ChromiumPortTest.TestLinuxPort.__init__):
3028         (ChromiumPortTest.TestWinPort.__init__):
3029         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
3030         (ChromiumWinTest):
3031         * Scripts/webkitpy/layout_tests/port/efl_unittest.py:
3032         (EflPortTest):
3033         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
3034         (GtkPortTest):
3035         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3036         (MacTest):
3037         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
3038         (MockDRTPortTest):
3039         (MockDRTPortTest.make_port):
3040         * Scripts/webkitpy/layout_tests/port/qt.py:
3041         (QtPort._port_flag_for_scripts):
3042         (QtPort.__init__):
3043         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
3044         (QtPortTest):
3045         (QtPortTest._assert_search_path):
3046         (QtPortTest.test_baseline_search_path):
3047         (QtPortTest.test_operating_system):
3048         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3049         (WebKitPortTest):
3050         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
3051         (WinPortTest):
3052         (WinPortTest._assert_search_path):
3053         (WinPortTest.test_baseline_search_path):
3054         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
3055         (get_test_config.TestMacPort.__init__):
3056
3057 2012-01-12  Dirk Pranke  <dpranke@chromium.org>
3058
3059         test-webkitpy: clean up logging and provide a real --help message
3060         https://bugs.webkit.org/show_bug.cgi?id=76142
3061
3062         Reviewed by Adam Barth.
3063
3064         This test is mostly more refactoring - it calls objects in
3065         unittest directly instead of routing through unittest.main() --
3066         in order to gain more control over the logging and to provide a
3067         correct --help message. This will also give us a better
3068         foundation for adding more feautures down the road.
3069
3070         This patch adds different levels of logging controlled by the
3071         '-s', '-q', and '-v' options (see --help for details), and
3072         removes the --verbose-logging option.
3073
3074         * Scripts/test-webkitpy:
3075         * Scripts/webkitpy/test/main.py:
3076         (Tester):
3077         (Tester.clean_packages):
3078         (Tester.__init__):
3079         (Tester.parse_args):
3080         (Tester.configure):
3081         (Tester._configure_logging):
3082         (Tester.find_modules):
3083         (Tester._exclude):
3084         (Tester._modules_from_paths):
3085         (Tester.run_tests):
3086
3087 2012-01-12  Simon Hausmann  <simon.hausmann@nokia.com>
3088
3089         Make the new WTF module build on Qt
3090         https://bugs.webkit.org/show_bug.cgi?id=76163
3091
3092         Reviewed by Tor Arne Vestbø.
3093
3094         * qmake/mkspecs/modules/wtf.prf: Pull in WTF from the new location.
3095
3096 2012-01-12  Vsevolod Vlasov  <vsevik@chromium.org>
3097
3098         Unreviewed, mac DumpRenderTree fix.
3099
3100         * DumpRenderTree/mac/UIDelegate.mm:
3101         (-[UIDelegate webView:addMessageToConsole:withSource:]):
3102
3103 2012-01-12  Vsevolod Vlasov  <vsevik@chromium.org>
3104
3105         WebKitTestRunner should not print console message line number when it is equal to zero.
3106         https://bugs.webkit.org/show_bug.cgi?id=76167
3107
3108         Reviewed by Csaba Osztrogonác.
3109
3110         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
3111         (WTR::InjectedBundlePage::willAddMessageToConsole):
3112
3113 2012-01-11  Vsevolod Vlasov  <vsevik@chromium.org>
3114
3115         Make default console messages line numbers consistent.
3116         https://bugs.webkit.org/show_bug.cgi?id=74075
3117
3118         Reviewed by Pavel Feldman.
3119
3120         Unset line numbers are not printed to console now.
3121         Added Source/WebCore/inspector to efl DumpRenderTree include path.
3122
3123         * DumpRenderTree/chromium/WebViewHost.cpp:
3124         (WebViewHost::didAddMessageToConsole):
3125         * DumpRenderTree/efl/CMakeLists.txt:
3126         * DumpRenderTree/efl/DumpRenderTreeView.cpp:
3127         (onConsoleMessage):
3128         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3129         (webViewConsoleMessage):
3130         * DumpRenderTree/mac/UIDelegate.mm:
3131         (-[UIDelegate webView:addMessageToConsole:withSource:]):
3132         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
3133         (WebCore::WebPage::javaScriptConsoleMessage):
3134         * DumpRenderTree/win/UIDelegate.cpp:
3135         (UIDelegate::webViewAddMessageToConsole):
3136         * DumpRenderTree/wx/DumpRenderTreeWx.cpp:
3137         (LayoutWebViewEventHandler::OnConsoleMessageEvent):
3138
3139 2012-01-10  Dirk Pranke  <dpranke@chromium.org>
3140
3141         test-webkitpy: push more logic into webkitpy.test.main, clean up code
3142         https://bugs.webkit.org/show_bug.cgi?id=76021
3143
3144         Reviewed by Eric Seidel.
3145
3146         I plan to add more functionality to test-webkitpy, but it is
3147         difficult to hack on now; this patch is the first of two that
3148         brings test-webkitpy more inline with current coding style by
3149         pushing more logic into the webkitpy.test.main.Tester class (so
3150         that it will be testable itself).
3151         
3152         There should be no functional changes in this patch.
3153
3154         * Scripts/test-webkitpy:
3155         (_path_from_webkit_root):
3156         * Scripts/webkitpy/test/main.py:
3157         (Tester):
3158         (Tester.init):
3159         (Tester.configure_logging):
3160         (Tester.configure_logging.filter):
3161         (Tester.clean_packages):
3162         (Tester.run_tests):
3163
3164 2012-01-11  Wei Jia  <wjia@chromium.org>
3165
3166         implement layout tests for <video> with media stream
3167         https://bugs.webkit.org/show_bug.cgi?id=74882
3168
3169         Reviewed by Kent Tamura.
3170
3171         * DumpRenderTree/DumpRenderTree.gypi:
3172         * DumpRenderTree/chromium/WebUserMediaClientMock.cpp: Added.
3173         (WebKit::WebUserMediaClientMock::create):
3174         (WebKit::WebUserMediaClientMock::IsMockStream):
3175         (WebKit::WebUserMediaClientMock::requestUserMedia):
3176         (WebKit::WebUserMediaClientMock::cancelUserMediaRequest):
3177         * DumpRenderTree/chromium/WebUserMediaClientMock.h: Added.
3178         (WebKit::WebUserMediaClientMock::~WebUserMediaClientMock):
3179         (WebKit::WebUserMediaClientMock::WebUserMediaClientMock):
3180         * DumpRenderTree/chromium/WebViewHost.cpp:
3181         (WebViewHost::userMediaClient):
3182         (WebViewHost::userMediaClientMock):
3183         (WebViewHost::createMediaPlayer):
3184         (WebViewHost::mediaStreamUtil):
3185         (WebViewHost::testMediaStreamClient):
3186         * DumpRenderTree/chromium/WebViewHost.h:
3187
3188 2012-01-11  Dirk Pranke  <dpranke@chromium.org>
3189
3190         Unreviewed, build fix.
3191
3192         Fix chromium-win's version remapping to work correctly on Win 7 SP1.
3193
3194         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
3195         (ChromiumWinPort.__init__):
3196
3197 2012-01-11  Ryosuke Niwa  <rniwa@webkit.org>
3198
3199         Sheriffbot build fix. Don't emit empty lines on IRC.
3200
3201         * Scripts/webkitpy/tool/bot/irc_command.py:
3202         (LastGreenRevision.execute):
3203
3204 2012-01-11  Ryosuke Niwa  <rniwa@webkit.org>
3205
3206         last-green-revision should report the revision that succeeded on all bots
3207         https://bugs.webkit.org/show_bug.cgi?id=76109
3208
3209         Reviewed by Adam Barth.
3210
3211         Add an algorithm to find the last known good revision among the specified bots. For a revision to be
3212         considered green by this algorithm, all matching builders must have a successful run at the revision
3213         or two consecutive successful runs before and after the revision.
3214
3215         Also fixed a bug in irc_command that the result wasn't posted on IRC properly and a bug in queries.py
3216         that resulted in an exception when the user didn't pass BUILDER_NAME argument.
3217
3218         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
3219         (BuildBot._revisions_for_builder):
3220         (BuildBot):
3221         (BuildBot._find_green_revision):
3222         (BuildBot.last_green_revision):
3223         * Scripts/webkitpy/common/net/buildbot/buildbot_mock.py:
3224         (MockBuildBot.last_green_revision):
3225         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
3226         (test_revisions_for_builder):
3227         (test_find_green_revision):
3228         (test_last_green_revision):
3229         * Scripts/webkitpy/tool/bot/irc_command.py:
3230         (LastGreenRevision.execute):
3231         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
3232         (SheriffIRCBotTest.test_lgr):
3233         * Scripts/webkitpy/tool/commands/queries.py:
3234         (LastGreenRevision):
3235
3236 2012-01-11  Dirk Pranke  <dpranke@chromium.org>
3237
3238         Unreviewed, fix build bustage on win32 introduced by r104725.
3239
3240         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
3241         (MockDRTPortTest.make_port):
3242         (MockChromiumDRTTest.test_pixeltest__fails):
3243
3244 2012-01-11  Jochen Eisinger  <jochen@chromium.org>
3245
3246         [Chromium] mimic the (old) behavior of the mac port for (re)setting the color profile
3247         https://bugs.webkit.org/show_bug.cgi?id=75618
3248
3249         Eventually, we should get away from setting the display's color profile, as the mac port is doing now.
3250
3251         Reviewed by Tony Chang.
3252
3253         * DumpRenderTree/chromium/LayoutTestHelper.mm:
3254         (installLayoutTestColorProfile):
3255         (restoreUserColorProfile):
3256         (main):
3257
3258 2012-01-11  Dirk Pranke  <dpranke@chromium.org>
3259
3260         test-webkitpy: fix -v and eliminate some spurious warnings
3261         https://bugs.webkit.org/show_bug.cgi?id=76099
3262
3263         Reviewed by Adam Barth.
3264
3265         when we added support for --xml output, we broke the -v flag.
3266
3267         Also, fix the code to swallow a harmless warning from
3268         VCSUtils.pm and fix a deprecated warning in jsonchecker.py.
3269
3270         test-webkitpy is now silent again except for the one known
3271         logging issue.
3272
3273         * Scripts/webkitpy/common/checkout/checkout.py:
3274         (Checkout.commit_message_for_this_commit):
3275         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
3276         (test_commit_message_for_this_commit):
3277         * Scripts/webkitpy/style/checkers/jsonchecker.py:
3278         (JSONChecker.check):
3279         (JSONChecker.line_number_from_json_exception):
3280         * Scripts/webkitpy/test/main.py:
3281         (Tester.run_tests):
3282
3283 2012-01-11  Dirk Pranke  <dpranke@chromium.org>
3284
3285         webkitpy: webkitpy.layout_tests.port tests should only use MockSystemHost, not MockHost
3286         https://bugs.webkit.org/show_bug.cgi?id=76084
3287
3288         Reviewed by Adam Barth.
3289
3290         This will help make it easier to catch layering violations. This
3291         change just refactors test code; there are no changes to
3292         production code.
3293
3294         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3295         (PortTest.make_port):
3296         (PortTest.test_find_no_paths_specified):
3297         (PortTest.test_find_one_test):
3298         (PortTest.test_find_glob):
3299         (PortTest.test_find_with_skipped_directories):
3300         (PortTest.test_find_with_skipped_directories_2):
3301         (PortTest.test_parse_reftest_list):
3302         (PortTest.test_httpd_returns_error_code):
3303         (PortTest.test_virtual_methods):
3304         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
3305         (ChromiumGpuTest.assert_port_works):
3306         (ChromiumGpuTest._assert_baseline_path):
3307         (ChromiumGpuTest.test_graphics_type):
3308         (ChromiumGpuTest.test_default_tests_paths.test_paths):
3309         (ChromiumGpuTest.test_test_files):
3310         (ChromiumGpuTest.test_test_files.test_paths):
3311         (ChromiumGpuTest):
3312         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
3313         (ChromiumLinuxPortTest.assert_architecture):
3314         (ChromiumLinuxPortTest.test_check_illegal_port_names):
3315         (ChromiumLinuxPortTest.test_determine_architecture_fails):
3316         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
3317         (DriverTest.make_port):
3318         (DriverTest):
3319         (DriverTest.assertVirtual):
3320         (DriverTest._assert_wrapper):
3321         (DriverTest.test_virtual_driver_methods):
3322         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
3323         (MockDRTPortTest.make_port):
3324         (MockDRTPortTest.test_port_name_in_constructor):
3325         (MockDRTTest.assertTest):
3326         (MockDRTTest.test_main):
3327         (MockChromiumDRTTest.test_pixeltest__fails):
3328         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
3329         (QtPortTest._assert_search_path):
3330         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3331         (TestWebKitPort.__init__):
3332         (WebKitPortUnitTests.test_default_options):
3333         (test_runtime_feature_list):
3334         (test_test_expectations):
3335
3336 2012-01-11  Ojan Vafai  <ojan@chromium.org>
3337
3338         Style skipped tests in the flakiness dashboard
3339         https://bugs.webkit.org/show_bug.cgi?id=76091
3340
3341         Reviewed by Tony Chang.
3342
3343         * TestResultServer/static-dashboards/dashboard_base.js:
3344         Removed "FLAKY" from this list since we don't ever mark individual runs
3345         as flaky. Also deleted an outdated line.
3346         * TestResultServer/static-dashboards/flakiness_dashboard.html:
3347
3348 2012-01-11  Ojan Vafai  <ojan@chromium.org>
3349
3350         Remove gtest normalization code from the test results server
3351         https://bugs.webkit.org/show_bug.cgi?id=76089
3352
3353         Reviewed by Tony Chang.
3354
3355         The normalization now happens on the gtest side.
3356
3357         * TestResultServer/model/jsonresults.py:
3358         (JsonResults.merge):
3359         * TestResultServer/model/jsonresults_unittest.py:
3360         (JsonResultsTest.test_gtest):
3361
3362 2012-01-11  Ojan Vafai  <ojan@chromium.org>
3363
3364         Be more aggressive about deleting unnecessary results from the test-results-server json files
3365         https://bugs.webkit.org/show_bug.cgi?id=76086
3366
3367         Reviewed by Tony Chang.
3368
3369         Remove any tests that only contain skip/pass/nodata values whose runtime
3370         is less than 5 seconds.
3371
3372         * TestResultServer/model/jsonresults.py:
3373         (JsonResults._normalize_results_json):
3374         * TestResultServer/model/jsonresults_unittest.py:
3375         (JsonResultsTest.test_merge_remove_test):
3376         (JsonResultsTest.test_merge_keep_test_with_all_pass_but_slow_time):
3377
3378 2012-01-11  Ryosuke Niwa  <rniwa@webkit.org>
3379
3380         last-green-revision should give us per-bot information
3381         https://bugs.webkit.org/show_bug.cgi?id=76011
3382
3383         Reviewed by Adam Barth.
3384
3385         Rewrote last-green-revision command. Instead of finding a revision for which all bots succeeded,
3386         we report the latest green run on each bot from the last 100 runs.
3387
3388         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
3389         (BuildBot._fetch_builder_page):
3390         (BuildBot):
3391         (BuildBot._green_revision_for_builder):
3392         (BuildBot.last_green_revision):
3393         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
3394         (test_green_revision_for_builder):
3395         (test_last_green_revision):
3396         * Scripts/webkitpy/tool/bot/irc_command.py:
3397         (LastGreenRevision.execute):
3398         * Scripts/webkitpy/tool/commands/queries.py:
3399         (LastGreenRevision.execute):
3400
3401 2012-01-11  Dirk Pranke  <dpranke@chromium.org>
3402
3403         webkitpy: clean up version detection in webkitpy.layout_tests.port
3404         https://bugs.webkit.org/show_bug.cgi?id=76016
3405
3406         Reviewed by Adam Barth.
3407
3408         This change updates the code in webkitpy.layout_tests.port to
3409         use the new os_name and os_version fields on host.platform. This
3410         gets layout_tests.port out of the version parsing business. Most
3411         of the diffs are just cleaning up test code to accomodate the
3412         changes.
3413
3414         * Scripts/webkitpy/common/system/executive_mock.py:
3415         (MockExecutive):
3416         (MockExecutive.ignore_error):
3417         (MockExecutive2):
3418         (MockExecutive2.ignore_error):
3419         * Scripts/webkitpy/common/system/platforminfo_mock.py:
3420         (MockPlatformInfo.is_mac):
3421         (MockPlatformInfo.is_linux):
3422         (MockPlatformInfo.is_win):
3423         * Scripts/webkitpy/common/system/systemhost_mock.py:
3424         (MockSystemHost.__init__):
3425         * Scripts/webkitpy/layout_tests/port/apple.py:
3426         (ApplePort.__init__):
3427         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
3428         (ChromiumMacPort.__init__):
3429         (ChromiumMacPort.check_wdiff):
3430         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
3431         (ChromiumMacPortTest):
3432         (ChromiumMacPortTest.assert_name):
3433         (ChromiumMacPortTest.test_versions):
3434         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3435         (ChromiumPortTest.test_all_test_configurations):
3436         (ChromiumPortTest.TestMacPort.__init__):
3437         (ChromiumPortTest.TestLinuxPort.__init__):
3438         (ChromiumPortTest.TestWinPort.__init__):
3439         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
3440         (ChromiumWinPort.__init__):
3441         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
3442         (ChromiumWinTest):
3443         (ChromiumWinTest.assert_name):
3444         (ChromiumWinTest.test_versions):
3445         (ChromiumWinTest.test_baseline_path):
3446         (ChromiumWinTest.test_build_path):
3447         * Scripts/webkitpy/layout_tests/port/factory.py:
3448         (PortFactory._port_name_from_arguments_and_options):
3449         (PortFactory.get):
3450         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
3451         (FactoryTest):
3452         (FactoryTest.setUp):
3453         (FactoryTest.assert_port):
3454         (FactoryTest.test_mac):
3455         (FactoryTest.test_win):
3456         (FactoryTest.test_google_chrome):
3457         (FactoryTest.test_gtk):
3458         (FactoryTest.test_qt):
3459         (FactoryTest.test_chromium_gpu):
3460         (FactoryTest.test_chromium_gpu_linux):
3461         (FactoryTest.test_chromium_gpu_mac):
3462         (FactoryTest.test_chromium_gpu_win):
3463         (FactoryTest.test_chromium_mac):
3464         (FactoryTest.test_chromium_linux):
3465         (FactoryTest.test_chromium_win):
3466         (FactoryTest.test_unknown_specified):
3467         (FactoryTest.test_unknown_default):
3468         * Scripts/webkitpy/layout_tests/port/google_chrome_unittest.py:
3469         (TestGoogleChromePort):
3470         (TestGoogleChromePort._verify_baseline_path):
3471         (TestGoogleChromePort._verify_expectations_overrides):
3472         (TestGoogleChromePort.test_get_google_chrome_port):
3473         * Scripts/webkitpy/layout_tests/port/mac.py:
3474         (MacPort):
3475         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3476         (MacTest):
3477         (assert_name):
3478         (test_tests_for_other_platforms):
3479         (test_version):
3480         (test_versions):
3481         (test_is_version_methods):
3482         (test_setup_environ_for_server):
3483         (_assert_search_path):
3484         (test_show_results_html_file):
3485         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
3486         (PortTestCase):
3487         (PortTestCase.make_port):
3488         * Scripts/webkitpy/layout_tests/port/win.py:
3489         (WinPort):
3490         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
3491         (WinPortTest):
3492         (WinPortTest.test_show_results_html_file):
3493         (WinPortTest._assert_search_path):
3494         (WinPortTest._assert_version):
3495
3496 2012-01-11  Hugo Parente Lima  <hugo.lima@openbossa.org>
3497
3498         [Qt][WK2] Remove workaround introduced into bug 75470
3499         https://bugs.webkit.org/show_bug.cgi?id=75889
3500
3501         Reviewed by Tor Arne Vestbø.
3502
3503         * MiniBrowser/qt/qml/BrowserWindow.qml:
3504
3505 2012-01-10  Dmitry Lomov  <dslomov@google.com>
3506
3507         [Chromium] Remove obsolete references to WebWorker class.
3508         https://bugs.webkit.org/show_bug.cgi?id=76020
3509
3510         Reviewed by David Levin.
3511
3512         * DumpRenderTree/chromium/TestWebWorker.h: Removed.
3513         * DumpRenderTree/chromium/WebViewHost.cpp:
3514         * DumpRenderTree/chromium/WebViewHost.h:
3515
3516 2012-01-11  Csaba Osztrogonác  <ossy@webkit.org>
3517
3518         [Qt] Some css3 filter tests are failing after r104698
3519         https://bugs.webkit.org/show_bug.cgi?id=76055
3520
3521         Reviewed by Simon Hausmann.
3522
3523         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
3524         (LayoutTestController::overridePreference): Add WebKitAcceleratedCompositingEnabled preference.
3525
3526 2012-01-11  Tony Chang  <tony@chromium.org>
3527
3528         Don't remove 0 byte files when running build-webkit --chromium
3529         https://bugs.webkit.org/show_bug.cgi?id=75997
3530
3531         Reviewed by Csaba Osztrogonác.
3532
3533         If QTDIR is set, isQt() is true even when running build-webkit --chromium.
3534
3535         * Scripts/webkitdirs.pm:
3536         (determineIsQt):
3537
3538 2012-01-10  David Kilzer  <ddkilzer@apple.com>
3539
3540         Add TestWebKitAPI tests for KURL
3541         <http://webkit.org/b/75774>
3542
3543         Reviewed by Adam Barth.
3544
3545         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Add a
3546         WebCore folder and KURL.cpp source file to the project.
3547         * TestWebKitAPI/Tests/WebCore/KURL.cpp: Added.
3548         (TestWebKitAPI::WebCore::SetUp): Initialize the main thread
3549         during setup since this test doesn't go through WebKit or
3550         WebKit2.
3551         (TestWebKitAPI::TEST_F): Add two basic tests.
3552
3553 2012-01-10  David Kilzer  <ddkilzer@apple.com>
3554
3555         Extract common code for comparing WTF::String objects into WTFStringUtilities.h
3556
3557         Reviewed by Adam Barth.
3558
3559         Needed for: <http://webkit.org/b/75774> Add TestWebKitAPI tests for KURL
3560
3561         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Add
3562         WTFStringUtilities.h to the project.
3563         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp: Update to use
3564         WTFStringUtilities.h.
3565         * TestWebKitAPI/WTFStringUtilities.h: Added.
3566         (WTF::operator<<):
3567
3568 2012-01-10  Vincent Scheib  <scheib@chromium.org>
3569
3570         Clarify run-webkit-tests options reset-results and new-baseline
3571         https://bugs.webkit.org/show_bug.cgi?id=75993
3572
3573         Reviewed by Dirk Pranke.
3574
3575         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3576         (parse_args):
3577
3578 2012-01-10  Simon Fraser  <simon.fraser@apple.com>
3579
3580         Disabled mock scrollbars should draw differently
3581         https://bugs.webkit.org/show_bug.cgi?id=75995
3582
3583         Reviewed by James Robinson.
3584
3585         When the scrollbar is disabled, paint the entire track of
3586         mock scrollbars with a lighter gray, and hide the thumb.
3587
3588         * DumpRenderTree/mac/DumpRenderTree.mm:
3589         (-[DRTMockScroller drawKnob]):
3590         (-[DRTMockScroller drawRect:]):
3591
3592 2012-01-10  James Robinson  <jamesr@chromium.org>
3593
3594         [chromium] Run webkit_unit_tests as part of webkit-patch build-and-test
3595         https://bugs.webkit.org/show_bug.cgi?id=75988
3596
3597         Reviewed by Adam Barth.
3598
3599         * Scripts/webkitpy/common/config/ports.py:
3600         (WebKitPort.run_webkit_unit_tests_command):
3601         (WebKitPort):
3602         (ChromiumPort.run_webkit_unit_tests_command):
3603         (ChromiumPort):
3604         * Scripts/webkitpy/common/config/ports_mock.py:
3605         (MockPort.run_webkit_unit_tests_command):
3606         * Scripts/webkitpy/tool/commands/download_unittest.py:
3607         (DownloadCommandsTest.test_build_and_test):
3608         (test_land):
3609         (test_land_red_builders):
3610         * Scripts/webkitpy/tool/steps/runtests.py:
3611         (RunTests.run):
3612         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
3613
3614 2012-01-09  Dirk Pranke  <dpranke@chromium.org>
3615
3616         webkitpy: add os_name, os_version to platforminfo
3617         https://bugs.webkit.org/show_bug.cgi?id=75931
3618
3619         Reviewed by Eric Siedel.
3620
3621         As a first step in cleaning up the version parsing logic in
3622         webkitpy.layout_tests.port, this adds common routines for
3623         getting the os_name and os_version fields to the PlatformInfo
3624         class. Nothing uses them yet but I've added FIXME's to some of
3625         the code that needs to be deleted in the port files.
3626
3627         * Scripts/webkitpy/common/system/platforminfo.py:
3628         (PlatformInfo.__init__):
3629         (PlatformInfo):
3630         (PlatformInfo._determine_os_name):
3631         (PlatformInfo._determine_mac_version):
3632         (PlatformInfo._determine_linux_version):
3633         (PlatformInfo._determine_win_version):
3634         (PlatformInfo.display_name):
3635         (PlatformInfo.total_bytes_memory):
3636         (PlatformInfo._compute_bytes_from_vm_stat_output):
3637         (PlatformInfo.free_bytes_memory):
3638         * Scripts/webkitpy/common/system/platforminfo_unittest.py: Added.
3639         (TestPlatformInfo):
3640         (TestPlatformInfo.make_info):
3641         (TestPlatformInfo.make_info.mock_run_command):
3642         (TestPlatformInfo.setUp):
3643         (TestPlatformInfo.tearDown):
3644         (TestPlatformInfo.test_basic):
3645         (TestPlatformInfo.integration_test_basic):
3646         (TestPlatformInfo.test_display_name_mac):
3647         (TestPlatformInfo.test_display_name_win32):
3648         (TestPlatformInfo.test_memory_mac):
3649         (TestPlatformInfo.test_memory_win32):
3650         (TestPlatformInfo.test_determine_os_name):
3651         (TestPlatformInfo.test_determine_mac_version):
3652         (TestPlatformInfo.test_determine_linux_version):
3653         (TestPlatformInfo.test_determine_win_version_from_tuple):
3654         (TestPlatformInfo.test_determine_win_version_from_cmd):
3655         * Scripts/webkitpy/layout_tests/port/apple.py:
3656         (ApplePort.__init__):
3657         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
3658         (ChromiumMacPort.__init__):
3659         * Scripts/webkitpy/layout_tests/port/mac.py:
3660         (MacPort):
3661         * Scripts/webkitpy/layout_tests/port/win.py:
3662         (WinPort):
3663
3664 2012-01-10  Adam Roben  <aroben@apple.com>
3665
3666         Make it possible to type data: URLs into MiniBrowser on Windows
3667
3668         Fixes <http://webkit.org/b/75084> Crash when trying to navigate to a data: URL in
3669         MiniBrowser on Windows
3670
3671         Reviewed by Darin Adler.
3672
3673         * MiniBrowser/win/BrowserWindow.cpp:
3674         (BrowserWindow::handleMessage): Instead of checking for an "http://" prefix when deciding
3675         whether to prepend "http://", just look for whether the typed URL contains a colon. If it
3676         does, we assume it already has a scheme and don't modify it.
3677
3678 2012-01-10  Adam Roben  <aroben@apple.com>
3679
3680         Make MiniBrowser automatically escape invalid URL characters typed in its URL field
3681
3682         Fixes <http://webkit.org/b/75086> Crash when typing a data: URL containing double-quotes in
3683         MiniBrowser on Windows
3684
3685         Reviewed by Darin Adler.
3686
3687         * MiniBrowser/win/BrowserView.cpp:
3688   &