https://bugs.webkit.org/show_bug.cgi?id=77383
[WebKit-https.git] / Tools / ChangeLog
1 2012-02-01  Beth Dakin  <bdakin@apple.com>
2
3         https://bugs.webkit.org/show_bug.cgi?id=77383
4         Add a different didFirstVisuallNonEmptyLayout heuristic to experiment with
5         -and corresponding-
6         <rdar://problem/10709560>
7
8         Reviewed by Sam Weinig.
9
10         WebKit2's WebLoaderClient has a temporary new function that must be accounted 
11         for.
12         * MiniBrowser/mac/BrowserWindowController.m:
13         (-[BrowserWindowController awakeFromNib]):
14         * WebKitTestRunner/TestController.cpp:
15         (WTR::TestController::initialize):
16
17 2012-02-01  Dirk Pranke  <dpranke@chromium.org>
18
19         TestWebKitAPI isn't being built on chromium bots any more
20         https://bugs.webkit.org/show_bug.cgi?id=77563
21
22         Reviewed by Dimitri Glazkov.
23
24         I accidentally dropped it in the refactoring in r105449.
25
26         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp: Added.
27
28 2012-02-01  Hans Wennborg  <hans@chromium.org>
29
30         Rename WebSpeechInputResult::set() to assign()
31         https://bugs.webkit.org/show_bug.cgi?id=77540
32
33         Reviewed by Darin Fisher.
34
35         * DumpRenderTree/chromium/MockWebSpeechInputController.cpp:
36         (MockWebSpeechInputController::addMockRecognitionResult):
37         (MockWebSpeechInputController::speechTaskFired):
38
39 2012-02-01  Carlos Garcia Campos  <cgarcia@igalia.com>
40
41         [GTK] WebKit1 API documentation is not generated when building with gtk-2.0
42         https://bugs.webkit.org/show_bug.cgi?id=77542
43
44         Reviewed by Martin Robinson.
45
46         * gtk/generate-gtkdoc: Check first whether there's
47         webkitgtk-3.0.pc and if it doesn't exist use webkitgtk-1.0.pc
48         instead.
49         (get_webkit2_options): Return just the options since the
50         pkg_config_path doesn't depend on options
51         (get_webkit1_options): Ditto.
52         (generate_doc): Helper function to create a generator and generate
53         documentation for the given pkg-config file with the given options.
54         * gtk/gtkdoc.py:
55         (GTKDoc.__init__): Don't use ** for args parameter, since it's
56         used as a dict and never expanded.
57         (PkgConfigGTKDoc.__init__): Ditto.
58
59 2012-02-01  Carlos Garcia Campos  <cgarcia@igalia.com>
60
61         [GTK] API documentation is not installed even when building with --enable-gtk-doc
62         https://bugs.webkit.org/show_bug.cgi?id=77094
63
64         Reviewed by Martin Robinson.
65
66         * GNUmakefile.am: Add install-data-local and uninstall-local rules
67         to install/uninstall WebKit1 and WebKit2 API documentation.
68
69 2012-02-01  Sheriff Bot  <webkit.review.bot@gmail.com>
70
71         Unreviewed, rolling out r106460.
72         http://trac.webkit.org/changeset/106460
73         https://bugs.webkit.org/show_bug.cgi?id=77552
74
75         It break unit tests if WEBKIT_TESTFONTS isn't defined.
76         (Requested by Ossy on #webkit).
77
78         * Scripts/webkitpy/layout_tests/port/qt.py:
79
80 2012-02-01  Nándor Huszka  <huszka.nandor@stud.u-szeged.hu>
81
82         [Qt] Automatic clean build feature always do clean build with --no-webkit2
83         https://bugs.webkit.org/show_bug.cgi?id=74519
84
85         Reviewed by Tor Arne Vestbø.
86
87         Set the PLUGIN_ARCHITECTURE_UNSUPPORTED define always.
88
89         * qmake/mkspecs/features/features.prf:
90
91 2012-02-01  Jesus Sanchez-Palencia  <jesus.palencia@openbossa.org>
92
93         [Qt][WK2] run-webkit-tests --qt  crashes if WEBKIT_TESTFONTS is not set
94         https://bugs.webkit.org/show_bug.cgi?id=77466
95
96         Reviewed by Kenneth Rohde Christiansen.
97
98         Replicate the behavior of old-run-webkit-tests and check if WEBKIT_TESTFONTS
99         is set or if we should raise an error.
100
101         * Scripts/webkitpy/layout_tests/port/qt.py:
102         (QtPort.setup_environ_for_server):
103
104 2012-02-01  Philippe Normand  <pnormand@igalia.com> and Sergio Villar Senin  <svillar@igalia.com>
105
106         Reviewed by Martin Robinson.
107
108         [GTK] Improve FrameLoader signals. Resource loading
109         https://bugs.webkit.org/show_bug.cgi?id=49543
110
111         Support for the new loader signals in DRT.
112
113         * DumpRenderTree/gtk/DumpRenderTree.cpp:
114         (urlPath):
115         (willSendRequestCallback):
116         (urlSuitableForTestResult):
117         (descriptionSuitableForTestResult):
118         (didReceiveResponse):
119         (didFinishLoading):
120         (didFailLoadingWithError):
121         (createWebView):
122
123 2012-02-01  Ryosuke Niwa  <rniwa@webkit.org>
124
125         run-perf-tests should restart DRT for each test
126         https://bugs.webkit.org/show_bug.cgi?id=77506
127
128         Reviewed by Adam Barth.
129
130         Always restart DRT when running performance tests in order to minimize the dependency between tests.
131
132         * Scripts/webkitpy/layout_tests/port/chromium.py:
133         (ChromiumDriver.stop):
134         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
135         (PerfTestsRunner._run_tests_set):
136         (PerfTestsRunner._run_single_test):
137         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
138         (create_runner):
139         (test_run_passing_test):
140         (test_run_silent_test):
141         (test_run_failed_test):
142         (test_run_tonguey_test):
143         (test_run_timeout_test):
144         (test_run_crash_test):
145         (test_run_test_set_kills_drt_per_run):
146         (test_run_test_set_kills_drt_per_run.TestDriverWithStopCount):
147         (test_run_test_set_kills_drt_per_run.TestDriverWithStopCount.__init__):
148         (test_run_test_set_kills_drt_per_run.TestDriverWithStopCount.stop):
149
150 2012-01-31  Hans Wennborg  <hans@chromium.org>
151
152         Speech Input: Report speech element rect relative to window rather than frame
153         https://bugs.webkit.org/show_bug.cgi?id=76443
154
155         Reviewed by Darin Fisher.
156
157         Update LayoutTestController to allow for retrieving the speech
158         element's position used in a request.
159
160         In Chromium this is handled by the MockWebSpeechInputController;
161         adding stubs for the other implementations.
162
163         * DumpRenderTree/LayoutTestController.cpp:
164         (setMockSpeechInputDumpRectCallback):
165         (LayoutTestController::staticFunctions):
166         * DumpRenderTree/LayoutTestController.h:
167         (LayoutTestController):
168         * DumpRenderTree/chromium/LayoutTestController.cpp:
169         (LayoutTestController::LayoutTestController):
170         (LayoutTestController::setMockSpeechInputDumpRect):
171         * DumpRenderTree/chromium/LayoutTestController.h:
172         (LayoutTestController):
173         * DumpRenderTree/chromium/MockWebSpeechInputController.cpp:
174         (MockWebSpeechInputController::setDumpRect):
175         (MockWebSpeechInputController::clearResults):
176         (MockWebSpeechInputController::startRecognition):
177         (MockWebSpeechInputController::MockWebSpeechInputController):
178         (makeRectResult):
179         (MockWebSpeechInputController::speechTaskFired):
180         * DumpRenderTree/chromium/MockWebSpeechInputController.h:
181         (MockWebSpeechInputController):
182         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
183         (LayoutTestController::setMockSpeechInputDumpRect):
184         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
185         (LayoutTestController::setMockSpeechInputDumpRect):
186         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
187         (LayoutTestController::setMockSpeechInputDumpRect):
188         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
189         (LayoutTestController::setMockSpeechInputDumpRect):
190         * DumpRenderTree/qt/LayoutTestControllerQt.h:
191         (LayoutTestController):
192         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
193         (LayoutTestController::setMockSpeechInputDumpRect):
194         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
195         (LayoutTestController::setMockSpeechInputDumpRect):
196
197 2012-01-31  Ryuan Choi  <ryuan.choi@samsung.com>
198
199         [EFL] Add basic DRT/Efl implementation to support viewport test.
200         https://bugs.webkit.org/show_bug.cgi?id=77320
201
202         Reviewed by Andreas Kling.
203
204         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
205         (LayoutTestController::dumpConfigurationForViewport): call dumpConfigurationForViewport().
206
207 2012-01-31  Alexey Proskuryakov  <ap@apple.com>
208
209         REGRESSION (WebKit2): event.keyCode is always zero when typing in Russian
210         https://bugs.webkit.org/show_bug.cgi?id=77473
211         <rdar://problem/10751357>
212
213         Reviewed by Darin Adler.
214
215         * DumpRenderTree/mac/EventSendingController.mm:
216         (-[EventSendingController keyDown:withModifiers:withLocation:]):
217         * WebKitTestRunner/mac/EventSenderProxy.mm:
218         (WTR::EventSenderProxy::keyDown):
219         Added a new special keyDown value to test this.
220
221 2012-01-31  Dirk Pranke  <dpranke@chromium.org>
222
223         nrwt: fix the mock port to work with DriverProxy, reftests
224         https://bugs.webkit.org/show_bug.cgi?id=77170
225
226         Reviewed by Eric Seidel.
227
228         The DryrunPort used for benchmarking and testing NRWT has been
229         broken since we added the DriverProxy code. It isn't really
230         needed any more, so I've removed it.
231
232         The MockDRT port was also broken by DriverProxy, and moreover,
233         never worked w/ reftests. Since we don't exercise this module in
234         the unit tests, the code had also bitrotted some. I have cleaned
235         it up and made it use DriverInput and DriverOutput for
236         consistency.
237
238         * Scripts/webkitpy/layout_tests/port/dryrun.py:
239         (DryRunPort.__init__):
240         (DryRunPort):
241         (DryRunPort.real_name):
242         (DryRunPort.create_driver):
243         (DryRunPort.driver_cmd_line):
244         (DryRunPort._driver_class):
245         (DryrunDriver):
246         (DryrunDriver.run_test):
247         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
248         (MockDRTPort.create_driver):
249         (MockChromiumDRT.run_one_test):
250
251 2012-01-31  Ojan Vafai  <ojan@chromium.org>
252
253         run-webkit-tests calls out to webkit-build-directory twice
254         https://bugs.webkit.org/show_bug.cgi?id=77248
255
256         Reviewed by Dirk Pranke.
257
258         * Scripts/webkit-build-directory:
259         If called with no arguments, print out both the top-level directory and the 
260         configuration directory
261
262         * Scripts/webkitpy/layout_tests/port/config.py:
263         (Config.build_directory):
264         When called with no arguments, get and cache both the top-level and configuration directories.
265
266         * Scripts/webkitpy/layout_tests/port/config_standalone.py:
267         (main):
268         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
269         (ConfigTest.assert_configuration):
270         (ConfigTest.test_build_directory.mock_webkit_build_directory):
271         (ConfigTest.test_build_directory):
272         (ConfigTest.test_default_configuration__notfound):
273         * Scripts/webkitpy/layout_tests/port/webkit.py:
274         (WebKitPort._build_path):
275         Cache the build directory in the options object so that other users
276         of the options object don't need to call out to webkit-build-directory.
277         Also, properly support the existing --build-directory argument.
278
279 2012-01-31  Ryosuke Niwa  <rniwa@webkit.org>
280
281         Bump up the timeout for performance tests since DOM/DOMTable.html
282         and Parser/html5-full-render.html are timing out on the bot.
283
284         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
285         (PerfTestsRunner._parse_args):
286
287 2012-01-31  Ehsan Akhgari  <ehsan.akhgari@gmail.com>
288
289         Fix the update_webgl_conformance_tests.py script in order to enable it
290         to detect the SCM being used correctly.
291         https://bugs.webkit.org/show_bug.cgi?id=77462
292
293         Reviewed by Kenneth Russell.
294
295         * Scripts/webkitpy/to_be_moved/update_webgl_conformance_tests.py:
296         (default_out_dir):
297
298 2012-01-31  Eric Seidel  <eric@webkit.org>
299
300         webkit-patch apply-from-bug should work regardless of your CWD
301         https://bugs.webkit.org/show_bug.cgi?id=77482
302
303         Reviewed by Adam Barth.
304
305         webkit-patch only ever works with patches made from the root
306         directory.  We should pass the CWD to svn-apply in the case
307         where we're running not from the webkit root.  I believe
308         this used to work in the past and regressed.
309
310         * Scripts/webkitpy/common/checkout/checkout.py:
311         (Checkout.apply_patch):
312         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
313         (CheckoutTest.test_apply_patch):
314
315 2012-01-31  Sheriff Bot  <webkit.review.bot@gmail.com>
316
317         Unreviewed, rolling out r106302.
318         http://trac.webkit.org/changeset/106302
319         https://bugs.webkit.org/show_bug.cgi?id=77472
320
321         Broke building DRT before running the tests (Requested by ojan
322         on #webkit).
323
324         * Scripts/webkit-build-directory:
325         * Scripts/webkitpy/layout_tests/port/config.py:
326         (Config.build_directory):
327         * Scripts/webkitpy/layout_tests/port/config_standalone.py:
328         (main):
329         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
330         (ConfigTest.assert_configuration):
331         (ConfigTest.test_build_directory.mock_webkit_build_directory):
332         (ConfigTest.test_build_directory):
333         (ConfigTest.test_default_configuration__notfound):
334         * Scripts/webkitpy/layout_tests/port/webkit.py:
335         (WebKitPort._build_path):
336
337 2012-01-31  Antoine Labour  <piman@chromium.org>
338
339         Merge WebGraphicsContext3D creation and initialization, and move it to
340         WebViewClient.
341         https://bugs.webkit.org/show_bug.cgi?id=76593
342
343         Reviewed by Darin Fisher.
344
345         * DumpRenderTree/chromium/TestWebPlugin.cpp:
346         (TestWebPlugin::TestWebPlugin):
347         (TestWebPlugin::initialize):
348         * DumpRenderTree/chromium/TestWebPlugin.h:
349         * DumpRenderTree/chromium/WebViewHost.cpp:
350         (WebViewHost::createGraphicsContext3D):
351         (WebViewHost::createPlugin):
352         * DumpRenderTree/chromium/WebViewHost.h:
353
354 2012-01-31  Gabor Rapcsanyi  <rgabor@webkit.org>
355
356         [GTK] build-webkit warning.
357         https://bugs.webkit.org/show_bug.cgi?id=77411
358
359         Reviewed by Philippe Normand.
360
361         * Scripts/webkitdirs.pm:
362         (isCrossCompilation):
363
364 2012-01-31  Dan Bernstein  <mitz@apple.com>
365
366         When generating derived sources, use the same compiler that is used to compile WebKitTestRunner.
367
368         Reviewed by Anders Carlsson.
369
370         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
371
372 2012-01-31  Csaba Osztrogonác  <ossy@webkit.org>
373
374         If the QTDIR environment variable is set, the build system fails to compile WinCairo
375         https://bugs.webkit.org/show_bug.cgi?id=77112
376
377         Reviewed by Daniel Bates.
378
379         * Scripts/webkitdirs.pm:
380         (determineIsQt):
381
382 2012-01-31  Adam Roben  <aroben@apple.com>
383
384         Only call -typingAttributes on WebViews, not WebHTMLViews
385
386         Looks like this case was just missed in r105908.
387
388         Fixes <http://webkit.org/b/77432> REGRESSION (r105908): WebKit1.InspectorBarTest is crashing
389
390         Reviewed by Antti Koivisto.
391
392         * TestWebKitAPI/Tests/mac/InspectorBar.mm:
393         (TestWebKitAPI::TEST): Call -typingAttributes on the WebView, just like we do 2 lines
394         earlier.
395
396 2012-01-31  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
397
398         [Qt] MiniBrowser: Show touch mocking indicators only when Ctrl is held down.
399         https://bugs.webkit.org/show_bug.cgi?id=77221
400
401         Reviewed by Kenneth Rohde Christiansen.
402
403         This makes sure that the indicator isn't in the way when doing clicks or dealing with the rest of the UI.
404
405         * MiniBrowser/qt/MiniBrowserApplication.cpp:
406         (MiniBrowserApplication::sendTouchEvent):
407
408 2012-01-26  Hans Wennborg  <hans@chromium.org>
409
410         Speech Input: move MockSpeechInputClient into Chromium DumpRenderTree implementation
411         https://bugs.webkit.org/show_bug.cgi?id=77083
412
413         Reviewed by Darin Fisher.
414
415         Add MockWebSpeechInputController which provides a mock implementation
416         of the WebSpeechInputController interface, and use that in
417         DumpRenderTree.
418
419         * DumpRenderTree/DumpRenderTree.gypi:
420         * DumpRenderTree/chromium/LayoutTestController.cpp:
421         (LayoutTestController::addMockSpeechInputResult):
422         * DumpRenderTree/chromium/LayoutTestController.h:
423         ():
424         * DumpRenderTree/chromium/MockWebSpeechInputController.cpp: Added.
425         (MockWebSpeechInputController::create):
426         (MockWebSpeechInputController::addMockRecognitionResult):
427         (MockWebSpeechInputController::clearResults):
428         (MockWebSpeechInputController::startRecognition):
429         (MockWebSpeechInputController::cancelRecognition):
430         (MockWebSpeechInputController::stopRecording):
431         (MockWebSpeechInputController::MockWebSpeechInputController):
432         (MockWebSpeechInputController::speechTaskFired):
433         (MockWebSpeechInputController::SpeechTask::SpeechTask):
434         (MockWebSpeechInputController::SpeechTask::stop):
435         (MockWebSpeechInputController::SpeechTask::runIfValid):
436         * DumpRenderTree/chromium/MockWebSpeechInputController.h: Added.
437         (WebKit):
438         (MockWebSpeechInputController):
439         (MockWebSpeechInputController::taskList):
440         (SpeechTask):
441         * DumpRenderTree/chromium/TestShell.cpp:
442         * DumpRenderTree/chromium/WebViewHost.cpp:
443         (WebViewHost::speechInputController):
444         * DumpRenderTree/chromium/WebViewHost.h:
445         (WebKit):
446         (WebViewHost::speechInputControllerMock):
447         (WebViewHost):
448         ():
449
450 2012-01-31  Hajime Morrita  <morrita@chromium.org>
451
452         [PerformanceTests] Add landing html for Dromaeo dom-query test
453         https://bugs.webkit.org/show_bug.cgi?id=77329
454
455         Reviewed by Ryosuke Niwa.
456
457         Added some more ignorable output patterns to allow warnings
458         from a third party test suite.
459
460         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
461         (PerfTestsRunner):
462         (PerfTestsRunner._should_ignore_line_in_parser_test_result):
463
464 2012-01-30  Kevin Ollivier  <kevino@theolliviers.com>
465
466         [wx] Unreviewed. Update download URLs to new domain.
467         
468         * waf/build/build_utils.py:
469         (update_wx_deps):
470
471 2012-01-30  Kevin Ollivier  <kevino@theolliviers.com>
472
473         [wx] Unreviewed. Build fix, add JavaScriptCore/tools
474         to the list of build directories.
475         
476         * waf/build/settings.py:
477
478 2012-01-30  Kentaro Hara  <haraken@chromium.org>
479
480         REGRESSION(r105797): prepare-ChangeLog for a .cpp file can
481         output an empty method name (i.e. "()")
482         https://bugs.webkit.org/show_bug.cgi?id=77336
483
484         Reviewed by Darin Adler.
485
486         r105797 tried to detect a change outside methods, but it causes a bug that
487         prepare-ChangeLog can output an empty method name, like this:
488
489             * foo/bar/baz.cpp:
490             (method1):
491             ():
492             (method2):
493
494         This is because the cpp parser in prepare-ChangeLog cannot distinguish
495         '{' as the beginning of a method with '{' as the beginning of an array definition
496         at the top level.
497
498             int a[] = { 1, 2, 3 };  // This '{' is the beginning of an array definition.
499
500             void func() { // This '{' is the beginning of a method.
501                 ...;
502             }
503
504         This patch fixes prepare-ChangeLog so that it skips an array definition at the top level.
505
506         * Scripts/prepare-ChangeLog:
507         (get_function_line_ranges_for_cpp): Modified as described above.
508         (generateFunctionLists): As a hack, modified so that prepare-ChangeLog does not output
509         an empty method name. Ideally this should not happen but may happen, since the
510         parsers are not perfect.
511         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests.cpp: Added test cases.
512         (NameSpace7):
513         (NameSpace8):
514         (Class109):
515         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests-expected.txt:
516
517 2012-01-30  Ojan Vafai  <ojan@chromium.org>
518
519         run-webkit-tests calls out to webkit-build-directory twice
520         https://bugs.webkit.org/show_bug.cgi?id=77248
521
522         Reviewed by Dirk Pranke.
523
524         This reduces to calls to webkit-build-directory to one call and
525         saves 500ms for run-webkit-tests over a single file.
526
527         * Scripts/webkit-build-directory:
528         If called with no arguments, print out both the top-level directory and the 
529         configuration directory.
530
531         * Scripts/webkit-build-directory:
532         * Scripts/webkitpy/layout_tests/port/config.py:
533         (Config.build_directory):
534         When called with no arguments, get and cache both the top-level and configuration directories.
535
536         * Scripts/webkitpy/layout_tests/port/config_standalone.py:
537         (main):
538         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
539         (ConfigTest.assert_configuration):
540         (ConfigTest.test_build_directory.mock_webkit_build_directory):
541         (ConfigTest.test_build_directory):
542         (ConfigTest.test_default_configuration__notfound):
543         * Scripts/webkitpy/layout_tests/port/webkit.py:
544         (WebKitPort._build_path):
545         Cache the root directory in the options object so that other users
546         of the options object don't need to call out to webkit-build-directory.
547
548 2012-01-30  Ojan Vafai  <ojan@chromium.org>
549
550         Parsing test_expecations.txt + Skipped lists takes too long
551         https://bugs.webkit.org/show_bug.cgi?id=77059
552
553         Reviewed by Dirk Pranke.
554
555         This saves ~100ms on the Apple Mac port.
556         -memoize a bunch of path methods.
557         -Avoid doing multiple disk accesses per line.
558         -Parse the skipped list directly instead of turning it into a test_expecations.txt
559         formatting string and parsing that.
560
561         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
562         (TestExpectationParser):
563         (TestExpectationParser.expectation_for_skipped_test):
564         (TestExpectationParser._parse_line):
565         (TestExpectationParser._collect_matching_tests):
566         (TestExpectations.__init__):
567         (TestExpectations._add_skipped_tests):
568         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
569         (test_add_skipped_tests):
570         (test_add_skipped_tests_duplicate):
571         * Scripts/webkitpy/layout_tests/port/base.py:
572         (Port):
573         (Port.test_isfile):
574         (Port.normalize_test_name):
575         (Port.layout_tests_dir):
576         (Port.abspath_for_test):
577         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
578
579 2012-01-30  Alexis Menard  <alexis.menard@openbossa.org>
580
581         Unreviewed. Add myself to CSS, GStreamer, Qt related watchlists.
582         
583         * Scripts/webkitpy/common/config/watchlist:
584
585 2012-01-30  Gavin Barraclough  <barraclough@apple.com>
586
587         Unreviewed build fix following bug#76855
588
589         * TestWebKitAPI/Tests/WTF/MetaAllocator.cpp:
590         ():
591         (TestWebKitAPI::TEST_F):
592
593 2012-01-26  No'am Rosenthal  <noam.rosenthal@nokia.com> and Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
594
595         [Qt] WKTR: Use a software rendering pipiline when running tests.
596         https://bugs.webkit.org/show_bug.cgi?id=76708
597
598         Reviewed by Kenneth Rohde Christiansen.
599
600         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
601         (WTR::WrapperWindow::WrapperWindow):
602         (WTR::PlatformWebView::PlatformWebView): Use software rendering of layers since the wrapping QQuickView isn't shown.
603
604 2012-01-30  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
605
606         [Qt] Store build config immediately instead of waiting for a successfull build
607
608         Otherwise we will not pick up changes to the config if the build is aborted,
609         for example due to build errors or if the user pressed Ctrl+Z. We now write
610         the new config before starting a build, but still defer to write the SVN
611         revision once the build completes successfully.
612
613         Reviewed by Ossy.
614
615         * Scripts/webkitdirs.pm:
616         (buildQMakeProjects):
617
618 2012-01-30  Alexander Færøy  <ahf@0x90.dk>
619
620         [Qt] Add "New Window" button to the Qt MiniBrowser
621         https://bugs.webkit.org/show_bug.cgi?id=77281
622
623         Reviewed by Tor Arne Vestbø.
624
625         * MiniBrowser/qt/BrowserWindow.cpp:
626         (BrowserWindow::BrowserWindow):
627         * MiniBrowser/qt/MiniBrowser.qrc:
628         * MiniBrowser/qt/icons/plus.png: Added.
629         * MiniBrowser/qt/qml/BrowserWindow.qml:
630
631 2012-01-30  Csaba Osztrogonác  <ossy@webkit.org>
632
633         Unreviewed rolling out r106222.
634
635         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
636         (SingleTestRunner.__init__):
637         * Scripts/webkitpy/layout_tests/controllers/worker.py:
638         (Worker.handle_test_list):
639         * Scripts/webkitpy/layout_tests/models/test_input.py:
640         (TestInput.__init__):
641         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
642         (_set_up_derived_options):
643         (parse_args):
644         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
645         (MainTest.test_repeat_each):
646         * WebKitTestRunner/TestInvocation.cpp:
647         (WTR::TestInvocation::setIsPixelTest):
648
649 2012-01-30  Hans Wennborg  <hans@chromium.org>
650
651         Unreviewed, rolling out r106219.
652         http://trac.webkit.org/changeset/106219
653         https://bugs.webkit.org/show_bug.cgi?id=77083
654
655         This broke Chromium's test_shell.
656
657         * DumpRenderTree/DumpRenderTree.gypi:
658         * DumpRenderTree/chromium/LayoutTestController.cpp:
659         (LayoutTestController::addMockSpeechInputResult):
660         * DumpRenderTree/chromium/LayoutTestController.h:
661         (WebKit):
662         (LayoutTestController):
663         ():
664         * DumpRenderTree/chromium/MockWebSpeechInputController.cpp: Removed.
665         * DumpRenderTree/chromium/MockWebSpeechInputController.h: Removed.
666         * DumpRenderTree/chromium/TestShell.cpp:
667         * DumpRenderTree/chromium/WebViewHost.cpp:
668         (WebViewHost::speechInputController):
669         * DumpRenderTree/chromium/WebViewHost.h:
670         (WebKit):
671         (WebViewHost::speechInputControllerMock):
672         (WebViewHost):
673         ():
674
675 2012-01-26  Hans Wennborg  <hans@chromium.org>
676
677         Speech Input: move MockSpeechInputClient into Chromium DumpRenderTree implementation
678         https://bugs.webkit.org/show_bug.cgi?id=77083
679
680         Reviewed by Darin Fisher.
681
682         Add MockWebSpeechInputController which provides a mock implementation
683         of the WebSpeechInputController interface, and use that in
684         DumpRenderTree.
685
686         * DumpRenderTree/DumpRenderTree.gypi:
687         * DumpRenderTree/chromium/LayoutTestController.cpp:
688         (LayoutTestController::addMockSpeechInputResult):
689         * DumpRenderTree/chromium/LayoutTestController.h:
690         ():
691         * DumpRenderTree/chromium/MockWebSpeechInputController.cpp: Added.
692         (MockWebSpeechInputController::create):
693         (MockWebSpeechInputController::addMockRecognitionResult):
694         (MockWebSpeechInputController::clearResults):
695         (MockWebSpeechInputController::startRecognition):
696         (MockWebSpeechInputController::cancelRecognition):
697         (MockWebSpeechInputController::stopRecording):
698         (MockWebSpeechInputController::MockWebSpeechInputController):
699         (MockWebSpeechInputController::speechTaskFired):
700         (MockWebSpeechInputController::SpeechTask::SpeechTask):
701         (MockWebSpeechInputController::SpeechTask::stop):
702         (MockWebSpeechInputController::SpeechTask::runIfValid):
703         * DumpRenderTree/chromium/MockWebSpeechInputController.h: Added.
704         (WebKit):
705         (MockWebSpeechInputController):
706         (MockWebSpeechInputController::taskList):
707         (SpeechTask):
708         * DumpRenderTree/chromium/TestShell.cpp:
709         * DumpRenderTree/chromium/WebViewHost.cpp:
710         (WebViewHost::speechInputController):
711         * DumpRenderTree/chromium/WebViewHost.h:
712         (WebKit):
713         (WebViewHost::speechInputControllerMock):
714         (WebViewHost):
715         ():
716
717 2012-01-29  Sheriff Bot  <webkit.review.bot@gmail.com>
718
719         Unreviewed, rolling out r106109.
720         http://trac.webkit.org/changeset/106109
721         https://bugs.webkit.org/show_bug.cgi?id=77302
722
723         It made tests crash (Requested by Ossy_weekend on #webkit).
724
725         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
726         (WTR::WrapperWindow::WrapperWindow):
727         (WTR::PlatformWebView::PlatformWebView):
728
729 2012-01-28  Nico Weber  <thakis@chromium.org>
730
731         [chromium] Add support for building standalone webkit/chromium checkouts with ninja.
732         https://bugs.webkit.org/show_bug.cgi?id=77243
733
734         Reviewed by Adam Barth.
735
736         * Scripts/update-webkit:
737         * Scripts/webkitdirs.pm:
738         (isChromiumNinja):
739         (determineIsChromiumNinja):
740         (buildChromiumNinja):
741         (buildChromium):
742
743 2012-01-27  Kentaro Hara  <haraken@chromium.org>
744
745         The cpp parser of prepare-ChangeLog treats if(...) {} as a method
746         https://bugs.webkit.org/show_bug.cgi?id=77241
747
748         Reviewed by Ryosuke Niwa.
749
750         The cpp parser of prepare-ChangeLog is wrong, and an inner {} block
751         in a method in some namespace or class is treated as a method.
752         This patch fixes the bug.
753
754             class C {
755                 void func()
756                 {
757                     if (1) {    // This should not be treated as a method.
758                     }
759                 }
760             };
761
762         * Scripts/prepare-ChangeLog:
763         (get_function_line_ranges_for_cpp):
764         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests-expected.txt:
765         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests.cpp:
766         (Class108):
767         (Class108::func35):
768
769 2012-01-27  Enrica Casucci  <enrica@apple.com>
770
771         Followup to http://trac.webkit.org/changeset/106171.
772         Dan Bernstein pointed out correctly that I should
773         have used copy instead of initWithString.
774
775         Reviewed by Dan Bernstein.
776
777         * DumpRenderTree/mac/DumpRenderTreePasteboard.m:
778         (-[LocalPasteboard initWithName:]):
779
780 2012-01-27  Enrica Casucci  <enrica@apple.com>
781
782         DRT pasteboard object should return the correct pasteboard name.
783         https://bugs.webkit.org/show_bug.cgi?id=77260
784         
785         Currently the name property always returns nil.
786         Adding a memeber variable to store the pasteboard name and return its
787         value from the name method instead of returning nil always.
788
789         Reviewed by Dan Bernstein.
790
791         * DumpRenderTree/mac/DumpRenderTreePasteboard.m:
792         (+[DumpRenderTreePasteboard _pasteboardWithName:]):
793         (-[LocalPasteboard initWithName:]):
794         (-[LocalPasteboard name]):
795
796 2012-01-27  Martin Robinson  <mrobinson@igalia.com>
797
798         [GTK] [WK2] Implement the policy client
799         https://bugs.webkit.org/show_bug.cgi?id=76343
800
801         Reviewed by Gustavo Noronha Silva.
802
803         * gtk/generate-gtkdoc:
804         (get_webkit2_options): Update the list of skipped source files to include all clients.
805
806 2012-01-27  Ada Chan  <adachan@apple.com>
807
808         Add test for WKBundleFrameGetParentFrame().
809         https://bugs.webkit.org/show_bug.cgi?id=77161
810
811         Reviewed by Anders Carlsson.
812
813         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Add ParentFrame.cpp and ParentFrame_Bundle.cpp.
814         * TestWebKitAPI/Tests/WebKit2/ParentFrame.cpp: Added.
815         (TestWebKitAPI):
816         (TestWebKitAPI::didReceiveMessageFromInjectedBundle): Get the check result from the injected bundle.
817         (TestWebKitAPI::setInjectedBundleClient):
818         (TestWebKitAPI::TEST): Load simple-iframe.html and then wait for the injected bundle to post result.
819         Make sure the check is successful.
820         * TestWebKitAPI/Tests/WebKit2/ParentFrame_Bundle.cpp: Added.
821         (TestWebKitAPI):
822         (ParentFrameTest):
823         (TestWebKitAPI::ParentFrameTest::ParentFrameTest):
824         (TestWebKitAPI::didFinishLoadForFrame): If the frame is a subframe, store it off for checking later.
825         If the frame is the main frame, check whether it's indeed the subframe's parent frame and post the result
826         to the test controller.
827         (TestWebKitAPI::ParentFrameTest::didCreatePage): Set the page loader client on this page.
828
829 2012-01-27  Zan Dobersek  <zandobersek@gmail.com>
830
831         [Gtk] DumpRenderTree lacks --no-timeout command line option
832         https://bugs.webkit.org/show_bug.cgi?id=77196
833
834         Reviewed by Martin Robinson.
835
836         Add the --no-timeout command line option and reorganize
837         watchdog code to make running DumpRenderTree without
838         timeout possible.
839
840         * DumpRenderTree/gtk/DumpRenderTree.cpp:
841         (setWaitToDumpWatchdog):
842         (shouldSetWaitToDumpWatchdog):
843         (initializeGlobalsFromCommandLineOptions):
844         * DumpRenderTree/gtk/DumpRenderTreeGtk.h:
845         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
846         (waitToDumpWatchdogFired):
847         (LayoutTestController::setWaitToDump):
848
849 2012-01-27  Andras Becsi  <andras.becsi@nokia.com>
850
851         [Qt][WK2] Enable the automatic touch->mouse mocking on the new Qt5 hashes
852         https://bugs.webkit.org/show_bug.cgi?id=77122
853
854         Reviewed by Kenneth Rohde Christiansen.
855
856         * MiniBrowser/qt/MiniBrowserApplication.cpp:
857         (MiniBrowserApplication::MiniBrowserApplication):
858
859 2012-01-27  Zeno Albisser  <zeno@webkit.org>
860
861         [Qt][Mac] Build fails after adding ICU support (r105997).
862         https://bugs.webkit.org/show_bug.cgi?id=77118
863
864         Define WTF_USE_ICU_UNICODE if we are on mac.
865
866         Reviewed by Tor Arne Vestbø.
867
868         * qmake/mkspecs/features/features.prf:
869
870 2012-01-27  Roland Takacs  <takacs.roland@stud.u-szeged.hu>
871
872         Update committers.py with new contributors.
873         https://bugs.webkit.org/show_bug.cgi?id=77207
874
875         I added six new contributors into committers.py:
876
877         - Adam Kallai
878         - Balazs Ankes
879         - Janos Badics
880         - Nandor Huszka
881         - Roland Takacs
882         - Szilard Ledan-Muntean
883
884         * Scripts/webkitpy/common/config/committers.py:
885
886 2012-01-27  Vsevolod Vlasov  <vsevik@chromium.org>
887
888         Add Vsevolod Vlasov to reviewers list.
889
890         Reviewed by Pavel Feldman.
891
892         * Scripts/webkitpy/common/config/committers.py:
893
894 2012-01-26  No'am Rosenthal  <noam.rosenthal@nokia.com> and Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
895
896         [Qt] WKTR: Use a software rendering pipiline when running tests.
897         https://bugs.webkit.org/show_bug.cgi?id=76708
898
899         Reviewed by Kenneth Rohde Christiansen.
900
901         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
902         (WTR::WrapperWindow::WrapperWindow):
903         (WTR::PlatformWebView::PlatformWebView): Use software rendering of layers since the wrapping QQuickView isn't shown.
904
905 2012-01-26  Kevin Ollivier  <kevino@theolliviers.com>
906
907         [wx] Unreviewed. Build fix, add WebCore/testing/js
908         to the list of source directories we build.
909
910         * waf/build/settings.py:
911
912 2012-01-26  Filip Pizlo  <fpizlo@apple.com>
913
914         Tools/Scripts/commit-log-editor is broken due to $_ getting clobbered
915         https://bugs.webkit.org/show_bug.cgi?id=77177
916
917         Reviewed by Jon Honeycutt.
918
919         * Scripts/commit-log-editor:
920
921 2012-01-26  Ojan Vafai  <ojan@chromium.org>
922
923         Decrease sleep time when killing server_process on Mac.
924         https://bugs.webkit.org/show_bug.cgi?id=77169
925
926         Reviewed by Eric Seidel.
927
928         This shaves another ~70ms off single test runtime.
929
930         * Scripts/webkitpy/layout_tests/port/server_process.py:
931         (ServerProcess.stop):
932
933 2012-01-26  Ojan Vafai  <ojan@chromium.org>
934
935         Remove builder name so http://trac.webkit.org/changeset/106035 actually works.
936         https://bugs.webkit.org/show_bug.cgi?id=77164
937
938         Reviewed by Dirk Pranke.
939
940         http://trac.webkit.org/changeset/106035 depends on builder_name being
941         falsey by default. There's other code that requires it to be iterable
942         by default. So, make it be the empty string.
943
944         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
945         (parse_args):
946
947 2012-01-26  Ryosuke Niwa  <rniwa@webkit.org>
948
949         run-perf-test should include relative file paths in test names
950         https://bugs.webkit.org/show_bug.cgi?id=77144
951
952         Reviewed by Adam Birth.
953
954         Include all directory names in the test name separated by "/".
955         e.g. Bindings/event-target-wrapper.html translates to Bindings/event-target-wrapper
956
957         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
958         (PerfTestsRunner):
959         (PerfTestsRunner._process_parser_test_result):
960         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
961         (test_run_test_set_with_json_output):
962         (test_run_test_set_with_json_source):
963
964 2012-01-26  Dirk Pranke  <dpranke@chromium.org>
965
966         webkitpy: re-land cleanup of test scaffolding
967         https://bugs.webkit.org/show_bug.cgi?id=77153
968
969         Reviewed by Eric Seidel.
970
971         This is an attempt to reland the change in bug 76973 / 
972         r105935. My earlier attempt to manipulate PYTHONPATH
973         seems to only work some of the time, for reasons that
974         elude me. Directly adding Tools/Scripts to sys.path from
975         inside the test script appears to work (more?) reliably.
976
977         Also, it seemed like the files didn't actually get deleted
978         the last time; I don't know why.
979
980         * Scripts/test-webkitpy:
981         * Scripts/webkitpy/common/system/executive_unittest.py:
982         (command_line):
983         (ExecutiveTest):
984         (ExecutiveTest.test_run_command_args_type):
985         (ExecutiveTest.test_run_command_with_unicode):
986         (ExecutiveTest.test_running_pids):
987         (main):
988         * Scripts/webkitpy/common/system/fileutils.py: Removed.
989         * Scripts/webkitpy/test/cat.py: Removed.
990         * Scripts/webkitpy/test/cat_unittest.py: Removed.
991         * Scripts/webkitpy/test/echo.py: Removed.
992         * Scripts/webkitpy/test/echo_unittest.py: Removed.
993
994 2012-01-26  Ojan Vafai  <ojan@chromium.org>
995
996         run-webkit-tests calls "nm" when it doesn't need to
997         https://bugs.webkit.org/show_bug.cgi?id=77143
998
999         Reviewed by Dirk Pranke.
1000
1001         If we are not running any test that can be skipped by checking
1002         the DRT feature list or looking at the list of symbols, avoid calling
1003         out to DRT or calling "nm".
1004
1005         This saves ~700ms on my Mac Pro. In cases where you still have to call out
1006         to "nm" it's only a couple ms slower.
1007
1008         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1009         (TestExpectations.__init__):
1010         * Scripts/webkitpy/layout_tests/port/base.py:
1011         (Port.skipped_tests):
1012         * Scripts/webkitpy/layout_tests/port/webkit.py:
1013         (WebKitPort._should_check_for_missing_features):
1014         (WebKitPort):
1015         (WebKitPort._should_check_for_missing_symbols):
1016         (WebKitPort._skipped_tests_for_unsupported_features):
1017         (WebKitPort.skipped_layout_tests):
1018         (WebKitPort.skipped_tests):
1019         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1020         (WebKitPortTest.test_skipped_directories_for_symbols):
1021         (test_skipped_directories_for_features):
1022         (test_skipped_directories_for_features_no_matching_tests_in_test_list):
1023         (test_skipped_layout_tests):
1024
1025 2012-01-26  Benjamin Poulain  <bpoulain@apple.com>
1026
1027         Support a suffix on ChangeLog filenames based on a configuration file
1028         https://bugs.webkit.org/show_bug.cgi?id=76956
1029
1030         Reviewed by Kentaro Hara.
1031
1032         Modify the tools used to generate, edit and resolve ChangeLogs to support
1033         the addition of a suffix to the ChangeLog filename.
1034
1035         This patch makes it easier to support alternate ChangeLog files in branches.
1036
1037         * Scripts/VCSUtils.pm:
1038         (changeLogSuffix):
1039         (changeLogFileName):
1040         * Scripts/commit-log-editor:
1041         * Scripts/prepare-ChangeLog:
1042         (getLatestChangeLogs):
1043         (generateNewChangeLogs):
1044         (generateFileList):
1045         * Scripts/resolve-ChangeLogs:
1046         (findChangeLog):
1047
1048 2012-01-26  Sheriff Bot  <webkit.review.bot@gmail.com>
1049
1050         Unreviewed, rolling out r106038.
1051         http://trac.webkit.org/changeset/106038
1052         https://bugs.webkit.org/show_bug.cgi?id=77142
1053
1054         Caused a bunch of skipped tests to not be skipped. (Requested
1055         by ojan on #webkit).
1056
1057         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1058         (TestExpectationParser):
1059         (TestExpectationParser.parse):
1060         (TestExpectationParser._parse_line):
1061         (TestExpectationParser._collect_matching_tests):
1062         (TestExpectations.__init__):
1063         (TestExpectations._add_skipped_tests):
1064         * Scripts/webkitpy/layout_tests/port/base.py:
1065         (Port.test_dirs):
1066         (Port.normalize_test_name):
1067         (Port.update_baseline):
1068         (Port.layout_tests_dir):
1069         (Port.relative_perf_test_filename):
1070         (Port.abspath_for_test):
1071
1072 2012-01-25  Ojan Vafai  <ojan@chromium.org>
1073
1074         Parsing test_expecations.txt + Skipped lists takes too long
1075         https://bugs.webkit.org/show_bug.cgi?id=77059
1076
1077         Reviewed by Dirk Pranke.
1078
1079         This saves ~100ms on the Apple Mac port.
1080         -memoize a bunch of path methods.
1081         -Avoid doing multiple disk accesses per line.
1082         -Parse the skipped list directly instead of turning it into a test_expecations.txt
1083         formatting string and parsing that.
1084
1085         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1086         (TestExpectationParser):
1087         (TestExpectationParser.parse_skipped_tests):
1088         (TestExpectationParser._parse_line):
1089         (TestExpectationParser._collect_matching_tests):
1090         (TestExpectations.__init__):
1091         (TestExpectations._add_skipped_tests):
1092         * Scripts/webkitpy/layout_tests/port/base.py:
1093         (Port):
1094         (Port.test_isfile):
1095         (Port.normalize_test_name):
1096         (Port.layout_tests_dir):
1097         (Port.abspath_for_test):
1098
1099 2012-01-25  Ojan Vafai  <ojan@chromium.org>
1100
1101         Only store the SVN revision in the summarized results if we're on a builder
1102         https://bugs.webkit.org/show_bug.cgi?id=76976
1103
1104         Reviewed by Dirk Pranke.
1105
1106         This shaves another ~130ms off the single test run-webkit-tests runtime.
1107
1108         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1109         (summarize_results):
1110         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1111         (ResultSummaryTest.get_result):
1112         (ResultSummaryTest):
1113         (ResultSummaryTest.get_result_summary):
1114         (ResultSummaryTest.get_unexpected_results):
1115         (ResultSummaryTest.test_no_svn_revision):
1116         (ResultSummaryTest.test_svn_revision):
1117         * Scripts/webkitpy/layout_tests/port/base.py:
1118         (Port.use_trac_links_in_results_html):
1119
1120 2012-01-26  Luciano Wolf  <luciano.wolf@openbossa.org>
1121
1122         [Qt][WK2] fast/loader tests failed after r94178
1123         https://bugs.webkit.org/show_bug.cgi?id=67388
1124
1125         Reviewed by Simon Hausmann.
1126
1127         Applying the same fix used for Qt-WebKit1 from bug #67254 -
1128         [Qt][DRT] Normalize file:///tmp/LayoutTests in LayoutTestController::pathToLocalResource()
1129         https://bugs.webkit.org/show_bug.cgi?id=67254
1130         Translate file:///tmp/LayoutTests/* urls into the repository LayoutTests directory, which is
1131         derived from the running location of DumpRenderTree binary.
1132
1133         * WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
1134         (WTR::LayoutTestController::pathToLocalResource):
1135
1136 2012-01-26  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
1137
1138         Implement overridePreference for boolean preferences in WebKitTestRunner
1139         https://bugs.webkit.org/show_bug.cgi?id=77033
1140
1141         Reviewed by Alexey Proskuryakov.
1142
1143         For now overridePreference supports boolean values.
1144
1145         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
1146         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1147         (WTR::LayoutTestController::setXSSAuditorEnabled):
1148         (WTR::LayoutTestController::overridePreference):
1149         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1150         * WebKitTestRunner/TestController.cpp:
1151         (WTR::TestController::resetStateToConsistentValues): Reset the overrides
1152         explicitly and properly disable the use of PageCache when resetting the
1153         state. Tests that use PageCache can now enable it via overridePreference.
1154
1155 2012-01-26  Alexander Færøy  <alexander.faeroy@nokia.com>
1156
1157         [Qt] Support opening multiple URL's from the command line in the Qt MiniBrowser
1158         https://bugs.webkit.org/show_bug.cgi?id=77100
1159
1160         Reviewed by Simon Hausmann.
1161
1162         This patch makes the MiniBrowser read multiple URL's from the command
1163         line such that there will be one browser window opened for each URL.
1164
1165         * MiniBrowser/qt/MiniBrowserApplication.cpp:
1166         (MiniBrowserApplication::handleUserOptions):
1167
1168 2012-01-26  Alexander Færøy  <alexander.faeroy@nokia.com>
1169
1170         [Qt] The Qt MiniBrowser crashes when given multiple URL's
1171         https://bugs.webkit.org/show_bug.cgi?id=77009
1172
1173         Reviewed by Simon Hausmann.
1174
1175         This patch fixes a crash in BrowserWindow::newWindow() where a
1176         BrowserWindow-object is constructed with a 0 pointer as the
1177         WindowOptions parameter which leads to a crash.
1178
1179         * MiniBrowser/qt/BrowserWindow.cpp:
1180         (BrowserWindow::BrowserWindow):
1181         (BrowserWindow::newWindow):
1182         * MiniBrowser/qt/BrowserWindow.h:
1183
1184 2012-01-26  Jesus Sanchez-Palencia  <jesus.palencia@openbossa.org>
1185
1186         [Qt] Use ICU if available
1187         https://bugs.webkit.org/show_bug.cgi?id=76821
1188
1189         Reviewed by Simon Hausmann.
1190
1191         Adding unicode defines for a Qt 5 based build.
1192
1193         * qmake/mkspecs/features/features.prf:
1194
1195 2012-01-26  Nándor Huszka  <huszka.nandor@stud.u-szeged.hu>
1196
1197         [WK2] WebKitTestRunner needs layoutTestController.setGeolocationPermission
1198         https://bugs.webkit.org/show_bug.cgi?id=42545
1199
1200         Reviewed by Zoltan Herczeg.
1201
1202         Implement the setGeolocationPermission method.
1203
1204         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
1205         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1206         (WTR::LayoutTestController::setGeolocationPermission):
1207         (WTR):
1208         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1209         (LayoutTestController):
1210
1211 2012-01-26  Balazs Kelemen  <kbalazs@webkit.org>
1212
1213         [NRWT] Support --ignore-metrics
1214         https://bugs.webkit.org/show_bug.cgi?id=76278
1215
1216         Reviewed by Tony Chang.
1217
1218         Port the feature from ORWT.
1219
1220         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
1221         (SingleTestRunner._run_compare_test):
1222         * Scripts/webkitpy/layout_tests/port/driver.py:
1223         (DriverOutput):
1224         (DriverOutput.strip_metrics):
1225         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
1226         (DriverOutputTest):
1227         (DriverOutputTest.test_strip_metrics):
1228         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1229         (_set_up_derived_options):
1230         (parse_args):
1231
1232 2012-01-26  Sheriff Bot  <webkit.review.bot@gmail.com>
1233
1234         Unreviewed, rolling out r105935 and r105954.
1235         http://trac.webkit.org/changeset/105935
1236         http://trac.webkit.org/changeset/105954
1237         https://bugs.webkit.org/show_bug.cgi?id=77080
1238
1239         test-webkitpy is still broken (Requested by abarth on
1240         #webkit).
1241
1242         * Scripts/test-webkitpy:
1243         * Scripts/webkitpy/common/system/executive_unittest.py:
1244         (never_ending_command):
1245         (ExecutiveTest.test_run_command_args_type):
1246         (ExecutiveTest.test_run_command_with_unicode):
1247         (ExecutiveTest.test_running_pids):
1248         * Scripts/webkitpy/common/system/fileutils.py:
1249         (make_stdout_binary):
1250         * Scripts/webkitpy/test/cat.py:
1251         (command_arguments):
1252         (main):
1253         * Scripts/webkitpy/test/cat_unittest.py:
1254         (CatTest):
1255         (CatTest.assert_cat):
1256         (CatTest.test_basic):
1257         (CatTest.test_no_newline):
1258         (CatTest.test_unicode):
1259         (CatTest.test_as_command):
1260         * Scripts/webkitpy/test/echo.py:
1261         (command_arguments):
1262         (main):
1263         * Scripts/webkitpy/test/echo_unittest.py:
1264         (EchoTest):
1265         (EchoTest.test_basic):
1266         (EchoTest.test_no_newline):
1267         (EchoTest.test_unicode):
1268         (EchoTest.test_argument_order):
1269         (EchoTest.test_empty_arguments):
1270         (EchoTest.test_no_arguments):
1271         (EchoTest.test_as_command):
1272
1273 2012-01-26  Gabor Rapcsanyi  <rgabor@webkit.org>
1274
1275         [GTK] Detect cross-compilation in webkitdirs.pm to not generate gtkdoc
1276         https://bugs.webkit.org/show_bug.cgi?id=76138
1277
1278         Reviewed by Martin Robinson.
1279
1280         * Scripts/webkitdirs.pm:
1281         (isCrossCompilation):
1282         (buildAutotoolsProject):
1283
1284 2012-01-25  Adam Barth  <abarth@webkit.org>
1285
1286         Kentaro is a reviewer now.  Yay!
1287
1288         * Scripts/webkitpy/common/config/committers.py:
1289
1290 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
1291
1292         nrwt: running w/ no arguments fails
1293         https://bugs.webkit.org/show_bug.cgi?id=77064
1294
1295         Unreviewed, build fix.
1296
1297         This change fixes a regression introduced in r105944
1298         where running without --platform <foo> or one of the other
1299         platform switches would cause a null deref :(.
1300
1301         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1302         (main):
1303
1304 2012-01-25  Eric Seidel  <eric@webkit.org>
1305
1306         Automatically CC someone from Apple on any changes to Obj-C public API
1307         https://bugs.webkit.org/show_bug.cgi?id=77053
1308
1309         Reviewed by Timothy Hatcher.
1310
1311         This has long been de-facto practice, seems now that we have
1312         fancy watchlist support, we should just make the tools
1313         do this automatically.
1314
1315         * Scripts/webkitpy/common/config/watchlist:
1316
1317 2012-01-25  Antoine Labour  <piman@chromium.org>
1318
1319         Updating status in committers.py.
1320
1321         Unreviewed.
1322
1323         * Scripts/webkitpy/common/config/committers.py:
1324
1325 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
1326
1327         test-webkitpy: executive_unittest fails if PYTHONPATH isn't set
1328         https://bugs.webkit.org/show_bug.cgi?id=77058
1329
1330         Reviewed by Eric Seidel.
1331
1332         r105935 broke executive_unittest by changing which script we
1333         were running and not ensuring that the PYTHONPATH was set up
1334         properly.
1335
1336         This change also fixes a regression introduced recently where
1337         test-webkitpy wasn't figuring out the path to Tools/Scripts
1338         right, either (so running ./test-webkitpy would fail).
1339
1340         * Scripts/test-webkitpy:
1341         * Scripts/webkitpy/common/system/executive_unittest.py:
1342         (script_dir):
1343         (ExecutiveTest.setUp):
1344         (ExecutiveTest):
1345         (ExecutiveTest.tearDown):
1346
1347 2012-01-25  Adam Barth  <abarth@webkit.org>
1348
1349         Disable tests on the mac-ews while we sort out hte stability issues.
1350
1351         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
1352         (MacEWS):
1353
1354 2012-01-23  Dirk Pranke  <dpranke@chromium.org>
1355
1356         investigate use of 'mac' and 'win' as fully-specified port names for the apple ports
1357         https://bugs.webkit.org/show_bug.cgi?id=76475
1358
1359         Reviewed by Eric Seidel.
1360
1361         This patch cleans up the internal usage of port names for the
1362         Apple mac and Apple win ports so that 'mac' and 'win are never
1363         considered "fully specified" port names: they are still legal
1364         input to the PortFactory.get() routine, but only if run on the given
1365         platform (i.e. trying to PortFactory.get('mac') on windows or
1366         linux will return an error, since we don't know which version of
1367         the port is desired.
1368
1369         This also cleans up handling of the webkit2 variants, so that
1370         the full port names consistently follow the
1371         <operating_system>-<version>[-<wk2>] convention.
1372
1373         Lastly this patch adds some assertions and removes some
1374         try/catch logic to catch more programming and usage errors;
1375         previously mac-tiger might've been allowed -- and it would've
1376         translated into using just the baselines in platform/mac -- but
1377         now it should fail.
1378
1379         * Scripts/webkitpy/layout_tests/port/apple.py:
1380         (ApplePort.determine_full_port_name):
1381         (ApplePort.__init__):
1382         (ApplePort):
1383         (ApplePort._port_name_with_version):
1384         * Scripts/webkitpy/layout_tests/port/builders.py:
1385         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
1386         (FactoryTest.test_mac):
1387         (FactoryTest.test_win):
1388         * Scripts/webkitpy/layout_tests/port/mac.py:
1389         (MacPort.baseline_search_path):
1390         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1391         (MacTest.test_versions):
1392         * Scripts/webkitpy/layout_tests/port/win.py:
1393         (WinPort.baseline_search_path):
1394         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
1395         (WinTest.test_versions):
1396
1397 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
1398
1399         webkitpy: print nicer errors while linting expectations files, remove redundant tests
1400         https://bugs.webkit.org/show_bug.cgi?id=76955
1401
1402         Reviewed by Adam Barth.
1403
1404         This change cleans up the errors that are printed so that the
1405         are more "quickfix" compatible (path:lineno).
1406
1407         Also, there were a bunch of redundant tests between
1408         layout_tests.models.test_expectations_unittest and
1409         style.checkers.test_expectations_unittest. The latter should
1410         just expect the former to do most of the testing.
1411
1412         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1413         (TestExpectationParser._parse_modifiers):
1414         (TestExpectations._report_errors):
1415         (TestExpectations._add_skipped_tests):
1416         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1417         (test_parse_error_nonfatal):
1418         (SemanticTests.test_bad_bugid):
1419         * Scripts/webkitpy/style/checkers/test_expectations.py:
1420         (TestExpectationsChecker.__init__):
1421         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
1422         (TestExpectationsTestCase.assert_lines_lint):
1423         (TestExpectationsTestCase.test_valid_expectations):
1424         (TestExpectationsTestCase.test_invalid_expectations):
1425         (TestExpectationsTestCase.test_tab):
1426
1427 2012-01-25  Eric Seidel  <eric@webkit.org>
1428
1429         webkit-patch apply-* should always continue after failures
1430         https://bugs.webkit.org/show_bug.cgi?id=77057
1431
1432         Reviewed by Adam Barth.
1433
1434         As far as I can tell there is only one potential drawback to always
1435         forcing: that if you're somehow in the wrong directory it will create new
1436         directories for new files.  Since webkit-patch always cd's to the root
1437         it seems that's not a drawback.  The drawback of not using --force for
1438         svn-apply, is that it will stop after the first failure, which is rarely
1439         (if ever) the desired behavior.  This just removes the force option
1440         (which was strangely hidden behind --non-interactive).  This should
1441         make for a better user experiance.
1442
1443         * Scripts/webkitpy/common/checkout/checkout.py:
1444         (Checkout.apply_patch):
1445         * Scripts/webkitpy/common/checkout/checkout_mock.py:
1446         (MockCheckout.apply_patch):
1447         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1448         (CheckoutTest.test_chromium_deps):
1449         (CheckoutTest):
1450         (CheckoutTest.test_apply_patch):
1451         * Scripts/webkitpy/tool/commands/download_unittest.py:
1452         (DownloadCommandsTest._default_options):
1453         * Scripts/webkitpy/tool/steps/applypatch.py:
1454         (ApplyPatch.options):
1455         (ApplyPatch.run):
1456         * Scripts/webkitpy/tool/steps/options.py:
1457         (Options):
1458
1459 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
1460
1461         nrwt: should be able to run --platform test interactively
1462         https://bugs.webkit.org/show_bug.cgi?id=76959
1463
1464         Reviewed by Adam Barth.
1465
1466         As part of refactoring the Port interfaces, in r103254 I made
1467         passing a Host object to the Port mandatory; previously TestPort
1468         objects would create their own (mock) hosts. However,
1469         new-run-webkit-tests always passed a real Host, not a MockHost,
1470         and so you couldn't run new-run-webkit-tests --platform test
1471         interactively to debug test failures.
1472
1473         This change fixes that by creating a MockHost instead of a real
1474         host if you say --platform test*.
1475
1476         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1477         (main):
1478
1479 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
1480
1481         run-webkit-tests --lint-test-files should lint all the ports by default
1482         https://bugs.webkit.org/show_bug.cgi?id=76749
1483
1484         Reviewed by Ojan Vafai.
1485
1486         Currently run-webkit-tests --lint-test-files will only lint the
1487         test_expectations for the single port determined by the command
1488         line options. This is not obvious and can produce unintended results
1489         if you want to lint the chromium port (probably the normal case
1490         for using --lint-test-files) but don't specify a port name or
1491         --chromium.
1492
1493         I think we should probably lint *all* of the ports if no port
1494         name is supplied. This change implements that, and also moves
1495         the linting logic out of the Manager class, since this really
1496         has nothing to do with Managers. It has little to do with
1497         run-webkit-tests, but splitting it into a different script would
1498         introduce compatibility issues. It might make sense to do that
1499         anyway, in a separate patch and combining that with cleaning up
1500         the style checker to share more code.
1501
1502         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1503         (Manager._strip_test_dir_prefix):
1504         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1505         (lint):
1506         (run):
1507         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1508         (LintTest.test_all_configurations): added.
1509         (LintTest.test_lint_test_files): added.
1510         (LintTest.test_lint_test_files__errors): added.
1511         (MainTest.test_lint_test_files): removed.
1512         (MainTest.test_lint_test_files__errors): removed.
1513
1514 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
1515
1516         webkitpy: clean up a bunch of test scaffolding
1517         https://bugs.webkit.org/show_bug.cgi?id=76973
1518
1519         Reviewed by Eric Seidel.
1520
1521         There were several helper modules in the webkitpy.test package
1522         that appear to only be used by executive_unittest.py. I am
1523         rolling them all into that file to make that clearer.
1524
1525         * Scripts/webkitpy/common/system/executive_unittest.py:
1526         (ExecutiveTest.test_run_command_args_type):
1527         (ExecutiveTest.test_run_command_with_unicode):
1528         (ExecutiveTest.test_running_pids):
1529         (command_line):
1530         (main):
1531         * Scripts/webkitpy/common/system/fileutils.py: Removed.
1532         * Scripts/webkitpy/test/cat.py: Removed.
1533         * Scripts/webkitpy/test/cat_unittest.py: Removed.
1534         * Scripts/webkitpy/test/echo.py: Removed.
1535         * Scripts/webkitpy/test/echo_unittest.py: Removed.
1536
1537 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
1538
1539         webkitpy: clean up a bunch of test scaffolding
1540         https://bugs.webkit.org/show_bug.cgi?id=76973
1541
1542         Reviewed by Eric Seidel.
1543
1544         There were several helper modules in the webkitpy.test package
1545         that appear to only be used by executive_unittest.py. I am
1546         rolling them all into that file to make that clearer.
1547
1548         * Scripts/webkitpy/common/system/executive_unittest.py:
1549         (ExecutiveTest.test_run_command_args_type):
1550         (ExecutiveTest.test_run_command_with_unicode):
1551         (ExecutiveTest.test_running_pids):
1552         (command_line):
1553         (main):
1554         * Scripts/webkitpy/common/system/fileutils.py: Removed.
1555         * Scripts/webkitpy/test/cat.py: Removed.
1556         * Scripts/webkitpy/test/cat_unittest.py: Removed.
1557         * Scripts/webkitpy/test/echo.py: Removed.
1558         * Scripts/webkitpy/test/echo_unittest.py: Removed.
1559
1560 2012-01-25  Eric Seidel  <eric@webkit.org>
1561
1562         Remove Python 2.5 support from WebKit
1563         https://bugs.webkit.org/show_bug.cgi?id=71593
1564
1565         Reviewed by Adam Barth.
1566
1567         Just removing support from this one core file.
1568         I'll commit the rest of the (large) change once we've
1569         gone 24 hours and seen that no bots have exploded.
1570         We can roll out this one change if we break bots
1571         and then roll in the rest once those bots are fixed.
1572
1573         * Scripts/webkitpy/common/system/executive.py:
1574         (Executive.cpu_count):
1575
1576 2012-01-25  Hajime Morita  <morrita@google.com>
1577
1578         ENABLE_SHADOW_DOM should be available via build-webkit --shadow-dom
1579         https://bugs.webkit.org/show_bug.cgi?id=76863
1580
1581         Reviewed by Dimitri Glazkov.
1582
1583         Added a feature flag.
1584
1585         * Scripts/build-webkit:
1586         * qmake/mkspecs/features/features.prf:
1587
1588 2012-01-25  Martin Robinson  <mrobinson@igalia.com>
1589
1590         [WK2] Navigation events not triggered by a mouse event claim to be triggered by the left mouse button during policy decisions
1591         https://bugs.webkit.org/show_bug.cgi?id=76344
1592
1593         Reviewed by Sam Weinig.
1594
1595         Add a test verifying that the mouse button during policy decisions for
1596         events not triggered by mouse events is NoButton.
1597
1598         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
1599         (TestWebKitAPI::decidePolicyForNavigationAction): Added test.
1600         (TestWebKitAPI::decidePolicyForNewWindowAction): Added test.
1601
1602 2012-01-25  Ryosuke Niwa  <rniwa@webkit.org>
1603
1604         run-perf-tests should timeout sooner
1605         https://bugs.webkit.org/show_bug.cgi?id=77026
1606
1607         Reviewed by Adam Barth.
1608
1609         Bump off the timeout from 10 minutes to 4 minutes.
1610
1611         Also use float instead of str in test result values to compress the size of json files bots upload.
1612
1613         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1614         (PerfTestsRunner._parse_args):
1615         (PerfTestsRunner._process_chromium_style_test_result):
1616         (PerfTestsRunner._process_parser_test_result):
1617         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1618         (test_run_test_set_for_parser_tests):
1619         (test_run_test_set_with_json_output):
1620         (test_run_test_set_with_json_source):
1621
1622 2012-01-25  Enrica Casucci  <enrica@apple.com>
1623
1624         WebView should implement typingAttributes methods to work correctly with the Inspector bar in Mail.
1625         https://bugs.webkit.org/show_bug.cgi?id=76951
1626         <rdar://problem/9325158>
1627
1628         Reviewed by Alexey Proskuryakov.
1629
1630         * TestWebKitAPI/Tests/mac/InspectorBar.mm: Modified to reflect the changes to WebView
1631         and WebHTMLView.
1632         (TestWebKitAPI::TEST):
1633
1634 2012-01-23  MORITA Hajime  <morrita@google.com>
1635
1636         [Refactoring][Internals] Should have InternalSettings
1637         https://bugs.webkit.org/show_bug.cgi?id=76424
1638
1639         Reviewed by Kent Tamura.
1640
1641         * GNUmakefile.am: Added so new files.
1642
1643 2012-01-25  Takashi Toyoshima  <toyoshim@chromium.org>
1644
1645         webkitpy: new-run-webkit-websocket is broken
1646         https://bugs.webkit.org/show_bug.cgi?id=76999
1647
1648         Reviewed by Dirk Pranke.
1649
1650         r105674 broke new-run-webkit-websocket, since options.chromium
1651         doesn't exist any more. new-run-webkit-httpd was also broken
1652         and already fixed at r105835.
1653         * Scripts/new-run-webkit-websocketserver:
1654         (main):
1655
1656 2012-01-25  Nándor Huszka  <huszka.nandor@stud.u-szeged.hu>
1657
1658         [WK2] WebKitTestRunner needs layoutTestController.setPopupBlockingEnabled
1659         https://bugs.webkit.org/show_bug.cgi?id=63458
1660
1661         Reviewed by Adam Roben.
1662
1663         Implement the setPopupBlockingEnabled method.
1664
1665         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
1666         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1667         (WTR::LayoutTestController::setPopupBlockingEnabled):
1668         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1669
1670 2012-01-24  Balazs Kelemen  <kbalazs@webkit.org>
1671
1672         [Qt] Cleanup testfonts handling
1673         https://bugs.webkit.org/show_bug.cgi?id=76068
1674
1675         Reviewed by Simon Hausmann.
1676
1677         Factor test font initialization into it's own file  so test
1678         font users can share the implementation by adding the file
1679         to their build.
1680         Use fonts.conf from the test font directory to not require the
1681         current directory to be the WebKit root.
1682         Fix the -use-test-fonts option of QtTestBrowser by initializing
1683         the fonts earlier, before it is too late (namely before calling
1684         show on the main window).
1685
1686         * DumpRenderTree/qt/DumpRenderTree.pro:
1687         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1688         (WebCore::DumpRenderTree::open):
1689         * DumpRenderTree/qt/DumpRenderTreeQt.h:
1690         * DumpRenderTree/qt/QtInitializeTestFonts.cpp: Copied from Tools/WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp.
1691         (WebKit::initializeTestFonts):
1692         * DumpRenderTree/qt/QtInitializeTestFonts.h: Added.
1693         * DumpRenderTree/qt/fonts.conf: Removed. No need for this since
1694         we will use the one from the test fonts directory.
1695         * DumpRenderTree/qt/main.cpp:
1696         (main): Fix the ifdefs to make it clear that the code to
1697         set up the Dpi has not been ported to Qt5 yet.
1698         * QtTestBrowser/QtTestBrowser.pro:
1699         * QtTestBrowser/launcherwindow.h:
1700         (WindowOptions::WindowOptions):
1701         * QtTestBrowser/main.cpp:
1702         (launcherMain):
1703         (LauncherApplication::handleUserOptions):
1704         * WebKitTestRunner/InjectedBundle/Target.pri:
1705         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
1706         (WTR::activateFonts): Remove the code to set up the Dpi
1707         because it does not work with Qt5.
1708
1709 2012-01-25  Simon Hausmann  <simon.hausmann@nokia.com>
1710
1711         [Qt] build-webkit --qt requires a git-svn or SVN checkout.
1712
1713         Reviewed by Tor Arne Vestø.
1714
1715         Don't require it by making svnRevisionForDirectory return "unknown"
1716         instead of bailing out with an error.
1717
1718         * Scripts/VCSUtils.pm:
1719         (svnRevisionForDirectory):
1720
1721 2012-01-25  Adam Barth  <abarth@webkit.org>
1722
1723         Temporarily disable running the unit tests on the Chromium EWS.  These
1724         tests don't work properly (see
1725         https://bugs.webkit.org/show_bug.cgi?id=76979) and are currently
1726         blocking the EWS.
1727
1728         * Scripts/webkitpy/common/config/ports.py:
1729         (ChromiumPort):
1730
1731 2012-01-25  Adam Barth  <abarth@webkit.org>
1732
1733         Update results.html version of PixelZoomer
1734         https://bugs.webkit.org/show_bug.cgi?id=76965
1735
1736         Reviewed by Ojan Vafai.
1737
1738         Improve readability as suggested by Ojan.
1739
1740         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/pixelzoomer.js:
1741
1742 2012-01-25  Carlos Garcia Campos  <cgarcia@igalia.com>
1743
1744         [GTK] Page clients don't need to be GObjects anymore in WebKit2 GTK+ API
1745         https://bugs.webkit.org/show_bug.cgi?id=76899
1746
1747         Reviewed by Martin Robinson.
1748
1749         * gtk/generate-gtkdoc:
1750         (get_webkit2_options): WebKitWebLoaderClient has been renamed to
1751         WebKitLoaderClient.
1752
1753 2012-01-24  Ryosuke Niwa  <rniwa@webkit.org>
1754
1755         Build fix. Trigger Chromium Mac Release (Perf) on new commits.
1756
1757         * BuildSlaveSupport/build.webkit.org-config/config.json:
1758
1759 2012-01-23  Ryosuke Niwa  <rniwa@webkit.org>
1760
1761         WebKit needs public performance tests running on public bots
1762         https://bugs.webkit.org/show_bug.cgi?id=10266
1763
1764         Reviewed by Adam Barth.
1765
1766         Completes the initial implementation of WebKit Perf bots. Add a build step for run-perf-test
1767         and make google-mac-4 run it, and upload the results to webkit-perf.webkit.org.
1768
1769         * BuildSlaveSupport/build.webkit.org-config/config.json:
1770         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1771         (RunAndUploadPerfTests):
1772         (RunAndUploadPerfTests.start):
1773         (UploadPerfTestResults):
1774         (BuildAndPerfTestFactory):
1775         (BuildAndPerfTestFactory.__init__):
1776
1777 2012-01-24  Ojan Vafai  <ojan@chromium.org>
1778
1779         Initializing the browser property of the Bugzilla class takes too long
1780         https://bugs.webkit.org/show_bug.cgi?id=76960
1781
1782         Reviewed by Adam Barth.
1783
1784         Lazily initialize it. This saves ~150ms on a no-op run-webkit-tests call
1785         on my Mac Pro.
1786
1787         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
1788         (Bugzilla.__init__):
1789         (Bugzilla._get_browser):
1790         (Bugzilla):
1791         (Bugzilla._set_browser):
1792
1793 2012-01-24  Mark Rowe  <mrowe@apple.com>
1794
1795         Death to JavaScriptGlue.
1796
1797         Rubber-stamped by Geoffrey Garen.
1798
1799         * BuildSlaveSupport/build-launcher-app:
1800         (copyNightlyLauncher):
1801         * Scripts/build-webkit:
1802         * Scripts/do-file-rename:
1803         * Scripts/do-webcore-rename:
1804         * Scripts/webkitpy/common/config/build.py:
1805         (_should_file_trigger_build):
1806         * Scripts/webkitpy/common/config/build_unittest.py: Update the test to use another Mac-specific
1807         path in conjunction with a cross-platform path.
1808         (ShouldBuildTest):
1809
1810 2012-01-24  Kentaro Hara  <haraken@chromium.org>
1811
1812         webkitperl/prepare-ChangeLog_unittest/parser_unittests.pl should test stdout and stderr
1813         https://bugs.webkit.org/show_bug.cgi?id=75836
1814
1815         Reviewed by David Kilzer.
1816
1817         Previously parser_unittests.pl can just test the returned value of
1818         get_function_line_ranges() in prepare-ChangeLog. This patch fixes
1819         parser_unittests.pl so that it also tests the stdout and stderr of
1820         get_function_line_ranges().
1821
1822         This patch adds css_unittests_warning.css, which contains test cases
1823         for warning messages of the CSS parser of prepare-ChangeLog.
1824
1825         * Scripts/webkitperl/prepare-ChangeLog_unittest/parser_unittests.pl:
1826         (captureOutput): Captures the stdout and stderr of a given method.
1827         (convertAbsolutepathToWebKitPath): Converts absolute paths in a given
1828         string to relative paths.
1829
1830         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/css_unittests_warning-expected.txt: Added.
1831         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/css_unittests_warning.css: Added.
1832         (element1):
1833         (element2):
1834         (element3):
1835         (element4):
1836         ():
1837         (element5):
1838         (element6):
1839
1840         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests-expected.txt:
1841         Just updated the test results, i.e. added stdout and stderr results.
1842         No essential changes in the test results.
1843         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/css_unittests-expected.txt: Ditto.
1844         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/java_unittests-expected.txt: Ditto.
1845         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/javascript_unittests-expected.txt: Ditto.
1846         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/perl_unittests-expected.txt: Ditto.
1847         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/python_unittests-expected.txt: Ditto.
1848
1849 2012-01-24  Dirk Pranke  <dpranke@chromium.org>
1850
1851         webkitpy: new-run-webkit-httpd is broken
1852         https://bugs.webkit.org/show_bug.cgi?id=76963
1853
1854         Reviewed by Adam Barth.
1855
1856         r105674 broke new-run-webkit-httpd, since options.chromium
1857         doesn't exist any more. We didn't catch this because we don't
1858         run the integration tests by default anywhere :(.
1859
1860         * Scripts/new-run-webkit-httpd:
1861         (run):
1862         (main):
1863
1864 2012-01-24  Adam Barth  <abarth@webkit.org>
1865
1866         garden-o-matic pixel zoomer should let you look at every pixel of the image
1867         https://bugs.webkit.org/show_bug.cgi?id=76962
1868
1869         Reviewed by Ojan Vafai.
1870
1871         Previously, you couldn't get up to the right and bottom pixels.
1872
1873         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/pixelzoomer.js:
1874
1875 2012-01-24  Enrica Casucci  <enrica@apple.com>
1876
1877         WebHTMLView should implement typingAttributes methods to work correctly with the Inspector bar in Mail.
1878         https://bugs.webkit.org/show_bug.cgi?id=76951
1879         <rdar://problem/9325158>
1880
1881         Added test to verify that WebHTMLView supports the typingAttributes method and that it
1882         works as expected.
1883         
1884         Reviewed by Alexey Proskuryakov.
1885
1886         * TestWebKitAPI/Tests/mac/InspectorBar.mm:
1887         (TestWebKitAPI::TEST):
1888
1889 2012-01-24  James Robinson  <jamesr@chromium.org>
1890
1891         Put a space between the trac link and ellipsis in sheriffbot rollout message so it linkifies better in some IRC clients
1892         https://bugs.webkit.org/show_bug.cgi?id=76884
1893
1894         Reviewed by Adam Barth.
1895
1896         * Scripts/webkitpy/tool/bot/irc_command.py:
1897         (Rollout.execute):
1898         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
1899         (SheriffIRCBotTest.test_rollout):
1900         (SheriffIRCBotTest.test_revert):
1901         (SheriffIRCBotTest.test_multi_rollout):
1902         (SheriffIRCBotTest.test_rollout_with_r_in_svn_revision):
1903         (SheriffIRCBotTest.test_multi_rollout_with_r_in_svn_revision):
1904         (SheriffIRCBotTest.test_rollout_invalidate_reason):
1905         (test_multi_rollout_invalidate_reason):
1906
1907 2012-01-24  Kentaro Hara  <haraken@chromium.org>
1908
1909         The cpp parser of prepare-ChangeLog cannot detect a change in classes and namespaces
1910         https://bugs.webkit.org/show_bug.cgi?id=75531
1911
1912         Reviewed by David Kilzer.
1913
1914         Previously, the cpp parser of prepare-ChangeLog could not detect a change
1915         outside methods. Consider the following cpp file.
1916
1917             namespace N {
1918             int a;     // this change does not appear on ChangeLog.
1919             class C {
1920                 int b;     // this change does not appear on ChangeLog.
1921                 void f()
1922                 {
1923                     int c;     // this change appears on ChangeLog.
1924                 }
1925                 int d;     // this change does not appear on ChangeLog.
1926             };
1927             int e;     // this change does not appear on ChangeLog.
1928             };
1929
1930         The previous prepare-ChangeLog outputs just methods in which a change is found:
1931             (N::C::f):
1932
1933         This patch fixes prepare-ChangeLog so that it outputs namespaces, classes
1934         and methods in which a change is found:
1935             (N):
1936             (N::C):
1937             (N::C::f):
1938
1939         * Scripts/prepare-ChangeLog:
1940         (get_function_line_ranges_for_cpp): Modified as described above.
1941
1942         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests.cpp: Added test cases.
1943         (Class104):
1944         (Class105):
1945         (Class106):
1946         (Class106::func32):
1947         (Class106::func33):
1948         (NameSpace3):
1949         (NameSpace4):
1950         (NameSpace5):
1951         (NameSpace6):
1952         (Class107):
1953         (NameSpace5::NameSpace6::Class107::func34):
1954         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests-expected.txt:
1955
1956 2012-01-24  Kentaro Hara  <haraken@chromium.org>
1957
1958         prepare-ChangeLog outputs warnings for top-level { ... } blocks
1959         https://bugs.webkit.org/show_bug.cgi?id=75943
1960
1961         Reviewed by David Kilzer.
1962
1963         Currently, prepare-ChangeLog outputs warnings for top-level { ... } blocks,
1964         because it does not distinguish '}' at the end of a subroutine
1965         with '}' at the end of a if or for statement.
1966
1967         foo.pl:
1968             if (1) {
1969             }
1970             for (@v) {
1971             }
1972
1973         output:
1974             nested functions found at top-level at foo.pl:2
1975             nested functions found at top-level at foo.pl:4
1976
1977         This patch fixes the bug.
1978
1979         * Scripts/prepare-ChangeLog:
1980         (get_function_line_ranges_for_perl):
1981         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/perl_unittests.pl:
1982         Added test cases. No change in the test results.
1983
1984 2012-01-24  Dirk Pranke  <dpranke@chromium.org>
1985
1986         r105674 broke check-webkit-style for chromium's test_expectations.
1987
1988         Unreviewed, build fix.
1989         
1990         We now need to pass in a platform name of 'chromium' if we want
1991         the chromium port; options.chromium is no longer recognized. I
1992         will follow this patch up with a separate patch that has a test,
1993         but this at least fixes the build.
1994
1995         * Scripts/webkitpy/style/checkers/test_expectations.py:
1996         (TestExpectationsChecker._determine_port_from_exepectations_path):
1997
1998 2012-01-20  Ojan Vafai  <ojan@chromium.org>
1999
2000         check-webkit-style of the chromium test_expectations.txt file takes too long
2001         https://bugs.webkit.org/show_bug.cgi?id=76745
2002
2003         Reviewed by Dimitri Glazkov.
2004
2005         When in lint mode, have TestExpectations test all configurations instead
2006         of looping over each configuration. This also has the benefit of making
2007         the error output considerably more concise.
2008
2009         Also, got rid of the double-printing of errors when linting through check-webkit-style.
2010
2011         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2012         (Manager.lint):
2013         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2014         (TestExpectations._report_errors):
2015         (TestExpectations._add_expectations):
2016         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
2017         (test_parse_error_nonfatal):
2018         (test_error_on_different_platform):
2019         * Scripts/webkitpy/style/checkers/test_expectations.py:
2020         (TestExpectationsChecker.check_test_expectations):
2021         (TestExpectationsChecker.check):
2022         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2023         (TestExpectationsTestCase.test_determine_port_from_exepectations_path):
2024         (TestExpectationsTestCase.assert_lines_lint):
2025
2026 2012-01-24  Ryosuke Niwa  <rniwa@webkit.org>
2027
2028         Another build fix attempt after r105543.
2029
2030         * Scripts/webkitpy/common/net/file_uploader.py:
2031         (FileUploader.upload_as_multipart_form_data):
2032
2033 2012-01-24  Ryosuke Niwa  <rniwa@webkit.org>
2034
2035         Build fix attempt after r105543.
2036
2037         * Scripts/webkitpy/common/net/file_uploader.py:
2038         (FileUploader.upload_as_multipart_form_data):
2039
2040 2012-01-24  Vincent Scheib  <scheib@chromium.org>
2041
2042         [Chromium] Add Pointer Lock test hooks and mock implementation to DumpRenderTree
2043         https://bugs.webkit.org/show_bug.cgi?id=76411
2044
2045         Reviewed by Adam Barth.
2046
2047         Breaking up https://bugs.webkit.org/show_bug.cgi?id=75762
2048         into a series of patches, this change adds a trivial
2049         pointer lock implementation to the Chromium port's DumpRenderTree
2050         and exposes LayoutTestController hooks to force failure conditions.
2051
2052         * DumpRenderTree/chromium/LayoutTestController.cpp:
2053         (LayoutTestController::LayoutTestController):
2054         (LayoutTestController::didLosePointerLock):
2055         (LayoutTestController::setPointerLockWillFailAsynchronously):
2056         (LayoutTestController::setPointerLockWillFailSynchronously):
2057         * DumpRenderTree/chromium/LayoutTestController.h:
2058         * DumpRenderTree/chromium/WebViewHost.cpp:
2059         (WebViewHost::requestPointerLock):
2060         (WebViewHost::requestPointerUnlock):
2061         (WebViewHost::isPointerLocked):
2062         (WebViewHost::didCompletePointerLock):
2063         (WebViewHost::didNotCompletePointerLock):
2064         (WebViewHost::didLosePointerLock):
2065         (WebViewHost::reset):
2066         * DumpRenderTree/chromium/WebViewHost.h:
2067         (WebViewHost::setPointerLockWillFailAsynchronously):
2068         (WebViewHost::setPointerLockWillFailSynchronously):
2069
2070 2012-01-24  Mario Sanchez Prada  <msanchez@igalia.com>
2071
2072         Unreviewed, skip WK2 API test timing out on some GTK bots.
2073
2074         WebKit2APITests/TestDownloads is consistently timing out on the
2075         32bit release and 64bit debug bots.
2076
2077         * Scripts/run-gtk-tests:
2078         (TestRunner): Skip WebKit2APITests/TestDownloads.
2079
2080 2012-01-24  Andras Becsi  <andras.becsi@nokia.com>
2081
2082         [Qt] Fix the build with the newes Qt5 hashes
2083         https://bugs.webkit.org/show_bug.cgi?id=76657
2084
2085         Reviewed by Simon Hausmann.
2086
2087         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2088         (MiniBrowserApplication::MiniBrowserApplication): Disable the automatic touch->mouse event synthesis
2089         for now to prevent an infinite loop where Qt always creates mouse events out of our touch mock events.
2090
2091 2012-01-24  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2092
2093         [Qt] Distinguish between Cygwin and cmd.exe in the way we set env variables
2094
2095         We can't use win32-msvc* to decide whether or not to use (set FOO=bar), as
2096         building with MinGW inside cmd.exe will require the same construct. Instead
2097         we assume that when the first character in PWD is a '/' we're running under
2098         Cygwin, and use the normal unix way of setting environment variables.
2099
2100         Reviewed by Simon Hausmann.
2101
2102         * qmake/mkspecs/features/default_pre.prf:
2103
2104 2012-01-24  Simon Hausmann  <simon.hausmann@nokia.com>
2105
2106         Some qmake build files must include Qt5's "quick" module.
2107         https://bugs.webkit.org/show_bug.cgi?id=76671
2108
2109         Unreviewed build fix.
2110
2111         * MiniBrowser/qt/MiniBrowser.pro:
2112         * WebKitTestRunner/Target.pri:
2113
2114 2012-01-24  Andras Becsi  <andras.becsi@nokia.com>
2115
2116         [Qt] Fix the event timestamps for touch mocking
2117
2118         Reviewed by Simon Hausmann.
2119
2120         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2121         (MiniBrowserApplication::sendTouchEvent): We should not set QEvent::None as the
2122         timestamp argument in the QWindowSystemInterface::handleTouchEvent call.
2123
2124 2012-01-24  Carlos Garcia Campos  <cgarcia@igalia.com>
2125
2126         [GTK] Implement DownloadClient in WebKit2 GTK+ API
2127         https://bugs.webkit.org/show_bug.cgi?id=72952
2128
2129         Reviewed by Martin Robinson.
2130
2131         * gtk/generate-gtkdoc:
2132         (get_webkit2_options): Ignore WebKitDownloadClient.
2133
2134 2012-01-23  Zan Dobersek  <zandobersek@gmail.com>
2135
2136         [GTK] editing/deleting/5408255.html results are incorrect
2137         https://bugs.webkit.org/show_bug.cgi?id=53644
2138
2139         Reviewed by Martin Robinson.
2140
2141         WEBKIT_TOP_LEVEL environment variable is now set directly in either
2142         WebKitTestRunner or DumpRenderTree through usage of a compilation-time
2143         macro. This way both tools can be run outside the test harness without
2144         the need to manually set the environment variable.
2145
2146         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2147         (getTopLevelPath):
2148         * GNUmakefile.am:
2149         * Scripts/webkitpy/layout_tests/port/gtk.py:
2150         (GtkPort.setup_environ_for_server):
2151         * WebKitTestRunner/GNUmakefile.am:
2152         * WebKitTestRunner/InjectedBundle/gtk/InjectedBundleGtk.cpp:
2153         (WTR::InjectedBundle::platformInitialize):
2154
2155 2012-01-23  Dmitry Lomov  <dslomov@google.com>
2156
2157         [Chromium] Implement layoutTestController.workerThreadCount in DRT
2158         https://bugs.webkit.org/show_bug.cgi?id=74653.
2159
2160         Reviewed by Darin Fisher.
2161
2162         * DumpRenderTree/chromium/LayoutTestController.cpp:
2163         (LayoutTestController::LayoutTestController):
2164         (LayoutTestController::workerThreadCount):
2165         * DumpRenderTree/chromium/LayoutTestController.h:
2166
2167 2012-01-23  Dirk Pranke  <dpranke@chromium.org>
2168
2169         nrwt: make --chromium work like --qt
2170         https://bugs.webkit.org/show_bug.cgi?id=76875
2171
2172         Reviewed by Adam Barth.
2173
2174         --chromium used to have to be handled differently from --qt
2175         due to the way the PortFactory was implemented; there's not
2176         really a good reason for that any more so this patch makes
2177         things slightly more consistent and eliminates the
2178         options.chromium flag (--chromium is now truly a synonym for
2179         --platform chromium).
2180
2181         * Scripts/webkitpy/layout_tests/port/factory.py:
2182         (PortFactory._default_port):
2183         (PortFactory.get):
2184         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
2185         (FactoryTest.setUp):
2186         (FactoryTest.test_chromium_mac):
2187         (FactoryTest.test_chromium_linux):
2188         (FactoryTest.test_chromium_win):
2189         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2190         (parse_args):
2191
2192 2012-01-23  Dirk Pranke  <dpranke@chromium.org>
2193
2194         run-webkit-tests needs to propagate --chromium
2195         https://bugs.webkit.org/show_bug.cgi?id=76870
2196
2197         Reviewed by Eric Seidel.
2198
2199         run-webkit-tests removes '--chromium' argument from @ARGV when
2200         determining which port to run, which means that that doesn't
2201         propagate to new-run-webkit-tests. That's bad (and is handled
2202         for the other ports by re-adding the flag, but apparently we're
2203         just now noticing for Chromium).
2204
2205         * Scripts/run-webkit-tests:
2206
2207 2012-01-23  Brian Weinstein  <bweinstein@apple.com>
2208
2209         More build fixing after r105646.
2210
2211         * TestWebKitAPI/Tests/WTF/RedBlackTree.cpp:
2212         (TestWebKitAPI::TestNode::key):
2213
2214 2012-01-23  Gavin Barraclough  <barraclough@apple.com>
2215
2216         Unreviewed build fix - r105646 broke this.
2217
2218         * TestWebKitAPI/Tests/WTF/RedBlackTree.cpp:
2219         (TestWebKitAPI::TestNode::TestNode):
2220         (TestWebKitAPI::TestNode::key):
2221         (TestWebKitAPI::RedBlackTreeTest::assertEqual):
2222         (TestWebKitAPI::RedBlackTreeTest::assertSameValuesForKey):
2223         (TestWebKitAPI::RedBlackTreeTest::testDriver):
2224
2225 2012-01-23  Sheriff Bot  <webkit.review.bot@gmail.com>
2226
2227         Unreviewed, rolling out r105658.
2228         http://trac.webkit.org/changeset/105658
2229         https://bugs.webkit.org/show_bug.cgi?id=76883
2230
2231         We want this eventually, but not right at this moment
2232         (Requested by abarth on #webkit).
2233
2234         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
2235
2236 2012-01-23  Adam Barth  <abarth@webkit.org>
2237
2238         garden-o-matic should support Chromium Mac Lion
2239         https://bugs.webkit.org/show_bug.cgi?id=76880
2240
2241         Reviewed by Eric Seidel.
2242
2243         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
2244
2245 2012-01-23  Emil A Eklund  <eae@chromium.org>
2246
2247         check-webkit-style whitespace/operators triggers on overloaded division operator
2248         https://bugs.webkit.org/show_bug.cgi?id=76650
2249
2250         Reviewed by Darin Adler.
2251
2252         * Scripts/webkitpy/style/checkers/cpp.py:
2253         (check_spacing):
2254         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
2255         (CppStyleTest.test_operator_methods):
2256
2257 2012-01-23  Ryosuke Niwa  <rniwa@webkit.org>
2258
2259         run-perf-tests should report server-side errors
2260         https://bugs.webkit.org/show_bug.cgi?id=76802
2261
2262         Reviewed by Tony Chang.
2263
2264         Report errors whenever server's response doesn't match "OK".
2265
2266         * Scripts/webkitpy/common/net/file_uploader.py:
2267         (FileUploader.upload_single_text_file):
2268         (FileUploader.upload_as_multipart_form_data):
2269         (FileUploader._upload_data.callback):
2270         (FileUploader):
2271         (FileUploader._upload_data):
2272         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2273         (PerfTestsRunner._upload_json):
2274         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2275         (test_upload_json):
2276         (test_upload_json.MockFileUploader.upload_single_text_file):
2277
2278 2012-01-22  Ryosuke Niwa  <rniwa@webkit.org>
2279
2280         run-perf-tests ignore Skipped list on chromium
2281         https://bugs.webkit.org/show_bug.cgi?id=76764
2282
2283         Reviewed by Dirk Pranke.
2284
2285         Move skipped_perf_tests from WebKit port to Base port so that Chromium port
2286         can also find skipped list. Chromium port only uses test_expectations.txt for
2287         layout tests but performacne tests don't use test_expectations.txt so Chromium port
2288         also needs to use Skipped list.
2289
2290         * Scripts/webkitpy/layout_tests/port/base.py:
2291         (Port._tests_from_skipped_file_contents):
2292         (Port):
2293         (Port._expectations_from_skipped_files):
2294         (Port.skipped_perf_tests):
2295         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2296         (PortTest.test_skipped_perf_tests):
2297         (PortTest.test_skipped_perf_tests.add_text_file):
2298         * Scripts/webkitpy/layout_tests/port/webkit.py:
2299         (WebKitPort._skipped_tests_for_unsupported_features):
2300         (WebKitPort._skipped_file_search_paths):
2301         (WebKitPort.skipped_layout_tests):
2302
2303 2012-01-23  Andy Estes  <aestes@apple.com>
2304
2305         Fix the build after r105635.
2306
2307         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp:
2308         (TestWebKitAPI::TEST):
2309
2310 2012-01-23  Xianzhu Wang  <wangxianzhu@chromium.org>
2311
2312         Basic enhancements to StringBuilder
2313         https://bugs.webkit.org/show_bug.cgi?id=67081
2314
2315         Reviewed by Darin Adler.
2316
2317         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp:
2318         (TestWebKitAPI::TEST):
2319
2320 2012-01-23  Mario Sanchez Prada  <msanchez@igalia.com>
2321
2322         [GTK] run-gtk-tests randomly fails while running the xprop comand
2323         https://bugs.webkit.org/show_bug.cgi?id=76817
2324
2325         Reviewed by Gustavo Noronha Silva.
2326
2327         No need to use xprop to remove the AT_SPI_BUS property since
2328         run-gtk-tests will always run new instances of Xvfb.
2329
2330         * Scripts/run-gtk-tests:
2331         (TestRunner): Unskip TestWebKitAccessibility.
2332         (TestRunner.run): Uncomment lines for launching the accessibility
2333         bus and registry daemon, and remove lines for running xprop.
2334
2335 2012-01-23  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2336
2337         [Qt] Don't warn about override and final being C++11 extensions
2338
2339         Clang will emit a warning when these extensions are used without passing
2340         --std=c++11, but we use feature checking to decide if we have the right
2341         extensions, so we can safely ignore these warnings. The XCode and Windows
2342         project files have the same workaround.
2343
2344         The reason for adding the flag to QMAKE_OBJECTIVE_CFLAGS as well is that
2345         we only have one extra compiler for Objective-C, which is also used for
2346         Objective-C++ sources, so we have to pass the flag, even if it doesn't
2347         make sense for Objective-C.
2348
2349         Reviewed by Simon Hausmann.
2350
2351         * qmake/mkspecs/features/unix/default_post.prf:
2352
2353 2012-01-20  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2354
2355         [Qt] Change how build-webkit decides when to do full incremental builds
2356
2357         Instead of relying on update-webkit (which isn't run on the bots) to
2358         decide when to do a full incremental build (make qmake), we let the
2359         build-webkit script itself check the current SVN revision against the
2360         previous build (by storing it in .webkit.config).
2361
2362         If the two differ we assume a full incremental build is needed, since
2363         the new revisions might have introduced problematic things like new
2364         Q_OBJECT macros. If not, we assume the developer is doing changes
2365         locally, and revert to doing a plain 'make'.
2366
2367         In addition, when the build fails in the latter case, we inform the
2368         developer of possible pitfalls and how to manually run 'make qmake'.
2369
2370         Reviewed by Ossy.
2371
2372         * Scripts/update-webkit:
2373         * Scripts/webkitdirs.pm:
2374         (buildQMakeProjects):
2375
2376 2012-01-23  Mario Sanchez Prada  <msanchez@igalia.com>
2377
2378         [GTK] run-gtk-tests randomly fails while running the xprop comand
2379         https://bugs.webkit.org/show_bug.cgi?id=76817
2380
2381         Reviewed by Philippe Normand.
2382
2383         Temporarily comment the lines related to launching the ATSPI bus
2384         and registry daemon, which are making the bots to fail randomly.
2385
2386         * Scripts/run-gtk-tests:
2387         (TestRunner): Skip TestWebKitAccessibility, as it won't run
2388         properly if the ATSPI infrastructure is not properly initialized.
2389         (TestRunner.run): Comment lines related to initialization of ATSPI.
2390
2391 2012-01-22  Ryosuke Niwa  <rniwa@webkit.org>
2392
2393         Windows python test build fix.
2394
2395         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2396         (PerfTestsRunner.__init__):
2397
2398 2012-01-22  Sergio Villar Senin  <svillar@igalia.com>
2399
2400         [GTK] DumpRenderTree converts "file:///" to a path differently
2401         https://bugs.webkit.org/show_bug.cgi?id=76631
2402
2403         Reviewed by Martin Robinson.
2404
2405         DumpRenderTree should print "/" as the last path component if the
2406         path is a lone slash instead of empty output.
2407
2408         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2409         (webViewConsoleMessage):
2410
2411 2012-01-20  David Levin  <levin@chromium.org>
2412
2413         The leaky DC check should use adoptPtr as a signal instead of OwnPtr to get less false positives.
2414         https://bugs.webkit.org/show_bug.cgi?id=76752
2415
2416         Reviewed by Eric Seidel.
2417
2418         * Scripts/webkitpy/style/checkers/cpp.py:
2419         (check_for_leaky_patterns):
2420         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
2421         (LeakyPatternTest.test_create_dc):
2422         (LeakyPatternTest.test_create_compatible_dc):
2423
2424 2012-01-20  Shawn Singh  <shawnsingh@chromium.org>
2425
2426         Changed my status in commiters.py
2427
2428         Unreviewed.
2429
2430         * Scripts/webkitpy/common/config/committers.py:
2431
2432 2012-01-20  Ojan Vafai  <ojan@chromium.org>
2433
2434         run-webkit-tests --lint-test-files crawls the whole LayoutTests subtree
2435         https://bugs.webkit.org/show_bug.cgi?id=76748
2436
2437         Reviewed by Ryosuke Niwa.
2438
2439         It crawls the whole subtree and then doesn't use the data. Cutting this out
2440         saves 4 seconds warm and 17 seconds cold on my Mac Pro.
2441
2442         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2443         (run):
2444
2445 2012-01-20  Adam Barth  <abarth@webkit.org>
2446
2447         Another tiny tweak to the garden-o-matic CSS.  This makes things line
2448         up slightly nicer when there aren't any failures.
2449
2450         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
2451         (#summary):
2452
2453 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
2454
2455         run-perf-tests should support --test-results-server option
2456         https://bugs.webkit.org/show_bug.cgi?id=76680
2457
2458         Reviewed by Adam Barth.
2459
2460         Add --test-results-server, --builder-name, and --build-number options to run-perf-tests
2461         to be used by perf bots. Also refactor file_uploader as needed.
2462
2463         * Scripts/webkitpy/common/net/file_uploader.py:
2464         (FileUploader.__init__):
2465         (FileUploader.upload_single_file):
2466         (FileUploader.upload_as_multipart_form_data):
2467         (FileUploader):
2468         (FileUploader._upload_data):
2469         (FileUploader._upload_data.callback):
2470         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2471         (JSONResultsGeneratorBase.upload_json_files):
2472         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2473         (PerfTestsRunner):
2474         (PerfTestsRunner._parse_args):
2475         (PerfTestsRunner.run):
2476         (PerfTestsRunner._generate_json):
2477         (PerfTestsRunner._upload_json):
2478         (PerfTestsRunner._run_tests_set):
2479         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2480         (create_runner):
2481         (test_run_with_upload_json):
2482         (test_run_with_upload_json.mock_upload_json):
2483         (test_upload_json):
2484         (test_upload_json.MockFileUploader):
2485         (test_upload_json.MockFileUploader.__init__):
2486         (test_upload_json.MockFileUploader.upload_single_file):
2487         (test_parse_args):
2488
2489 2012-01-20  Adam Barth  <abarth@webkit.org>
2490
2491         The party time image overlaps real content!  This patch makes the
2492         image centered vertically.
2493
2494         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
2495         (#onebar.partytime #summary):
2496
2497 2012-01-19  Ojan Vafai  <ojan@chromium.org>
2498
2499         Refactor TestExpectationsParser in preparation for caching the results
2500         https://bugs.webkit.org/show_bug.cgi?id=76669
2501
2502         Reviewed by Dimitri Glazkov.
2503
2504         Make everything private expect for the parse method. 
2505         Eventually, we'll need the expectations lines to not be modified
2506         outside of TestExpectationsParser so we can cache the results.
2507         This makes check-webkit-style of the chromium test_expectations.txt file
2508         go from ~17 seconds to ~12 seconds on my Mac Pro.
2509
2510         This patch is just a refactor in preparation, so no new tests.
2511
2512         * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor_unittest.py:
2513         (TestExpectationEditorTests.make_parsed_expectation_lines):
2514         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2515         (TestExpectationParser.parse):
2516         (TestExpectationParser):
2517         (TestExpectationParser._parse_line):
2518         (TestExpectationParser._tokenize):
2519         (TestExpectationParser._tokenize_list):
2520         (TestExpectationsModel._clear_expectations_for_test):
2521         (TestExpectations.__init__):
2522         (TestExpectations._add_expectations):
2523         (TestExpectations._add_skipped_tests):
2524         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
2525         (TestExpectationParserTests.test_tokenize_blank):
2526         (TestExpectationParserTests.test_tokenize_missing_colon):
2527         (TestExpectationParserTests.test_tokenize_extra_colon):
2528         (TestExpectationParserTests.test_tokenize_empty_comment):
2529         (TestExpectationParserTests.test_tokenize_comment):
2530         (TestExpectationParserTests.test_tokenize_missing_equal):
2531         (TestExpectationParserTests.test_tokenize_extra_equal):
2532         (TestExpectationParserTests.test_tokenize_valid):
2533         (TestExpectationParserTests.test_tokenize_valid_with_comment):
2534         (TestExpectationParserTests.test_tokenize_valid_with_multiple_modifiers):
2535         (TestExpectationParserTests.test_parse_empty_string):
2536         (TestExpectationSerializerTests.assert_round_trip):
2537         (TestExpectationSerializerTests.assert_list_round_trip):
2538         * Scripts/webkitpy/tool/commands/expectations.py:
2539         (OptimizeExpectations.execute):
2540         * Scripts/webkitpy/tool/servers/gardeningserver.py:
2541         (GardeningExpectationsUpdater.update_expectations):
2542
2543 2012-01-20  Adam Barth  <abarth@webkit.org>
2544
2545         Follow-up to previous patch: don't produce NaN when the revision number
2546         is missing.
2547
2548         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
2549         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model_unittests.js:
2550
2551 2012-01-20  Adam Barth  <abarth@webkit.org>
2552
2553         Garden-o-matic should tell me which revisions have been checked by all the bots
2554         https://bugs.webkit.org/show_bug.cgi?id=76722
2555
2556         Reviewed by Dimitri Glazkov.
2557
2558         This information helps me know when it's safe for me to roll a
2559         candidate revision.
2560
2561         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
2562         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
2563         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model_unittests.js:
2564
2565 2012-01-20  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2566
2567         build-webkit: Don't spit out congratulations message on Ctrl+C
2568
2569         Reviewed by Andreas Kling.
2570
2571         * Scripts/build-webkit:
2572
2573 2012-01-20  Philippe Normand  <pnormand@igalia.com>
2574
2575         [GTK] Allow extra jhbuild modulesets
2576         https://bugs.webkit.org/show_bug.cgi?id=76691
2577
2578         Reviewed by Gustavo Noronha Silva.
2579
2580         * gtk/jhbuildrc: Extra modulesets and modules can be added in
2581         JHBuild using the WEBKIT_EXTRA_MODULESETS and WEBKIT_EXTRA_MODULES
2582         env variables respectively. Use comma-separated values. Example:
2583         WEBKIT_EXTRA_MODULES=file:///path/to/module.set,file:///other/path.
2584
2585 2012-01-20  Mario Sanchez Prada  <msanchez@igalia.com>
2586
2587         [GTK] Expose accessibility hierarchy in WebKit2 to ATK/AT-SPI based ATs
2588         https://bugs.webkit.org/show_bug.cgi?id=72589
2589
2590         Reviewed by Martin Robinson.
2591
2592         Ensure the accessibility infrastructure is available when running
2593         the unit tests, otherwise WK2's accessibility tests won't work.
2594
2595         * Scripts/run-gtk-tests:
2596         (TestRunner):
2597         (TestRunner._lookup_atspi2_binary): New. Looks for the directory
2598         where at-spi2's stuff is located (like at-spi-bus-launcher and
2599         at-spi2-registryd).
2600         (TestRunner.run): Initialize the accessibility infrastructure.
2601
2602         * gtk/generate-gtkdoc:
2603         (get_webkit2_options): Add 'WebKitWebViewBaseAccessible.*'.
2604
2605         * gtk/jhbuild.modules: Added at-spi2-core and at-spi2-atk.
2606
2607 2012-01-20  Hayato Ito  <hayato@chromium.org>
2608
2609         Add ShadowRoot.idl which is enabled by newly introduced SHADOW_DOM flag.
2610         https://bugs.webkit.org/show_bug.cgi?id=76353
2611
2612         Reviewed by Hajime Morita.
2613
2614         Add ShadowRoot.idl, which is enabled only on chromium port since this is
2615         under development feature.
2616         ShadowRoot.idl contains minimum API so that we can test it.
2617         Other APIs should be added on other changes so that we can isolate issues.
2618
2619         * DumpRenderTree/chromium/TestShell.cpp:
2620         (TestShell::TestShell):
2621
2622 2012-01-20  Mario Sanchez Prada  <msanchez@igalia.com>
2623
2624         [GTK] GTK's DRT not to log events for already defunct objects
2625         https://bugs.webkit.org/show_bug.cgi?id=76620
2626
2627         Reviewed by Martin Robinson.
2628
2629         Do not log 'state-change:defunct' events.
2630
2631         * DumpRenderTree/gtk/AccessibilityCallbacks.cpp:
2632         (printAccessibilityEvent): Early return if a
2633         'state-change:defunct' signal is passed.
2634         (axObjectEventListener): Pass the signal name and value in
2635         separate parameters to printAccessibilityEvent.
2636
2637 2012-01-19  Gustavo Noronha Silva  <gns@gnome.org>
2638
2639         Unreviewed build fix. Wrong variable being used as the cwd for git
2640         remote update command.
2641
2642         * gtk/run-with-jhbuild:
2643         (update_jhbuild):
2644
2645 2012-01-19  Gustavo Noronha Silva  <gns@gnome.org>
2646
2647         [GTK] ensure the jhbuild used by webkit is as up-to-date as needed
2648         https://bugs.webkit.org/show_bug.cgi?id=76585
2649
2650         Reviewed by Martin Robinson.
2651
2652         This allows us to specify a minimum version of jhbuild required,
2653         while also providing stability against changes done to the master
2654         branch, by letting us choose what is the exact version that will
2655         be used.
2656
2657         * gtk/run-with-jhbuild: refactored for more hackability and easier
2658         reading
2659         (jhbuild_installed): checks whether jhbuild is installed
2660         (jhbuild_cloned): checks whether jhbuild has been cloned
2661         (jhbuild_at_expected_revision): checks whether jhbuild is at the
2662         expected revision
2663         (update_jhbuild): brings jhbuild to the expected revision
2664         (clone_jhbuild): clones jhbuild
2665         (install_jhbuild): build and installs jhbuild
2666         (update_webkitgtk_libs): runs update-webkitgtk-libs
2667         (ensure_jhbuild): high-level logic to decide whether and which of
2668         the steps above are executed
2669
2670 2012-01-19  Alexey Proskuryakov  <ap@apple.com>
2671
2672         [WK2] DumpRenderTree converts "file:///" to a path differently
2673         https://bugs.webkit.org/show_bug.cgi?id=76653
2674
2675         Reviewed by John Sullivan.
2676
2677         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp: (WTR::lastFileURLPathComponent):
2678         Do print a slash if there are no path components.
2679
2680 2012-01-19  Eric Seidel  <eric@webkit.org>
2681
2682         webkit-patch cannot rollout patches without changelog/bug number
2683         https://bugs.webkit.org/show_bug.cgi?id=75962
2684
2685         Reviewed by Adam Barth.
2686
2687         * Scripts/webkitpy/tool/commands/download.py:
2688         (AbstractRolloutPrepCommand._prepare_state):
2689         * Scripts/webkitpy/tool/commands/download_unittest.py:
2690         (AbstractRolloutPrepCommandTest.test_prepare_state):
2691
2692 2012-01-19  Alexey Proskuryakov  <ap@apple.com>
2693
2694         [WK2] fast/url/degenerate-file-base.html fails
2695         https://bugs.webkit.org/show_bug.cgi?id=76619
2696
2697         Reviewed by Adam Roben.
2698
2699         A URL generated form local path for loading had an extra slash, making document.URL
2700         incorrect in all WebKit2 tests (file://localhost//path/to/test.html).
2701
2702         * WebKitTestRunner/TestInvocation.cpp: (WTR::createWKURL):
2703
2704 2012-01-18  Ojan Vafai  <ojan@chromium.org>
2705
2706         check-webkit-style of the chromium test_expectations.txt file doesn't test all chromium ports
2707         https://bugs.webkit.org/show_bug.cgi?id=76510
2708
2709         Reviewed by Adam Barth.
2710
2711         Unlike other style rules, if you get the syntax of the test_expectations.txt
2712         file wrong, the layout tests won't run. Also, this check is simple and only
2713         slows down committing if you actually modify one of the test_expectations.txt files.
2714
2715         * Scripts/webkitpy/style/checkers/test_expectations.py:
2716         (TestExpectationsChecker.check_test_expectations):
2717         (TestExpectationsChecker.check):
2718         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2719         (TestExpectationsTestCase.test_check_covers_all_configurations):
2720         (TestExpectationsTestCase.test_check_covers_all_configurations.mock_check_test_expectations):
2721         (TestExpectationsTestCase):
2722         (assert_lines_lint):
2723
2724 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
2725
2726         Build fix. Revert an unintentional change to WebKitDriver in r105443.
2727
2728         * Scripts/webkitpy/layout_tests/port/driver.py:
2729         (DriverProxy.__init__):
2730
2731 2012-01-19  Dirk Pranke  <dpranke@chromium.org>
2732
2733         remove the duplicated build rules from WebKit.gyp, Tools.gyp
2734         https://bugs.webkit.org/show_bug.cgi?id=73384
2735
2736         Reviewed by Tony Chang.
2737
2738         This patch removes the no-longer-necessary conditional logic for
2739         build_webkit_exes_from_webkit_gyp; we now always build the exes
2740         from their own dedicated gyp files.
2741
2742         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2743
2744 2012-01-19  Alexey Proskuryakov  <ap@apple.com>
2745
2746         make-new-script-test should work with HTTP tests
2747         https://bugs.webkit.org/show_bug.cgi?id=76603
2748
2749         Reviewed by Adam Roben.
2750
2751         * Scripts/make-new-script-test:
2752         (makePathToSharedSources): Renamed from makeTestRelativePathToSharedSources(), since it
2753         now also creates absolute paths for HTTP.
2754
2755 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
2756
2757         Some perf tests time out when ran by run-perf-tests
2758         https://bugs.webkit.org/show_bug.cgi?id=76612
2759
2760         Reviewed by Dirk Pranke and Eric Seidel.
2761
2762         Always pass --no-timeout to DumpRenderTree from run-perf-tests.
2763         Otherwise some tests such as Parser/xml-parser.html will timeout.
2764
2765         --no-timeout option is currently supported by Chromium and Mac ports.
2766
2767         * Scripts/webkitpy/layout_tests/port/base.py:
2768         (Port.to.create_driver):
2769         * Scripts/webkitpy/layout_tests/port/chromium.py:
2770         (ChromiumDriver.__init__):
2771         (ChromiumDriver._wrapper_options):
2772         * Scripts/webkitpy/layout_tests/port/driver.py:
2773         (Driver.__init__):
2774         (DriverProxy.__init__):
2775         * Scripts/webkitpy/layout_tests/port/webkit.py:
2776         (WebKitDriver.__init__):
2777         (WebKitDriver.cmd_line):
2778         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2779         (WebKitDriverTest.test_read_binary_block):
2780         (WebKitDriverTest):
2781         (WebKitDriverTest.test_no_timeout):
2782         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2783         (get_tests_run.RecordingTestDriver.__init__):
2784         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2785         (PerfTestsRunner._parse_args):
2786         (PerfTestsRunner._run_tests_set):
2787         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2788         (create_runner):
2789
2790 2012-01-19  Ben Wells  <benwells@chromium.org>
2791
2792         Add support for window.print to chromium DRT
2793         https://bugs.webkit.org/show_bug.cgi?id=76479
2794
2795         When used in a test window.print goes into print mode and then
2796         straight out again. This will allow the afterprint event to be
2797         tested by having the afterprint event happen before the dump
2798         happens.
2799
2800         Reviewed by Mihai Parparita.
2801
2802         * DumpRenderTree/chromium/WebViewHost.cpp:
2803         (WebViewHost::printPage):
2804         * DumpRenderTree/chromium/WebViewHost.h:
2805
2806 2012-01-18  Enrica Casucci  <enrica@apple.com>
2807
2808         editingAttributedStringFromRange in WebHTMLConverter does not handle NSUnderlineStyleAttributeName.
2809         https://bugs.webkit.org/show_bug.cgi?id=76588
2810         <rdar://problem/9325183>
2811
2812         Added test.
2813         
2814         Reviewed by Dan Bernstein.
2815
2816         * TestWebKitAPI/Tests/mac/InspectorBar.mm:
2817         (TestWebKitAPI::TEST):
2818
2819 2012-01-19  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2820
2821         [Qt] Do full incremental builds until bots use update-webkit
2822
2823         Rubber-stamped by Simon Hausmann.
2824
2825         * Scripts/webkitdirs.pm:
2826         (buildQMakeProjects):
2827
2828 2012-01-19  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2829
2830         Fix the Qt build
2831
2832         Unreviewed build fix.
2833
2834         * Scripts/webkitdirs.pm:
2835         (buildQMakeProjects):
2836
2837 2012-01-19  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2838
2839         [Qt] Make build-webkit always do safe incremental builds after update-webkit
2840
2841         When building against Qt5 with GCC we would just run 'make' in the build dir,
2842         and rely on the dependency files output by GCC (-MD) to handle dependency
2843         tracking, but that fails for special-cases like adding a Q_OBJECT macro
2844         to a header.
2845
2846         To guarantee that an incrmental build will work, we have to run 'make qmake',
2847         which we now do on every build-webkit that's followed by a successful run
2848         of update-webkit. The reasoning is that update-webkit can result in such
2849         potential corner-cases being applied, and since we can't know for sure
2850         unless we inspect the diff and account for all the corner cases we assume
2851         the worst and always run 'make qmake'.
2852
2853         After a succesful run of build-webkit we proceed to do just 'make' for any
2854         subsequent runs, since we assume that the developer knows what kind of
2855         changes he/she is doing, and when a 'make qmake' is needed.
2856
2857         Reviewed by Simon Hausmann.
2858
2859         * Scripts/update-webkit:
2860         * Scripts/webkitdirs.pm:
2861         (determineCurrentSVNRevision):
2862         (currentSVNRevision):
2863         (buildQMakeProjects):
2864         * qmake/mkspecs/features/default_post.prf:
2865
2866 2012-01-18  Ryosuke Niwa  <rniwa@webkit.org>
2867
2868         run-perf-tests should support Skipped list
2869         https://bugs.webkit.org/show_bug.cgi?id=76594
2870
2871         Reviewed by Adam Barth.
2872
2873         Add a support for Skipped list in run-perf-tests; also skip files in resources directories.
2874
2875         * Scripts/webkitpy/layout_tests/port/base.py:
2876         (Port.perf_tests_dir):
2877         (Port.skipped_perf_tests):
2878         (Port.skips_perf_test):
2879         (Port.relative_perf_test_filename):
2880         * Scripts/webkitpy/layout_tests/port/test.py:
2881         (TestPort.perf_tests_dir):
2882         * Scripts/webkitpy/layout_tests/port/webkit.py:
2883         (WebKitPort._expectations_from_skipped_files):
2884         (WebKitPort):
2885         (WebKitPort.skipped_layout_tests):
2886         (WebKitPort.skipped_perf_tests):
2887         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2888         (PerfTestsRunner._collect_tests):
2889         (PerfTestsRunner.run):
2890         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2891         (test_run_test_set_with_json_output):
2892         (test_run_test_set_with_json_source):
2893         (test_collect_tests_with_skipped_list):
2894         (test_collect_tests_with_skipped_list.add_file):
2895
2896 2012-01-18  Dirk Pranke  <dpranke@chromium.org>
2897
2898         Fix path to chromium_src_dir introduced in previous change.
2899
2900         Unreviewed, build fix.
2901
2902         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2903
2904 2012-01-18  Alexey Proskuryakov  <ap@apple.com>
2905
2906         Need infrastructure to test Content-Disposition filename encoding support
2907         https://bugs.webkit.org/show_bug.cgi?id=76572
2908         <rdar://problem/6421825>
2909
2910         Reviewed by Sam Weinig.
2911
2912         Made policy delegate log suggested file name when resource has Content-Disposition: attachment.
2913
2914         WebKitTestRunner does not have a policy delegate yet, bug 42546.
2915
2916         * DumpRenderTree/mac/PolicyDelegate.mm:
2917         (dispositionTypeFromContentDispositionHeader):
2918         (-[PolicyDelegate webView:decidePolicyForMIMEType:request:frame:decisionListener:]):
2919
2920 2012-01-18  Ojan Vafai  <ojan@chromium.org>
2921
2922         Fix tests from http://trac.webkit.org/changeset/105332.
2923
2924         The code assumed that if you passed an options object with a configuration
2925         attribute that it would have a valid value. A bunch of the testing mocks
2926         do things like having a configuration value of None.
2927
2928         * Scripts/webkitpy/layout_tests/port/base.py:
2929         (Port.__init__):
2930
2931 2012-01-18  Ojan Vafai  <ojan@chromium.org>
2932
2933         Shave 0.5 seconds off check-webkit-style runtime for test_expectations.txt
2934         https://bugs.webkit.org/show_bug.cgi?id=76576
2935
2936         Reviewed by Adam Barth.
2937
2938         Avoid calling default_configuration in base.py. We don't need it for
2939         checking test_expectations.txt style. It takes ~0.5 seconds to run on my
2940         Mac Pro. It's the call to "perl Tools/Scripts/webkit-build-directory --top-level"
2941         from common.executive.
2942
2943         At some point someone should probably look into why that call is so slow
2944         since it's on the critical path for run-webkit-tests and build-webkit startup.
2945
2946         * Scripts/webkitpy/layout_tests/port/base.py:
2947         (Port.__init__):
2948         * Scripts/webkitpy/style/checkers/test_expectations.py:
2949         (TestExpectationsChecker._determine_port_from_exepectations_path):
2950
2951 2012-01-18  Dirk Pranke  <dpranke@chromium.org>
2952
2953         [chromium] move Tools.gyp, switch build-webkit --chromium to All.gyp
2954         https://bugs.webkit.org/show_bug.cgi?id=76505
2955
2956         Reviewed by Tony Chang.
2957
2958         Update chromium build script to use All.gyp instead of WebKit.gyp
2959
2960         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp: Renamed from Source/WebKit/chromium/Tools.gyp.
2961         * Scripts/webkitdirs.pm:
2962
2963 2012-01-18  Sam Weinig  <sam@webkit.org>
2964
2965         Don’t hardcode the path to ‘make‘.
2966
2967         Reviewed by Dan Bernstein.
2968
2969         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
2970
2971 2012-01-18  Joshua Bell  <jsbell@chromium.org>
2972
2973         Adding myself to committers.py
2974         https://bugs.webkit.org/show_bug.cgi?id=76569
2975
2976         Unreviewed.
2977
2978         * Scripts/webkitpy/common/config/committers.py:
2979
2980 2012-01-18  Adam Barth  <abarth@webkit.org>
2981
2982         Up the cap on test failures tolerated by the EWS so the mac-ews can run to completion
2983         https://bugs.webkit.org/show_bug.cgi?id=76567
2984
2985         Reviewed by Eric Seidel.
2986
2987         Current the AppleMac port has slightly more failures than we allow on
2988         the EWS, causing the mac-ews to spin.  This patch ups the limit a bit,
2989         which should hopefully let the mac-ews run to completion.  Once
2990         AppleMac has fewer failures, we can lower the cap again.
2991
2992         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
2993         (LayoutTestResultsReaderTest.test_layout_test_results):
2994         * Scripts/webkitpy/tool/steps/runtests.py:
2995         (RunTests):
2996         * Scripts/webkitpy/tool/steps/steps_unittest.py:
2997
2998 2012-01-18  Eric Carlson  <eric.carlson@apple.com>
2999
3000         Provide access to user's list of preferred languages
3001         https://bugs.webkit.org/show_bug.cgi?id=76138
3002
3003         Reviewed by Alexey Proskuryakov.
3004
3005         * GNUmakefile.am: Add JSInternalsCustom.cpp.
3006
3007 2012-01-18  Adam Barth  <abarth@webkit.org>
3008
3009         Add Chromium Linux GPU 32 to the list of allowed Chromium Linux
3010         configurations.
3011
3012         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
3013         (ChromiumLinuxPort.__init__):
3014
3015 2012-01-18  Dominic Mazzoni  <dmazzoni@google.com>
3016
3017         Accessibility: Chromium needs methods to scroll an object into view or to a specific location.
3018         https://bugs.webkit.org/show_bug.cgi?id=73460
3019
3020         Reviewed by Chris Fleizach.
3021
3022         * DumpRenderTree/AccessibilityUIElement.h:
3023         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
3024         (AccessibilityUIElement::AccessibilityUIElement):
3025         (AccessibilityUIElement::scrollToMakeVisibleCallback):
3026         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocusCallback):
3027         (AccessibilityUIElement::scrollToGlobalPointCallback):
3028         * DumpRenderTree/chromium/AccessibilityUIElement.h:
3029         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3030         (AccessibilityUIElement::scrollToMakeVisible):
3031         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
3032         (AccessibilityUIElement::scrollToGlobalPoint):
3033         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
3034         (AccessibilityUIElement::scrollToMakeVisible):
3035         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
3036         (AccessibilityUIElement::scrollToGlobalPoint):
3037         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
3038         (AccessibilityUIElement::scrollToMakeVisible):
3039         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
3040         (AccessibilityUIElement::scrollToGlobalPoint):
3041
3042 2012-01-17  Enrica Casucci  <enrica@apple.com>
3043
3044         Missing NSForegroundColorAttributeName should be treated as black instead of transparent.
3045         https://bugs.webkit.org/show_bug.cgi?id=76490
3046         <rdar://problem/9460733>
3047
3048         Added TestWebKitAPI test.
3049         
3050         Reviewed by Dan Bernstein.
3051
3052         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added new test file.
3053         * TestWebKitAPI/Tests/mac/InspectorBar.mm: Added test that calls
3054         changeAttributes to the WebView.
3055         
3056 2012-01-18  Balazs Kelemen  <kbalazs@webkit.org>
3057
3058         [Qt] Consolidate layout test crash logging
3059         https://bugs.webkit.org/show_bug.cgi?id=75088
3060
3061         Reviewed by Simon Hausmann.
3062
3063         Move backtrace generating logic into WTFReportBacktrace
3064         and add a way to deinstall signal handlers if we know
3065         that we have already printed the backtrace.
3066
3067         * DumpRenderTree/qt/main.cpp:
3068         (crashHandler):
3069         (setupSignalHandlers):
3070         (WTFCrashHook):
3071         (main):
3072         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
3073         (WTR::crashHandler):
3074         (WTR::setupSignalHandlers):
3075         (WTR::crashHook):
3076         (WTR::InjectedBundle::platformInitialize):
3077
3078 2012-01-18  Ilya Tikhonovsky  <loislo@chromium.org>
3079
3080         Unreviewed build fix after r105256.
3081
3082         There was a cyclic dependency between self._port and self._host assigments.
3083
3084         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3085         (PerfTestsRunner.__init__):
3086
3087 2012-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3088
3089         Add new watchlist rule for the Qt build system
3090
3091         Reviewed by Simon Hausmann.
3092
3093         * Scripts/webkitpy/common/config/watchlist:
3094
3095 2012-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3096
3097         [Qt] Move OTHER_FILES from WebKit.pro to Tools.pro
3098
3099         Reviewed by Simon Hausmann.
3100
3101         * Tools.pro:
3102
3103 2012-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3104
3105         [Qt] Allow building only a subset of the WebKit sub-projects
3106
3107         Fixes build-jsc, broken in r104825.
3108
3109         We no longer need the intermediate buildQMakeQtProject(), as we're the
3110         only port using qmake, and buildQMakeProject is now highly Qt specific.
3111
3112         <http://webkit.org/b/76179>
3113
3114         Reviewed by Simon Hausmann.
3115
3116         * Scripts/build-jsc:
3117         * Scripts/build-webkit:
3118         * Scripts/webkitdirs.pm:
3119         (buildQMakeProjects):
3120
3121 2012-01-18  Ryosuke Niwa  <rniwa@webkit.org>
3122
3123         run-perf-tests should generate a json file that summaries the result
3124         https://bugs.webkit.org/show_bug.cgi?id=76504
3125
3126         Reviewed by Adam Barth.
3127
3128         Add an ability to generate a json file to run-perf-test in the preparation for perf bots.
3129         New option --outout-json-path specifies the json file's path, and --source-json-path specifies
3130         another json file to be merged into the generated json file.
3131
3132         Also fixed a bug that --build wasn't set by default.
3133
3134         * Scripts/webkitpy/layout_tests/port/test.py:
3135         (TestPort.webkit_base):
3136         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3137         (PerfTestsRunner.__init__):
3138         (PerfTestsRunner):
3139         (PerfTestsRunner._parse_args):
3140         (PerfTestsRunner.run):
3141         (PerfTestsRunner._generate_json_if_specified):
3142         (PerfTestsRunner._process_chromium_style_test_result):
3143         (PerfTestsRunner._process_parser_test_result):
3144         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
3145         (MainTest):
3146         (create_runner):
3147         (test_run_test_set):
3148         (test_run_test_set_for_parser_tests):
3149         (test_run_test_set_with_summary_json):
3150         (test_collect_tests):
3151         (test_parse_args):
3152
3153 2012-01-17  Sheriff Bot  <webkit.review.bot@gmail.com>
3154
3155         Unreviewed, rolling out r105244.
3156         http://trac.webkit.org/changeset/105244
3157         https://bugs.webkit.org/show_bug.cgi?id=76518
3158
3159         broke Chromium Mac (Requested by rolandsteiner on #webkit).
3160
3161         * DumpRenderTree/AccessibilityUIElement.h:
3162         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
3163         (AccessibilityUIElement::AccessibilityUIElement):
3164         * DumpRenderTree/chromium/AccessibilityUIElement.h:
3165         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3166         (AccessibilityUIElement::removeSelection):
3167         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
3168         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
3169         (AccessibilityUIElement::removeSelection):
3170
3171 2012-01-17  Dominic Mazzoni  <dmazzoni@google.com>
3172
3173         Accessibility: Chromium needs methods to scroll an object into view or to a specific location.
3174         https://bugs.webkit.org/show_bug.cgi?id=73460
3175
3176         Reviewed by Chris Fleizach.
3177
3178         * DumpRenderTree/AccessibilityUIElement.h:
3179         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
3180         (AccessibilityUIElement::AccessibilityUIElement):
3181         (AccessibilityUIElement::scrollToMakeVisibleCallback):
3182         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocusCallback):
3183         (AccessibilityUIElement::scrollToGlobalPointCallback):
3184         * DumpRenderTree/chromium/AccessibilityUIElement.h:
3185         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3186         (AccessibilityUIElement::scrollToMakeVisible):
3187         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
3188         (AccessibilityUIElement::scrollToGlobalPoint):
3189         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
3190         (AccessibilityUIElement::scrollToMakeVisible):
3191         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
3192         (AccessibilityUIElement::scrollToGlobalPoint):
3193         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
3194         (AccessibilityUIElement::scrollToMakeVisible):
3195         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
3196         (AccessibilityUIElement::scrollToGlobalPoint):
3197
3198 2012-01-17  Adam Barth  <abarth@webkit.org>
3199
3200         Enable tests on the mac-ews
3201         https://bugs.webkit.org/show_bug.cgi?id=76502
3202
3203         Reviewed by Simon Fraser.
3204
3205         Apple now has five machines running the mac-ews.  Lucas Forschler has
3206         asked that we try enabling tests to see if we have enough bandwidth.
3207
3208         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
3209         (MacEWS):
3210
3211 2012-01-17  Sanjoy Das  <sanjoy@playingwithpointers.com>
3212
3213         Add a check-webkit-style exception for GDBInterface.cpp
3214         https://bugs.webkit.org/show_bug.cgi?id=76187
3215
3216         Add an exception for Source/JavaScriptCore/jit/GDBInterface.cpp so
3217         that it does not fail the style check.
3218
3219         Reviewed by Martin Robinson.
3220
3221         * Scripts/webkitpy/style/checker.py: Tell the style checker to not run
3222         readability/naming tests on
3223         Source/JavaScriptCore/jit/GDBInterface.cpp.
3224         * Scripts/webkitpy/style/checker_unittest.py:
3225         (GlobalVariablesTest.test_path_rules_specifier): Add a test for
3226         the above exception.
3227
3228 2012-01-17  Hajime Morrita  <morrita@chromium.org>
3229
3230         https://bugs.webkit.org/show_bug.cgi?id=76245
3231         [Mac][DRT] should allow query string for the file URL.
3232
3233         Reviewed by Tony Chang.
3234
3235         Taught DRT about file:// as a url instead of a file path.
3236         We are now able to use a path file:///foo/bar.html?baz as a url
3237         with which DRT opens "/foo/bar.html" with "baz" as a query string.
3238
3239         * DumpRenderTree/mac/DumpRenderTree.mm:
3240         (runTest):
3241
3242 2012-01-17  Ojan Vafai  <ojan@chromium.org>
3243
3244         Check style on test_expectations.txt files before commit
3245         https://bugs.webkit.org/show_bug.cgi?id=76484
3246
3247         Reviewed by Adam Barth.
3248
3249         Unlike other style rules, if you get the syntax of the test_expectations.txt
3250         file wrong, the layout tests won't run. Also, this check is simple and only
3251         slows down committing if you actually modify one of the test_expectations.txt files.
3252
3253         * Scripts/webkitpy/tool/steps/commit.py:
3254         (Commit._check_test_expectations):
3255         (Commit.run):
3256         * Scripts/webkitpy/tool/steps/commit_unittest.py: Added.
3257         (CommitTest):
3258         (CommitTest.test_check_test_expectations):
3259
3260 2012-01-17  Viatcheslav Ostapenko  <ostapenko.viatcheslav@nokia.com>
3261
3262         [Qt] Debug build fails with debug qt5
3263         https://bugs.webkit.org/show_bug.cgi?id=76463
3264
3265         Reviewed by Simon Hausmann.
3266
3267         Add QNetworkCookie include in order to satisfy sizeof in QTypeInfo in debug builds.
3268
3269         * QtTestBrowser/cookiejar.h:
3270
3271 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
3272
3273         webkitpy: fix json import on linux
3274         https://bugs.webkit.org/show_bug.cgi?id=76481
3275
3276         Reviewed by Adam Barth.
3277
3278         For some reason, the version of simplejson autoinstalled on my linux
3279         box seems to be broken. We should really only be using
3280         simplejson if json isn't available anyway, so this change looks
3281         for json first.
3282
3283         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
3284         (JSONGeneratorTest.test_test_timings_trie):
3285
3286 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
3287
3288         Fix failures in test-webkitpy caused by r105177.
3289
3290         Unreviewed, build fix.
3291
3292         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
3293         (ChromiumLinuxPort._determine_driver_path_statically):
3294         (ChromiumLinuxPort._static_build_path):
3295
3296 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
3297
3298         webkitpy: make PortFactory.get() be fully data-driven
3299         https://bugs.webkit.org/show_bug.cgi?id=76360
3300
3301         Reviewed by Eric Seidel.
3302
3303         Refactor PortFactory.get() to be fully data-driven.
3304
3305         * Scripts/webkitpy/layout_tests/port/factory.py:
3306         (PortFactory):
3307         (PortFactory._default_port):
3308         (PortFactory.get):
3309         (PortFactory.get.in):
3310         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
3311         (FactoryTest.assert_port):
3312
3313 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
3314
3315         webkitpy: add determine_full_port_name(), clean up port.__init__()
3316         https://bugs.webkit.org/show_bug.cgi?id=76357
3317
3318         Reviewed by Adam Barth.
3319
3320         Each Port class is now required to implement a 'port_name' field
3321         that will match the start of all ports constructed by this class and
3322         a 'determine_full_port_name' field that will make the port
3323         fully-qualified.
3324
3325         * Scripts/webkitpy/layout_tests/port/apple.py:
3326         (ApplePort.__init__):
3327         * Scripts/webkitpy/layout_tests/port/base.py:
3328         (Port):
3329         (Port.determine_port_name):
3330         (Port.__init__):
3331         * Scripts/webkitpy/layout_tests/port/chromium.py:
3332         (ChromiumPort.for):
3333         (ChromiumPort.__init__):
3334         (ChromiumPort):
3335         (ChromiumPort._chromium_base_dir):
3336         (ChromiumPort.path_from_chromium_base):
3337         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
3338         (ChromiumGpuLinuxPort):
3339         (ChromiumGpuLinuxPort.__init__):
3340         (ChromiumGpuMacPort):
3341         (ChromiumGpuMacPort.__init__):
3342         (ChromiumGpuWinPort):
3343         (ChromiumGpuWinPort.__init__):
3344         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
3345         (ChromiumLinuxPort):
3346         (ChromiumLinuxPort.determine_port_name):
3347         (ChromiumLinuxPort.__init__):
3348         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
3349         (ChromiumMacPort):
3350         (ChromiumMacPort.determine_port_name):
3351         (ChromiumMacPort.__init__):
3352         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
3353         (ChromiumWinPort):
3354         (ChromiumWinPort.determine_port_name):
3355         (ChromiumWinPort.__init__):
3356         * Scripts/webkitpy/layout_tests/port/dryrun.py:
3357         (DryRunPort):
3358         (DryRunPort.determine_port_name):
3359         (DryRunPort.__init__):
3360         * Scripts/webkitpy/layout_tests/port/efl.py:
3361         (EflPort):
3362         * Scripts/webkitpy/layout_tests/port/factory.py:
3363         (PortFactory):
3364         (PortFactory._default_port):
3365         (PortFactory.get):
3366         (PortFactory.get.in):
3367         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
3368         (FactoryTest.assert_port):
3369         * Scripts/webkitpy/layout_tests/port/google_chrome.py:
3370         (GoogleChromeLinux32Port):
3371         (GoogleChromeLinux32Port.determine_port_name):
3372         (GoogleChromeLinux64Port):
3373         (GoogleChromeLinux64Port.determine_port_name):
3374         (GoogleChromeMacPort):
3375         (GoogleChromeMacPort.determine_port_name):
3376         (GoogleChromeWinPort):
3377         (GoogleChromeWinPort.determine_port_name):
3378         * Scripts/webkitpy/layout_tests/port/gtk.py:
3379         (GtkPort):
3380         * Scripts/webkitpy/layout_tests/port/mac.py:
3381         (MacPort):
3382         (MacPort.determine_port_name):
3383         (MacPort.__init__):
3384         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3385         (test_versions):
3386         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
3387         (MockDRTPort):
3388         (MockDRTPort.determine_port_name):
3389         (MockDRTPort.__init__):
3390         (MockDRT):
3391         (MockDRT.determine_port_name):
3392         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
3393         (PortTestCase.make_port):
3394         * Scripts/webkitpy/layout_tests/port/qt.py:
3395         (QtPort):
3396         (QtPort.determine_port_name):
3397         (QtPort.__init__):
3398         * Scripts/webkitpy/layout_tests/port/test.py:
3399         (TestPort):
3400         (TestPort.determine_port_name):
3401         (TestPort.__init__):
3402         * Scripts/webkitpy/layout_tests/port/webkit.py:
3403         (WebKitPort.__init__):
3404         * Scripts/webkitpy/layout_tests/port/win.py:
3405         (WinPort):
3406         (WinPort.determine_port_name):
3407         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
3408         (get_test_baselines.AllPlatformsPort.__init__):
3409         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
3410         (get_test_config.TestMacPort):
3411         (get_test_config):
3412
3413 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
3414
3415         webkitpy: clean up port code in preparation for static port names
3416         https://bugs.webkit.org/show_bug.cgi?id=76356
3417
3418         Reviewed by Adam Barth.
3419
3420         To fix bug 76215 and be able to determine appropriate port names
3421         in webkitpy without actually constructing Port objects, we need
3422         to shuffle some logic in the chromium ports to be able to figure
3423         out whether we built DRT in 32 or 64-bit mode.
3424
3425         * Scripts/webkitpy/layout_tests/port/chromium.py:
3426         (ChromiumPort):
3427         (ChromiumPort._chromium_base_dir):
3428         (ChromiumPort.__init__):
3429         (ChromiumPort.path_from_chromium_base):
3430         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
3431         (ChromiumLinuxPort):
3432         (ChromiumLinuxPort._determine_driver_path_statically):
3433         (ChromiumLinuxPort._static_build_path):
3434         (ChromiumLinuxPort._determine_architecture):
3435         (ChromiumLinuxPort.__init__):
3436         (ChromiumLinuxPort._build_path):
3437
3438 2012-01-17  Hao Zheng  <zhenghao@chromium.org>
3439
3440         Do not remove pixel failures for ref tests.
3441         https://bugs.webkit.org/show_bug.cgi?id=76243
3442
3443         Reviewed by Tony Chang.
3444
3445         'NRWT --no-pixel-tests' complains when reftests are expected to
3446         be image mismatch.
3447
3448         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3449         (Manager._update_summary_with_result):
3450         * Scripts/webkitpy/layout_tests/models/test_failures.py:
3451         (is_reftest_failure):
3452         (determine_result_type):
3453
3454 2012-01-17  Balazs Ankes  <Ankes.Balazs@stud.u-szeged.hu>
3455
3456         REGRESSION(r81225): ORWT should ignore reftests
3457         https://bugs.webkit.org/show_bug.cgi?id=67936
3458
3459         Reviewed by Csaba Osztrogonác.
3460
3461         * Scripts/old-run-webkit-tests:
3462         (isUsedInReftest): $filename should be the first parameter
3463         (findTestsToRun): run test if it isn't reftest
3464
3465 2012-01-17  Gustavo Noronha Silva  <gns@gnome.org>
3466
3467         Unreviewed. Fix python unit test I broke with r105142
3468
3469         * Scripts/webkitpy/common/config/ports_unittest.py:
3470         (WebKitPortTest.test_gtk_port):
3471
3472 2012-01-17  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3473
3474         [Qt] Don't set the 'primary' flag when mocking touch-points in MiniBrowser
3475
3476         Qt 5 no longer has that flag.
3477
3478         Reviewed by Simon Hausmann.
3479
3480         * MiniBrowser/qt/MiniBrowserApplication.cpp:
3481         (MiniBrowserApplication::notify):
3482
3483 2012-01-16  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
3484
3485         GTK+ EWS needs to run update-webkitgtk-libs after applying a patch
3486         https://bugs.webkit.org/show_bug.cgi?id=75857
3487
3488         Reviewed by Adam Barth.
3489
3490         * Scripts/webkitdirs.pm:
3491         (buildAutotoolsProject): move cleaning the build to happen early,
3492         and handle --update-gtk right after that
3493         * Scripts/webkitpy/common/config/ports.py:
3494         (GtkPort.build_webkit_command): pass --update-gtk as an option to
3495         build-webkit, so that it is used in the EWS
3496
3497 2012-01-16  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3498
3499         [Qt] Don't add _debug postfix to targets when building debug only
3500
3501         Reviewed by Simon Hausmann.
3502
3503         * qmake/mkspecs/features/default_post.prf:
3504         * qmake/mkspecs/features/functions.prf:
3505
3506 2012-01-16  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3507
3508         [Qt] Add more files to project-files for easier navigation in Qt Creator
3509
3510         Reviewed by Simon Hausmann.
3511
3512         * Tools.pro:
3513         * qmake/config.tests/README: Added.
3514
3515 2012-01-17  Philippe Normand  <pnormand@igalia.com>
3516
3517         [GTK] fast/dom/constructed-objects-prototypes.html fails
3518         https://bugs.webkit.org/show_bug.cgi?id=75432
3519
3520         Reviewed by Martin Robinson.
3521
3522         * Scripts/build-webkit: Enable video-track in GTK.
3523
3524 2012-01-16  Takashi Toyoshima  <toyoshim@chromium.org>
3525
3526         Add toyoshim as a committer.
3527         https://bugs.webkit.org/show_bug.cgi?id=76390
3528
3529         Reviewed by Hajime Morita.
3530
3531         * Scripts/webkitpy/common/config/committers.py:
3532
3533 2012-01-16  Brent Fulgham  <bfulgham@webkit.org>
3534
3535         [WinCairo] Build fix after r104155. Several debug targets were not
3536         properly including the '*Common' property sheets.
3537
3538         * DumpRenderTree/win/DumpRenderTreeLauncherDebugCairoCFLite.vsprops:
3539         * DumpRenderTree/win/ImageDiffLauncherDebugCairoCFLite.vsprops:
3540         * MiniBrowser/Configurations/MiniBrowserLauncherDebugCairoCFLite.vsprops:
3541         * WebKitTestRunner/win/WebKitTestRunnerLauncherDebugCairoCFLite.vsprops:
3542         * WinLauncher/WinLauncherLauncherDebugCairoCFLite.vsprops:
3543
3544 2012-01-16  Csaba Osztrogonác  <ossy@webkit.org>
3545
3546         REGRESSION(r105072): It broke committers_unittest.py
3547         https://bugs.webkit.org/show_bug.cgi?id=76396
3548
3549         Reviewed by Adam Barth.
3550
3551         * Scripts/webkitpy/common/config/committers_unittest.py:
3552         (CommittersTest._assert_fuzz_match): It should handle more names.
3553         (CommittersTest.test_contributors_by_fuzzy_match_with_legacy_names): Fix the testcase.
3554
3555 2012-01-16  Mario Sanchez Prada  <msanchez@igalia.com>
3556
3557         [GTK] Include gtk+3 in the jhbuild modules
3558         https://bugs.webkit.org/show_bug.cgi?id=76284
3559
3560         Reviewed by Philippe Normand.
3561
3562         * gtk/jhbuild.modules: Add some gtk+3 to the jhbuild moduleset.
3563
3564 2012-01-16  Dan Winship  <danw@gnome.org>
3565
3566         Add myself as a committer
3567
3568         Unreviewed
3569
3570         * Scripts/webkitpy/common/config/committers.py:
3571
3572 2012-01-16  Raphael Kubo da Costa  <kubo@profusion.mobi>
3573
3574         Unreviewed; fix the build when --as-needed and --no-copy-dt-needed-entries are passed to the linker.
3575
3576         Some recent versions of a few Linux distributions have started passing
3577         --as-needed and --no-copy-dt-needed-entries by default to the linker,
3578         which broke the build when SHARED_CORE was turned on, as libwtf_efl.so
3579         needed a few symbols from libjavascriptcore_efl.so.
3580
3581         * DumpRenderTree/efl/CMakeLists.txt: Explicitly pass WTF_LIBRARY_NAME
3582         before JavaScriptCore_LIBRARY_NAME in DumpRenderTree_LIBRARIES so the
3583         libraries are searched in the right order by the linker.
3584
3585 2012-01-16  Philippe Normand  <pnormand@igalia.com>
3586
3587         [GTK] set audio resources path in NRTW
3588         https://bugs.webkit.org/show_bug.cgi?id=76380
3589
3590         Reviewed by Martin Robinson.
3591
3592         * Scripts/webkitpy/layout_tests/port/gtk.py:
3593         (GtkPort.setup_environ_for_server): Set $AUDIO_RESOURCES_PATH so
3594         the uninstalled audio resources are used during webaudio tests.
3595
3596 2012-01-16  Carlos Garcia Campos  <cgarcia@igalia.com>
3597
3598         [GTK] ldflags and cflags should take precedence over the existing flags in gtkdoc.py
3599         https://bugs.webkit.org/show_bug.cgi?id=76369
3600
3601         Reviewed by Xan Lopez.
3602
3603         If there's an installed version of webkit and there are new
3604         symbols in the build tree, gtkdoc-scanobj fails due to undefined
3605         references, because the env ldflags are given before the local
3606         ones.
3607
3608         * gtk/gtkdoc.py:
3609         (GTKDoc._run_gtkdoc_scangobj):
3610
3611 2012-01-15  Pablo Flouret  <pablof@motorola.com>
3612
3613         Fix compilation errors on build-webkit --debug --no-video on mac.
3614         https://bugs.webkit.org/show_bug.cgi?id=75867
3615
3616         Reviewed by Philippe Normand.
3617
3618         Add --[no-]video-track to build-webkit, since video track feature
3619         depends on video.
3620
3621         * Scripts/build-webkit:
3622
3623 2012-01-14  James Robinson  <jamesr@chromium.org>
3624
3625         [chromium] Failing webkit_unit_tests does not make the cr-linux EWS bubble red
3626         https://bugs.webkit.org/show_bug.cgi?id=76313
3627
3628         Reviewed by Adam Barth.
3629
3630         The exit code of this script should be the exit code of webkit_unit_tests. Tested manually.
3631
3632         * Scripts/run-chromium-webkit-unit-tests:
3633
3634 2012-01-14  Sheriff Bot  <webkit.review.bot@gmail.com>
3635
3636         Unreviewed, rolling out r105017.
3637         http://trac.webkit.org/changeset/105017
3638         https://bugs.webkit.org/show_bug.cgi?id=76333
3639
3640         broke the build (Requested by philn-tp on #webkit).
3641
3642         * gtk/jhbuild.modules:
3643
3644 2012-01-14  Mario Sanchez Prada  <msanchez@igalia.com>
3645
3646         [GTK] Include gtk+3 in the jhbuild modules
3647         https://bugs.webkit.org/show_bug.cgi?id=76284
3648
3649         Reviewed by Philippe Normand.
3650
3651         * gtk/jhbuild.modules: Add some gtk+3 to the jhbuild moduleset.
3652
3653 2012-01-13  Adam Barth  <abarth@webkit.org>
3654
3655         Remove --dry-run support from webkit-patch
3656         https://bugs.webkit.org/show_bug.cgi?id=76300
3657
3658         Reviewed by Dirk Pranke.
3659
3660         This is the rest of the patch from Bug 76300.  I landed it separtely
3661         because main.py had some stray conflict markers that I needed to clean
3662         up first.
3663
3664         * Scripts/webkitpy/test/main.py:
3665         (Tester.parse_args):
3666         (Tester._run_tests):
3667
3668 2012-01-13  Adam Barth  <abarth@webkit.org>
3669
3670         test-webkitpy does not run
3671         https://bugs.webkit.org/show_bug.cgi?id=76318
3672
3673         Unreviewed.
3674
3675         Looks like some conflict markers got left in the file.
3676
3677         * Scripts/webkitpy/test/main.py:
3678         (Tester.run):
3679
3680 2012-01-13  Adam Barth  <abarth@webkit.org>
3681
3682         Remove --dry-run support from webkit-patch
3683         https://bugs.webkit.org/show_bug.cgi?id=76300
3684
3685         Reviewed by Dirk Pranke.
3686
3687         We don't use this functionality for anything, and it doesn't really
3688         work.  Someone in #webkit tried using it and and got sad/confused.  We
3689         should just remove it.
3690