Bump up the timeout for performance tests since DOM/DOMTable.html
[WebKit-https.git] / Tools / ChangeLog
1 2012-01-31  Ryosuke Niwa  <rniwa@webkit.org>
2
3         Bump up the timeout for performance tests since DOM/DOMTable.html
4         and Parser/html5-full-render.html are timing out on the bot.
5
6         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
7         (PerfTestsRunner._parse_args):
8
9 2012-01-31  Ehsan Akhgari  <ehsan.akhgari@gmail.com>
10
11         Fix the update_webgl_conformance_tests.py script in order to enable it
12         to detect the SCM being used correctly.
13         https://bugs.webkit.org/show_bug.cgi?id=77462
14
15         Reviewed by Kenneth Russell.
16
17         * Scripts/webkitpy/to_be_moved/update_webgl_conformance_tests.py:
18         (default_out_dir):
19
20 2012-01-31  Eric Seidel  <eric@webkit.org>
21
22         webkit-patch apply-from-bug should work regardless of your CWD
23         https://bugs.webkit.org/show_bug.cgi?id=77482
24
25         Reviewed by Adam Barth.
26
27         webkit-patch only ever works with patches made from the root
28         directory.  We should pass the CWD to svn-apply in the case
29         where we're running not from the webkit root.  I believe
30         this used to work in the past and regressed.
31
32         * Scripts/webkitpy/common/checkout/checkout.py:
33         (Checkout.apply_patch):
34         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
35         (CheckoutTest.test_apply_patch):
36
37 2012-01-31  Sheriff Bot  <webkit.review.bot@gmail.com>
38
39         Unreviewed, rolling out r106302.
40         http://trac.webkit.org/changeset/106302
41         https://bugs.webkit.org/show_bug.cgi?id=77472
42
43         Broke building DRT before running the tests (Requested by ojan
44         on #webkit).
45
46         * Scripts/webkit-build-directory:
47         * Scripts/webkitpy/layout_tests/port/config.py:
48         (Config.build_directory):
49         * Scripts/webkitpy/layout_tests/port/config_standalone.py:
50         (main):
51         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
52         (ConfigTest.assert_configuration):
53         (ConfigTest.test_build_directory.mock_webkit_build_directory):
54         (ConfigTest.test_build_directory):
55         (ConfigTest.test_default_configuration__notfound):
56         * Scripts/webkitpy/layout_tests/port/webkit.py:
57         (WebKitPort._build_path):
58
59 2012-01-31  Antoine Labour  <piman@chromium.org>
60
61         Merge WebGraphicsContext3D creation and initialization, and move it to
62         WebViewClient.
63         https://bugs.webkit.org/show_bug.cgi?id=76593
64
65         Reviewed by Darin Fisher.
66
67         * DumpRenderTree/chromium/TestWebPlugin.cpp:
68         (TestWebPlugin::TestWebPlugin):
69         (TestWebPlugin::initialize):
70         * DumpRenderTree/chromium/TestWebPlugin.h:
71         * DumpRenderTree/chromium/WebViewHost.cpp:
72         (WebViewHost::createGraphicsContext3D):
73         (WebViewHost::createPlugin):
74         * DumpRenderTree/chromium/WebViewHost.h:
75
76 2012-01-31  Gabor Rapcsanyi  <rgabor@webkit.org>
77
78         [GTK] build-webkit warning.
79         https://bugs.webkit.org/show_bug.cgi?id=77411
80
81         Reviewed by Philippe Normand.
82
83         * Scripts/webkitdirs.pm:
84         (isCrossCompilation):
85
86 2012-01-31  Dan Bernstein  <mitz@apple.com>
87
88         When generating derived sources, use the same compiler that is used to compile WebKitTestRunner.
89
90         Reviewed by Anders Carlsson.
91
92         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
93
94 2012-01-31  Csaba Osztrogonác  <ossy@webkit.org>
95
96         If the QTDIR environment variable is set, the build system fails to compile WinCairo
97         https://bugs.webkit.org/show_bug.cgi?id=77112
98
99         Reviewed by Daniel Bates.
100
101         * Scripts/webkitdirs.pm:
102         (determineIsQt):
103
104 2012-01-31  Adam Roben  <aroben@apple.com>
105
106         Only call -typingAttributes on WebViews, not WebHTMLViews
107
108         Looks like this case was just missed in r105908.
109
110         Fixes <http://webkit.org/b/77432> REGRESSION (r105908): WebKit1.InspectorBarTest is crashing
111
112         Reviewed by Antti Koivisto.
113
114         * TestWebKitAPI/Tests/mac/InspectorBar.mm:
115         (TestWebKitAPI::TEST): Call -typingAttributes on the WebView, just like we do 2 lines
116         earlier.
117
118 2012-01-31  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
119
120         [Qt] MiniBrowser: Show touch mocking indicators only when Ctrl is held down.
121         https://bugs.webkit.org/show_bug.cgi?id=77221
122
123         Reviewed by Kenneth Rohde Christiansen.
124
125         This makes sure that the indicator isn't in the way when doing clicks or dealing with the rest of the UI.
126
127         * MiniBrowser/qt/MiniBrowserApplication.cpp:
128         (MiniBrowserApplication::sendTouchEvent):
129
130 2012-01-26  Hans Wennborg  <hans@chromium.org>
131
132         Speech Input: move MockSpeechInputClient into Chromium DumpRenderTree implementation
133         https://bugs.webkit.org/show_bug.cgi?id=77083
134
135         Reviewed by Darin Fisher.
136
137         Add MockWebSpeechInputController which provides a mock implementation
138         of the WebSpeechInputController interface, and use that in
139         DumpRenderTree.
140
141         * DumpRenderTree/DumpRenderTree.gypi:
142         * DumpRenderTree/chromium/LayoutTestController.cpp:
143         (LayoutTestController::addMockSpeechInputResult):
144         * DumpRenderTree/chromium/LayoutTestController.h:
145         ():
146         * DumpRenderTree/chromium/MockWebSpeechInputController.cpp: Added.
147         (MockWebSpeechInputController::create):
148         (MockWebSpeechInputController::addMockRecognitionResult):
149         (MockWebSpeechInputController::clearResults):
150         (MockWebSpeechInputController::startRecognition):
151         (MockWebSpeechInputController::cancelRecognition):
152         (MockWebSpeechInputController::stopRecording):
153         (MockWebSpeechInputController::MockWebSpeechInputController):
154         (MockWebSpeechInputController::speechTaskFired):
155         (MockWebSpeechInputController::SpeechTask::SpeechTask):
156         (MockWebSpeechInputController::SpeechTask::stop):
157         (MockWebSpeechInputController::SpeechTask::runIfValid):
158         * DumpRenderTree/chromium/MockWebSpeechInputController.h: Added.
159         (WebKit):
160         (MockWebSpeechInputController):
161         (MockWebSpeechInputController::taskList):
162         (SpeechTask):
163         * DumpRenderTree/chromium/TestShell.cpp:
164         * DumpRenderTree/chromium/WebViewHost.cpp:
165         (WebViewHost::speechInputController):
166         * DumpRenderTree/chromium/WebViewHost.h:
167         (WebKit):
168         (WebViewHost::speechInputControllerMock):
169         (WebViewHost):
170         ():
171
172 2012-01-31  Hajime Morrita  <morrita@chromium.org>
173
174         [PerformanceTests] Add landing html for Dromaeo dom-query test
175         https://bugs.webkit.org/show_bug.cgi?id=77329
176
177         Reviewed by Ryosuke Niwa.
178
179         Added some more ignorable output patterns to allow warnings
180         from a third party test suite.
181
182         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
183         (PerfTestsRunner):
184         (PerfTestsRunner._should_ignore_line_in_parser_test_result):
185
186 2012-01-30  Kevin Ollivier  <kevino@theolliviers.com>
187
188         [wx] Unreviewed. Update download URLs to new domain.
189         
190         * waf/build/build_utils.py:
191         (update_wx_deps):
192
193 2012-01-30  Kevin Ollivier  <kevino@theolliviers.com>
194
195         [wx] Unreviewed. Build fix, add JavaScriptCore/tools
196         to the list of build directories.
197         
198         * waf/build/settings.py:
199
200 2012-01-30  Kentaro Hara  <haraken@chromium.org>
201
202         REGRESSION(r105797): prepare-ChangeLog for a .cpp file can
203         output an empty method name (i.e. "()")
204         https://bugs.webkit.org/show_bug.cgi?id=77336
205
206         Reviewed by Darin Adler.
207
208         r105797 tried to detect a change outside methods, but it causes a bug that
209         prepare-ChangeLog can output an empty method name, like this:
210
211             * foo/bar/baz.cpp:
212             (method1):
213             ():
214             (method2):
215
216         This is because the cpp parser in prepare-ChangeLog cannot distinguish
217         '{' as the beginning of a method with '{' as the beginning of an array definition
218         at the top level.
219
220             int a[] = { 1, 2, 3 };  // This '{' is the beginning of an array definition.
221
222             void func() { // This '{' is the beginning of a method.
223                 ...;
224             }
225
226         This patch fixes prepare-ChangeLog so that it skips an array definition at the top level.
227
228         * Scripts/prepare-ChangeLog:
229         (get_function_line_ranges_for_cpp): Modified as described above.
230         (generateFunctionLists): As a hack, modified so that prepare-ChangeLog does not output
231         an empty method name. Ideally this should not happen but may happen, since the
232         parsers are not perfect.
233         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests.cpp: Added test cases.
234         (NameSpace7):
235         (NameSpace8):
236         (Class109):
237         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests-expected.txt:
238
239 2012-01-30  Ojan Vafai  <ojan@chromium.org>
240
241         run-webkit-tests calls out to webkit-build-directory twice
242         https://bugs.webkit.org/show_bug.cgi?id=77248
243
244         Reviewed by Dirk Pranke.
245
246         This reduces to calls to webkit-build-directory to one call and
247         saves 500ms for run-webkit-tests over a single file.
248
249         * Scripts/webkit-build-directory:
250         If called with no arguments, print out both the top-level directory and the 
251         configuration directory.
252
253         * Scripts/webkit-build-directory:
254         * Scripts/webkitpy/layout_tests/port/config.py:
255         (Config.build_directory):
256         When called with no arguments, get and cache both the top-level and configuration directories.
257
258         * Scripts/webkitpy/layout_tests/port/config_standalone.py:
259         (main):
260         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
261         (ConfigTest.assert_configuration):
262         (ConfigTest.test_build_directory.mock_webkit_build_directory):
263         (ConfigTest.test_build_directory):
264         (ConfigTest.test_default_configuration__notfound):
265         * Scripts/webkitpy/layout_tests/port/webkit.py:
266         (WebKitPort._build_path):
267         Cache the root directory in the options object so that other users
268         of the options object don't need to call out to webkit-build-directory.
269
270 2012-01-30  Ojan Vafai  <ojan@chromium.org>
271
272         Parsing test_expecations.txt + Skipped lists takes too long
273         https://bugs.webkit.org/show_bug.cgi?id=77059
274
275         Reviewed by Dirk Pranke.
276
277         This saves ~100ms on the Apple Mac port.
278         -memoize a bunch of path methods.
279         -Avoid doing multiple disk accesses per line.
280         -Parse the skipped list directly instead of turning it into a test_expecations.txt
281         formatting string and parsing that.
282
283         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
284         (TestExpectationParser):
285         (TestExpectationParser.expectation_for_skipped_test):
286         (TestExpectationParser._parse_line):
287         (TestExpectationParser._collect_matching_tests):
288         (TestExpectations.__init__):
289         (TestExpectations._add_skipped_tests):
290         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
291         (test_add_skipped_tests):
292         (test_add_skipped_tests_duplicate):
293         * Scripts/webkitpy/layout_tests/port/base.py:
294         (Port):
295         (Port.test_isfile):
296         (Port.normalize_test_name):
297         (Port.layout_tests_dir):
298         (Port.abspath_for_test):
299         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
300
301 2012-01-30  Alexis Menard  <alexis.menard@openbossa.org>
302
303         Unreviewed. Add myself to CSS, GStreamer, Qt related watchlists.
304         
305         * Scripts/webkitpy/common/config/watchlist:
306
307 2012-01-30  Gavin Barraclough  <barraclough@apple.com>
308
309         Unreviewed build fix following bug#76855
310
311         * TestWebKitAPI/Tests/WTF/MetaAllocator.cpp:
312         ():
313         (TestWebKitAPI::TEST_F):
314
315 2012-01-26  No'am Rosenthal  <noam.rosenthal@nokia.com> and Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
316
317         [Qt] WKTR: Use a software rendering pipiline when running tests.
318         https://bugs.webkit.org/show_bug.cgi?id=76708
319
320         Reviewed by Kenneth Rohde Christiansen.
321
322         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
323         (WTR::WrapperWindow::WrapperWindow):
324         (WTR::PlatformWebView::PlatformWebView): Use software rendering of layers since the wrapping QQuickView isn't shown.
325
326 2012-01-30  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
327
328         [Qt] Store build config immediately instead of waiting for a successfull build
329
330         Otherwise we will not pick up changes to the config if the build is aborted,
331         for example due to build errors or if the user pressed Ctrl+Z. We now write
332         the new config before starting a build, but still defer to write the SVN
333         revision once the build completes successfully.
334
335         Reviewed by Ossy.
336
337         * Scripts/webkitdirs.pm:
338         (buildQMakeProjects):
339
340 2012-01-30  Alexander Færøy  <ahf@0x90.dk>
341
342         [Qt] Add "New Window" button to the Qt MiniBrowser
343         https://bugs.webkit.org/show_bug.cgi?id=77281
344
345         Reviewed by Tor Arne Vestbø.
346
347         * MiniBrowser/qt/BrowserWindow.cpp:
348         (BrowserWindow::BrowserWindow):
349         * MiniBrowser/qt/MiniBrowser.qrc:
350         * MiniBrowser/qt/icons/plus.png: Added.
351         * MiniBrowser/qt/qml/BrowserWindow.qml:
352
353 2012-01-30  Csaba Osztrogonác  <ossy@webkit.org>
354
355         Unreviewed rolling out r106222.
356
357         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
358         (SingleTestRunner.__init__):
359         * Scripts/webkitpy/layout_tests/controllers/worker.py:
360         (Worker.handle_test_list):
361         * Scripts/webkitpy/layout_tests/models/test_input.py:
362         (TestInput.__init__):
363         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
364         (_set_up_derived_options):
365         (parse_args):
366         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
367         (MainTest.test_repeat_each):
368         * WebKitTestRunner/TestInvocation.cpp:
369         (WTR::TestInvocation::setIsPixelTest):
370
371 2012-01-30  Hans Wennborg  <hans@chromium.org>
372
373         Unreviewed, rolling out r106219.
374         http://trac.webkit.org/changeset/106219
375         https://bugs.webkit.org/show_bug.cgi?id=77083
376
377         This broke Chromium's test_shell.
378
379         * DumpRenderTree/DumpRenderTree.gypi:
380         * DumpRenderTree/chromium/LayoutTestController.cpp:
381         (LayoutTestController::addMockSpeechInputResult):
382         * DumpRenderTree/chromium/LayoutTestController.h:
383         (WebKit):
384         (LayoutTestController):
385         ():
386         * DumpRenderTree/chromium/MockWebSpeechInputController.cpp: Removed.
387         * DumpRenderTree/chromium/MockWebSpeechInputController.h: Removed.
388         * DumpRenderTree/chromium/TestShell.cpp:
389         * DumpRenderTree/chromium/WebViewHost.cpp:
390         (WebViewHost::speechInputController):
391         * DumpRenderTree/chromium/WebViewHost.h:
392         (WebKit):
393         (WebViewHost::speechInputControllerMock):
394         (WebViewHost):
395         ():
396
397 2012-01-26  Hans Wennborg  <hans@chromium.org>
398
399         Speech Input: move MockSpeechInputClient into Chromium DumpRenderTree implementation
400         https://bugs.webkit.org/show_bug.cgi?id=77083
401
402         Reviewed by Darin Fisher.
403
404         Add MockWebSpeechInputController which provides a mock implementation
405         of the WebSpeechInputController interface, and use that in
406         DumpRenderTree.
407
408         * DumpRenderTree/DumpRenderTree.gypi:
409         * DumpRenderTree/chromium/LayoutTestController.cpp:
410         (LayoutTestController::addMockSpeechInputResult):
411         * DumpRenderTree/chromium/LayoutTestController.h:
412         ():
413         * DumpRenderTree/chromium/MockWebSpeechInputController.cpp: Added.
414         (MockWebSpeechInputController::create):
415         (MockWebSpeechInputController::addMockRecognitionResult):
416         (MockWebSpeechInputController::clearResults):
417         (MockWebSpeechInputController::startRecognition):
418         (MockWebSpeechInputController::cancelRecognition):
419         (MockWebSpeechInputController::stopRecording):
420         (MockWebSpeechInputController::MockWebSpeechInputController):
421         (MockWebSpeechInputController::speechTaskFired):
422         (MockWebSpeechInputController::SpeechTask::SpeechTask):
423         (MockWebSpeechInputController::SpeechTask::stop):
424         (MockWebSpeechInputController::SpeechTask::runIfValid):
425         * DumpRenderTree/chromium/MockWebSpeechInputController.h: Added.
426         (WebKit):
427         (MockWebSpeechInputController):
428         (MockWebSpeechInputController::taskList):
429         (SpeechTask):
430         * DumpRenderTree/chromium/TestShell.cpp:
431         * DumpRenderTree/chromium/WebViewHost.cpp:
432         (WebViewHost::speechInputController):
433         * DumpRenderTree/chromium/WebViewHost.h:
434         (WebKit):
435         (WebViewHost::speechInputControllerMock):
436         (WebViewHost):
437         ():
438
439 2012-01-29  Sheriff Bot  <webkit.review.bot@gmail.com>
440
441         Unreviewed, rolling out r106109.
442         http://trac.webkit.org/changeset/106109
443         https://bugs.webkit.org/show_bug.cgi?id=77302
444
445         It made tests crash (Requested by Ossy_weekend on #webkit).
446
447         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
448         (WTR::WrapperWindow::WrapperWindow):
449         (WTR::PlatformWebView::PlatformWebView):
450
451 2012-01-28  Nico Weber  <thakis@chromium.org>
452
453         [chromium] Add support for building standalone webkit/chromium checkouts with ninja.
454         https://bugs.webkit.org/show_bug.cgi?id=77243
455
456         Reviewed by Adam Barth.
457
458         * Scripts/update-webkit:
459         * Scripts/webkitdirs.pm:
460         (isChromiumNinja):
461         (determineIsChromiumNinja):
462         (buildChromiumNinja):
463         (buildChromium):
464
465 2012-01-27  Kentaro Hara  <haraken@chromium.org>
466
467         The cpp parser of prepare-ChangeLog treats if(...) {} as a method
468         https://bugs.webkit.org/show_bug.cgi?id=77241
469
470         Reviewed by Ryosuke Niwa.
471
472         The cpp parser of prepare-ChangeLog is wrong, and an inner {} block
473         in a method in some namespace or class is treated as a method.
474         This patch fixes the bug.
475
476             class C {
477                 void func()
478                 {
479                     if (1) {    // This should not be treated as a method.
480                     }
481                 }
482             };
483
484         * Scripts/prepare-ChangeLog:
485         (get_function_line_ranges_for_cpp):
486         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests-expected.txt:
487         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests.cpp:
488         (Class108):
489         (Class108::func35):
490
491 2012-01-27  Enrica Casucci  <enrica@apple.com>
492
493         Followup to http://trac.webkit.org/changeset/106171.
494         Dan Bernstein pointed out correctly that I should
495         have used copy instead of initWithString.
496
497         Reviewed by Dan Bernstein.
498
499         * DumpRenderTree/mac/DumpRenderTreePasteboard.m:
500         (-[LocalPasteboard initWithName:]):
501
502 2012-01-27  Enrica Casucci  <enrica@apple.com>
503
504         DRT pasteboard object should return the correct pasteboard name.
505         https://bugs.webkit.org/show_bug.cgi?id=77260
506         
507         Currently the name property always returns nil.
508         Adding a memeber variable to store the pasteboard name and return its
509         value from the name method instead of returning nil always.
510
511         Reviewed by Dan Bernstein.
512
513         * DumpRenderTree/mac/DumpRenderTreePasteboard.m:
514         (+[DumpRenderTreePasteboard _pasteboardWithName:]):
515         (-[LocalPasteboard initWithName:]):
516         (-[LocalPasteboard name]):
517
518 2012-01-27  Martin Robinson  <mrobinson@igalia.com>
519
520         [GTK] [WK2] Implement the policy client
521         https://bugs.webkit.org/show_bug.cgi?id=76343
522
523         Reviewed by Gustavo Noronha Silva.
524
525         * gtk/generate-gtkdoc:
526         (get_webkit2_options): Update the list of skipped source files to include all clients.
527
528 2012-01-27  Ada Chan  <adachan@apple.com>
529
530         Add test for WKBundleFrameGetParentFrame().
531         https://bugs.webkit.org/show_bug.cgi?id=77161
532
533         Reviewed by Anders Carlsson.
534
535         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Add ParentFrame.cpp and ParentFrame_Bundle.cpp.
536         * TestWebKitAPI/Tests/WebKit2/ParentFrame.cpp: Added.
537         (TestWebKitAPI):
538         (TestWebKitAPI::didReceiveMessageFromInjectedBundle): Get the check result from the injected bundle.
539         (TestWebKitAPI::setInjectedBundleClient):
540         (TestWebKitAPI::TEST): Load simple-iframe.html and then wait for the injected bundle to post result.
541         Make sure the check is successful.
542         * TestWebKitAPI/Tests/WebKit2/ParentFrame_Bundle.cpp: Added.
543         (TestWebKitAPI):
544         (ParentFrameTest):
545         (TestWebKitAPI::ParentFrameTest::ParentFrameTest):
546         (TestWebKitAPI::didFinishLoadForFrame): If the frame is a subframe, store it off for checking later.
547         If the frame is the main frame, check whether it's indeed the subframe's parent frame and post the result
548         to the test controller.
549         (TestWebKitAPI::ParentFrameTest::didCreatePage): Set the page loader client on this page.
550
551 2012-01-27  Zan Dobersek  <zandobersek@gmail.com>
552
553         [Gtk] DumpRenderTree lacks --no-timeout command line option
554         https://bugs.webkit.org/show_bug.cgi?id=77196
555
556         Reviewed by Martin Robinson.
557
558         Add the --no-timeout command line option and reorganize
559         watchdog code to make running DumpRenderTree without
560         timeout possible.
561
562         * DumpRenderTree/gtk/DumpRenderTree.cpp:
563         (setWaitToDumpWatchdog):
564         (shouldSetWaitToDumpWatchdog):
565         (initializeGlobalsFromCommandLineOptions):
566         * DumpRenderTree/gtk/DumpRenderTreeGtk.h:
567         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
568         (waitToDumpWatchdogFired):
569         (LayoutTestController::setWaitToDump):
570
571 2012-01-27  Andras Becsi  <andras.becsi@nokia.com>
572
573         [Qt][WK2] Enable the automatic touch->mouse mocking on the new Qt5 hashes
574         https://bugs.webkit.org/show_bug.cgi?id=77122
575
576         Reviewed by Kenneth Rohde Christiansen.
577
578         * MiniBrowser/qt/MiniBrowserApplication.cpp:
579         (MiniBrowserApplication::MiniBrowserApplication):
580
581 2012-01-27  Zeno Albisser  <zeno@webkit.org>
582
583         [Qt][Mac] Build fails after adding ICU support (r105997).
584         https://bugs.webkit.org/show_bug.cgi?id=77118
585
586         Define WTF_USE_ICU_UNICODE if we are on mac.
587
588         Reviewed by Tor Arne Vestbø.
589
590         * qmake/mkspecs/features/features.prf:
591
592 2012-01-27  Roland Takacs  <takacs.roland@stud.u-szeged.hu>
593
594         Update committers.py with new contributors.
595         https://bugs.webkit.org/show_bug.cgi?id=77207
596
597         I added six new contributors into committers.py:
598
599         - Adam Kallai
600         - Balazs Ankes
601         - Janos Badics
602         - Nandor Huszka
603         - Roland Takacs
604         - Szilard Ledan-Muntean
605
606         * Scripts/webkitpy/common/config/committers.py:
607
608 2012-01-27  Vsevolod Vlasov  <vsevik@chromium.org>
609
610         Add Vsevolod Vlasov to reviewers list.
611
612         Reviewed by Pavel Feldman.
613
614         * Scripts/webkitpy/common/config/committers.py:
615
616 2012-01-26  No'am Rosenthal  <noam.rosenthal@nokia.com> and Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
617
618         [Qt] WKTR: Use a software rendering pipiline when running tests.
619         https://bugs.webkit.org/show_bug.cgi?id=76708
620
621         Reviewed by Kenneth Rohde Christiansen.
622
623         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
624         (WTR::WrapperWindow::WrapperWindow):
625         (WTR::PlatformWebView::PlatformWebView): Use software rendering of layers since the wrapping QQuickView isn't shown.
626
627 2012-01-26  Kevin Ollivier  <kevino@theolliviers.com>
628
629         [wx] Unreviewed. Build fix, add WebCore/testing/js
630         to the list of source directories we build.
631
632         * waf/build/settings.py:
633
634 2012-01-26  Filip Pizlo  <fpizlo@apple.com>
635
636         Tools/Scripts/commit-log-editor is broken due to $_ getting clobbered
637         https://bugs.webkit.org/show_bug.cgi?id=77177
638
639         Reviewed by Jon Honeycutt.
640
641         * Scripts/commit-log-editor:
642
643 2012-01-26  Ojan Vafai  <ojan@chromium.org>
644
645         Decrease sleep time when killing server_process on Mac.
646         https://bugs.webkit.org/show_bug.cgi?id=77169
647
648         Reviewed by Eric Seidel.
649
650         This shaves another ~70ms off single test runtime.
651
652         * Scripts/webkitpy/layout_tests/port/server_process.py:
653         (ServerProcess.stop):
654
655 2012-01-26  Ojan Vafai  <ojan@chromium.org>
656
657         Remove builder name so http://trac.webkit.org/changeset/106035 actually works.
658         https://bugs.webkit.org/show_bug.cgi?id=77164
659
660         Reviewed by Dirk Pranke.
661
662         http://trac.webkit.org/changeset/106035 depends on builder_name being
663         falsey by default. There's other code that requires it to be iterable
664         by default. So, make it be the empty string.
665
666         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
667         (parse_args):
668
669 2012-01-26  Ryosuke Niwa  <rniwa@webkit.org>
670
671         run-perf-test should include relative file paths in test names
672         https://bugs.webkit.org/show_bug.cgi?id=77144
673
674         Reviewed by Adam Birth.
675
676         Include all directory names in the test name separated by "/".
677         e.g. Bindings/event-target-wrapper.html translates to Bindings/event-target-wrapper
678
679         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
680         (PerfTestsRunner):
681         (PerfTestsRunner._process_parser_test_result):
682         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
683         (test_run_test_set_with_json_output):
684         (test_run_test_set_with_json_source):
685
686 2012-01-26  Dirk Pranke  <dpranke@chromium.org>
687
688         webkitpy: re-land cleanup of test scaffolding
689         https://bugs.webkit.org/show_bug.cgi?id=77153
690
691         Reviewed by Eric Seidel.
692
693         This is an attempt to reland the change in bug 76973 / 
694         r105935. My earlier attempt to manipulate PYTHONPATH
695         seems to only work some of the time, for reasons that
696         elude me. Directly adding Tools/Scripts to sys.path from
697         inside the test script appears to work (more?) reliably.
698
699         Also, it seemed like the files didn't actually get deleted
700         the last time; I don't know why.
701
702         * Scripts/test-webkitpy:
703         * Scripts/webkitpy/common/system/executive_unittest.py:
704         (command_line):
705         (ExecutiveTest):
706         (ExecutiveTest.test_run_command_args_type):
707         (ExecutiveTest.test_run_command_with_unicode):
708         (ExecutiveTest.test_running_pids):
709         (main):
710         * Scripts/webkitpy/common/system/fileutils.py: Removed.
711         * Scripts/webkitpy/test/cat.py: Removed.
712         * Scripts/webkitpy/test/cat_unittest.py: Removed.
713         * Scripts/webkitpy/test/echo.py: Removed.
714         * Scripts/webkitpy/test/echo_unittest.py: Removed.
715
716 2012-01-26  Ojan Vafai  <ojan@chromium.org>
717
718         run-webkit-tests calls "nm" when it doesn't need to
719         https://bugs.webkit.org/show_bug.cgi?id=77143
720
721         Reviewed by Dirk Pranke.
722
723         If we are not running any test that can be skipped by checking
724         the DRT feature list or looking at the list of symbols, avoid calling
725         out to DRT or calling "nm".
726
727         This saves ~700ms on my Mac Pro. In cases where you still have to call out
728         to "nm" it's only a couple ms slower.
729
730         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
731         (TestExpectations.__init__):
732         * Scripts/webkitpy/layout_tests/port/base.py:
733         (Port.skipped_tests):
734         * Scripts/webkitpy/layout_tests/port/webkit.py:
735         (WebKitPort._should_check_for_missing_features):
736         (WebKitPort):
737         (WebKitPort._should_check_for_missing_symbols):
738         (WebKitPort._skipped_tests_for_unsupported_features):
739         (WebKitPort.skipped_layout_tests):
740         (WebKitPort.skipped_tests):
741         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
742         (WebKitPortTest.test_skipped_directories_for_symbols):
743         (test_skipped_directories_for_features):
744         (test_skipped_directories_for_features_no_matching_tests_in_test_list):
745         (test_skipped_layout_tests):
746
747 2012-01-26  Benjamin Poulain  <bpoulain@apple.com>
748
749         Support a suffix on ChangeLog filenames based on a configuration file
750         https://bugs.webkit.org/show_bug.cgi?id=76956
751
752         Reviewed by Kentaro Hara.
753
754         Modify the tools used to generate, edit and resolve ChangeLogs to support
755         the addition of a suffix to the ChangeLog filename.
756
757         This patch makes it easier to support alternate ChangeLog files in branches.
758
759         * Scripts/VCSUtils.pm:
760         (changeLogSuffix):
761         (changeLogFileName):
762         * Scripts/commit-log-editor:
763         * Scripts/prepare-ChangeLog:
764         (getLatestChangeLogs):
765         (generateNewChangeLogs):
766         (generateFileList):
767         * Scripts/resolve-ChangeLogs:
768         (findChangeLog):
769
770 2012-01-26  Sheriff Bot  <webkit.review.bot@gmail.com>
771
772         Unreviewed, rolling out r106038.
773         http://trac.webkit.org/changeset/106038
774         https://bugs.webkit.org/show_bug.cgi?id=77142
775
776         Caused a bunch of skipped tests to not be skipped. (Requested
777         by ojan on #webkit).
778
779         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
780         (TestExpectationParser):
781         (TestExpectationParser.parse):
782         (TestExpectationParser._parse_line):
783         (TestExpectationParser._collect_matching_tests):
784         (TestExpectations.__init__):
785         (TestExpectations._add_skipped_tests):
786         * Scripts/webkitpy/layout_tests/port/base.py:
787         (Port.test_dirs):
788         (Port.normalize_test_name):
789         (Port.update_baseline):
790         (Port.layout_tests_dir):
791         (Port.relative_perf_test_filename):
792         (Port.abspath_for_test):
793
794 2012-01-25  Ojan Vafai  <ojan@chromium.org>
795
796         Parsing test_expecations.txt + Skipped lists takes too long
797         https://bugs.webkit.org/show_bug.cgi?id=77059
798
799         Reviewed by Dirk Pranke.
800
801         This saves ~100ms on the Apple Mac port.
802         -memoize a bunch of path methods.
803         -Avoid doing multiple disk accesses per line.
804         -Parse the skipped list directly instead of turning it into a test_expecations.txt
805         formatting string and parsing that.
806
807         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
808         (TestExpectationParser):
809         (TestExpectationParser.parse_skipped_tests):
810         (TestExpectationParser._parse_line):
811         (TestExpectationParser._collect_matching_tests):
812         (TestExpectations.__init__):
813         (TestExpectations._add_skipped_tests):
814         * Scripts/webkitpy/layout_tests/port/base.py:
815         (Port):
816         (Port.test_isfile):
817         (Port.normalize_test_name):
818         (Port.layout_tests_dir):
819         (Port.abspath_for_test):
820
821 2012-01-25  Ojan Vafai  <ojan@chromium.org>
822
823         Only store the SVN revision in the summarized results if we're on a builder
824         https://bugs.webkit.org/show_bug.cgi?id=76976
825
826         Reviewed by Dirk Pranke.
827
828         This shaves another ~130ms off the single test run-webkit-tests runtime.
829
830         * Scripts/webkitpy/layout_tests/controllers/manager.py:
831         (summarize_results):
832         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
833         (ResultSummaryTest.get_result):
834         (ResultSummaryTest):
835         (ResultSummaryTest.get_result_summary):
836         (ResultSummaryTest.get_unexpected_results):
837         (ResultSummaryTest.test_no_svn_revision):
838         (ResultSummaryTest.test_svn_revision):
839         * Scripts/webkitpy/layout_tests/port/base.py:
840         (Port.use_trac_links_in_results_html):
841
842 2012-01-26  Luciano Wolf  <luciano.wolf@openbossa.org>
843
844         [Qt][WK2] fast/loader tests failed after r94178
845         https://bugs.webkit.org/show_bug.cgi?id=67388
846
847         Reviewed by Simon Hausmann.
848
849         Applying the same fix used for Qt-WebKit1 from bug #67254 -
850         [Qt][DRT] Normalize file:///tmp/LayoutTests in LayoutTestController::pathToLocalResource()
851         https://bugs.webkit.org/show_bug.cgi?id=67254
852         Translate file:///tmp/LayoutTests/* urls into the repository LayoutTests directory, which is
853         derived from the running location of DumpRenderTree binary.
854
855         * WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
856         (WTR::LayoutTestController::pathToLocalResource):
857
858 2012-01-26  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
859
860         Implement overridePreference for boolean preferences in WebKitTestRunner
861         https://bugs.webkit.org/show_bug.cgi?id=77033
862
863         Reviewed by Alexey Proskuryakov.
864
865         For now overridePreference supports boolean values.
866
867         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
868         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
869         (WTR::LayoutTestController::setXSSAuditorEnabled):
870         (WTR::LayoutTestController::overridePreference):
871         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
872         * WebKitTestRunner/TestController.cpp:
873         (WTR::TestController::resetStateToConsistentValues): Reset the overrides
874         explicitly and properly disable the use of PageCache when resetting the
875         state. Tests that use PageCache can now enable it via overridePreference.
876
877 2012-01-26  Alexander Færøy  <alexander.faeroy@nokia.com>
878
879         [Qt] Support opening multiple URL's from the command line in the Qt MiniBrowser
880         https://bugs.webkit.org/show_bug.cgi?id=77100
881
882         Reviewed by Simon Hausmann.
883
884         This patch makes the MiniBrowser read multiple URL's from the command
885         line such that there will be one browser window opened for each URL.
886
887         * MiniBrowser/qt/MiniBrowserApplication.cpp:
888         (MiniBrowserApplication::handleUserOptions):
889
890 2012-01-26  Alexander Færøy  <alexander.faeroy@nokia.com>
891
892         [Qt] The Qt MiniBrowser crashes when given multiple URL's
893         https://bugs.webkit.org/show_bug.cgi?id=77009
894
895         Reviewed by Simon Hausmann.
896
897         This patch fixes a crash in BrowserWindow::newWindow() where a
898         BrowserWindow-object is constructed with a 0 pointer as the
899         WindowOptions parameter which leads to a crash.
900
901         * MiniBrowser/qt/BrowserWindow.cpp:
902         (BrowserWindow::BrowserWindow):
903         (BrowserWindow::newWindow):
904         * MiniBrowser/qt/BrowserWindow.h:
905
906 2012-01-26  Jesus Sanchez-Palencia  <jesus.palencia@openbossa.org>
907
908         [Qt] Use ICU if available
909         https://bugs.webkit.org/show_bug.cgi?id=76821
910
911         Reviewed by Simon Hausmann.
912
913         Adding unicode defines for a Qt 5 based build.
914
915         * qmake/mkspecs/features/features.prf:
916
917 2012-01-26  Nándor Huszka  <huszka.nandor@stud.u-szeged.hu>
918
919         [WK2] WebKitTestRunner needs layoutTestController.setGeolocationPermission
920         https://bugs.webkit.org/show_bug.cgi?id=42545
921
922         Reviewed by Zoltan Herczeg.
923
924         Implement the setGeolocationPermission method.
925
926         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
927         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
928         (WTR::LayoutTestController::setGeolocationPermission):
929         (WTR):
930         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
931         (LayoutTestController):
932
933 2012-01-26  Balazs Kelemen  <kbalazs@webkit.org>
934
935         [NRWT] Support --ignore-metrics
936         https://bugs.webkit.org/show_bug.cgi?id=76278
937
938         Reviewed by Tony Chang.
939
940         Port the feature from ORWT.
941
942         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
943         (SingleTestRunner._run_compare_test):
944         * Scripts/webkitpy/layout_tests/port/driver.py:
945         (DriverOutput):
946         (DriverOutput.strip_metrics):
947         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
948         (DriverOutputTest):
949         (DriverOutputTest.test_strip_metrics):
950         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
951         (_set_up_derived_options):
952         (parse_args):
953
954 2012-01-26  Sheriff Bot  <webkit.review.bot@gmail.com>
955
956         Unreviewed, rolling out r105935 and r105954.
957         http://trac.webkit.org/changeset/105935
958         http://trac.webkit.org/changeset/105954
959         https://bugs.webkit.org/show_bug.cgi?id=77080
960
961         test-webkitpy is still broken (Requested by abarth on
962         #webkit).
963
964         * Scripts/test-webkitpy:
965         * Scripts/webkitpy/common/system/executive_unittest.py:
966         (never_ending_command):
967         (ExecutiveTest.test_run_command_args_type):
968         (ExecutiveTest.test_run_command_with_unicode):
969         (ExecutiveTest.test_running_pids):
970         * Scripts/webkitpy/common/system/fileutils.py:
971         (make_stdout_binary):
972         * Scripts/webkitpy/test/cat.py:
973         (command_arguments):
974         (main):
975         * Scripts/webkitpy/test/cat_unittest.py:
976         (CatTest):
977         (CatTest.assert_cat):
978         (CatTest.test_basic):
979         (CatTest.test_no_newline):
980         (CatTest.test_unicode):
981         (CatTest.test_as_command):
982         * Scripts/webkitpy/test/echo.py:
983         (command_arguments):
984         (main):
985         * Scripts/webkitpy/test/echo_unittest.py:
986         (EchoTest):
987         (EchoTest.test_basic):
988         (EchoTest.test_no_newline):
989         (EchoTest.test_unicode):
990         (EchoTest.test_argument_order):
991         (EchoTest.test_empty_arguments):
992         (EchoTest.test_no_arguments):
993         (EchoTest.test_as_command):
994
995 2012-01-26  Gabor Rapcsanyi  <rgabor@webkit.org>
996
997         [GTK] Detect cross-compilation in webkitdirs.pm to not generate gtkdoc
998         https://bugs.webkit.org/show_bug.cgi?id=76138
999
1000         Reviewed by Martin Robinson.
1001
1002         * Scripts/webkitdirs.pm:
1003         (isCrossCompilation):
1004         (buildAutotoolsProject):
1005
1006 2012-01-25  Adam Barth  <abarth@webkit.org>
1007
1008         Kentaro is a reviewer now.  Yay!
1009
1010         * Scripts/webkitpy/common/config/committers.py:
1011
1012 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
1013
1014         nrwt: running w/ no arguments fails
1015         https://bugs.webkit.org/show_bug.cgi?id=77064
1016
1017         Unreviewed, build fix.
1018
1019         This change fixes a regression introduced in r105944
1020         where running without --platform <foo> or one of the other
1021         platform switches would cause a null deref :(.
1022
1023         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1024         (main):
1025
1026 2012-01-25  Eric Seidel  <eric@webkit.org>
1027
1028         Automatically CC someone from Apple on any changes to Obj-C public API
1029         https://bugs.webkit.org/show_bug.cgi?id=77053
1030
1031         Reviewed by Timothy Hatcher.
1032
1033         This has long been de-facto practice, seems now that we have
1034         fancy watchlist support, we should just make the tools
1035         do this automatically.
1036
1037         * Scripts/webkitpy/common/config/watchlist:
1038
1039 2012-01-25  Antoine Labour  <piman@chromium.org>
1040
1041         Updating status in committers.py.
1042
1043         Unreviewed.
1044
1045         * Scripts/webkitpy/common/config/committers.py:
1046
1047 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
1048
1049         test-webkitpy: executive_unittest fails if PYTHONPATH isn't set
1050         https://bugs.webkit.org/show_bug.cgi?id=77058
1051
1052         Reviewed by Eric Seidel.
1053
1054         r105935 broke executive_unittest by changing which script we
1055         were running and not ensuring that the PYTHONPATH was set up
1056         properly.
1057
1058         This change also fixes a regression introduced recently where
1059         test-webkitpy wasn't figuring out the path to Tools/Scripts
1060         right, either (so running ./test-webkitpy would fail).
1061
1062         * Scripts/test-webkitpy:
1063         * Scripts/webkitpy/common/system/executive_unittest.py:
1064         (script_dir):
1065         (ExecutiveTest.setUp):
1066         (ExecutiveTest):
1067         (ExecutiveTest.tearDown):
1068
1069 2012-01-25  Adam Barth  <abarth@webkit.org>
1070
1071         Disable tests on the mac-ews while we sort out hte stability issues.
1072
1073         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
1074         (MacEWS):
1075
1076 2012-01-23  Dirk Pranke  <dpranke@chromium.org>
1077
1078         investigate use of 'mac' and 'win' as fully-specified port names for the apple ports
1079         https://bugs.webkit.org/show_bug.cgi?id=76475
1080
1081         Reviewed by Eric Seidel.
1082
1083         This patch cleans up the internal usage of port names for the
1084         Apple mac and Apple win ports so that 'mac' and 'win are never
1085         considered "fully specified" port names: they are still legal
1086         input to the PortFactory.get() routine, but only if run on the given
1087         platform (i.e. trying to PortFactory.get('mac') on windows or
1088         linux will return an error, since we don't know which version of
1089         the port is desired.
1090
1091         This also cleans up handling of the webkit2 variants, so that
1092         the full port names consistently follow the
1093         <operating_system>-<version>[-<wk2>] convention.
1094
1095         Lastly this patch adds some assertions and removes some
1096         try/catch logic to catch more programming and usage errors;
1097         previously mac-tiger might've been allowed -- and it would've
1098         translated into using just the baselines in platform/mac -- but
1099         now it should fail.
1100
1101         * Scripts/webkitpy/layout_tests/port/apple.py:
1102         (ApplePort.determine_full_port_name):
1103         (ApplePort.__init__):
1104         (ApplePort):
1105         (ApplePort._port_name_with_version):
1106         * Scripts/webkitpy/layout_tests/port/builders.py:
1107         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
1108         (FactoryTest.test_mac):
1109         (FactoryTest.test_win):
1110         * Scripts/webkitpy/layout_tests/port/mac.py:
1111         (MacPort.baseline_search_path):
1112         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1113         (MacTest.test_versions):
1114         * Scripts/webkitpy/layout_tests/port/win.py:
1115         (WinPort.baseline_search_path):
1116         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
1117         (WinTest.test_versions):
1118
1119 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
1120
1121         webkitpy: print nicer errors while linting expectations files, remove redundant tests
1122         https://bugs.webkit.org/show_bug.cgi?id=76955
1123
1124         Reviewed by Adam Barth.
1125
1126         This change cleans up the errors that are printed so that the
1127         are more "quickfix" compatible (path:lineno).
1128
1129         Also, there were a bunch of redundant tests between
1130         layout_tests.models.test_expectations_unittest and
1131         style.checkers.test_expectations_unittest. The latter should
1132         just expect the former to do most of the testing.
1133
1134         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1135         (TestExpectationParser._parse_modifiers):
1136         (TestExpectations._report_errors):
1137         (TestExpectations._add_skipped_tests):
1138         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1139         (test_parse_error_nonfatal):
1140         (SemanticTests.test_bad_bugid):
1141         * Scripts/webkitpy/style/checkers/test_expectations.py:
1142         (TestExpectationsChecker.__init__):
1143         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
1144         (TestExpectationsTestCase.assert_lines_lint):
1145         (TestExpectationsTestCase.test_valid_expectations):
1146         (TestExpectationsTestCase.test_invalid_expectations):
1147         (TestExpectationsTestCase.test_tab):
1148
1149 2012-01-25  Eric Seidel  <eric@webkit.org>
1150
1151         webkit-patch apply-* should always continue after failures
1152         https://bugs.webkit.org/show_bug.cgi?id=77057
1153
1154         Reviewed by Adam Barth.
1155
1156         As far as I can tell there is only one potential drawback to always
1157         forcing: that if you're somehow in the wrong directory it will create new
1158         directories for new files.  Since webkit-patch always cd's to the root
1159         it seems that's not a drawback.  The drawback of not using --force for
1160         svn-apply, is that it will stop after the first failure, which is rarely
1161         (if ever) the desired behavior.  This just removes the force option
1162         (which was strangely hidden behind --non-interactive).  This should
1163         make for a better user experiance.
1164
1165         * Scripts/webkitpy/common/checkout/checkout.py:
1166         (Checkout.apply_patch):
1167         * Scripts/webkitpy/common/checkout/checkout_mock.py:
1168         (MockCheckout.apply_patch):
1169         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1170         (CheckoutTest.test_chromium_deps):
1171         (CheckoutTest):
1172         (CheckoutTest.test_apply_patch):
1173         * Scripts/webkitpy/tool/commands/download_unittest.py:
1174         (DownloadCommandsTest._default_options):
1175         * Scripts/webkitpy/tool/steps/applypatch.py:
1176         (ApplyPatch.options):
1177         (ApplyPatch.run):
1178         * Scripts/webkitpy/tool/steps/options.py:
1179         (Options):
1180
1181 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
1182
1183         nrwt: should be able to run --platform test interactively
1184         https://bugs.webkit.org/show_bug.cgi?id=76959
1185
1186         Reviewed by Adam Barth.
1187
1188         As part of refactoring the Port interfaces, in r103254 I made
1189         passing a Host object to the Port mandatory; previously TestPort
1190         objects would create their own (mock) hosts. However,
1191         new-run-webkit-tests always passed a real Host, not a MockHost,
1192         and so you couldn't run new-run-webkit-tests --platform test
1193         interactively to debug test failures.
1194
1195         This change fixes that by creating a MockHost instead of a real
1196         host if you say --platform test*.
1197
1198         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1199         (main):
1200
1201 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
1202
1203         run-webkit-tests --lint-test-files should lint all the ports by default
1204         https://bugs.webkit.org/show_bug.cgi?id=76749
1205
1206         Reviewed by Ojan Vafai.
1207
1208         Currently run-webkit-tests --lint-test-files will only lint the
1209         test_expectations for the single port determined by the command
1210         line options. This is not obvious and can produce unintended results
1211         if you want to lint the chromium port (probably the normal case
1212         for using --lint-test-files) but don't specify a port name or
1213         --chromium.
1214
1215         I think we should probably lint *all* of the ports if no port
1216         name is supplied. This change implements that, and also moves
1217         the linting logic out of the Manager class, since this really
1218         has nothing to do with Managers. It has little to do with
1219         run-webkit-tests, but splitting it into a different script would
1220         introduce compatibility issues. It might make sense to do that
1221         anyway, in a separate patch and combining that with cleaning up
1222         the style checker to share more code.
1223
1224         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1225         (Manager._strip_test_dir_prefix):
1226         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1227         (lint):
1228         (run):
1229         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1230         (LintTest.test_all_configurations): added.
1231         (LintTest.test_lint_test_files): added.
1232         (LintTest.test_lint_test_files__errors): added.
1233         (MainTest.test_lint_test_files): removed.
1234         (MainTest.test_lint_test_files__errors): removed.
1235
1236 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
1237
1238         webkitpy: clean up a bunch of test scaffolding
1239         https://bugs.webkit.org/show_bug.cgi?id=76973
1240
1241         Reviewed by Eric Seidel.
1242
1243         There were several helper modules in the webkitpy.test package
1244         that appear to only be used by executive_unittest.py. I am
1245         rolling them all into that file to make that clearer.
1246
1247         * Scripts/webkitpy/common/system/executive_unittest.py:
1248         (ExecutiveTest.test_run_command_args_type):
1249         (ExecutiveTest.test_run_command_with_unicode):
1250         (ExecutiveTest.test_running_pids):
1251         (command_line):
1252         (main):
1253         * Scripts/webkitpy/common/system/fileutils.py: Removed.
1254         * Scripts/webkitpy/test/cat.py: Removed.
1255         * Scripts/webkitpy/test/cat_unittest.py: Removed.
1256         * Scripts/webkitpy/test/echo.py: Removed.
1257         * Scripts/webkitpy/test/echo_unittest.py: Removed.
1258
1259 2012-01-25  Dirk Pranke  <dpranke@chromium.org>
1260
1261         webkitpy: clean up a bunch of test scaffolding
1262         https://bugs.webkit.org/show_bug.cgi?id=76973
1263
1264         Reviewed by Eric Seidel.
1265
1266         There were several helper modules in the webkitpy.test package
1267         that appear to only be used by executive_unittest.py. I am
1268         rolling them all into that file to make that clearer.
1269
1270         * Scripts/webkitpy/common/system/executive_unittest.py:
1271         (ExecutiveTest.test_run_command_args_type):
1272         (ExecutiveTest.test_run_command_with_unicode):
1273         (ExecutiveTest.test_running_pids):
1274         (command_line):
1275         (main):
1276         * Scripts/webkitpy/common/system/fileutils.py: Removed.
1277         * Scripts/webkitpy/test/cat.py: Removed.
1278         * Scripts/webkitpy/test/cat_unittest.py: Removed.
1279         * Scripts/webkitpy/test/echo.py: Removed.
1280         * Scripts/webkitpy/test/echo_unittest.py: Removed.
1281
1282 2012-01-25  Eric Seidel  <eric@webkit.org>
1283
1284         Remove Python 2.5 support from WebKit
1285         https://bugs.webkit.org/show_bug.cgi?id=71593
1286
1287         Reviewed by Adam Barth.
1288
1289         Just removing support from this one core file.
1290         I'll commit the rest of the (large) change once we've
1291         gone 24 hours and seen that no bots have exploded.
1292         We can roll out this one change if we break bots
1293         and then roll in the rest once those bots are fixed.
1294
1295         * Scripts/webkitpy/common/system/executive.py:
1296         (Executive.cpu_count):
1297
1298 2012-01-25  Hajime Morita  <morrita@google.com>
1299
1300         ENABLE_SHADOW_DOM should be available via build-webkit --shadow-dom
1301         https://bugs.webkit.org/show_bug.cgi?id=76863
1302
1303         Reviewed by Dimitri Glazkov.
1304
1305         Added a feature flag.
1306
1307         * Scripts/build-webkit:
1308         * qmake/mkspecs/features/features.prf:
1309
1310 2012-01-25  Martin Robinson  <mrobinson@igalia.com>
1311
1312         [WK2] Navigation events not triggered by a mouse event claim to be triggered by the left mouse button during policy decisions
1313         https://bugs.webkit.org/show_bug.cgi?id=76344
1314
1315         Reviewed by Sam Weinig.
1316
1317         Add a test verifying that the mouse button during policy decisions for
1318         events not triggered by mouse events is NoButton.
1319
1320         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
1321         (TestWebKitAPI::decidePolicyForNavigationAction): Added test.
1322         (TestWebKitAPI::decidePolicyForNewWindowAction): Added test.
1323
1324 2012-01-25  Ryosuke Niwa  <rniwa@webkit.org>
1325
1326         run-perf-tests should timeout sooner
1327         https://bugs.webkit.org/show_bug.cgi?id=77026
1328
1329         Reviewed by Adam Barth.
1330
1331         Bump off the timeout from 10 minutes to 4 minutes.
1332
1333         Also use float instead of str in test result values to compress the size of json files bots upload.
1334
1335         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1336         (PerfTestsRunner._parse_args):
1337         (PerfTestsRunner._process_chromium_style_test_result):
1338         (PerfTestsRunner._process_parser_test_result):
1339         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1340         (test_run_test_set_for_parser_tests):
1341         (test_run_test_set_with_json_output):
1342         (test_run_test_set_with_json_source):
1343
1344 2012-01-25  Enrica Casucci  <enrica@apple.com>
1345
1346         WebView should implement typingAttributes methods to work correctly with the Inspector bar in Mail.
1347         https://bugs.webkit.org/show_bug.cgi?id=76951
1348         <rdar://problem/9325158>
1349
1350         Reviewed by Alexey Proskuryakov.
1351
1352         * TestWebKitAPI/Tests/mac/InspectorBar.mm: Modified to reflect the changes to WebView
1353         and WebHTMLView.
1354         (TestWebKitAPI::TEST):
1355
1356 2012-01-23  MORITA Hajime  <morrita@google.com>
1357
1358         [Refactoring][Internals] Should have InternalSettings
1359         https://bugs.webkit.org/show_bug.cgi?id=76424
1360
1361         Reviewed by Kent Tamura.
1362
1363         * GNUmakefile.am: Added so new files.
1364
1365 2012-01-25  Takashi Toyoshima  <toyoshim@chromium.org>
1366
1367         webkitpy: new-run-webkit-websocket is broken
1368         https://bugs.webkit.org/show_bug.cgi?id=76999
1369
1370         Reviewed by Dirk Pranke.
1371
1372         r105674 broke new-run-webkit-websocket, since options.chromium
1373         doesn't exist any more. new-run-webkit-httpd was also broken
1374         and already fixed at r105835.
1375         * Scripts/new-run-webkit-websocketserver:
1376         (main):
1377
1378 2012-01-25  Nándor Huszka  <huszka.nandor@stud.u-szeged.hu>
1379
1380         [WK2] WebKitTestRunner needs layoutTestController.setPopupBlockingEnabled
1381         https://bugs.webkit.org/show_bug.cgi?id=63458
1382
1383         Reviewed by Adam Roben.
1384
1385         Implement the setPopupBlockingEnabled method.
1386
1387         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
1388         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1389         (WTR::LayoutTestController::setPopupBlockingEnabled):
1390         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1391
1392 2012-01-24  Balazs Kelemen  <kbalazs@webkit.org>
1393
1394         [Qt] Cleanup testfonts handling
1395         https://bugs.webkit.org/show_bug.cgi?id=76068
1396
1397         Reviewed by Simon Hausmann.
1398
1399         Factor test font initialization into it's own file  so test
1400         font users can share the implementation by adding the file
1401         to their build.
1402         Use fonts.conf from the test font directory to not require the
1403         current directory to be the WebKit root.
1404         Fix the -use-test-fonts option of QtTestBrowser by initializing
1405         the fonts earlier, before it is too late (namely before calling
1406         show on the main window).
1407
1408         * DumpRenderTree/qt/DumpRenderTree.pro:
1409         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1410         (WebCore::DumpRenderTree::open):
1411         * DumpRenderTree/qt/DumpRenderTreeQt.h:
1412         * DumpRenderTree/qt/QtInitializeTestFonts.cpp: Copied from Tools/WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp.
1413         (WebKit::initializeTestFonts):
1414         * DumpRenderTree/qt/QtInitializeTestFonts.h: Added.
1415         * DumpRenderTree/qt/fonts.conf: Removed. No need for this since
1416         we will use the one from the test fonts directory.
1417         * DumpRenderTree/qt/main.cpp:
1418         (main): Fix the ifdefs to make it clear that the code to
1419         set up the Dpi has not been ported to Qt5 yet.
1420         * QtTestBrowser/QtTestBrowser.pro:
1421         * QtTestBrowser/launcherwindow.h:
1422         (WindowOptions::WindowOptions):
1423         * QtTestBrowser/main.cpp:
1424         (launcherMain):
1425         (LauncherApplication::handleUserOptions):
1426         * WebKitTestRunner/InjectedBundle/Target.pri:
1427         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
1428         (WTR::activateFonts): Remove the code to set up the Dpi
1429         because it does not work with Qt5.
1430
1431 2012-01-25  Simon Hausmann  <simon.hausmann@nokia.com>
1432
1433         [Qt] build-webkit --qt requires a git-svn or SVN checkout.
1434
1435         Reviewed by Tor Arne Vestø.
1436
1437         Don't require it by making svnRevisionForDirectory return "unknown"
1438         instead of bailing out with an error.
1439
1440         * Scripts/VCSUtils.pm:
1441         (svnRevisionForDirectory):
1442
1443 2012-01-25  Adam Barth  <abarth@webkit.org>
1444
1445         Temporarily disable running the unit tests on the Chromium EWS.  These
1446         tests don't work properly (see
1447         https://bugs.webkit.org/show_bug.cgi?id=76979) and are currently
1448         blocking the EWS.
1449
1450         * Scripts/webkitpy/common/config/ports.py:
1451         (ChromiumPort):
1452
1453 2012-01-25  Adam Barth  <abarth@webkit.org>
1454
1455         Update results.html version of PixelZoomer
1456         https://bugs.webkit.org/show_bug.cgi?id=76965
1457
1458         Reviewed by Ojan Vafai.
1459
1460         Improve readability as suggested by Ojan.
1461
1462         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/pixelzoomer.js:
1463
1464 2012-01-25  Carlos Garcia Campos  <cgarcia@igalia.com>
1465
1466         [GTK] Page clients don't need to be GObjects anymore in WebKit2 GTK+ API
1467         https://bugs.webkit.org/show_bug.cgi?id=76899
1468
1469         Reviewed by Martin Robinson.
1470
1471         * gtk/generate-gtkdoc:
1472         (get_webkit2_options): WebKitWebLoaderClient has been renamed to
1473         WebKitLoaderClient.
1474
1475 2012-01-24  Ryosuke Niwa  <rniwa@webkit.org>
1476
1477         Build fix. Trigger Chromium Mac Release (Perf) on new commits.
1478
1479         * BuildSlaveSupport/build.webkit.org-config/config.json:
1480
1481 2012-01-23  Ryosuke Niwa  <rniwa@webkit.org>
1482
1483         WebKit needs public performance tests running on public bots
1484         https://bugs.webkit.org/show_bug.cgi?id=10266
1485
1486         Reviewed by Adam Barth.
1487
1488         Completes the initial implementation of WebKit Perf bots. Add a build step for run-perf-test
1489         and make google-mac-4 run it, and upload the results to webkit-perf.webkit.org.
1490
1491         * BuildSlaveSupport/build.webkit.org-config/config.json:
1492         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1493         (RunAndUploadPerfTests):
1494         (RunAndUploadPerfTests.start):
1495         (UploadPerfTestResults):
1496         (BuildAndPerfTestFactory):
1497         (BuildAndPerfTestFactory.__init__):
1498
1499 2012-01-24  Ojan Vafai  <ojan@chromium.org>
1500
1501         Initializing the browser property of the Bugzilla class takes too long
1502         https://bugs.webkit.org/show_bug.cgi?id=76960
1503
1504         Reviewed by Adam Barth.
1505
1506         Lazily initialize it. This saves ~150ms on a no-op run-webkit-tests call
1507         on my Mac Pro.
1508
1509         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
1510         (Bugzilla.__init__):
1511         (Bugzilla._get_browser):
1512         (Bugzilla):
1513         (Bugzilla._set_browser):
1514
1515 2012-01-24  Mark Rowe  <mrowe@apple.com>
1516
1517         Death to JavaScriptGlue.
1518
1519         Rubber-stamped by Geoffrey Garen.
1520
1521         * BuildSlaveSupport/build-launcher-app:
1522         (copyNightlyLauncher):
1523         * Scripts/build-webkit:
1524         * Scripts/do-file-rename:
1525         * Scripts/do-webcore-rename:
1526         * Scripts/webkitpy/common/config/build.py:
1527         (_should_file_trigger_build):
1528         * Scripts/webkitpy/common/config/build_unittest.py: Update the test to use another Mac-specific
1529         path in conjunction with a cross-platform path.
1530         (ShouldBuildTest):
1531
1532 2012-01-24  Kentaro Hara  <haraken@chromium.org>
1533
1534         webkitperl/prepare-ChangeLog_unittest/parser_unittests.pl should test stdout and stderr
1535         https://bugs.webkit.org/show_bug.cgi?id=75836
1536
1537         Reviewed by David Kilzer.
1538
1539         Previously parser_unittests.pl can just test the returned value of
1540         get_function_line_ranges() in prepare-ChangeLog. This patch fixes
1541         parser_unittests.pl so that it also tests the stdout and stderr of
1542         get_function_line_ranges().
1543
1544         This patch adds css_unittests_warning.css, which contains test cases
1545         for warning messages of the CSS parser of prepare-ChangeLog.
1546
1547         * Scripts/webkitperl/prepare-ChangeLog_unittest/parser_unittests.pl:
1548         (captureOutput): Captures the stdout and stderr of a given method.
1549         (convertAbsolutepathToWebKitPath): Converts absolute paths in a given
1550         string to relative paths.
1551
1552         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/css_unittests_warning-expected.txt: Added.
1553         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/css_unittests_warning.css: Added.
1554         (element1):
1555         (element2):
1556         (element3):
1557         (element4):
1558         ():
1559         (element5):
1560         (element6):
1561
1562         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests-expected.txt:
1563         Just updated the test results, i.e. added stdout and stderr results.
1564         No essential changes in the test results.
1565         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/css_unittests-expected.txt: Ditto.
1566         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/java_unittests-expected.txt: Ditto.
1567         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/javascript_unittests-expected.txt: Ditto.
1568         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/perl_unittests-expected.txt: Ditto.
1569         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/python_unittests-expected.txt: Ditto.
1570
1571 2012-01-24  Dirk Pranke  <dpranke@chromium.org>
1572
1573         webkitpy: new-run-webkit-httpd is broken
1574         https://bugs.webkit.org/show_bug.cgi?id=76963
1575
1576         Reviewed by Adam Barth.
1577
1578         r105674 broke new-run-webkit-httpd, since options.chromium
1579         doesn't exist any more. We didn't catch this because we don't
1580         run the integration tests by default anywhere :(.
1581
1582         * Scripts/new-run-webkit-httpd:
1583         (run):
1584         (main):
1585
1586 2012-01-24  Adam Barth  <abarth@webkit.org>
1587
1588         garden-o-matic pixel zoomer should let you look at every pixel of the image
1589         https://bugs.webkit.org/show_bug.cgi?id=76962
1590
1591         Reviewed by Ojan Vafai.
1592
1593         Previously, you couldn't get up to the right and bottom pixels.
1594
1595         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/pixelzoomer.js:
1596
1597 2012-01-24  Enrica Casucci  <enrica@apple.com>
1598
1599         WebHTMLView should implement typingAttributes methods to work correctly with the Inspector bar in Mail.
1600         https://bugs.webkit.org/show_bug.cgi?id=76951
1601         <rdar://problem/9325158>
1602
1603         Added test to verify that WebHTMLView supports the typingAttributes method and that it
1604         works as expected.
1605         
1606         Reviewed by Alexey Proskuryakov.
1607
1608         * TestWebKitAPI/Tests/mac/InspectorBar.mm:
1609         (TestWebKitAPI::TEST):
1610
1611 2012-01-24  James Robinson  <jamesr@chromium.org>
1612
1613         Put a space between the trac link and ellipsis in sheriffbot rollout message so it linkifies better in some IRC clients
1614         https://bugs.webkit.org/show_bug.cgi?id=76884
1615
1616         Reviewed by Adam Barth.
1617
1618         * Scripts/webkitpy/tool/bot/irc_command.py:
1619         (Rollout.execute):
1620         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
1621         (SheriffIRCBotTest.test_rollout):
1622         (SheriffIRCBotTest.test_revert):
1623         (SheriffIRCBotTest.test_multi_rollout):
1624         (SheriffIRCBotTest.test_rollout_with_r_in_svn_revision):
1625         (SheriffIRCBotTest.test_multi_rollout_with_r_in_svn_revision):
1626         (SheriffIRCBotTest.test_rollout_invalidate_reason):
1627         (test_multi_rollout_invalidate_reason):
1628
1629 2012-01-24  Kentaro Hara  <haraken@chromium.org>
1630
1631         The cpp parser of prepare-ChangeLog cannot detect a change in classes and namespaces
1632         https://bugs.webkit.org/show_bug.cgi?id=75531
1633
1634         Reviewed by David Kilzer.
1635
1636         Previously, the cpp parser of prepare-ChangeLog could not detect a change
1637         outside methods. Consider the following cpp file.
1638
1639             namespace N {
1640             int a;     // this change does not appear on ChangeLog.
1641             class C {
1642                 int b;     // this change does not appear on ChangeLog.
1643                 void f()
1644                 {
1645                     int c;     // this change appears on ChangeLog.
1646                 }
1647                 int d;     // this change does not appear on ChangeLog.
1648             };
1649             int e;     // this change does not appear on ChangeLog.
1650             };
1651
1652         The previous prepare-ChangeLog outputs just methods in which a change is found:
1653             (N::C::f):
1654
1655         This patch fixes prepare-ChangeLog so that it outputs namespaces, classes
1656         and methods in which a change is found:
1657             (N):
1658             (N::C):
1659             (N::C::f):
1660
1661         * Scripts/prepare-ChangeLog:
1662         (get_function_line_ranges_for_cpp): Modified as described above.
1663
1664         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests.cpp: Added test cases.
1665         (Class104):
1666         (Class105):
1667         (Class106):
1668         (Class106::func32):
1669         (Class106::func33):
1670         (NameSpace3):
1671         (NameSpace4):
1672         (NameSpace5):
1673         (NameSpace6):
1674         (Class107):
1675         (NameSpace5::NameSpace6::Class107::func34):
1676         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/cpp_unittests-expected.txt:
1677
1678 2012-01-24  Kentaro Hara  <haraken@chromium.org>
1679
1680         prepare-ChangeLog outputs warnings for top-level { ... } blocks
1681         https://bugs.webkit.org/show_bug.cgi?id=75943
1682
1683         Reviewed by David Kilzer.
1684
1685         Currently, prepare-ChangeLog outputs warnings for top-level { ... } blocks,
1686         because it does not distinguish '}' at the end of a subroutine
1687         with '}' at the end of a if or for statement.
1688
1689         foo.pl:
1690             if (1) {
1691             }
1692             for (@v) {
1693             }
1694
1695         output:
1696             nested functions found at top-level at foo.pl:2
1697             nested functions found at top-level at foo.pl:4
1698
1699         This patch fixes the bug.
1700
1701         * Scripts/prepare-ChangeLog:
1702         (get_function_line_ranges_for_perl):
1703         * Scripts/webkitperl/prepare-ChangeLog_unittest/resources/perl_unittests.pl:
1704         Added test cases. No change in the test results.
1705
1706 2012-01-24  Dirk Pranke  <dpranke@chromium.org>
1707
1708         r105674 broke check-webkit-style for chromium's test_expectations.
1709
1710         Unreviewed, build fix.
1711         
1712         We now need to pass in a platform name of 'chromium' if we want
1713         the chromium port; options.chromium is no longer recognized. I
1714         will follow this patch up with a separate patch that has a test,
1715         but this at least fixes the build.
1716
1717         * Scripts/webkitpy/style/checkers/test_expectations.py:
1718         (TestExpectationsChecker._determine_port_from_exepectations_path):
1719
1720 2012-01-20  Ojan Vafai  <ojan@chromium.org>
1721
1722         check-webkit-style of the chromium test_expectations.txt file takes too long
1723         https://bugs.webkit.org/show_bug.cgi?id=76745
1724
1725         Reviewed by Dimitri Glazkov.
1726
1727         When in lint mode, have TestExpectations test all configurations instead
1728         of looping over each configuration. This also has the benefit of making
1729         the error output considerably more concise.
1730
1731         Also, got rid of the double-printing of errors when linting through check-webkit-style.
1732
1733         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1734         (Manager.lint):
1735         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1736         (TestExpectations._report_errors):
1737         (TestExpectations._add_expectations):
1738         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1739         (test_parse_error_nonfatal):
1740         (test_error_on_different_platform):
1741         * Scripts/webkitpy/style/checkers/test_expectations.py:
1742         (TestExpectationsChecker.check_test_expectations):
1743         (TestExpectationsChecker.check):
1744         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
1745         (TestExpectationsTestCase.test_determine_port_from_exepectations_path):
1746         (TestExpectationsTestCase.assert_lines_lint):
1747
1748 2012-01-24  Ryosuke Niwa  <rniwa@webkit.org>
1749
1750         Another build fix attempt after r105543.
1751
1752         * Scripts/webkitpy/common/net/file_uploader.py:
1753         (FileUploader.upload_as_multipart_form_data):
1754
1755 2012-01-24  Ryosuke Niwa  <rniwa@webkit.org>
1756
1757         Build fix attempt after r105543.
1758
1759         * Scripts/webkitpy/common/net/file_uploader.py:
1760         (FileUploader.upload_as_multipart_form_data):
1761
1762 2012-01-24  Vincent Scheib  <scheib@chromium.org>
1763
1764         [Chromium] Add Pointer Lock test hooks and mock implementation to DumpRenderTree
1765         https://bugs.webkit.org/show_bug.cgi?id=76411
1766
1767         Reviewed by Adam Barth.
1768
1769         Breaking up https://bugs.webkit.org/show_bug.cgi?id=75762
1770         into a series of patches, this change adds a trivial
1771         pointer lock implementation to the Chromium port's DumpRenderTree
1772         and exposes LayoutTestController hooks to force failure conditions.
1773
1774         * DumpRenderTree/chromium/LayoutTestController.cpp:
1775         (LayoutTestController::LayoutTestController):
1776         (LayoutTestController::didLosePointerLock):
1777         (LayoutTestController::setPointerLockWillFailAsynchronously):
1778         (LayoutTestController::setPointerLockWillFailSynchronously):
1779         * DumpRenderTree/chromium/LayoutTestController.h:
1780         * DumpRenderTree/chromium/WebViewHost.cpp:
1781         (WebViewHost::requestPointerLock):
1782         (WebViewHost::requestPointerUnlock):
1783         (WebViewHost::isPointerLocked):
1784         (WebViewHost::didCompletePointerLock):
1785         (WebViewHost::didNotCompletePointerLock):
1786         (WebViewHost::didLosePointerLock):
1787         (WebViewHost::reset):
1788         * DumpRenderTree/chromium/WebViewHost.h:
1789         (WebViewHost::setPointerLockWillFailAsynchronously):
1790         (WebViewHost::setPointerLockWillFailSynchronously):
1791
1792 2012-01-24  Mario Sanchez Prada  <msanchez@igalia.com>
1793
1794         Unreviewed, skip WK2 API test timing out on some GTK bots.
1795
1796         WebKit2APITests/TestDownloads is consistently timing out on the
1797         32bit release and 64bit debug bots.
1798
1799         * Scripts/run-gtk-tests:
1800         (TestRunner): Skip WebKit2APITests/TestDownloads.
1801
1802 2012-01-24  Andras Becsi  <andras.becsi@nokia.com>
1803
1804         [Qt] Fix the build with the newes Qt5 hashes
1805         https://bugs.webkit.org/show_bug.cgi?id=76657
1806
1807         Reviewed by Simon Hausmann.
1808
1809         * MiniBrowser/qt/MiniBrowserApplication.cpp:
1810         (MiniBrowserApplication::MiniBrowserApplication): Disable the automatic touch->mouse event synthesis
1811         for now to prevent an infinite loop where Qt always creates mouse events out of our touch mock events.
1812
1813 2012-01-24  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1814
1815         [Qt] Distinguish between Cygwin and cmd.exe in the way we set env variables
1816
1817         We can't use win32-msvc* to decide whether or not to use (set FOO=bar), as
1818         building with MinGW inside cmd.exe will require the same construct. Instead
1819         we assume that when the first character in PWD is a '/' we're running under
1820         Cygwin, and use the normal unix way of setting environment variables.
1821
1822         Reviewed by Simon Hausmann.
1823
1824         * qmake/mkspecs/features/default_pre.prf:
1825
1826 2012-01-24  Simon Hausmann  <simon.hausmann@nokia.com>
1827
1828         Some qmake build files must include Qt5's "quick" module.
1829         https://bugs.webkit.org/show_bug.cgi?id=76671
1830
1831         Unreviewed build fix.
1832
1833         * MiniBrowser/qt/MiniBrowser.pro:
1834         * WebKitTestRunner/Target.pri:
1835
1836 2012-01-24  Andras Becsi  <andras.becsi@nokia.com>
1837
1838         [Qt] Fix the event timestamps for touch mocking
1839
1840         Reviewed by Simon Hausmann.
1841
1842         * MiniBrowser/qt/MiniBrowserApplication.cpp:
1843         (MiniBrowserApplication::sendTouchEvent): We should not set QEvent::None as the
1844         timestamp argument in the QWindowSystemInterface::handleTouchEvent call.
1845
1846 2012-01-24  Carlos Garcia Campos  <cgarcia@igalia.com>
1847
1848         [GTK] Implement DownloadClient in WebKit2 GTK+ API
1849         https://bugs.webkit.org/show_bug.cgi?id=72952
1850
1851         Reviewed by Martin Robinson.
1852
1853         * gtk/generate-gtkdoc:
1854         (get_webkit2_options): Ignore WebKitDownloadClient.
1855
1856 2012-01-23  Zan Dobersek  <zandobersek@gmail.com>
1857
1858         [GTK] editing/deleting/5408255.html results are incorrect
1859         https://bugs.webkit.org/show_bug.cgi?id=53644
1860
1861         Reviewed by Martin Robinson.
1862
1863         WEBKIT_TOP_LEVEL environment variable is now set directly in either
1864         WebKitTestRunner or DumpRenderTree through usage of a compilation-time
1865         macro. This way both tools can be run outside the test harness without
1866         the need to manually set the environment variable.
1867
1868         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1869         (getTopLevelPath):
1870         * GNUmakefile.am:
1871         * Scripts/webkitpy/layout_tests/port/gtk.py:
1872         (GtkPort.setup_environ_for_server):
1873         * WebKitTestRunner/GNUmakefile.am:
1874         * WebKitTestRunner/InjectedBundle/gtk/InjectedBundleGtk.cpp:
1875         (WTR::InjectedBundle::platformInitialize):
1876
1877 2012-01-23  Dmitry Lomov  <dslomov@google.com>
1878
1879         [Chromium] Implement layoutTestController.workerThreadCount in DRT
1880         https://bugs.webkit.org/show_bug.cgi?id=74653.
1881
1882         Reviewed by Darin Fisher.
1883
1884         * DumpRenderTree/chromium/LayoutTestController.cpp:
1885         (LayoutTestController::LayoutTestController):
1886         (LayoutTestController::workerThreadCount):
1887         * DumpRenderTree/chromium/LayoutTestController.h:
1888
1889 2012-01-23  Dirk Pranke  <dpranke@chromium.org>
1890
1891         nrwt: make --chromium work like --qt
1892         https://bugs.webkit.org/show_bug.cgi?id=76875
1893
1894         Reviewed by Adam Barth.
1895
1896         --chromium used to have to be handled differently from --qt
1897         due to the way the PortFactory was implemented; there's not
1898         really a good reason for that any more so this patch makes
1899         things slightly more consistent and eliminates the
1900         options.chromium flag (--chromium is now truly a synonym for
1901         --platform chromium).
1902
1903         * Scripts/webkitpy/layout_tests/port/factory.py:
1904         (PortFactory._default_port):
1905         (PortFactory.get):
1906         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
1907         (FactoryTest.setUp):
1908         (FactoryTest.test_chromium_mac):
1909         (FactoryTest.test_chromium_linux):
1910         (FactoryTest.test_chromium_win):
1911         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1912         (parse_args):
1913
1914 2012-01-23  Dirk Pranke  <dpranke@chromium.org>
1915
1916         run-webkit-tests needs to propagate --chromium
1917         https://bugs.webkit.org/show_bug.cgi?id=76870
1918
1919         Reviewed by Eric Seidel.
1920
1921         run-webkit-tests removes '--chromium' argument from @ARGV when
1922         determining which port to run, which means that that doesn't
1923         propagate to new-run-webkit-tests. That's bad (and is handled
1924         for the other ports by re-adding the flag, but apparently we're
1925         just now noticing for Chromium).
1926
1927         * Scripts/run-webkit-tests:
1928
1929 2012-01-23  Brian Weinstein  <bweinstein@apple.com>
1930
1931         More build fixing after r105646.
1932
1933         * TestWebKitAPI/Tests/WTF/RedBlackTree.cpp:
1934         (TestWebKitAPI::TestNode::key):
1935
1936 2012-01-23  Gavin Barraclough  <barraclough@apple.com>
1937
1938         Unreviewed build fix - r105646 broke this.
1939
1940         * TestWebKitAPI/Tests/WTF/RedBlackTree.cpp:
1941         (TestWebKitAPI::TestNode::TestNode):
1942         (TestWebKitAPI::TestNode::key):
1943         (TestWebKitAPI::RedBlackTreeTest::assertEqual):
1944         (TestWebKitAPI::RedBlackTreeTest::assertSameValuesForKey):
1945         (TestWebKitAPI::RedBlackTreeTest::testDriver):
1946
1947 2012-01-23  Sheriff Bot  <webkit.review.bot@gmail.com>
1948
1949         Unreviewed, rolling out r105658.
1950         http://trac.webkit.org/changeset/105658
1951         https://bugs.webkit.org/show_bug.cgi?id=76883
1952
1953         We want this eventually, but not right at this moment
1954         (Requested by abarth on #webkit).
1955
1956         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1957
1958 2012-01-23  Adam Barth  <abarth@webkit.org>
1959
1960         garden-o-matic should support Chromium Mac Lion
1961         https://bugs.webkit.org/show_bug.cgi?id=76880
1962
1963         Reviewed by Eric Seidel.
1964
1965         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
1966
1967 2012-01-23  Emil A Eklund  <eae@chromium.org>
1968
1969         check-webkit-style whitespace/operators triggers on overloaded division operator
1970         https://bugs.webkit.org/show_bug.cgi?id=76650
1971
1972         Reviewed by Darin Adler.
1973
1974         * Scripts/webkitpy/style/checkers/cpp.py:
1975         (check_spacing):
1976         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1977         (CppStyleTest.test_operator_methods):
1978
1979 2012-01-23  Ryosuke Niwa  <rniwa@webkit.org>
1980
1981         run-perf-tests should report server-side errors
1982         https://bugs.webkit.org/show_bug.cgi?id=76802
1983
1984         Reviewed by Tony Chang.
1985
1986         Report errors whenever server's response doesn't match "OK".
1987
1988         * Scripts/webkitpy/common/net/file_uploader.py:
1989         (FileUploader.upload_single_text_file):
1990         (FileUploader.upload_as_multipart_form_data):
1991         (FileUploader._upload_data.callback):
1992         (FileUploader):
1993         (FileUploader._upload_data):
1994         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
1995         (PerfTestsRunner._upload_json):
1996         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
1997         (test_upload_json):
1998         (test_upload_json.MockFileUploader.upload_single_text_file):
1999
2000 2012-01-22  Ryosuke Niwa  <rniwa@webkit.org>
2001
2002         run-perf-tests ignore Skipped list on chromium
2003         https://bugs.webkit.org/show_bug.cgi?id=76764
2004
2005         Reviewed by Dirk Pranke.
2006
2007         Move skipped_perf_tests from WebKit port to Base port so that Chromium port
2008         can also find skipped list. Chromium port only uses test_expectations.txt for
2009         layout tests but performacne tests don't use test_expectations.txt so Chromium port
2010         also needs to use Skipped list.
2011
2012         * Scripts/webkitpy/layout_tests/port/base.py:
2013         (Port._tests_from_skipped_file_contents):
2014         (Port):
2015         (Port._expectations_from_skipped_files):
2016         (Port.skipped_perf_tests):
2017         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2018         (PortTest.test_skipped_perf_tests):
2019         (PortTest.test_skipped_perf_tests.add_text_file):
2020         * Scripts/webkitpy/layout_tests/port/webkit.py:
2021         (WebKitPort._skipped_tests_for_unsupported_features):
2022         (WebKitPort._skipped_file_search_paths):
2023         (WebKitPort.skipped_layout_tests):
2024
2025 2012-01-23  Andy Estes  <aestes@apple.com>
2026
2027         Fix the build after r105635.
2028
2029         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp:
2030         (TestWebKitAPI::TEST):
2031
2032 2012-01-23  Xianzhu Wang  <wangxianzhu@chromium.org>
2033
2034         Basic enhancements to StringBuilder
2035         https://bugs.webkit.org/show_bug.cgi?id=67081
2036
2037         Reviewed by Darin Adler.
2038
2039         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp:
2040         (TestWebKitAPI::TEST):
2041
2042 2012-01-23  Mario Sanchez Prada  <msanchez@igalia.com>
2043
2044         [GTK] run-gtk-tests randomly fails while running the xprop comand
2045         https://bugs.webkit.org/show_bug.cgi?id=76817
2046
2047         Reviewed by Gustavo Noronha Silva.
2048
2049         No need to use xprop to remove the AT_SPI_BUS property since
2050         run-gtk-tests will always run new instances of Xvfb.
2051
2052         * Scripts/run-gtk-tests:
2053         (TestRunner): Unskip TestWebKitAccessibility.
2054         (TestRunner.run): Uncomment lines for launching the accessibility
2055         bus and registry daemon, and remove lines for running xprop.
2056
2057 2012-01-23  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2058
2059         [Qt] Don't warn about override and final being C++11 extensions
2060
2061         Clang will emit a warning when these extensions are used without passing
2062         --std=c++11, but we use feature checking to decide if we have the right
2063         extensions, so we can safely ignore these warnings. The XCode and Windows
2064         project files have the same workaround.
2065
2066         The reason for adding the flag to QMAKE_OBJECTIVE_CFLAGS as well is that
2067         we only have one extra compiler for Objective-C, which is also used for
2068         Objective-C++ sources, so we have to pass the flag, even if it doesn't
2069         make sense for Objective-C.
2070
2071         Reviewed by Simon Hausmann.
2072
2073         * qmake/mkspecs/features/unix/default_post.prf:
2074
2075 2012-01-20  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2076
2077         [Qt] Change how build-webkit decides when to do full incremental builds
2078
2079         Instead of relying on update-webkit (which isn't run on the bots) to
2080         decide when to do a full incremental build (make qmake), we let the
2081         build-webkit script itself check the current SVN revision against the
2082         previous build (by storing it in .webkit.config).
2083
2084         If the two differ we assume a full incremental build is needed, since
2085         the new revisions might have introduced problematic things like new
2086         Q_OBJECT macros. If not, we assume the developer is doing changes
2087         locally, and revert to doing a plain 'make'.
2088
2089         In addition, when the build fails in the latter case, we inform the
2090         developer of possible pitfalls and how to manually run 'make qmake'.
2091
2092         Reviewed by Ossy.
2093
2094         * Scripts/update-webkit:
2095         * Scripts/webkitdirs.pm:
2096         (buildQMakeProjects):
2097
2098 2012-01-23  Mario Sanchez Prada  <msanchez@igalia.com>
2099
2100         [GTK] run-gtk-tests randomly fails while running the xprop comand
2101         https://bugs.webkit.org/show_bug.cgi?id=76817
2102
2103         Reviewed by Philippe Normand.
2104
2105         Temporarily comment the lines related to launching the ATSPI bus
2106         and registry daemon, which are making the bots to fail randomly.
2107
2108         * Scripts/run-gtk-tests:
2109         (TestRunner): Skip TestWebKitAccessibility, as it won't run
2110         properly if the ATSPI infrastructure is not properly initialized.
2111         (TestRunner.run): Comment lines related to initialization of ATSPI.
2112
2113 2012-01-22  Ryosuke Niwa  <rniwa@webkit.org>
2114
2115         Windows python test build fix.
2116
2117         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2118         (PerfTestsRunner.__init__):
2119
2120 2012-01-22  Sergio Villar Senin  <svillar@igalia.com>
2121
2122         [GTK] DumpRenderTree converts "file:///" to a path differently
2123         https://bugs.webkit.org/show_bug.cgi?id=76631
2124
2125         Reviewed by Martin Robinson.
2126
2127         DumpRenderTree should print "/" as the last path component if the
2128         path is a lone slash instead of empty output.
2129
2130         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2131         (webViewConsoleMessage):
2132
2133 2012-01-20  David Levin  <levin@chromium.org>
2134
2135         The leaky DC check should use adoptPtr as a signal instead of OwnPtr to get less false positives.
2136         https://bugs.webkit.org/show_bug.cgi?id=76752
2137
2138         Reviewed by Eric Seidel.
2139
2140         * Scripts/webkitpy/style/checkers/cpp.py:
2141         (check_for_leaky_patterns):
2142         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
2143         (LeakyPatternTest.test_create_dc):
2144         (LeakyPatternTest.test_create_compatible_dc):
2145
2146 2012-01-20  Shawn Singh  <shawnsingh@chromium.org>
2147
2148         Changed my status in commiters.py
2149
2150         Unreviewed.
2151
2152         * Scripts/webkitpy/common/config/committers.py:
2153
2154 2012-01-20  Ojan Vafai  <ojan@chromium.org>
2155
2156         run-webkit-tests --lint-test-files crawls the whole LayoutTests subtree
2157         https://bugs.webkit.org/show_bug.cgi?id=76748
2158
2159         Reviewed by Ryosuke Niwa.
2160
2161         It crawls the whole subtree and then doesn't use the data. Cutting this out
2162         saves 4 seconds warm and 17 seconds cold on my Mac Pro.
2163
2164         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2165         (run):
2166
2167 2012-01-20  Adam Barth  <abarth@webkit.org>
2168
2169         Another tiny tweak to the garden-o-matic CSS.  This makes things line
2170         up slightly nicer when there aren't any failures.
2171
2172         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
2173         (#summary):
2174
2175 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
2176
2177         run-perf-tests should support --test-results-server option
2178         https://bugs.webkit.org/show_bug.cgi?id=76680
2179
2180         Reviewed by Adam Barth.
2181
2182         Add --test-results-server, --builder-name, and --build-number options to run-perf-tests
2183         to be used by perf bots. Also refactor file_uploader as needed.
2184
2185         * Scripts/webkitpy/common/net/file_uploader.py:
2186         (FileUploader.__init__):
2187         (FileUploader.upload_single_file):
2188         (FileUploader.upload_as_multipart_form_data):
2189         (FileUploader):
2190         (FileUploader._upload_data):
2191         (FileUploader._upload_data.callback):
2192         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2193         (JSONResultsGeneratorBase.upload_json_files):
2194         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2195         (PerfTestsRunner):
2196         (PerfTestsRunner._parse_args):
2197         (PerfTestsRunner.run):
2198         (PerfTestsRunner._generate_json):
2199         (PerfTestsRunner._upload_json):
2200         (PerfTestsRunner._run_tests_set):
2201         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2202         (create_runner):
2203         (test_run_with_upload_json):
2204         (test_run_with_upload_json.mock_upload_json):
2205         (test_upload_json):
2206         (test_upload_json.MockFileUploader):
2207         (test_upload_json.MockFileUploader.__init__):
2208         (test_upload_json.MockFileUploader.upload_single_file):
2209         (test_parse_args):
2210
2211 2012-01-20  Adam Barth  <abarth@webkit.org>
2212
2213         The party time image overlaps real content!  This patch makes the
2214         image centered vertically.
2215
2216         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css:
2217         (#onebar.partytime #summary):
2218
2219 2012-01-19  Ojan Vafai  <ojan@chromium.org>
2220
2221         Refactor TestExpectationsParser in preparation for caching the results
2222         https://bugs.webkit.org/show_bug.cgi?id=76669
2223
2224         Reviewed by Dimitri Glazkov.
2225
2226         Make everything private expect for the parse method. 
2227         Eventually, we'll need the expectations lines to not be modified
2228         outside of TestExpectationsParser so we can cache the results.
2229         This makes check-webkit-style of the chromium test_expectations.txt file
2230         go from ~17 seconds to ~12 seconds on my Mac Pro.
2231
2232         This patch is just a refactor in preparation, so no new tests.
2233
2234         * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor_unittest.py:
2235         (TestExpectationEditorTests.make_parsed_expectation_lines):
2236         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2237         (TestExpectationParser.parse):
2238         (TestExpectationParser):
2239         (TestExpectationParser._parse_line):
2240         (TestExpectationParser._tokenize):
2241         (TestExpectationParser._tokenize_list):
2242         (TestExpectationsModel._clear_expectations_for_test):
2243         (TestExpectations.__init__):
2244         (TestExpectations._add_expectations):
2245         (TestExpectations._add_skipped_tests):
2246         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
2247         (TestExpectationParserTests.test_tokenize_blank):
2248         (TestExpectationParserTests.test_tokenize_missing_colon):
2249         (TestExpectationParserTests.test_tokenize_extra_colon):
2250         (TestExpectationParserTests.test_tokenize_empty_comment):
2251         (TestExpectationParserTests.test_tokenize_comment):
2252         (TestExpectationParserTests.test_tokenize_missing_equal):
2253         (TestExpectationParserTests.test_tokenize_extra_equal):
2254         (TestExpectationParserTests.test_tokenize_valid):
2255         (TestExpectationParserTests.test_tokenize_valid_with_comment):
2256         (TestExpectationParserTests.test_tokenize_valid_with_multiple_modifiers):
2257         (TestExpectationParserTests.test_parse_empty_string):
2258         (TestExpectationSerializerTests.assert_round_trip):
2259         (TestExpectationSerializerTests.assert_list_round_trip):
2260         * Scripts/webkitpy/tool/commands/expectations.py:
2261         (OptimizeExpectations.execute):
2262         * Scripts/webkitpy/tool/servers/gardeningserver.py:
2263         (GardeningExpectationsUpdater.update_expectations):
2264
2265 2012-01-20  Adam Barth  <abarth@webkit.org>
2266
2267         Follow-up to previous patch: don't produce NaN when the revision number
2268         is missing.
2269
2270         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
2271         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model_unittests.js:
2272
2273 2012-01-20  Adam Barth  <abarth@webkit.org>
2274
2275         Garden-o-matic should tell me which revisions have been checked by all the bots
2276         https://bugs.webkit.org/show_bug.cgi?id=76722
2277
2278         Reviewed by Dimitri Glazkov.
2279
2280         This information helps me know when it's safe for me to roll a
2281         candidate revision.
2282
2283         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
2284         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
2285         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model_unittests.js:
2286
2287 2012-01-20  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2288
2289         build-webkit: Don't spit out congratulations message on Ctrl+C
2290
2291         Reviewed by Andreas Kling.
2292
2293         * Scripts/build-webkit:
2294
2295 2012-01-20  Philippe Normand  <pnormand@igalia.com>
2296
2297         [GTK] Allow extra jhbuild modulesets
2298         https://bugs.webkit.org/show_bug.cgi?id=76691
2299
2300         Reviewed by Gustavo Noronha Silva.
2301
2302         * gtk/jhbuildrc: Extra modulesets and modules can be added in
2303         JHBuild using the WEBKIT_EXTRA_MODULESETS and WEBKIT_EXTRA_MODULES
2304         env variables respectively. Use comma-separated values. Example:
2305         WEBKIT_EXTRA_MODULES=file:///path/to/module.set,file:///other/path.
2306
2307 2012-01-20  Mario Sanchez Prada  <msanchez@igalia.com>
2308
2309         [GTK] Expose accessibility hierarchy in WebKit2 to ATK/AT-SPI based ATs
2310         https://bugs.webkit.org/show_bug.cgi?id=72589
2311
2312         Reviewed by Martin Robinson.
2313
2314         Ensure the accessibility infrastructure is available when running
2315         the unit tests, otherwise WK2's accessibility tests won't work.
2316
2317         * Scripts/run-gtk-tests:
2318         (TestRunner):
2319         (TestRunner._lookup_atspi2_binary): New. Looks for the directory
2320         where at-spi2's stuff is located (like at-spi-bus-launcher and
2321         at-spi2-registryd).
2322         (TestRunner.run): Initialize the accessibility infrastructure.
2323
2324         * gtk/generate-gtkdoc:
2325         (get_webkit2_options): Add 'WebKitWebViewBaseAccessible.*'.
2326
2327         * gtk/jhbuild.modules: Added at-spi2-core and at-spi2-atk.
2328
2329 2012-01-20  Hayato Ito  <hayato@chromium.org>
2330
2331         Add ShadowRoot.idl which is enabled by newly introduced SHADOW_DOM flag.
2332         https://bugs.webkit.org/show_bug.cgi?id=76353
2333
2334         Reviewed by Hajime Morita.
2335
2336         Add ShadowRoot.idl, which is enabled only on chromium port since this is
2337         under development feature.
2338         ShadowRoot.idl contains minimum API so that we can test it.
2339         Other APIs should be added on other changes so that we can isolate issues.
2340
2341         * DumpRenderTree/chromium/TestShell.cpp:
2342         (TestShell::TestShell):
2343
2344 2012-01-20  Mario Sanchez Prada  <msanchez@igalia.com>
2345
2346         [GTK] GTK's DRT not to log events for already defunct objects
2347         https://bugs.webkit.org/show_bug.cgi?id=76620
2348
2349         Reviewed by Martin Robinson.
2350
2351         Do not log 'state-change:defunct' events.
2352
2353         * DumpRenderTree/gtk/AccessibilityCallbacks.cpp:
2354         (printAccessibilityEvent): Early return if a
2355         'state-change:defunct' signal is passed.
2356         (axObjectEventListener): Pass the signal name and value in
2357         separate parameters to printAccessibilityEvent.
2358
2359 2012-01-19  Gustavo Noronha Silva  <gns@gnome.org>
2360
2361         Unreviewed build fix. Wrong variable being used as the cwd for git
2362         remote update command.
2363
2364         * gtk/run-with-jhbuild:
2365         (update_jhbuild):
2366
2367 2012-01-19  Gustavo Noronha Silva  <gns@gnome.org>
2368
2369         [GTK] ensure the jhbuild used by webkit is as up-to-date as needed
2370         https://bugs.webkit.org/show_bug.cgi?id=76585
2371
2372         Reviewed by Martin Robinson.
2373
2374         This allows us to specify a minimum version of jhbuild required,
2375         while also providing stability against changes done to the master
2376         branch, by letting us choose what is the exact version that will
2377         be used.
2378
2379         * gtk/run-with-jhbuild: refactored for more hackability and easier
2380         reading
2381         (jhbuild_installed): checks whether jhbuild is installed
2382         (jhbuild_cloned): checks whether jhbuild has been cloned
2383         (jhbuild_at_expected_revision): checks whether jhbuild is at the
2384         expected revision
2385         (update_jhbuild): brings jhbuild to the expected revision
2386         (clone_jhbuild): clones jhbuild
2387         (install_jhbuild): build and installs jhbuild
2388         (update_webkitgtk_libs): runs update-webkitgtk-libs
2389         (ensure_jhbuild): high-level logic to decide whether and which of
2390         the steps above are executed
2391
2392 2012-01-19  Alexey Proskuryakov  <ap@apple.com>
2393
2394         [WK2] DumpRenderTree converts "file:///" to a path differently
2395         https://bugs.webkit.org/show_bug.cgi?id=76653
2396
2397         Reviewed by John Sullivan.
2398
2399         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp: (WTR::lastFileURLPathComponent):
2400         Do print a slash if there are no path components.
2401
2402 2012-01-19  Eric Seidel  <eric@webkit.org>
2403
2404         webkit-patch cannot rollout patches without changelog/bug number
2405         https://bugs.webkit.org/show_bug.cgi?id=75962
2406
2407         Reviewed by Adam Barth.
2408
2409         * Scripts/webkitpy/tool/commands/download.py:
2410         (AbstractRolloutPrepCommand._prepare_state):
2411         * Scripts/webkitpy/tool/commands/download_unittest.py:
2412         (AbstractRolloutPrepCommandTest.test_prepare_state):
2413
2414 2012-01-19  Alexey Proskuryakov  <ap@apple.com>
2415
2416         [WK2] fast/url/degenerate-file-base.html fails
2417         https://bugs.webkit.org/show_bug.cgi?id=76619
2418
2419         Reviewed by Adam Roben.
2420
2421         A URL generated form local path for loading had an extra slash, making document.URL
2422         incorrect in all WebKit2 tests (file://localhost//path/to/test.html).
2423
2424         * WebKitTestRunner/TestInvocation.cpp: (WTR::createWKURL):
2425
2426 2012-01-18  Ojan Vafai  <ojan@chromium.org>
2427
2428         check-webkit-style of the chromium test_expectations.txt file doesn't test all chromium ports
2429         https://bugs.webkit.org/show_bug.cgi?id=76510
2430
2431         Reviewed by Adam Barth.
2432
2433         Unlike other style rules, if you get the syntax of the test_expectations.txt
2434         file wrong, the layout tests won't run. Also, this check is simple and only
2435         slows down committing if you actually modify one of the test_expectations.txt files.
2436
2437         * Scripts/webkitpy/style/checkers/test_expectations.py:
2438         (TestExpectationsChecker.check_test_expectations):
2439         (TestExpectationsChecker.check):
2440         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2441         (TestExpectationsTestCase.test_check_covers_all_configurations):
2442         (TestExpectationsTestCase.test_check_covers_all_configurations.mock_check_test_expectations):
2443         (TestExpectationsTestCase):
2444         (assert_lines_lint):
2445
2446 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
2447
2448         Build fix. Revert an unintentional change to WebKitDriver in r105443.
2449
2450         * Scripts/webkitpy/layout_tests/port/driver.py:
2451         (DriverProxy.__init__):
2452
2453 2012-01-19  Dirk Pranke  <dpranke@chromium.org>
2454
2455         remove the duplicated build rules from WebKit.gyp, Tools.gyp
2456         https://bugs.webkit.org/show_bug.cgi?id=73384
2457
2458         Reviewed by Tony Chang.
2459
2460         This patch removes the no-longer-necessary conditional logic for
2461         build_webkit_exes_from_webkit_gyp; we now always build the exes
2462         from their own dedicated gyp files.
2463
2464         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2465
2466 2012-01-19  Alexey Proskuryakov  <ap@apple.com>
2467
2468         make-new-script-test should work with HTTP tests
2469         https://bugs.webkit.org/show_bug.cgi?id=76603
2470
2471         Reviewed by Adam Roben.
2472
2473         * Scripts/make-new-script-test:
2474         (makePathToSharedSources): Renamed from makeTestRelativePathToSharedSources(), since it
2475         now also creates absolute paths for HTTP.
2476
2477 2012-01-19  Ryosuke Niwa  <rniwa@webkit.org>
2478
2479         Some perf tests time out when ran by run-perf-tests
2480         https://bugs.webkit.org/show_bug.cgi?id=76612
2481
2482         Reviewed by Dirk Pranke and Eric Seidel.
2483
2484         Always pass --no-timeout to DumpRenderTree from run-perf-tests.
2485         Otherwise some tests such as Parser/xml-parser.html will timeout.
2486
2487         --no-timeout option is currently supported by Chromium and Mac ports.
2488
2489         * Scripts/webkitpy/layout_tests/port/base.py:
2490         (Port.to.create_driver):
2491         * Scripts/webkitpy/layout_tests/port/chromium.py:
2492         (ChromiumDriver.__init__):
2493         (ChromiumDriver._wrapper_options):
2494         * Scripts/webkitpy/layout_tests/port/driver.py:
2495         (Driver.__init__):
2496         (DriverProxy.__init__):
2497         * Scripts/webkitpy/layout_tests/port/webkit.py:
2498         (WebKitDriver.__init__):
2499         (WebKitDriver.cmd_line):
2500         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2501         (WebKitDriverTest.test_read_binary_block):
2502         (WebKitDriverTest):
2503         (WebKitDriverTest.test_no_timeout):
2504         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2505         (get_tests_run.RecordingTestDriver.__init__):
2506         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2507         (PerfTestsRunner._parse_args):
2508         (PerfTestsRunner._run_tests_set):
2509         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2510         (create_runner):
2511
2512 2012-01-19  Ben Wells  <benwells@chromium.org>
2513
2514         Add support for window.print to chromium DRT
2515         https://bugs.webkit.org/show_bug.cgi?id=76479
2516
2517         When used in a test window.print goes into print mode and then
2518         straight out again. This will allow the afterprint event to be
2519         tested by having the afterprint event happen before the dump
2520         happens.
2521
2522         Reviewed by Mihai Parparita.
2523
2524         * DumpRenderTree/chromium/WebViewHost.cpp:
2525         (WebViewHost::printPage):
2526         * DumpRenderTree/chromium/WebViewHost.h:
2527
2528 2012-01-18  Enrica Casucci  <enrica@apple.com>
2529
2530         editingAttributedStringFromRange in WebHTMLConverter does not handle NSUnderlineStyleAttributeName.
2531         https://bugs.webkit.org/show_bug.cgi?id=76588
2532         <rdar://problem/9325183>
2533
2534         Added test.
2535         
2536         Reviewed by Dan Bernstein.
2537
2538         * TestWebKitAPI/Tests/mac/InspectorBar.mm:
2539         (TestWebKitAPI::TEST):
2540
2541 2012-01-19  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2542
2543         [Qt] Do full incremental builds until bots use update-webkit
2544
2545         Rubber-stamped by Simon Hausmann.
2546
2547         * Scripts/webkitdirs.pm:
2548         (buildQMakeProjects):
2549
2550 2012-01-19  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2551
2552         Fix the Qt build
2553
2554         Unreviewed build fix.
2555
2556         * Scripts/webkitdirs.pm:
2557         (buildQMakeProjects):
2558
2559 2012-01-19  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2560
2561         [Qt] Make build-webkit always do safe incremental builds after update-webkit
2562
2563         When building against Qt5 with GCC we would just run 'make' in the build dir,
2564         and rely on the dependency files output by GCC (-MD) to handle dependency
2565         tracking, but that fails for special-cases like adding a Q_OBJECT macro
2566         to a header.
2567
2568         To guarantee that an incrmental build will work, we have to run 'make qmake',
2569         which we now do on every build-webkit that's followed by a successful run
2570         of update-webkit. The reasoning is that update-webkit can result in such
2571         potential corner-cases being applied, and since we can't know for sure
2572         unless we inspect the diff and account for all the corner cases we assume
2573         the worst and always run 'make qmake'.
2574
2575         After a succesful run of build-webkit we proceed to do just 'make' for any
2576         subsequent runs, since we assume that the developer knows what kind of
2577         changes he/she is doing, and when a 'make qmake' is needed.
2578
2579         Reviewed by Simon Hausmann.
2580
2581         * Scripts/update-webkit:
2582         * Scripts/webkitdirs.pm:
2583         (determineCurrentSVNRevision):
2584         (currentSVNRevision):
2585         (buildQMakeProjects):
2586         * qmake/mkspecs/features/default_post.prf:
2587
2588 2012-01-18  Ryosuke Niwa  <rniwa@webkit.org>
2589
2590         run-perf-tests should support Skipped list
2591         https://bugs.webkit.org/show_bug.cgi?id=76594
2592
2593         Reviewed by Adam Barth.
2594
2595         Add a support for Skipped list in run-perf-tests; also skip files in resources directories.
2596
2597         * Scripts/webkitpy/layout_tests/port/base.py:
2598         (Port.perf_tests_dir):
2599         (Port.skipped_perf_tests):
2600         (Port.skips_perf_test):
2601         (Port.relative_perf_test_filename):
2602         * Scripts/webkitpy/layout_tests/port/test.py:
2603         (TestPort.perf_tests_dir):
2604         * Scripts/webkitpy/layout_tests/port/webkit.py:
2605         (WebKitPort._expectations_from_skipped_files):
2606         (WebKitPort):
2607         (WebKitPort.skipped_layout_tests):
2608         (WebKitPort.skipped_perf_tests):
2609         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2610         (PerfTestsRunner._collect_tests):
2611         (PerfTestsRunner.run):
2612         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2613         (test_run_test_set_with_json_output):
2614         (test_run_test_set_with_json_source):
2615         (test_collect_tests_with_skipped_list):
2616         (test_collect_tests_with_skipped_list.add_file):
2617
2618 2012-01-18  Dirk Pranke  <dpranke@chromium.org>
2619
2620         Fix path to chromium_src_dir introduced in previous change.
2621
2622         Unreviewed, build fix.
2623
2624         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2625
2626 2012-01-18  Alexey Proskuryakov  <ap@apple.com>
2627
2628         Need infrastructure to test Content-Disposition filename encoding support
2629         https://bugs.webkit.org/show_bug.cgi?id=76572
2630         <rdar://problem/6421825>
2631
2632         Reviewed by Sam Weinig.
2633
2634         Made policy delegate log suggested file name when resource has Content-Disposition: attachment.
2635
2636         WebKitTestRunner does not have a policy delegate yet, bug 42546.
2637
2638         * DumpRenderTree/mac/PolicyDelegate.mm:
2639         (dispositionTypeFromContentDispositionHeader):
2640         (-[PolicyDelegate webView:decidePolicyForMIMEType:request:frame:decisionListener:]):
2641
2642 2012-01-18  Ojan Vafai  <ojan@chromium.org>
2643
2644         Fix tests from http://trac.webkit.org/changeset/105332.
2645
2646         The code assumed that if you passed an options object with a configuration
2647         attribute that it would have a valid value. A bunch of the testing mocks
2648         do things like having a configuration value of None.
2649
2650         * Scripts/webkitpy/layout_tests/port/base.py:
2651         (Port.__init__):
2652
2653 2012-01-18  Ojan Vafai  <ojan@chromium.org>
2654
2655         Shave 0.5 seconds off check-webkit-style runtime for test_expectations.txt
2656         https://bugs.webkit.org/show_bug.cgi?id=76576
2657
2658         Reviewed by Adam Barth.
2659
2660         Avoid calling default_configuration in base.py. We don't need it for
2661         checking test_expectations.txt style. It takes ~0.5 seconds to run on my
2662         Mac Pro. It's the call to "perl Tools/Scripts/webkit-build-directory --top-level"
2663         from common.executive.
2664
2665         At some point someone should probably look into why that call is so slow
2666         since it's on the critical path for run-webkit-tests and build-webkit startup.
2667
2668         * Scripts/webkitpy/layout_tests/port/base.py:
2669         (Port.__init__):
2670         * Scripts/webkitpy/style/checkers/test_expectations.py:
2671         (TestExpectationsChecker._determine_port_from_exepectations_path):
2672
2673 2012-01-18  Dirk Pranke  <dpranke@chromium.org>
2674
2675         [chromium] move Tools.gyp, switch build-webkit --chromium to All.gyp
2676         https://bugs.webkit.org/show_bug.cgi?id=76505
2677
2678         Reviewed by Tony Chang.
2679
2680         Update chromium build script to use All.gyp instead of WebKit.gyp
2681
2682         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp: Renamed from Source/WebKit/chromium/Tools.gyp.
2683         * Scripts/webkitdirs.pm:
2684
2685 2012-01-18  Sam Weinig  <sam@webkit.org>
2686
2687         Don’t hardcode the path to ‘make‘.
2688
2689         Reviewed by Dan Bernstein.
2690
2691         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
2692
2693 2012-01-18  Joshua Bell  <jsbell@chromium.org>
2694
2695         Adding myself to committers.py
2696         https://bugs.webkit.org/show_bug.cgi?id=76569
2697
2698         Unreviewed.
2699
2700         * Scripts/webkitpy/common/config/committers.py:
2701
2702 2012-01-18  Adam Barth  <abarth@webkit.org>
2703
2704         Up the cap on test failures tolerated by the EWS so the mac-ews can run to completion
2705         https://bugs.webkit.org/show_bug.cgi?id=76567
2706
2707         Reviewed by Eric Seidel.
2708
2709         Current the AppleMac port has slightly more failures than we allow on
2710         the EWS, causing the mac-ews to spin.  This patch ups the limit a bit,
2711         which should hopefully let the mac-ews run to completion.  Once
2712         AppleMac has fewer failures, we can lower the cap again.
2713
2714         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
2715         (LayoutTestResultsReaderTest.test_layout_test_results):
2716         * Scripts/webkitpy/tool/steps/runtests.py:
2717         (RunTests):
2718         * Scripts/webkitpy/tool/steps/steps_unittest.py:
2719
2720 2012-01-18  Eric Carlson  <eric.carlson@apple.com>
2721
2722         Provide access to user's list of preferred languages
2723         https://bugs.webkit.org/show_bug.cgi?id=76138
2724
2725         Reviewed by Alexey Proskuryakov.
2726
2727         * GNUmakefile.am: Add JSInternalsCustom.cpp.
2728
2729 2012-01-18  Adam Barth  <abarth@webkit.org>
2730
2731         Add Chromium Linux GPU 32 to the list of allowed Chromium Linux
2732         configurations.
2733
2734         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2735         (ChromiumLinuxPort.__init__):
2736
2737 2012-01-18  Dominic Mazzoni  <dmazzoni@google.com>
2738
2739         Accessibility: Chromium needs methods to scroll an object into view or to a specific location.
2740         https://bugs.webkit.org/show_bug.cgi?id=73460
2741
2742         Reviewed by Chris Fleizach.
2743
2744         * DumpRenderTree/AccessibilityUIElement.h:
2745         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
2746         (AccessibilityUIElement::AccessibilityUIElement):
2747         (AccessibilityUIElement::scrollToMakeVisibleCallback):
2748         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocusCallback):
2749         (AccessibilityUIElement::scrollToGlobalPointCallback):
2750         * DumpRenderTree/chromium/AccessibilityUIElement.h:
2751         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2752         (AccessibilityUIElement::scrollToMakeVisible):
2753         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2754         (AccessibilityUIElement::scrollToGlobalPoint):
2755         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2756         (AccessibilityUIElement::scrollToMakeVisible):
2757         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2758         (AccessibilityUIElement::scrollToGlobalPoint):
2759         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2760         (AccessibilityUIElement::scrollToMakeVisible):
2761         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2762         (AccessibilityUIElement::scrollToGlobalPoint):
2763
2764 2012-01-17  Enrica Casucci  <enrica@apple.com>
2765
2766         Missing NSForegroundColorAttributeName should be treated as black instead of transparent.
2767         https://bugs.webkit.org/show_bug.cgi?id=76490
2768         <rdar://problem/9460733>
2769
2770         Added TestWebKitAPI test.
2771         
2772         Reviewed by Dan Bernstein.
2773
2774         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added new test file.
2775         * TestWebKitAPI/Tests/mac/InspectorBar.mm: Added test that calls
2776         changeAttributes to the WebView.
2777         
2778 2012-01-18  Balazs Kelemen  <kbalazs@webkit.org>
2779
2780         [Qt] Consolidate layout test crash logging
2781         https://bugs.webkit.org/show_bug.cgi?id=75088
2782
2783         Reviewed by Simon Hausmann.
2784
2785         Move backtrace generating logic into WTFReportBacktrace
2786         and add a way to deinstall signal handlers if we know
2787         that we have already printed the backtrace.
2788
2789         * DumpRenderTree/qt/main.cpp:
2790         (crashHandler):
2791         (setupSignalHandlers):
2792         (WTFCrashHook):
2793         (main):
2794         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
2795         (WTR::crashHandler):
2796         (WTR::setupSignalHandlers):
2797         (WTR::crashHook):
2798         (WTR::InjectedBundle::platformInitialize):
2799
2800 2012-01-18  Ilya Tikhonovsky  <loislo@chromium.org>
2801
2802         Unreviewed build fix after r105256.
2803
2804         There was a cyclic dependency between self._port and self._host assigments.
2805
2806         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2807         (PerfTestsRunner.__init__):
2808
2809 2012-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2810
2811         Add new watchlist rule for the Qt build system
2812
2813         Reviewed by Simon Hausmann.
2814
2815         * Scripts/webkitpy/common/config/watchlist:
2816
2817 2012-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2818
2819         [Qt] Move OTHER_FILES from WebKit.pro to Tools.pro
2820
2821         Reviewed by Simon Hausmann.
2822
2823         * Tools.pro:
2824
2825 2012-01-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2826
2827         [Qt] Allow building only a subset of the WebKit sub-projects
2828
2829         Fixes build-jsc, broken in r104825.
2830
2831         We no longer need the intermediate buildQMakeQtProject(), as we're the
2832         only port using qmake, and buildQMakeProject is now highly Qt specific.
2833
2834         <http://webkit.org/b/76179>
2835
2836         Reviewed by Simon Hausmann.
2837
2838         * Scripts/build-jsc:
2839         * Scripts/build-webkit:
2840         * Scripts/webkitdirs.pm:
2841         (buildQMakeProjects):
2842
2843 2012-01-18  Ryosuke Niwa  <rniwa@webkit.org>
2844
2845         run-perf-tests should generate a json file that summaries the result
2846         https://bugs.webkit.org/show_bug.cgi?id=76504
2847
2848         Reviewed by Adam Barth.
2849
2850         Add an ability to generate a json file to run-perf-test in the preparation for perf bots.
2851         New option --outout-json-path specifies the json file's path, and --source-json-path specifies
2852         another json file to be merged into the generated json file.
2853
2854         Also fixed a bug that --build wasn't set by default.
2855
2856         * Scripts/webkitpy/layout_tests/port/test.py:
2857         (TestPort.webkit_base):
2858         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
2859         (PerfTestsRunner.__init__):
2860         (PerfTestsRunner):
2861         (PerfTestsRunner._parse_args):
2862         (PerfTestsRunner.run):
2863         (PerfTestsRunner._generate_json_if_specified):
2864         (PerfTestsRunner._process_chromium_style_test_result):
2865         (PerfTestsRunner._process_parser_test_result):
2866         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
2867         (MainTest):
2868         (create_runner):
2869         (test_run_test_set):
2870         (test_run_test_set_for_parser_tests):
2871         (test_run_test_set_with_summary_json):
2872         (test_collect_tests):
2873         (test_parse_args):
2874
2875 2012-01-17  Sheriff Bot  <webkit.review.bot@gmail.com>
2876
2877         Unreviewed, rolling out r105244.
2878         http://trac.webkit.org/changeset/105244
2879         https://bugs.webkit.org/show_bug.cgi?id=76518
2880
2881         broke Chromium Mac (Requested by rolandsteiner on #webkit).
2882
2883         * DumpRenderTree/AccessibilityUIElement.h:
2884         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
2885         (AccessibilityUIElement::AccessibilityUIElement):
2886         * DumpRenderTree/chromium/AccessibilityUIElement.h:
2887         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2888         (AccessibilityUIElement::removeSelection):
2889         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2890         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2891         (AccessibilityUIElement::removeSelection):
2892
2893 2012-01-17  Dominic Mazzoni  <dmazzoni@google.com>
2894
2895         Accessibility: Chromium needs methods to scroll an object into view or to a specific location.
2896         https://bugs.webkit.org/show_bug.cgi?id=73460
2897
2898         Reviewed by Chris Fleizach.
2899
2900         * DumpRenderTree/AccessibilityUIElement.h:
2901         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
2902         (AccessibilityUIElement::AccessibilityUIElement):
2903         (AccessibilityUIElement::scrollToMakeVisibleCallback):
2904         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocusCallback):
2905         (AccessibilityUIElement::scrollToGlobalPointCallback):
2906         * DumpRenderTree/chromium/AccessibilityUIElement.h:
2907         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2908         (AccessibilityUIElement::scrollToMakeVisible):
2909         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2910         (AccessibilityUIElement::scrollToGlobalPoint):
2911         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2912         (AccessibilityUIElement::scrollToMakeVisible):
2913         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2914         (AccessibilityUIElement::scrollToGlobalPoint):
2915         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
2916         (AccessibilityUIElement::scrollToMakeVisible):
2917         (AccessibilityUIElement::scrollToMakeVisibleWithSubFocus):
2918         (AccessibilityUIElement::scrollToGlobalPoint):
2919
2920 2012-01-17  Adam Barth  <abarth@webkit.org>
2921
2922         Enable tests on the mac-ews
2923         https://bugs.webkit.org/show_bug.cgi?id=76502
2924
2925         Reviewed by Simon Fraser.
2926
2927         Apple now has five machines running the mac-ews.  Lucas Forschler has
2928         asked that we try enabling tests to see if we have enough bandwidth.
2929
2930         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
2931         (MacEWS):
2932
2933 2012-01-17  Sanjoy Das  <sanjoy@playingwithpointers.com>
2934
2935         Add a check-webkit-style exception for GDBInterface.cpp
2936         https://bugs.webkit.org/show_bug.cgi?id=76187
2937
2938         Add an exception for Source/JavaScriptCore/jit/GDBInterface.cpp so
2939         that it does not fail the style check.
2940
2941         Reviewed by Martin Robinson.
2942
2943         * Scripts/webkitpy/style/checker.py: Tell the style checker to not run
2944         readability/naming tests on
2945         Source/JavaScriptCore/jit/GDBInterface.cpp.
2946         * Scripts/webkitpy/style/checker_unittest.py:
2947         (GlobalVariablesTest.test_path_rules_specifier): Add a test for
2948         the above exception.
2949
2950 2012-01-17  Hajime Morrita  <morrita@chromium.org>
2951
2952         https://bugs.webkit.org/show_bug.cgi?id=76245
2953         [Mac][DRT] should allow query string for the file URL.
2954
2955         Reviewed by Tony Chang.
2956
2957         Taught DRT about file:// as a url instead of a file path.
2958         We are now able to use a path file:///foo/bar.html?baz as a url
2959         with which DRT opens "/foo/bar.html" with "baz" as a query string.
2960
2961         * DumpRenderTree/mac/DumpRenderTree.mm:
2962         (runTest):
2963
2964 2012-01-17  Ojan Vafai  <ojan@chromium.org>
2965
2966         Check style on test_expectations.txt files before commit
2967         https://bugs.webkit.org/show_bug.cgi?id=76484
2968
2969         Reviewed by Adam Barth.
2970
2971         Unlike other style rules, if you get the syntax of the test_expectations.txt
2972         file wrong, the layout tests won't run. Also, this check is simple and only
2973         slows down committing if you actually modify one of the test_expectations.txt files.
2974
2975         * Scripts/webkitpy/tool/steps/commit.py:
2976         (Commit._check_test_expectations):
2977         (Commit.run):
2978         * Scripts/webkitpy/tool/steps/commit_unittest.py: Added.
2979         (CommitTest):
2980         (CommitTest.test_check_test_expectations):
2981
2982 2012-01-17  Viatcheslav Ostapenko  <ostapenko.viatcheslav@nokia.com>
2983
2984         [Qt] Debug build fails with debug qt5
2985         https://bugs.webkit.org/show_bug.cgi?id=76463
2986
2987         Reviewed by Simon Hausmann.
2988
2989         Add QNetworkCookie include in order to satisfy sizeof in QTypeInfo in debug builds.
2990
2991         * QtTestBrowser/cookiejar.h:
2992
2993 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
2994
2995         webkitpy: fix json import on linux
2996         https://bugs.webkit.org/show_bug.cgi?id=76481
2997
2998         Reviewed by Adam Barth.
2999
3000         For some reason, the version of simplejson autoinstalled on my linux
3001         box seems to be broken. We should really only be using
3002         simplejson if json isn't available anyway, so this change looks
3003         for json first.
3004
3005         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
3006         (JSONGeneratorTest.test_test_timings_trie):
3007
3008 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
3009
3010         Fix failures in test-webkitpy caused by r105177.
3011
3012         Unreviewed, build fix.
3013
3014         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
3015         (ChromiumLinuxPort._determine_driver_path_statically):
3016         (ChromiumLinuxPort._static_build_path):
3017
3018 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
3019
3020         webkitpy: make PortFactory.get() be fully data-driven
3021         https://bugs.webkit.org/show_bug.cgi?id=76360
3022
3023         Reviewed by Eric Seidel.
3024
3025         Refactor PortFactory.get() to be fully data-driven.
3026
3027         * Scripts/webkitpy/layout_tests/port/factory.py:
3028         (PortFactory):
3029         (PortFactory._default_port):
3030         (PortFactory.get):
3031         (PortFactory.get.in):
3032         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
3033         (FactoryTest.assert_port):
3034
3035 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
3036
3037         webkitpy: add determine_full_port_name(), clean up port.__init__()
3038         https://bugs.webkit.org/show_bug.cgi?id=76357
3039
3040         Reviewed by Adam Barth.
3041
3042         Each Port class is now required to implement a 'port_name' field
3043         that will match the start of all ports constructed by this class and
3044         a 'determine_full_port_name' field that will make the port
3045         fully-qualified.
3046
3047         * Scripts/webkitpy/layout_tests/port/apple.py:
3048         (ApplePort.__init__):
3049         * Scripts/webkitpy/layout_tests/port/base.py:
3050         (Port):
3051         (Port.determine_port_name):
3052         (Port.__init__):
3053         * Scripts/webkitpy/layout_tests/port/chromium.py:
3054         (ChromiumPort.for):
3055         (ChromiumPort.__init__):
3056         (ChromiumPort):
3057         (ChromiumPort._chromium_base_dir):
3058         (ChromiumPort.path_from_chromium_base):
3059         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
3060         (ChromiumGpuLinuxPort):
3061         (ChromiumGpuLinuxPort.__init__):
3062         (ChromiumGpuMacPort):
3063         (ChromiumGpuMacPort.__init__):
3064         (ChromiumGpuWinPort):
3065         (ChromiumGpuWinPort.__init__):
3066         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
3067         (ChromiumLinuxPort):
3068         (ChromiumLinuxPort.determine_port_name):
3069         (ChromiumLinuxPort.__init__):
3070         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
3071         (ChromiumMacPort):
3072         (ChromiumMacPort.determine_port_name):
3073         (ChromiumMacPort.__init__):
3074         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
3075         (ChromiumWinPort):
3076         (ChromiumWinPort.determine_port_name):
3077         (ChromiumWinPort.__init__):
3078         * Scripts/webkitpy/layout_tests/port/dryrun.py:
3079         (DryRunPort):
3080         (DryRunPort.determine_port_name):
3081         (DryRunPort.__init__):
3082         * Scripts/webkitpy/layout_tests/port/efl.py:
3083         (EflPort):
3084         * Scripts/webkitpy/layout_tests/port/factory.py:
3085         (PortFactory):
3086         (PortFactory._default_port):
3087         (PortFactory.get):
3088         (PortFactory.get.in):
3089         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
3090         (FactoryTest.assert_port):
3091         * Scripts/webkitpy/layout_tests/port/google_chrome.py:
3092         (GoogleChromeLinux32Port):
3093         (GoogleChromeLinux32Port.determine_port_name):
3094         (GoogleChromeLinux64Port):
3095         (GoogleChromeLinux64Port.determine_port_name):
3096         (GoogleChromeMacPort):
3097         (GoogleChromeMacPort.determine_port_name):
3098         (GoogleChromeWinPort):
3099         (GoogleChromeWinPort.determine_port_name):
3100         * Scripts/webkitpy/layout_tests/port/gtk.py:
3101         (GtkPort):
3102         * Scripts/webkitpy/layout_tests/port/mac.py:
3103         (MacPort):
3104         (MacPort.determine_port_name):
3105         (MacPort.__init__):
3106         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3107         (test_versions):
3108         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
3109         (MockDRTPort):
3110         (MockDRTPort.determine_port_name):
3111         (MockDRTPort.__init__):
3112         (MockDRT):
3113         (MockDRT.determine_port_name):
3114         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
3115         (PortTestCase.make_port):
3116         * Scripts/webkitpy/layout_tests/port/qt.py:
3117         (QtPort):
3118         (QtPort.determine_port_name):
3119         (QtPort.__init__):
3120         * Scripts/webkitpy/layout_tests/port/test.py:
3121         (TestPort):
3122         (TestPort.determine_port_name):
3123         (TestPort.__init__):
3124         * Scripts/webkitpy/layout_tests/port/webkit.py:
3125         (WebKitPort.__init__):
3126         * Scripts/webkitpy/layout_tests/port/win.py:
3127         (WinPort):
3128         (WinPort.determine_port_name):
3129         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
3130         (get_test_baselines.AllPlatformsPort.__init__):
3131         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
3132         (get_test_config.TestMacPort):
3133         (get_test_config):
3134
3135 2012-01-17  Dirk Pranke  <dpranke@chromium.org>
3136
3137         webkitpy: clean up port code in preparation for static port names
3138         https://bugs.webkit.org/show_bug.cgi?id=76356
3139
3140         Reviewed by Adam Barth.
3141
3142         To fix bug 76215 and be able to determine appropriate port names
3143         in webkitpy without actually constructing Port objects, we need
3144         to shuffle some logic in the chromium ports to be able to figure
3145         out whether we built DRT in 32 or 64-bit mode.
3146
3147         * Scripts/webkitpy/layout_tests/port/chromium.py:
3148         (ChromiumPort):
3149         (ChromiumPort._chromium_base_dir):
3150         (ChromiumPort.__init__):
3151         (ChromiumPort.path_from_chromium_base):
3152         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
3153         (ChromiumLinuxPort):
3154         (ChromiumLinuxPort._determine_driver_path_statically):
3155         (ChromiumLinuxPort._static_build_path):
3156         (ChromiumLinuxPort._determine_architecture):
3157         (ChromiumLinuxPort.__init__):
3158         (ChromiumLinuxPort._build_path):
3159
3160 2012-01-17  Hao Zheng  <zhenghao@chromium.org>
3161
3162         Do not remove pixel failures for ref tests.
3163         https://bugs.webkit.org/show_bug.cgi?id=76243
3164
3165         Reviewed by Tony Chang.
3166
3167         'NRWT --no-pixel-tests' complains when reftests are expected to
3168         be image mismatch.
3169
3170         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3171         (Manager._update_summary_with_result):
3172         * Scripts/webkitpy/layout_tests/models/test_failures.py:
3173         (is_reftest_failure):
3174         (determine_result_type):
3175
3176 2012-01-17  Balazs Ankes  <Ankes.Balazs@stud.u-szeged.hu>
3177
3178         REGRESSION(r81225): ORWT should ignore reftests
3179         https://bugs.webkit.org/show_bug.cgi?id=67936
3180
3181         Reviewed by Csaba Osztrogonác.
3182
3183         * Scripts/old-run-webkit-tests:
3184         (isUsedInReftest): $filename should be the first parameter
3185         (findTestsToRun): run test if it isn't reftest
3186
3187 2012-01-17  Gustavo Noronha Silva  <gns@gnome.org>
3188
3189         Unreviewed. Fix python unit test I broke with r105142
3190
3191         * Scripts/webkitpy/common/config/ports_unittest.py:
3192         (WebKitPortTest.test_gtk_port):
3193
3194 2012-01-17  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3195
3196         [Qt] Don't set the 'primary' flag when mocking touch-points in MiniBrowser
3197
3198         Qt 5 no longer has that flag.
3199
3200         Reviewed by Simon Hausmann.
3201
3202         * MiniBrowser/qt/MiniBrowserApplication.cpp:
3203         (MiniBrowserApplication::notify):
3204
3205 2012-01-16  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
3206
3207         GTK+ EWS needs to run update-webkitgtk-libs after applying a patch
3208         https://bugs.webkit.org/show_bug.cgi?id=75857
3209
3210         Reviewed by Adam Barth.
3211
3212         * Scripts/webkitdirs.pm:
3213         (buildAutotoolsProject): move cleaning the build to happen early,
3214         and handle --update-gtk right after that
3215         * Scripts/webkitpy/common/config/ports.py:
3216         (GtkPort.build_webkit_command): pass --update-gtk as an option to
3217         build-webkit, so that it is used in the EWS
3218
3219 2012-01-16  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3220
3221         [Qt] Don't add _debug postfix to targets when building debug only
3222
3223         Reviewed by Simon Hausmann.
3224
3225         * qmake/mkspecs/features/default_post.prf:
3226         * qmake/mkspecs/features/functions.prf:
3227
3228 2012-01-16  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3229
3230         [Qt] Add more files to project-files for easier navigation in Qt Creator
3231
3232         Reviewed by Simon Hausmann.
3233
3234         * Tools.pro:
3235         * qmake/config.tests/README: Added.
3236
3237 2012-01-17  Philippe Normand  <pnormand@igalia.com>
3238
3239         [GTK] fast/dom/constructed-objects-prototypes.html fails
3240         https://bugs.webkit.org/show_bug.cgi?id=75432
3241
3242         Reviewed by Martin Robinson.
3243
3244         * Scripts/build-webkit: Enable video-track in GTK.
3245
3246 2012-01-16  Takashi Toyoshima  <toyoshim@chromium.org>
3247
3248         Add toyoshim as a committer.
3249         https://bugs.webkit.org/show_bug.cgi?id=76390
3250
3251         Reviewed by Hajime Morita.
3252
3253         * Scripts/webkitpy/common/config/committers.py:
3254
3255 2012-01-16  Brent Fulgham  <bfulgham@webkit.org>
3256
3257         [WinCairo] Build fix after r104155. Several debug targets were not
3258         properly including the '*Common' property sheets.
3259
3260         * DumpRenderTree/win/DumpRenderTreeLauncherDebugCairoCFLite.vsprops:
3261         * DumpRenderTree/win/ImageDiffLauncherDebugCairoCFLite.vsprops:
3262         * MiniBrowser/Configurations/MiniBrowserLauncherDebugCairoCFLite.vsprops:
3263         * WebKitTestRunner/win/WebKitTestRunnerLauncherDebugCairoCFLite.vsprops:
3264         * WinLauncher/WinLauncherLauncherDebugCairoCFLite.vsprops:
3265
3266 2012-01-16  Csaba Osztrogonác  <ossy@webkit.org>
3267
3268         REGRESSION(r105072): It broke committers_unittest.py
3269         https://bugs.webkit.org/show_bug.cgi?id=76396
3270
3271         Reviewed by Adam Barth.
3272
3273         * Scripts/webkitpy/common/config/committers_unittest.py:
3274         (CommittersTest._assert_fuzz_match): It should handle more names.
3275         (CommittersTest.test_contributors_by_fuzzy_match_with_legacy_names): Fix the testcase.
3276
3277 2012-01-16  Mario Sanchez Prada  <msanchez@igalia.com>
3278
3279         [GTK] Include gtk+3 in the jhbuild modules
3280         https://bugs.webkit.org/show_bug.cgi?id=76284
3281
3282         Reviewed by Philippe Normand.
3283
3284         * gtk/jhbuild.modules: Add some gtk+3 to the jhbuild moduleset.
3285
3286 2012-01-16  Dan Winship  <danw@gnome.org>
3287
3288         Add myself as a committer
3289
3290         Unreviewed
3291
3292         * Scripts/webkitpy/common/config/committers.py:
3293
3294 2012-01-16  Raphael Kubo da Costa  <kubo@profusion.mobi>
3295
3296         Unreviewed; fix the build when --as-needed and --no-copy-dt-needed-entries are passed to the linker.
3297
3298         Some recent versions of a few Linux distributions have started passing
3299         --as-needed and --no-copy-dt-needed-entries by default to the linker,
3300         which broke the build when SHARED_CORE was turned on, as libwtf_efl.so
3301         needed a few symbols from libjavascriptcore_efl.so.
3302
3303         * DumpRenderTree/efl/CMakeLists.txt: Explicitly pass WTF_LIBRARY_NAME
3304         before JavaScriptCore_LIBRARY_NAME in DumpRenderTree_LIBRARIES so the
3305         libraries are searched in the right order by the linker.
3306
3307 2012-01-16  Philippe Normand  <pnormand@igalia.com>
3308
3309         [GTK] set audio resources path in NRTW
3310         https://bugs.webkit.org/show_bug.cgi?id=76380
3311
3312         Reviewed by Martin Robinson.
3313
3314         * Scripts/webkitpy/layout_tests/port/gtk.py:
3315         (GtkPort.setup_environ_for_server): Set $AUDIO_RESOURCES_PATH so
3316         the uninstalled audio resources are used during webaudio tests.
3317
3318 2012-01-16  Carlos Garcia Campos  <cgarcia@igalia.com>
3319
3320         [GTK] ldflags and cflags should take precedence over the existing flags in gtkdoc.py
3321         https://bugs.webkit.org/show_bug.cgi?id=76369
3322
3323         Reviewed by Xan Lopez.
3324
3325         If there's an installed version of webkit and there are new
3326         symbols in the build tree, gtkdoc-scanobj fails due to undefined
3327         references, because the env ldflags are given before the local
3328         ones.
3329
3330         * gtk/gtkdoc.py:
3331         (GTKDoc._run_gtkdoc_scangobj):
3332
3333 2012-01-15  Pablo Flouret  <pablof@motorola.com>
3334
3335         Fix compilation errors on build-webkit --debug --no-video on mac.
3336         https://bugs.webkit.org/show_bug.cgi?id=75867
3337
3338         Reviewed by Philippe Normand.
3339
3340         Add --[no-]video-track to build-webkit, since video track feature
3341         depends on video.
3342
3343         * Scripts/build-webkit:
3344
3345 2012-01-14  James Robinson  <jamesr@chromium.org>
3346
3347         [chromium] Failing webkit_unit_tests does not make the cr-linux EWS bubble red
3348         https://bugs.webkit.org/show_bug.cgi?id=76313
3349
3350         Reviewed by Adam Barth.
3351
3352         The exit code of this script should be the exit code of webkit_unit_tests. Tested manually.
3353
3354         * Scripts/run-chromium-webkit-unit-tests:
3355
3356 2012-01-14  Sheriff Bot  <webkit.review.bot@gmail.com>
3357
3358         Unreviewed, rolling out r105017.
3359         http://trac.webkit.org/changeset/105017
3360         https://bugs.webkit.org/show_bug.cgi?id=76333
3361
3362         broke the build (Requested by philn-tp on #webkit).
3363
3364         * gtk/jhbuild.modules:
3365
3366 2012-01-14  Mario Sanchez Prada  <msanchez@igalia.com>
3367
3368         [GTK] Include gtk+3 in the jhbuild modules
3369         https://bugs.webkit.org/show_bug.cgi?id=76284
3370
3371         Reviewed by Philippe Normand.
3372
3373         * gtk/jhbuild.modules: Add some gtk+3 to the jhbuild moduleset.
3374
3375 2012-01-13  Adam Barth  <abarth@webkit.org>
3376
3377         Remove --dry-run support from webkit-patch
3378         https://bugs.webkit.org/show_bug.cgi?id=76300
3379
3380         Reviewed by Dirk Pranke.
3381
3382         This is the rest of the patch from Bug 76300.  I landed it separtely
3383         because main.py had some stray conflict markers that I needed to clean
3384         up first.
3385
3386         * Scripts/webkitpy/test/main.py:
3387         (Tester.parse_args):
3388         (Tester._run_tests):
3389
3390 2012-01-13  Adam Barth  <abarth@webkit.org>
3391
3392         test-webkitpy does not run
3393         https://bugs.webkit.org/show_bug.cgi?id=76318
3394
3395         Unreviewed.
3396
3397         Looks like some conflict markers got left in the file.
3398
3399         * Scripts/webkitpy/test/main.py:
3400         (Tester.run):
3401
3402 2012-01-13  Adam Barth  <abarth@webkit.org>
3403
3404         Remove --dry-run support from webkit-patch
3405         https://bugs.webkit.org/show_bug.cgi?id=76300
3406
3407         Reviewed by Dirk Pranke.
3408
3409         We don't use this functionality for anything, and it doesn't really
3410         work.  Someone in #webkit tried using it and and got sad/confused.  We
3411         should just remove it.
3412
3413         * Scripts/webkitpy/common/checkout/scm/git.py:
3414         (Git.push_local_commits_to_server):
3415         * Scripts/webkitpy/common/checkout/scm/scm.py:
3416         (SCM.__init__):
3417         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
3418         (_shared_test_commit_with_message):
3419         (GitSVNTest.test_commit_text_parsing):
3420         * Scripts/webkitpy/common/checkout/scm/svn.py:
3421         (SVN.commit_with_message):
3422         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
3423         (Bugzilla.__init__):
3424         (Bugzilla.quips):
3425         (Bugzilla.authenticate):
3426         (Bugzilla.add_attachment_to_bug):
3427         (Bugzilla.add_patch_to_bug):
3428         (Bugzilla.create_bug):
3429         (Bugzilla.clear_attachment_flags):
3430         (Bugzilla.set_flag_on_attachment):
3431         (Bugzilla.obsolete_attachment):
3432         (Bugzilla.add_cc_to_bug):
3433         (Bugzilla.post_comment_to_bug):
3434         (Bugzilla.close_bug_as_fixed):
3435         (Bugzilla.reassign_bug):
3436         (reopen_bug):
3437         * Scripts/webkitpy/test/main.py:
3438         (Tester.parse_args):
3439         (Tester.run_tests):
3440
3441 2012-01-13  David Levin  <levin@chromium.org>
3442
3443         check-webkit-style: should encourage the use of Own* classes for Windows DC.
3444         https://bugs.webkit.org/show_bug.cgi?id=76227
3445
3446         Reviewed by Dirk Pranke.
3447
3448         * Scripts/webkitpy/style/checkers/cpp.py:
3449         (check_for_leaky_patterns): The new check.
3450         (process_line): Added a call to the new check.
3451         (CppChecker): Added the new error type.
3452         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
3453         (CppStyleTestBase):
3454         (CppStyleTestBase.perform_leaky_pattern_check):
3455         The check for only leaky pattern errors.
3456         (LeakyPatternTest): Test cases.
3457
3458 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
3459
3460         test-webkitpy: clean up handling of test directories, QueueStatusServer
3461         https://bugs.webkit.org/show_bug.cgi?id=76238
3462
3463         Reviewed by Adam Barth.
3464
3465         This patch cleans up a bunch of hacks related to how we find all
3466         of the python unit tests and set up the sys.path variable in
3467         test-webkitpy.
3468
3469         This patch changes how test-webkitpy works slightly in that we
3470         will now look for test files under Tools/Scripts and Source/WebKit2/Scripts
3471         rather than just Tools/Scripts/webkitpy and Source/WebKit2/Scripts/webkit.
3472         This greatly simplifies the mapping of paths to module names and
3473         appears to have no significant performance impact or other drawbacks.
3474
3475         There are other ways that we could remove all of the hacks and
3476         still have a more flexible mapping of paths to module names, but
3477         the ones I've thought of so far add a decent amount of complexity to
3478         the implementation and don't seem to buy us much.
3479         
3480         * QueueStatusServer/__init__.py:
3481         * Scripts/test-webkitpy:
3482         * Scripts/webkitpy/test/main.py:
3483         (Tester._modules_from_paths):
3484         (Tester._is_module):
3485
3486 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
3487
3488         test-webkitpy: should support classes and individual test names as well as modules
3489         https://bugs.webkit.org/show_bug.cgi?id=76233
3490
3491         Reviewed by Adam Barth.
3492
3493         The default unittest.main() supports a list of individual
3494         TestCase classes or individual tests as well as modules. This
3495         change adds support for the same to test-webkitpy, so now you
3496         can say
3497
3498           % test-wekbitpy webkitpy.test.cat_unittest.CatTest
3499
3500         and
3501
3502           % test-wekbitpy webkitpy.test.cat_unittest.CatTest.test_basic
3503
3504         in addition to
3505
3506           % test-wekbitpy webkitpy.test.cat_unittest
3507
3508         * Scripts/test-webkitpy:
3509         * Scripts/webkitpy/test/cat_unittest.py:
3510         * Scripts/webkitpy/test/main.py:
3511         (Tester.parse_args):
3512         (Tester.run):
3513         (Tester):
3514         (Tester._find_modules):
3515         (Tester._exclude):
3516         (Tester._run_tests):
3517         (Tester._is_module):
3518         (Tester._log_exception):
3519
3520 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
3521
3522         test-webkitpy: add basic code coverage support
3523         https://bugs.webkit.org/show_bug.cgi?id=76201
3524
3525         Reviewed by Adam Barth.
3526
3527         This test adds basic code coverage support to test-webkitpy; if
3528         you pass the -c flag, we will generate a new coverage file. We
3529         can start simply with this and add more features as necessary.
3530
3531         This patch requires coverage to be already installed on the
3532         machine. I am not auto-installing this for now.
3533
3534         * Scripts/webkitpy/test/main.py:
3535         (Tester.parse_args):
3536         (Tester.run_tests):
3537
3538 2012-01-13  Dirk Pranke  <dpranke@chromium.org>
3539
3540         StatusBubbleTest.test_build_bubble is failing
3541         https://bugs.webkit.org/show_bug.cgi?id=76222
3542
3543         Reviewed by Adam Barth.
3544
3545         * QueueStatusServer/handlers/statusbubble_unittest.py:
3546         (StatusBubbleTest.test_build_bubble):
3547
3548 2012-01-13  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
3549
3550         nrwt should use multiple http shards
3551         https://bugs.webkit.org/show_bug.cgi?id=75958
3552
3553         Add --max-locked-shards option to the nrwt for manual set
3554         the maximum number of the locked shards
3555
3556         Reviewed by Dirk Pranke.
3557
3558         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3559         (Manager._max_locked_shards):
3560         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3561         (ShardingTests.get_shards):
3562         (ShardingTests.test_multiple_locked_shards):
3563         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3564         (parse_args):
3565
3566 2012-01-13  Raphael Kubo da Costa  <kubo@profusion.mobi>
3567
3568         prepare-ChangeLog: Look for a text editor in a few more variables.
3569         https://bugs.webkit.org/show_bug.cgi?id=76289
3570
3571         Reviewed by Ryosuke Niwa.
3572
3573         A text editor was looked for only in the CHANGE_LOG_EDITOR and
3574         CHANGE_LOG_EDIT_APPLICATION environment variables before falling back
3575         to the OS X-specific open(1) command.
3576
3577         As there does not seem to be a common way of looking for a text editor
3578         among the scripts in Tools/Scripts right now, just look at $VISUAL and
3579         $EDITOR (set much more often than the other variables in Unix systems)
3580         before falling back to open(1).
3581
3582         * Scripts/prepare-ChangeLog:
3583         (openChangeLogs):
3584
3585 2012-01-13  Gabor Rapcsanyi  <rgabor@webkit.org>
3586
3587         [GTK] gtkdoc.py ignoring LDFLAGS and CFLAGS environment variables
3588         https://bugs.webkit.org/show_bug.cgi?id=76269
3589
3590         Reviewed by Martin Robinson.
3591
3592         * gtk/gtkdoc.py:
3593         (GTKDoc._run_gtkdoc_scangobj):
3594
3595 2012-01-13  Csaba Osztrogonác  <ossy@webkit.org>
3596
3597         [Qt][WK2]REGRESSION(r104881):It broke hundreds of tests
3598         https://bugs.webkit.org/show_bug.cgi?id=76247
3599
3600         Reviewed by Kenneth Rohde Christiansen.
3601
3602         * WebKitTestRunner/TestController.cpp:
3603         (WTR::TestController::resetStateToConsistentValues): Disable mock scrollbars on Qt by default. (workaround)
3604
3605 2012-01-13  Hajime Morrita  <morrita@chromium.org>
3606
3607         [Mac][DRT] Should support --no-timeout
3608         https://bugs.webkit.org/show_bug.cgi?id=76242
3609
3610         Reviewed by Darin Adler.
3611
3612         * DumpRenderTree/mac/DumpRenderTree.mm:
3613         (initializeGlobalsFromCommandLineOptions): Added "--no-timeout" to the options list.
3614         (setWaitToDumpWatchdog): Added to encapsulate the waitToDumpWatchdog global variable.
3615         (shouldSetWaitToDumpWatchdog): Added to check the useTimeoutWatchdog flag set by the option.
3616         * DumpRenderTree/mac/DumpRenderTreeMac.h:
3617         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
3618         (LayoutTestController::setWaitToDump): Now calls shouldSetWaitToDumpWatchdog() to check the flag.
3619
3620 2012-01-12  Simon Fraser  <simon.fraser@apple.com>
3621
3622         Mitigate scrollbar differences when running pixel tests
3623         https://bugs.webkit.org/show_bug.cgi?id=67217
3624
3625         Reviewed by Dan Bernstein.
3626         
3627         Enable mock scrollbars by default for tests.
3628
3629         * DumpRenderTree/mac/DumpRenderTree.mm:
3630         (resetDefaultsToConsistentValues):
3631         * WebKitTestRunner/TestController.cpp:
3632         (WTR::TestController::resetStateToConsistentValues):
3633
3634 2012-01-12  Ryosuke Niwa  <rniwa@webkit.org>
3635
3636         Need a script to run tests in PerformanceTests
3637         https://bugs.webkit.org/show_bug.cgi?id=76132
3638
3639         Reviewed by Adam Barth.
3640
3641         Add run-perf-tests to run performance tests using parser/resources/runner.js.
3642         Unfortunately, there isn't an easy way of telling which test uses which format
3643         so hard-code directory that uses Chromium perf-bot style (only inspector perf. tests for now).
3644
3645         All test outputs are re-formatted to match Chromium perf-bot style.
3646
3647         * Scripts/run-inspector-perf-tests.py:
3648         * Scripts/run-perf-tests: Added.
3649         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
3650         (PerfTestsRunner):
3651         (PerfTestsRunner.__init__):
3652         (PerfTestsRunner._collect_tests):
3653         (PerfTestsRunner.run):
3654         (PerfTestsRunner._print_status):
3655         (PerfTestsRunner._run_tests_set):
3656         (PerfTestsRunner._process_chromium_style_test_result):
3657         (PerfTestsRunner._should_ignore_line_in_parser_test_result):
3658         (PerfTestsRunner._process_parser_test_result):
3659         (PerfTestsRunner._run_single_test):
3660         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
3661         (MainTest.TestDriver.run_test):
3662         (create_runner):
3663         (run_test):
3664         (test_run_test_set):
3665         (test_run_test_set_for_parser_tests):
3666         (test_collect_tests):
3667
3668 2012-01-12  Dirk Pranke  <dpranke@chromium.org>
3669
3670         webkitpy: clean up port unit tests in preparation for making port_name mandatory
3671         https://bugs.webkit.org/show_bug.cgi?id=76123
3672
3673         Reviewed by Adam Barth.
3674
3675         This change updates the unit tests to be more consistent in how
3676         ports are created, in preparation for making port_name a
3677         mandatory parameter to the constructor and removing any version
3678         interpretation from it.
3679
3680         There should be no functional changes in this patch, but some
3681         tests that are no longer relevant have been removed or reworked.
3682
3683         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3684         (PortTest.test_get_option__default):
3685         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
3686         (ChromiumLinuxPortTest):
3687         (ChromiumLinuxPortTest.assert_architecture):
3688         (ChromiumLinuxPortTest.test_determine_architecture_fails):
3689         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py: