f6a07e1ca11ea34757df8e0cb3c694445184996e
[WebKit-https.git] / Tools / ChangeLog
1 2011-12-13  Peter Beverloo  <peter@chromium.org>
2
3         [Chromium] Pull in FreeType and decrease check-out complexity for Android
4         https://bugs.webkit.org/show_bug.cgi?id=74401
5
6         Decrease the check-out complexity by removing the need to first run
7         "update-webkit-chromium --chromium" prior to running it with
8         "--chromium-android", which was caused by an unversioned third_party/
9         directory being created. This moves the Android NDK to Source/WebKit/
10         chromium/, coming from third_party/ in that directory.
11
12         Reviewed by Adam Barth.
13
14         * Scripts/update-webkit-chromium:
15         * Scripts/webkitdirs.pm:
16         (buildChromiumMakefile):
17
18 2011-12-13  Andy Wingo  <wingo@igalia.com> and Martin Robinson <mrobinson@igalia.com>
19
20         build-jsc passing incorrect args to buildGtkProject
21         https://bugs.webkit.org/show_bug.cgi?id=74308
22
23         Reviewed by Gustavo Noronha Silva.
24
25         * Scripts/build-webkit: Instead of invoking buildGtkProject with
26         another argument about webkit2, unshift a --disable-webkit2
27         argument onto the buildArgs.
28         * Scripts/webkitdirs.pm: Remove $enableWebKit2 arg. Only save autogen.sh
29         arguments for WebKit to prevent unnecessary reconfiguration when build-webkit
30         and build-jsc are run one after the other.
31
32 2011-12-13  Csaba Osztrogonác  <ossy@webkit.org>
33
34         [Qt] REGRESSION(102679): It broke Qt5-WK1 build
35         https://bugs.webkit.org/show_bug.cgi?id=74413
36
37         Unreviewed buildfix.
38
39         * qmake/mkspecs/features/webcore.prf: Temporarily revert a part of r102679 until a proper fix.
40
41 2011-12-13  Csaba Osztrogonác  <ossy@webkit.org>
42
43         [Qt] Unreviewed speculativ debug buildfix after r102679.
44
45         * QtTestBrowser/QtTestBrowser.pro:
46         * QtTestBrowser/main.cpp:
47
48 2011-12-13  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
49
50         [Qt] Get rid of layering violations in includes
51
52         WebKit/qt/API and WebKit/qt/WebCoreSupport should not be included
53         in the webcore.prf, but rather in each target that specificly needs
54         headers in these location. We used to include them directly in webcore
55         since we had layering violations between WebCore and WebKit, but now
56         that they are gone there's no reason to do that.
57
58         Reviewed by Simon Hausmann.
59
60         * DumpRenderTree/qt/DumpRenderTree.pro:
61         * QtTestBrowser/launcherwindow.h:
62         * WebKitTestRunner/InjectedBundle/Target.pri:
63         * qmake/mkspecs/features/webcore.prf:
64         * qmake/mkspecs/features/webkit2.prf:
65
66 2011-12-13  János Badics  <dicska@gmail.com>
67
68         [NRWT] It crashes intermittently when there are WebProcess crashes
69         https://bugs.webkit.org/show_bug.cgi?id=73451
70
71         Use self.poll() instead of self._proc.poll() because it checks if self._proc is None or not.
72
73         Reviewed by Kenneth Rohde Christiansen.
74
75         * Scripts/webkitpy/layout_tests/port/server_process.py:
76         (ServerProcess._check_for_crash):
77
78 2011-12-13  Carlos Garcia Campos  <cgarcia@igalia.com>
79
80         Unreviewed. Fix GTK+ API docs build.
81
82         * gtk/generate-gtkdoc:
83         (get_webkit2_options): Ignore WebKitUIClient since it's private
84         API.
85
86 2011-12-13  Kenichi Ishibashi  <bashi@chromium.org>
87
88         Reviewed, removed one of my email addresses to use the apprpriate one.
89
90         * Scripts/webkitpy/common/config/committers.py:
91
92 2011-12-12  Ádám Kallai  <Kallai.Adam@stud.u-szeged.hu>
93
94         [Qt] Fix skip list search path and expected file search path in ORWT
95         https://bugs.webkit.org/show_bug.cgi?id=74281
96
97         Expected file search path fixed.
98
99         Print all used skipped list and expected search path in verbose mode similar to NRWT.
100
101         Reviewed by Csaba Osztrogonác.
102
103         * Scripts/old-run-webkit-tests:
104         (buildPlatformResultHierarchy):
105         (buildPlatformTestHierarchy):
106         (readSkippedFiles):
107
108 2011-12-12  Nándor Huszka  <huszka.nandor@stud.u-szeged.hu>
109
110         [Qt] Missing layoutTestController.dumpWillCacheResponse
111         https://bugs.webkit.org/show_bug.cgi?id=74287
112
113         Reviewed by Csaba Osztrogonác.
114
115         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
116         (LayoutTestController::reset):
117         (LayoutTestController::dumpWillCacheResponse):
118         * DumpRenderTree/qt/LayoutTestControllerQt.h:
119
120 2011-12-12  Martin Robinson  <mrobinson@igalia.com>
121
122         [GTK] platform/gtk/fonts/custom-font-missing-glyphs.html fails on 64-bit Debug
123         https://bugs.webkit.org/show_bug.cgi?id=73771
124
125         Reviewed by Gustavo Noronha Silva.
126
127         When using a fallback font during testing, always use DejaVu Sans. This prevents
128         falling back to DejaVu Serif on some systems.
129
130         * DumpRenderTree/gtk/fonts/fonts.conf: Fall back to DejaVu Sans.
131
132 2011-12-08  Pierre Rossi  <pierre.rossi@gmail.com>
133
134         Add myself as a committer
135
136         * Scripts/webkitpy/common/config/committers.py:
137
138 2011-12-12  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
139
140         [Qt] [WK2] Support customizing popup menus with QML
141         https://bugs.webkit.org/show_bug.cgi?id=73560
142
143         Reviewed by Tor Arne Vestbø.
144
145         Add an Item Selector to our WebView using the experimental API.
146
147         * MiniBrowser/qt/MiniBrowser.pro:
148         * MiniBrowser/qt/MiniBrowser.qrc:
149         * MiniBrowser/qt/qml/BrowserWindow.qml:
150         * MiniBrowser/qt/qml/ItemSelector.qml: Added.
151
152 2011-12-12  Alexander Færøy  <alexander.faeroy@nokia.com>
153
154         [Qt] MiniBrowser should only visualize mock touch points when in non-desktop mode
155         https://bugs.webkit.org/show_bug.cgi?id=74283
156
157         Reviewed by Simon Hausmann.
158
159         * MiniBrowser/qt/MiniBrowserApplication.cpp:
160         (MiniBrowserApplication::sendTouchEvent):
161
162 2011-12-12  Kentaro Hara  <haraken@chromium.org>
163
164         [Refactoring] In prepare-ChangeLog, move into a method the top-level code
165         to show ChangeLogs diff and open ChangeLogs in an editor
166         https://bugs.webkit.org/show_bug.cgi?id=74266
167
168         Reviewed by Ryosuke Niwa.
169
170         We are planning to write unit-tests for prepare-ChangeLog
171         in a run-leaks_unittest manner. This patch is one of the incremental
172         refactorings to remove all top-level code and global variables from
173         prepare-ChangeLog.
174
175         * Scripts/prepare-ChangeLog: Moved the top-level code to show ChangeLogs diff into printDiff().
176         Moved the top-level code to open ChangeLogs in an editor into openChangeLogs().
177         Renamed @changed_files to @changedFiles.
178         Renamed %conflict_files to %conflictFiles.
179         (generateFunctionLists):
180         (printDiff):
181         (openChangeLogs):
182
183 2011-12-11  Kentaro Hara  <haraken@chromium.org>
184
185         [Refactoring] Move top-level code to resolve conflicted ChangeLogs into a method
186         https://bugs.webkit.org/show_bug.cgi?id=74257
187
188         Reviewed by Ryosuke Niwa.
189
190         We are planning to write unit-tests for prepare-ChangeLog
191         in a run-leaks_unittest/ manner. This patch is one of the incremental
192         refactorings to remove all top-level code and global variables from
193         prepare-ChangeLog.
194
195         * Scripts/prepare-ChangeLog: Moved top-level code to get the latest ChangeLogs
196         into getLatestChangeLogs(), and moved top-level code to resolve conflicted ChangeLogs
197         into resolveConflictedChangeLogs().
198         (getLatestChangeLogs):
199         (resolveConflictedChangeLogs):
200
201 2011-12-11  Kentaro Hara  <haraken@chromium.org>
202
203         [Refactoring] Move top-level code to generate a new ChangeLog into a method
204         https://bugs.webkit.org/show_bug.cgi?id=74253
205
206         Reviewed by David Kilzer.
207
208         The objective is to make prepare-ChangeLog a loadable Perl module for unit testing,
209         which requires to remove top-level code and global variables. This patch is one of
210         the incremental refactorings for that.
211
212         * Scripts/prepare-ChangeLog: Moved top-level code to generate a new ChangeLog into generateNewChangeLogs().
213         (generateFunctionLists):
214         (findChangeLogs):
215         (generateNewChangeLogs):
216         (generateFileList): Removed an unnecessary variable $didChangeRegressionTests.
217
218 2011-12-10  Kentaro Hara  <haraken@chromium.org>
219
220         [Refactoring] In prepare-ChangeLog, move top-level code to find ChangeLogs into a method
221         https://bugs.webkit.org/show_bug.cgi?id=74175
222
223         Reviewed by Ryosuke Niwa.
224
225         The objective is to make prepare-ChangeLog a loadable Perl module for unit testing.
226         This requires to remove top-level code. This patch is one of the incremental refactorings
227         for that.
228
229         * Scripts/prepare-ChangeLog: Moved top-level code to find ChangeLogs into findChangeLogs().
230         (findChangeLogs):
231
232 2011-12-10  Andreas Kling  <kling@webkit.org>
233
234         Add my Apple e-mail to committers.py.
235
236         * Scripts/webkitpy/common/config/committers.py:
237
238 2011-12-09  Dirk Pranke  <dpranke@chromium.org>
239
240         chromium win build: revert build_webkit_exes_from_webkit_gyp change
241         https://bugs.webkit.org/show_bug.cgi?id=74225
242
243         This change sets build_webkit_exes_from_webkit_gyp back to 1;
244         the "All.gyp" solution doesn't work on windows.
245
246         * Scripts/webkitdirs.pm:
247         (buildChromium):
248
249 2011-12-09  Hugo Parente Lima  <hugo.lima@openbossa.org>
250
251         [Qt] Click's count is limited to three continuous clicks.
252         https://bugs.webkit.org/show_bug.cgi?id=45666
253
254         Reviewed by Kenneth Rohde Christiansen.
255
256         Update m_time at every call to leapForward, so double clicks
257         event are correctly sent by EventSender.
258
259         * WebKitTestRunner/qt/EventSenderProxyQt.cpp:
260         (WTR::EventSenderProxy::updateClickCountForButton):
261         (WTR::EventSenderProxy::leapForward):
262
263 2011-12-09  Kentaro Hara  <haraken@chromium.org>
264
265         [Refactoring] Reduce top-level code in prepare-ChangeLog
266         https://bugs.webkit.org/show_bug.cgi?id=74172
267
268         Reviewed by Ryosuke Niwa.
269
270         The objective is to make prepare-ChangeLog a loadable Perl module for unit testing.
271         This requires to remove top-level code. This patch is one of the incremental refactorings
272         for that.
273
274         * Scripts/prepare-ChangeLog: Moved some top-level code into generateFunctionLists().
275         (generateFunctionLists):
276
277 2011-12-09  Sam Weinig  <sam@webkit.org>
278
279         Expose a WKConnectionRef which represents the connection to/from the WebProcess/UIProcess
280         https://bugs.webkit.org/show_bug.cgi?id=74218
281
282         Reviewed by Anders Carlsson.
283
284         Add a test of the WKConnectionRef API.
285
286         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
287         * TestWebKitAPI/Tests/WebKit2/WKConnection.cpp: Added.
288         (TestWebKitAPI::didCreateConnection):
289         (TestWebKitAPI::connectionDidReceiveMessage):
290         (TestWebKitAPI::connectionDidClose):
291         (TestWebKitAPI::TEST):
292         * TestWebKitAPI/Tests/WebKit2/WKConnection_Bundle.cpp: Added.
293         (TestWebKitAPI::connectionDidReceiveMessage):
294         (TestWebKitAPI::WKConnectionTest::WKConnectionTest):
295         (TestWebKitAPI::WKConnectionTest::initialize):
296
297 2011-12-09  David Levin  <levin@chromium.org>
298
299         Hash* iterators should allow comparison between const and const versions.
300         https://bugs.webkit.org/show_bug.cgi?id=73370
301
302         Reviewed by Darin Adler.
303
304         * TestWebKitAPI/TestWebKitAPI.gypi: Add the new test file to the build.
305         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Ditto.
306         * TestWebKitAPI/win/TestWebKitAPI.vcproj: Ditto.
307         * TestWebKitAPI/Tests/WTF/HashMap.cpp: Added.
308         (TestWebKitAPI::TEST): Add a test for the new functionality.
309
310 2011-12-09  Dirk Pranke  <dpranke@chromium.org>
311
312         chromium: visual studio projects are busted when build_webkit_exes_from_webkit_gyp==0
313         https://bugs.webkit.org/show_bug.cgi?id=74212
314
315         Reviewed by Tony Chang.
316
317         The change introduced in r102201 to split the executables out of
318         WebKit.gyp apparently doesn't work right on Windows, since
319         WebKit.gyp and Tools.gyp refer to dependent projects by
320         different relative paths (it appears gyp doesn't normalize the
321         paths). This change moves Tools.gyp to be in the same directory
322         as WebKit.gyp (and All.gyp), which should fix this and
323         conveniently will also solve the problem of how to add Tools.gyp
324         to the chromium DEPS file.
325
326         * Tools.gyp: Renamed to ../Source/WebKit/chromium/Tools.gyp
327
328 2011-12-09  Kentaro Hara  <haraken@chromium.org>
329
330         [Refactoring] In prepare-ChangeLog, move top-level code to fetch a bug description from URL into a method
331         https://bugs.webkit.org/show_bug.cgi?id=74173
332
333         Reviewed by Ryosuke Niwa.
334
335         The objective is to make prepare-ChangeLog a loadable Perl module for unit testing.
336         This requires to remove top-level code. This patch is one of the incremental refactorings
337         for that.
338
339         * Scripts/prepare-ChangeLog: Moved top-level code to fetch a bug description from URL into fetchBugDescriptionFromURL().
340         (fetchBugDescriptionFromURL):
341
342 2011-12-09  Tony Chang  <tony@chromium.org>
343
344         Switch the chromium mac bots to using skia test results
345         https://bugs.webkit.org/show_bug.cgi?id=74201
346
347         Reviewed by Ryosuke Niwa.
348
349         This is reverting http://trac.webkit.org/changeset/97781 and changing
350         the name in master.cfg to no longer have cg in the name.
351
352         * BuildSlaveSupport/build.webkit.org-config/config.json:
353         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
354         (appendCustomBuildFlags):
355         (RunWebKitTests.start):
356
357 2011-12-09  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
358
359         [Qt] Remove a wrongly placed ASSERT.
360
361         Reviewed by Tor Arne Vestbø.
362
363         * MiniBrowser/qt/MiniBrowserApplication.cpp:
364         (MiniBrowserApplication::notify):
365
366 2011-12-09  Cary Clark  <caryclark@google.com>
367
368         layout_test update to accommodate new use_skia=1 default on Chromium-Mac
369         https://bugs.webkit.org/show_bug.cgi?id=74118
370
371         Reviewed by Adam Barth.
372
373         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
374         (get):
375         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
376         (ChromiumMacPort.__init__):
377         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
378         (ChromiumMacPortTest.test_versions):
379         (ChromiumMacPortTest.test_graphics_type):
380
381 2011-12-09  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
382
383         [Qt] Use new qmake option in Qt5 to prevent dependency issues with GNUmake
384
385         The GNUmake + gcc_MD_depends options together produce .d files using GCC, but
386         this failed when moving files in the source tree, as the .d file would contain
387         stale dependencies on the old source file location. By keeping the object files
388         (and hence the depdendency files) in a proper tree structure inside the objects
389         directory we can prevent this failure.
390
391         A side benefit is that we can also have objects with the same name inside the
392         same project.
393
394         Reviewed by Simon Hausmann.
395
396         * qmake/mkspecs/features/unix/default_pre.prf:
397
398 2011-12-09  Csaba Osztrogonác  <ossy@webkit.org>
399
400         Unreviewed. Update unit test after r102412.
401
402         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
403         (IRCCommandTest.test_whois):
404
405 2011-12-09  Kentaro Hara  <haraken@chromium.org>
406
407         REGRESSION(r102105): run-bindings-tests should output full diff results
408         https://bugs.webkit.org/show_bug.cgi?id=74166
409
410         Reviewed by Adam Barth.
411
412         Since r102105, the output of run-bindings-tests is truncated to 500 characters.
413         run-bindings-tests should output full diff results.
414
415         * Scripts/webkitpy/bindings/main.py: Replaced e.message_with_output() with e.output.
416         (BindingsTests.generate_from_idl):
417         (BindingsTests.generate_supplemental_dependency):
418         (BindingsTests.detect_changes):
419
420 2011-12-09  Simon Hausmann  <simon.hausmann@nokia.com>
421
422         [Qt] Fix compilation with newly separated Qt 5 QtQuick module
423         https://bugs.webkit.org/show_bug.cgi?id=73990
424
425         Reviewed by Kenneth Rohde Christiansen, tweaked by Csaba Osztrogonác.
426
427         Fix QtQuick includes after move from QtDeclarative to QtQuick.
428
429         * MiniBrowser/qt/BrowserWindow.h:
430         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
431
432 2011-12-08  Adam Klein  <adamk@chromium.org>
433
434         [MutationObservers] V8LazyEventHandler breaks microtask delivery semantics
435         https://bugs.webkit.org/show_bug.cgi?id=73492
436
437         Reviewed by Adam Barth.
438
439         Add eventSender.scheduleAsynchronousKeyDown, needed to cause a
440         keypress event without any script on the stack.
441
442         * DumpRenderTree/chromium/EventSender.cpp:
443         (EventSender::EventSender):
444         (EventSender::keyDown):
445         (KeyDownTask::KeyDownTask):
446         (KeyDownTask::runIfValid):
447         (EventSender::scheduleAsynchronousKeyDown):
448         * DumpRenderTree/chromium/EventSender.h:
449
450 2011-12-08  Hayato Ito  <hayato@chromium.org>
451
452         [gdb] Pretty printer for a 8-bit version of WTF::StringImpl and LChar*.
453         https://bugs.webkit.org/show_bug.cgi?id=73980
454
455         Reviewed by Tony Chang.
456
457         * gdb/webkit.py:
458         (guess_string_length):
459         (ustring_to_string):
460         (lstring_to_string):
461         (LCharStringPrinter):
462         (LCharStringPrinter.to_string):
463         (WTFStringImplPrinter.get_length):
464         (WTFStringImplPrinter.to_string):
465         (WTFStringImplPrinter.is_8bit):
466         (WTFStringPrinter.stringimpl_ptr):
467         (WTFStringPrinter.get_length):
468         (WTFStringPrinter.to_string):
469         (add_pretty_printers.lookup_function):
470
471 2011-12-08  Pierre Rossi  <pierre.rossi@gmail.com>
472
473         Drop ENABLE_NO_LISTBOX_RENDERING, and make it a runtime decision.
474         https://bugs.webkit.org/show_bug.cgi?id=73916
475
476         Reviewed by Kent Tamura.
477
478         * qmake/mkspecs/features/features.prf:
479
480 2011-12-08  Adrienne Walker  <enne@google.com>
481
482         Unreviewed contributor additions to committers.py.
483
484         * Scripts/webkitpy/common/config/committers.py:
485
486 2011-12-08  David Reveman  <reveman@chromium.org>
487
488         [Chromium] Add per-tile painting flag to DumpRenderTree and rename AcceleratedDrawing to AcceleratedPainting in chromium specific code.
489         https://bugs.webkit.org/show_bug.cgi?id=74017
490
491         Reviewed by James Robinson.
492         
493         Add support for command line flag enable-per-tile-painting in DumpRenderTree.
494
495         * DumpRenderTree/chromium/DumpRenderTree.cpp:
496         (main):
497         * DumpRenderTree/chromium/TestShell.cpp:
498         (TestShell::TestShell):
499         (TestShell::resetWebSettings):
500         * DumpRenderTree/chromium/TestShell.h:
501         (TestShell::setAcceleratedPaintingEnabled):
502         (TestShell::setPerTilePaintingEnabled):
503         * DumpRenderTree/chromium/WebPreferences.cpp:
504         (WebPreferences::reset):
505         (WebPreferences::applyTo):
506         * DumpRenderTree/chromium/WebPreferences.h:
507         * Scripts/webkitpy/layout_tests/port/chromium.py:
508         (ChromiumDriver._wrapper_options):
509         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
510         (parse_args):
511
512 2011-12-08  Dirk Pranke  <dpranke@chromium.org>
513
514         create a "SystemHost" object for webkitpy to slim down the Host object
515         https://bugs.webkit.org/show_bug.cgi?id=72680
516
517         Reviewed by Eric Siedel.
518
519         This creates a SystemHost object that other objects that should
520         only have dependencies on functionality in webkitpy.common.system
521         can use; this will allow them to not require a full Host.
522
523         * Scripts/webkitpy/common/host.py:
524         * Scripts/webkitpy/common/host_mock.py:
525         * Scripts/webkitpy/common/system/systemhost.py: Added.
526         * Scripts/webkitpy/common/system/systemhost_mock.py: Added.
527
528 2011-12-08  Chris Fleizach  <cfleizach@apple.com>
529
530         AX: platform/mac/accessibility/search-with-frames.html crashes
531         https://bugs.webkit.org/show_bug.cgi?id=74093
532
533         Reviewed by Darin Adler.
534
535         Take care of the case when an incoming element is null.
536
537         * DumpRenderTree/AccessibilityUIElement.cpp:
538         (uiElementForSearchPredicateCallback):
539         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
540         (AccessibilityUIElement::uiElementForSearchPredicate):
541         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
542         (WTR::AccessibilityUIElement::uiElementForSearchPredicate):
543
544 2011-12-08  Alexander Færøy  <ahf@0x90.dk>
545
546         [Qt] Visualize mock point ID's in the MiniBrowser
547         https://bugs.webkit.org/show_bug.cgi?id=74098
548
549         Reviewed by Tor Arne Vestbø.
550
551         * MiniBrowser/qt/BrowserWindow.cpp:
552         (BrowserWindow::updateVisualMockTouchPoints):
553         * MiniBrowser/qt/qml/MockTouchPoint.qml:
554
555 2011-12-08  Dominic Mazzoni  <dmazzoni@google.com>
556
557         Accessibility: AccessibilityController should support listening to notifications on all elements.
558         https://bugs.webkit.org/show_bug.cgi?id=72866
559
560         Adds addNotificationListener and removeNotificationListener methods
561         to AccessibilityController, to listen to notifications on any element
562         rather than a specific element. Mac (DRT and WKTR) and Chromium (DRT)
563         implementations are all contained in this change.
564
565         On Mac, refactors AccessibilityNotificationHandler into its own
566         source file that can be used by both AccessibilityController and
567         AccessibilityUIElement. (Both DRT and WKTR.)
568
569         Reviewed by Chris Fleizach.
570
571         * DumpRenderTree/AccessibilityController.cpp:
572         (addNotificationListenerCallback):
573         (removeNotificationListenerCallback):
574         (AccessibilityController::getJSClass):
575         * DumpRenderTree/AccessibilityController.h:
576         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
577         * DumpRenderTree/chromium/AccessibilityController.cpp:
578         (AccessibilityController::AccessibilityController):
579         (AccessibilityController::notificationReceived):
580         (AccessibilityController::addNotificationListenerCallback):
581         (AccessibilityController::removeNotificationListenerCallback):
582         * DumpRenderTree/chromium/AccessibilityController.h:
583         * DumpRenderTree/gtk/AccessibilityControllerGtk.cpp:
584         (AccessibilityController::addNotificationListener):
585         (AccessibilityController::removeNotificationListener):
586         * DumpRenderTree/mac/AccessibilityControllerMac.mm:
587         (AccessibilityController::~AccessibilityController):
588         (AccessibilityController::addNotificationListener):
589         (AccessibilityController::removeNotificationListener):
590         * DumpRenderTree/mac/AccessibilityNotificationHandler.h: Added.
591         * DumpRenderTree/mac/AccessibilityNotificationHandler.mm: Added.
592         (-[NSString createJSStringRef]):
593         (-[AccessibilityNotificationHandler init]):
594         (-[AccessibilityNotificationHandler setPlatformElement:]):
595         (-[AccessibilityNotificationHandler dealloc]):
596         (-[AccessibilityNotificationHandler setCallback:]):
597         (-[AccessibilityNotificationHandler startObserving]):
598         (-[AccessibilityNotificationHandler _notificationReceived:]):
599         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
600         (AccessibilityUIElement::addNotificationListener):
601         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
602         (notificationListenerProc):
603         (AccessibilityController::addNotificationListener):
604         (AccessibilityController::removeNotificationListener):
605         (AccessibilityController::winNotificationReceived):
606         (AccessibilityController::winAddNotificationListener):
607         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
608         (AccessibilityUIElement::addNotificationListener):
609         * WebKitTestRunner/InjectedBundle/AccessibilityController.cpp:
610         (WTR::AccessibilityController::addNotificationListener):
611         (WTR::AccessibilityController::removeNotificationListener):
612         * WebKitTestRunner/InjectedBundle/AccessibilityController.h:
613         (WTR::AccessibilityController::logAccessibilityEvents):
614         * WebKitTestRunner/InjectedBundle/AccessibilityTextMarker.cpp:
615         (WTR::AccessibilityTextMarker::AccessibilityTextMarker):
616         * WebKitTestRunner/InjectedBundle/AccessibilityTextMarkerRange.cpp:
617         (WTR::AccessibilityTextMarkerRange::AccessibilityTextMarkerRange):
618         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityController.idl:
619         * WebKitTestRunner/InjectedBundle/mac/AccessibilityControllerMac.mm: Added.
620         (WTR::AccessibilityController::addNotificationListener):
621         (WTR::AccessibilityController::removeNotificationListener):
622         * WebKitTestRunner/InjectedBundle/mac/AccessibilityNotificationHandler.h: Added.
623         * WebKitTestRunner/InjectedBundle/mac/AccessibilityNotificationHandler.mm: Added.
624         (-[NSString createJSStringRef]):
625         (-[AccessibilityNotificationHandler init]):
626         (-[AccessibilityNotificationHandler setPlatformElement:]):
627         (-[AccessibilityNotificationHandler dealloc]):
628         (-[AccessibilityNotificationHandler setCallback:]):
629         (-[AccessibilityNotificationHandler startObserving]):
630         (-[AccessibilityNotificationHandler _notificationReceived:]):
631         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
632         (WTR::AccessibilityUIElement::AccessibilityUIElement):
633         (WTR::AccessibilityUIElement::verticalScrollbar):
634         (WTR::AccessibilityUIElement::addNotificationListener):
635         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
636
637 2011-12-08  Csaba Osztrogonác  <ossy@webkit.org>
638
639         [Qt][WK2] Unreviewed buildfix after r102352.
640
641         * qmake/mkspecs/features/webkit2.prf:
642
643 2011-12-08  Adam Roben  <aroben@apple.com>
644
645         Make finding crash logs by PID work on Snow Leopard
646
647         Crash Reporter on Snow Leopard doesn't set the "app_description" extended attribute that we
648         were using to check the PID of the crashed process. (Crash Reporter on Lion does.) Now we
649         read the first line of the crash log and pull the PID out of there.
650
651         Fixes <http://webkit.org/b/74094> "webkit-patch crash-log <process> <PID>" doesn't work on
652         Snow Leopard
653
654         Reviewed by Darin Adler.
655
656         * Scripts/webkitpy/common/system/crashlogs.py:
657         (CrashLogs._find_newest_log_darwin): Read the first line of the crash log to get the process
658         name and PID for the log, rather than checking the Lion-only app_description extended
659         attribute. We now also reject the log if the process name doesn't match what we expect.
660
661         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
662         (make_mock_crash_report_darwin): Added. Returns a mock crash report given a process name and
663         PID.
664         (CrashLogsTest.assertLinesEqual): Added. This is a compatibility shim around
665         unittest.TestCase.assertMultiLineEqual, which didn't exist before Python 2.7.
666         (CrashLogsTest.test_find_log_darwin): Made the mock crash reports more believable by using
667         the new make_mock_crash_report_darwin function. Also added tests for files that can't be
668         read, files that actually contain a crash log for some other process, and files that are
669         misformatted.
670
671         * Scripts/webkitpy/common/system/filesystem.py: Removed FileSystem.getxattr, which is now
672         unused.
673
674         * Scripts/webkitpy/common/system/filesystem_mock.py:
675         (MockFileSystem.__init__): Removed xattr support, which is now unused.
676         (MockFileSystem.open_text_file_for_reading): Changed to actually pass the file's data to
677         ReadableTextFileObject. This function was completely broken before!
678         (ReadableBinaryFileObject.__init__): Made the data parameter non-optional to try to prevent
679         bugs like the above.
680         (ReadableTextFileObject.__init__): Added. Wraps the data in a StringIO object so we can do
681         things like readline() and seek().
682
683         (ReadableTextFileObject.close):
684         (ReadableTextFileObject.read):
685         Added. We override the superclass implementation to account for using a StringIO object.
686
687         (ReadableTextFileObject.readline):
688         (ReadableTextFileObject.seek):
689         Added. These just call through to StringIO.
690
691         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
692         (MainTest.test_crash_log):
693         (MainTest.test_web_process_crash_log):
694         Changed to use make_mock_crash_report_darwin.
695
696 2011-12-06  Jon Lee  <jonlee@apple.com>
697
698         [WK2] Add permissions support (73960)
699         https://bugs.webkit.org/show_bug.cgi?id=73960
700         <rdar://problem/10357008>
701
702         Reviewed by Darin Adler.
703
704         Add null function item to comply with added callback to WKPageUIClient.
705
706         * MiniBrowser/mac/BrowserWindowController.m:
707         (-[BrowserWindowController awakeFromNib]):
708         * WebKitTestRunner/TestController.cpp:
709         (WTR::TestController::createOtherPage):
710         (WTR::TestController::initialize):
711
712 2011-12-08  Alexander Færøy  <alexander.faeroy@nokia.com>
713
714         [Qt] Visualize mock points in the Qt MiniBrowser
715         https://bugs.webkit.org/show_bug.cgi?id=74074
716
717         Reviewed by Tor Arne Vestbø.
718
719         * MiniBrowser/qt/BrowserWindow.cpp:
720         (BrowserWindow::updateVisualMockTouchPoints):
721         * MiniBrowser/qt/BrowserWindow.h:
722         * MiniBrowser/qt/MiniBrowser.qrc:
723         * MiniBrowser/qt/MiniBrowserApplication.cpp:
724         (MiniBrowserApplication::notify):
725         (MiniBrowserApplication::sendTouchEvent):
726         * MiniBrowser/qt/MiniBrowserApplication.h:
727         * MiniBrowser/qt/icons/touchpoint.png: Added.
728         * MiniBrowser/qt/qml/MockTouchPoint.qml: Added.
729
730 2011-12-08  Simon Hausmann  <simon.hausmann@nokia.com>
731
732         [Qt] WebKitTestRunner crashes with debug builds
733         https://bugs.webkit.org/show_bug.cgi?id=74090
734
735         Reviewed by Tor Arne Vestbø.
736
737         * WebKitTestRunner/qt/TestControllerQt.cpp:
738         (WTR::TestController::platformInitialize): Call QQuickWebView::platformInitialize
739         to ensure WTF/JSC threading is initialized.
740
741 2011-12-08  Dirk Pranke  <dpranke@chromium.org>
742
743         fix build-webkit --chromium after breakage in r102201
744         https://bugs.webkit.org/show_bug.cgi?id=74031
745
746         Reviewed by Tony Chang.
747
748         * Scripts/webkitdirs.pm:
749         (buildChromium):
750
751 2011-12-08  Kentaro Hara  <haraken@chromium.org>
752
753         Unreviewed, rolling out r102280.
754         http://trac.webkit.org/changeset/102280
755         https://bugs.webkit.org/show_bug.cgi?id=74031
756
757         Chromium/Windows build is broken
758
759         * Scripts/webkitdirs.pm:
760         (buildChromium):
761
762 2011-12-08  Nayan Kumar K  <nayankk@motorola.com>
763
764         Adding myself to committers list.
765
766         Unreviewed.
767
768         * Scripts/webkitpy/common/config/committers.py:
769
770 2011-12-07  Michael Saboff  <msaboff@apple.com>
771
772         StringBuilderTest.Append and StringBuilderTest.ToStringPreserveCapacity are failing.
773         https://bugs.webkit.org/show_bug.cgi?id=73995
774
775         Reenabled failing tests that the code part of the patch fixes.
776
777         Reviewed by Geoffrey Garen.
778
779         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp:
780         (TestWebKitAPI::TEST):
781
782 2011-12-07  MORITA Hajime  <morrita@google.com>
783
784         [filter-build-webkit] should not emit reset color when --no-color is given
785         https://bugs.webkit.org/show_bug.cgi?id=73992
786
787         Reviewed by Daniel Bates.
788
789         Stop printing control sequences when --no-color is given.
790         After this change, the filter runs nicely even on environments
791         like Emacs compilation-mode which cannot understand the control sequence.
792
793         * Scripts/filter-build-webkit:
794         (printLine):
795
796 2011-12-07  Eric Seidel  <eric@webkit.org>
797
798         Use free memory to determine if we have space for DRT instances instead of total memory
799         https://bugs.webkit.org/show_bug.cgi?id=74021
800
801         Reviewed by Adam Barth.
802
803         This will hopefully make the Mac buildbots behave better.
804
805         I also reduced the expected memory needed per DRT now that
806         we're properly accounting for the system memory in our calculations.
807
808         * Scripts/webkitpy/common/system/platforminfo.py:
809         (PlatformInfo.total_bytes_memory):
810         (PlatformInfo._compute_free_bytes_from_vm_stat_output):
811         * Scripts/webkitpy/common/system/platforminfo_mock.py:
812         (MockPlatformInfo.total_bytes_memory):
813         * Scripts/webkitpy/layout_tests/port/base.py:
814         (Port.default_child_processes):
815         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
816         (PortTest.test_default_child_processes):
817
818 2011-12-07  Dirk Pranke  <dpranke@chromium.org>
819
820         fix build-webkit --chromium after breakage in r102201
821         https://bugs.webkit.org/show_bug.cgi?id=74031
822
823         Reviewed by Tony Chang.
824
825         r102201 introduced a change to the chromium build that splits
826         the executables out of WebKit.gyp into other gyp files, and adds
827         a new "All.gyp" that builds everything. However, I forgot to
828         update the build script to use All.gyp.
829
830         * Scripts/webkitdirs.pm:
831         (buildChromium):
832
833 2011-12-07  Tom Zakrajsek  <tomz@codeaurora.org>
834
835         Moved myself to the committers list.
836
837         * Scripts/webkitpy/common/config/committers.py:
838
839 2011-11-30  Simon Hausmann  <simon.hausmann@nokia.com>
840
841         [Qt] Fix debug shlib build with V8.
842
843         Reviewed by Tor Arne Vestbø.
844
845         * qmake/mkspecs/features/qtwebkit.prf: Don't link jscore with
846         v8 builds.
847
848 2011-11-30  Simon Hausmann  <simon.hausmann@nokia.com>
849
850         [Qt] V8 build fixes.
851
852         Reviewed by Tor Arne Vestbø.
853
854         * qmake/mkspecs/features/javascriptcore.prf: Don't add
855         Source/Thirdparty to the include search path here.
856         * qmake/mkspecs/features/webcore.prf: Instead it belongs here, because it's
857         used when building WebCore as well as using it, because header files like
858         GraphicsContext3D.h include stuff from ThirdParty.
859
860 2011-12-07  Peter Beverloo  <peter@chromium.org>
861
862         [Chromium] Fix project file generation for the Android bot after r102201
863         https://bugs.webkit.org/show_bug.cgi?id=73996
864
865         The gyp refactoring in r102201 was didn't explicitly exclude
866         the Android operating system in Tools/Tools.gyp, which broke
867         as gcc_version is not defined for that configuration.
868
869         Reviewed by Andreas Kling.
870
871         * Tools.gyp:
872
873 2011-12-07  Vsevolod Vlasov  <vsevik@chromium.org>
874
875         Unreviewed. Disabled failing tests on chromium.
876
877         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp:
878         (TestWebKitAPI::TEST):
879
880 2011-12-06  Dirk Pranke  <dpranke@chromium.org>
881
882         binaries in webkit.gyp should be split into a new file to break circular dependencies on webkit_support
883         https://bugs.webkit.org/show_bug.cgi?id=73382
884
885         Reviewed by Tony Chang.
886
887         XCode (and hence GYP) can't handle circular dependencies between
888         files (and projects). Accordingly, although the binaries in
889         WebKit.gyp depend on webkit_support and webkit_support depends
890         on webkit, we don't mention the latter dependency in the file.
891         To fix this, this change moves the binaries like DRT into a new
892         GYP file, and modifies gyp_webkit to build from that file
893         instead of WebKit.gyp.
894
895         * Tools.gyp: Added.
896
897 2011-12-06  Adam Roben  <aroben@apple.com>
898
899         Add a pid parameter to CrashLogs.find_newest_log
900
901         When a PID is passed, only crash logs corresponding to a crashed process with that PID are
902         considered.
903
904         Fixes <http://webkit.org/b/73942> webkitpy provides no way to specify the PID of the crashed
905         process whose crash log you want to find
906
907         Reviewed by Dirk Pranke.
908
909         * Scripts/webkitpy/common/system/crashlogs.py:
910         (CrashLogs.find_newest_log): Added an optional pid parameter.
911         (CrashLogs._find_newest_log_darwin): Added a pid parameter. When specified, we look at each
912         candidate log's app_description extended attribute to see if that log corresponds to a
913         crashed process with the specified PID.
914
915         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
916         (CrashLogsTest.test_find_newest_log_darwin): Added a few more mock crash logs with various
917         contents and PIDs. Added tests that show that the pid parameter to find_newest_log is
918         respected. Also fixed an erroneous use of assertTrue instead of assertEqual.
919
920         * Scripts/webkitpy/common/system/filesystem.py:
921         (FileSystem.getxattr): Added. On Darwin, calls through to the xattr module. On other
922         systems, raises a KeyError, mimicking the behavior on Darwin when the given attribute isn't
923         present.
924
925         * Scripts/webkitpy/common/system/filesystem_mock.py:
926         (MockFileSystem.__init__): Added an optional xattrs parameter. Documented other parameters.
927         (MockFileSystem.getxattr): Added. Just pulls the given attribute out of the xattrs
928         dictionary.
929
930         * Scripts/webkitpy/tool/commands/queries.py:
931         (CrashLog): Added help text. Added an optional PID parameter.
932         (CrashLog.execute): Pass the PID, if any, through to find_newest_log.
933
934 2011-12-06  Adam Roben  <aroben@apple.com>
935
936         Wait for Crash Reporter to finish even when it lets the crashed process die quickly
937
938         NRWT was only waiting for Crash Reporter in cases where it was keeping the crashed process
939         alive beyond the normal timeout limit. In cases where the crashed process was able to die
940         faster, NRWT would assume that Crash Reporter had finished even though it often was still
941         running, which would lead to an incorrect crash log being picked up.
942
943         Part of <http://webkit.org/b/71380> NRWT incorrectly associates crash logs with tests
944
945         Reviewed by Dirk Pranke.
946
947         * Scripts/webkitpy/layout_tests/port/server_process.py:
948         (ServerProcess._reset):
949         (ServerProcess.write):
950         (ServerProcess._check_for_crash): Changed to use new set_crashed function instead of setting
951         .crashed directly. Added wait_for_crash_reporter parameter, which we pass along to
952         set_crashed.
953
954         (ServerProcess._handle_timeout): Fixed a logic error that would cause .crashed and
955         .timed_out both to be set to True in cases where Crash Reporter took a long time to run. Now
956         we bail out of handling the failure as a timeout if we find out that the process in fact
957         crashed. We tell _check_for_crash not to wait for Crash Reporter because we've already done
958         so.
959         (ServerProcess.set_crashed): Added. When the process crashes, we wait for Crash Reporter to
960         finish running (unless directed otherwise) so we can be sure the crash log has been saved to
961         disk.
962
963         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
964         (TrivialMockPort.is_crash_reporter): Added.
965
966         * Scripts/webkitpy/layout_tests/port/webkit.py:
967         (WebKitDriver._check_for_driver_crash): Changed to use set_crashed. (This also fixed a typo
968         that would have partially broken crash detection on Windows, if NRWT worked on Windows.)
969
970 2011-12-06  Adam Barth  <abarth@webkit.org>
971
972         NRWT fails on unreleased versions of Mac OS X
973         https://bugs.webkit.org/show_bug.cgi?id=72748
974
975         Reviewed by Dirk Pranke.
976
977         Hopefully this patch will make NRWT work on future versions on Mac OS X.
978         Most of the infrastructure was already in place, but there were some
979         asserts that failed.  I've tested this patch by haxoring the version
980         detection logic to simulate an unknown version.  We might need to
981         iterate based on feedback from folks with a real future version.
982
983         * Scripts/webkitpy/layout_tests/port/apple.py:
984         (ApplePort.__init__):
985             - mac-future is an allowed port_name that's used internally by
986               webkitpy to represent future versions of Mac OS X.
987         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
988         (test_versions):
989             - Test that we don't throw exceptions when dealing with future
990               versions of Mac OS X.
991         * Scripts/webkitpy/layout_tests/port/webkit.py:
992         (WebKitPort._expectations_from_skipped_files):
993             - Demote this logging message to "debug".  It's expected that
994               mac-future doesn't have a Skipped file.
995         * Scripts/webkitpy/layout_tests/port/mac.py:
996         (MacPort.__init__):
997         * Scripts/webkitpy/layout_tests/port/win.py:
998         (WinPort.__init__):
999             - Re-order some code in these constructors so that
1000               self._operating_system is available to the ApplePort constructor.
1001
1002 2011-12-06  Tommy Widenflycht  <tommyw@google.com>
1003
1004         Added myself as a committer.
1005
1006         * Scripts/webkitpy/common/config/committers.py:
1007
1008 2011-12-06  Peter Beverloo  <peter@chromium.org>
1009
1010         Added myself as a committer.
1011
1012         * Scripts/webkitpy/common/config/committers.py:
1013
1014 2011-12-06  Kentaro Hara  <haraken@chromium.org>
1015
1016         [Refactoring] In webkitpy/bindings/main.py, replace subprocess.Popen() with Executive
1017         https://bugs.webkit.org/show_bug.cgi?id=73637
1018
1019         Reviewed by Adam Barth.
1020
1021         This patch replaces subprocess.Popen() with Executive, similar to other
1022         Python scripts in webkitpy/. I manually confirmed that run-bindings-tests
1023         works well for cases where no exception is raised, no exception is raised
1024         but diff is found, and ScriptError is raised.
1025
1026         * Scripts/run-bindings-tests:
1027         (main):
1028         * Scripts/webkitpy/bindings/main.py: Replaces subprocess.Popen() with Executive.
1029         (BindingsTests.__init__):
1030         (BindingsTests.generate_from_idl):
1031         (BindingsTests.generate_supplemental_dependency):
1032         (BindingsTests.detect_changes):
1033
1034 2011-12-05  Alexander Færøy  <alexander.faeroy@nokia.com>
1035
1036         [Qt] MiniBrowser should default to touch behavior
1037         https://bugs.webkit.org/show_bug.cgi?id=73863
1038
1039         Reviewed by Chang Shu.
1040
1041         MiniBrowser currently defaults to the traditional desktop behaviour
1042         instead of touch behaviour which is prefered.
1043
1044         * MiniBrowser/qt/BrowserWindow.cpp:
1045         (BrowserWindow::BrowserWindow):
1046         * MiniBrowser/qt/MiniBrowserApplication.cpp:
1047         (printHelp):
1048         (MiniBrowserApplication::handleUserOptions):
1049         * MiniBrowser/qt/MiniBrowserApplication.h:
1050         (WindowOptions::WindowOptions):
1051         (WindowOptions::setUseTraditionalDesktopBehavior):
1052         (WindowOptions::useTraditionalDesktopBehavior):
1053
1054 2011-12-05  Raphael Kubo da Costa  <kubo@profusion.mobi>
1055
1056         webkit.py: Really fix the printing of StringImpl now that 8-bit strings are present.
1057         https://bugs.webkit.org/show_bug.cgi?id=73878
1058
1059         Reviewed by Tony Chang.
1060
1061         r98624 turned StringImpl::m_data into StringImpl::{m_data8,m_data16}.
1062         r98785 made webkit.py always use m_data16, which does not work when
1063         the string is an 8-bit string (such as KURL::string()).
1064
1065         I was not able to directly call StringImpl::is8Bit() in the Python
1066         code, so I just reproduced the implementation there.
1067
1068         * gdb/webkit.py:
1069         (WTFStringPrinter.is_8bit): Reproduced StringImpl::is8Bit().
1070         (WTFStringPrinter.to_string):
1071         (JSCUStringPrinter.is_8bit): Reproduced StringImpl::is8Bit().
1072         (JSCUStringPrinter.to_string):
1073
1074 2011-12-05  Eric Seidel  <eric@webkit.org>
1075
1076         Up the "expected ram for each DRT instance" to 400MB instead of 300MB
1077         in hopes of making the Snow Leopard bot stop timing out.
1078
1079         Unreviewed.
1080
1081         * Scripts/webkitpy/layout_tests/port/base.py:
1082         (Port.default_child_processes):
1083         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1084         (PortTest.test_default_child_processes):
1085
1086 2011-12-05  Eric Seidel  <eric@webkit.org>
1087
1088         NRWT creates too many workers on RAM-limited machines
1089         https://bugs.webkit.org/show_bug.cgi?id=73847
1090
1091         Reviewed by Dirk Pranke.
1092
1093         Make NRWT restrict its number of workers when there isn't enough
1094         RAM to support them.
1095         Also re-enable parallel testing on the Mac bots which were
1096         failing due to not enough ram.
1097
1098         * Scripts/run-webkit-tests:
1099         (platformIsReadyForParallelTesting):
1100         * Scripts/webkitpy/common/host.py:
1101         (Host.__init__):
1102         * Scripts/webkitpy/common/system/executive_mock.py:
1103         (MockExecutive.cpu_count):
1104         * Scripts/webkitpy/common/system/platforminfo.py:
1105         (PlatformInfo.__init__):
1106         (PlatformInfo.display_name):
1107         * Scripts/webkitpy/common/system/platforminfo_mock.py:
1108         (MockPlatformInfo.display_name):
1109         * Scripts/webkitpy/layout_tests/port/base.py:
1110         (Port.default_child_processes):
1111         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1112         (PortTest.test_default_child_processes):
1113         (PortTest.test_pretty_patch_os_error):
1114
1115 2011-12-05  Hugo Parente Lima  <hugo.lima@openbossa.org>
1116
1117         Last mouse event could be lost if a JS call to eventSender.leapForward is made
1118         https://bugs.webkit.org/show_bug.cgi?id=73366
1119
1120         Reviewed by Kenneth Rohde Christiansen.
1121
1122         Empty the entire event queue when requested besides using
1123         sendEvent instead of postEvent to make sure all events were
1124         delivered.
1125
1126         * WebKitTestRunner/qt/EventSenderProxyQt.cpp:
1127         (WTR::EventSenderProxy::EventSenderProxy):
1128         (WTR::EventSenderProxy::sendOrQueueEvent):
1129         (WTR::EventSenderProxy::replaySavedEvents):
1130
1131 2011-12-05  Adam Barth  <abarth@webkit.org>
1132
1133         [Chromium] Remove no-longer-used forwarding headers in the WebKit API
1134         https://bugs.webkit.org/show_bug.cgi?id=73845
1135
1136         Reviewed by Darin Fisher.
1137
1138         Update includes to point to the new header locations.
1139
1140         * DumpRenderTree/chromium/AccessibilityController.cpp:
1141         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
1142         * DumpRenderTree/chromium/CppBoundClass.cpp:
1143         * DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
1144         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
1145         * DumpRenderTree/chromium/DRTDevToolsClient.h:
1146         * DumpRenderTree/chromium/EventSender.cpp:
1147         * DumpRenderTree/chromium/EventSender.h:
1148         * DumpRenderTree/chromium/GamepadController.h:
1149         * DumpRenderTree/chromium/LayoutTestController.cpp:
1150         * DumpRenderTree/chromium/LayoutTestController.h:
1151         * DumpRenderTree/chromium/MockSpellCheck.cpp:
1152         * DumpRenderTree/chromium/NotificationPresenter.cpp:
1153         * DumpRenderTree/chromium/PlainTextController.cpp:
1154         * DumpRenderTree/chromium/Task.cpp:
1155         * DumpRenderTree/chromium/TestNavigationController.h:
1156         * DumpRenderTree/chromium/TestShell.cpp:
1157         * DumpRenderTree/chromium/TestWebPlugin.cpp:
1158         * DumpRenderTree/chromium/TestWebPlugin.h:
1159         * DumpRenderTree/chromium/TextInputController.cpp:
1160         * DumpRenderTree/chromium/WebPermissions.cpp:
1161         * DumpRenderTree/chromium/WebPreferences.h:
1162         * DumpRenderTree/chromium/WebThemeEngineDRTMac.h:
1163         * DumpRenderTree/chromium/WebThemeEngineDRTMac.mm:
1164         * DumpRenderTree/chromium/WebThemeEngineDRTWin.cpp:
1165         * DumpRenderTree/chromium/WebThemeEngineDRTWin.h:
1166         * DumpRenderTree/chromium/WebViewHost.cpp:
1167
1168 2011-12-05  Ryosuke Niwa  <rniwa@webkit.org>
1169
1170         Add Aharon Lanin to the contributors list.
1171
1172         * Scripts/webkitpy/common/config/committers.py:
1173
1174 2011-12-05  Jarred Nicholls  <jarred@sencha.com>
1175
1176         Moving myself from contributors to committers.
1177
1178         * Scripts/webkitpy/common/config/committers.py:
1179
1180 2011-12-05  Raphael Kubo da Costa  <kubo@profusion.mobi>
1181
1182         Add myself as a committer.
1183
1184         * Scripts/webkitpy/common/config/committers.py:
1185
1186 2011-12-05  Michael Brüning  <michael.bruning@nokia.com>
1187
1188         [Qt] Rename QQuickWebView::canStop property to just loading 
1189         https://bugs.webkit.org/show_bug.cgi?id=73813
1190
1191         Reviewed by Kenneth Rohde Christiansen.
1192
1193         Replaced references to WebView.canStop with WebView.loading.
1194
1195         * MiniBrowser/qt/qml/BrowserWindow.qml:
1196
1197 2011-12-05  Dirk Pranke  <dpranke@chromium.org>
1198
1199         run-webkit-tests: limit --child-processes properly on apple mac
1200         https://bugs.webkit.org/show_bug.cgi?id=73856
1201
1202         Reviewed by Adam Roben.
1203
1204         Fix typos in r102013 so that we correctly pass --child-processes=1 on
1205         the apple mac bots temporarily.
1206
1207         * Scripts/run-webkit-tests:
1208         (platformIsReadyForParallelTesting):
1209
1210 2011-12-05  Ryosuke Niwa  <rniwa@webkit.org>
1211
1212         Temporarily set --child-process=1 on Mac because 20+ tests are timing out
1213         https://bugs.webkit.org/show_bug.cgi?id=73840
1214
1215         Reviewed by Simon Fraser.
1216
1217         Temporarily go back to single process mode on Mac.
1218
1219         * Scripts/run-webkit-tests:
1220         (platformIsReadyForParallelTesting):
1221
1222 2011-12-05  Adam Roben  <aroben@apple.com>
1223
1224         Redeploy apple-macpro-7 and apple-xserve-1
1225
1226         These slaves have been sitting idle since we removed our Leopard builders in r97496.
1227
1228         Fixes <http://webkit.org/b/73829> REGRESSION (r97496): apple-macpro-7 and apple-xserve-1 are
1229         sitting idle
1230
1231         Reviewed by Sam Weinig.
1232
1233         * BuildSlaveSupport/build.webkit.org-config/config.json: Re-added apple-macpro-7 and
1234         apple-xserve-1. The former is now doing "SnowLeopard Intel Debug (WebKit2 Tests)", which
1235         allows apple-xserve-9 to be a build-only slave, and the latter is doing "Lion Intel Debug
1236         (WebKit2 Tests)".
1237
1238 2011-12-05  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1239
1240         [Qt] Make the UI of the MiniBrowser slightly nicer
1241
1242         Rubberstamped by Simon Hausmann.
1243
1244         - Add proper padding
1245         - Show feedback when pressing on the toolbar icons
1246         - Make the text entry better looking
1247
1248         * MiniBrowser/qt/qml/BrowserWindow.qml:
1249
1250 2011-12-05  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
1251
1252         NRWT should handle duplicated expectations
1253         https://bugs.webkit.org/show_bug.cgi?id=69750
1254
1255         Reviewed by Dirk Pranke.
1256
1257         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1258         (TestExpectations.__init__):
1259         (TestExpectations._report_errors):
1260         (TestExpectations._add_expectations):
1261         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
1262         (test_add_skipped_tests):
1263         * Scripts/webkitpy/layout_tests/port/base.py:
1264         (Port.skipped_tests):
1265         * Scripts/webkitpy/layout_tests/port/webkit.py:
1266         (WebKitPort.test_expectations):
1267         (WebKitPort.skipped_tests):
1268         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1269         (test_test_expectations):
1270
1271 2011-12-04  Eric Seidel  <eric@webkit.org>
1272
1273         After success with Mac, Win and Gtk, turning on parallel testing
1274         by default for all platforms (except Qt).
1275
1276         We'll add more platforms to the blacklist if we see problems.
1277
1278         Unreviewed.
1279
1280         * Scripts/run-webkit-tests:
1281         (platformIsReadyForParallelTesting):
1282
1283 2011-12-04  Sam Weinig  <sam@webkit.org>
1284
1285         Add tests for RetainPtr::adoptNS/adoptCF
1286         https://bugs.webkit.org/show_bug.cgi?id=73470
1287
1288         Reviewed by Adam Roben and Darin Adler.
1289
1290         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1291         * TestWebKitAPI/Tests/WTF/cf/RetainPtr.cpp: Added.
1292         (TestWebKitAPI::TEST):
1293         Test that adopting using adoptCF creates a RetainPtr with a
1294         value whose retain count is 1.
1295
1296         * TestWebKitAPI/Tests/WTF/ns: Added.
1297         * TestWebKitAPI/Tests/WTF/ns/RetainPtr.mm: Added.
1298         (TestWebKitAPI::TEST):
1299         Test that adopting using adoptNS creates a RetainPtr with a
1300         value whose retain count is 1.
1301
1302 2011-12-04  Martin Robinson  <mrobinson@igalia.com>
1303
1304         [GTK] Include cairo, pixman and gnome-icon-theme in the jhbuild modules
1305         https://bugs.webkit.org/show_bug.cgi?id=73748
1306
1307         Reviewed by Gustavo Noronha Silva.
1308
1309         * gtk/jhbuild.modules: Add some new modules the jhbuild moduleset.
1310
1311 2011-12-04  Philippe Normand  <pnormand@igalia.com>
1312
1313         Another GTK build fix after r101922.
1314
1315         Rubber-stamped by Martin Robinson.
1316
1317         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1318         (LayoutTestController::pathToLocalResource): Plug possible String ref leak.
1319
1320 2011-12-04  Xan Lopez  <xlopez@igalia.com>
1321
1322         Fix compiler warning in LayoutTestController
1323
1324         Reviewed by Philippe Normand.
1325
1326         * DumpRenderTree/LayoutTestController.cpp:
1327         (LayoutTestController::LayoutTestController): fix warning.
1328
1329 2011-12-04  Alexandre Mazari  <scaroo@gmail.com>
1330
1331         Add missing methods used by fast/notifications tests to LayoutTestController
1332         https://bugs.webkit.org/show_bug.cgi?id=63616
1333
1334         Reviewed by Martin Robinson.
1335
1336         Add missing simulateDesktopNotificationClick method used by
1337         fast/notifications. Provide a default implementation for
1338         areDesktopNotificationPermissionRequestsIgnored and
1339         ignoreDesktopNotificationPermissionRequests.
1340
1341         * DumpRenderTree/LayoutTestController.cpp:
1342         (LayoutTestController::LayoutTestController): call into the port implementation.
1343         (simulateDesktopNotificationClickCallback): ditto.
1344         (ignoreDesktopNotificationPermissionRequestsCallback): default implementation.
1345         (LayoutTestController::staticFunctions): declare new methods.
1346         (LayoutTestController::ignoreDesktopNotificationPermissionRequests): default implementation.
1347         * DumpRenderTree/LayoutTestController.h:
1348         (LayoutTestController::areDesktopNotificationPermissionRequestsIgnored):
1349         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1350         (LayoutTestController::simulateDesktopNotificationClick): dummy implementation.
1351         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1352         (LayoutTestController::simulateDesktopNotificationClick): ditto.
1353         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
1354         (LayoutTestController::simulateDesktopNotificationClick): ditto.
1355         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1356         (LayoutTestController::simulateDesktopNotificationClick): ditto.
1357         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
1358         (LayoutTestController::simulateDesktopNotificationClick): ditto.
1359
1360 2011-12-03  Mario Sanchez Prada  <msanchez@igalia.com>
1361
1362         [GTK] Don't log document events in DRT
1363         https://bugs.webkit.org/show_bug.cgi?id=73746
1364
1365         Reviewed by Martin Robinson.
1366
1367         Stop connecting globally to document-related events.
1368
1369         * DumpRenderTree/gtk/AccessibilityCallbacks.cpp:
1370         (connectAccessibilityCallbacks): Removed connections.
1371         (disconnectAccessibilityCallbacks): Removed disconnections.
1372
1373 2011-12-03  Dan Bernstein  <mitz@apple.com>
1374
1375         Added a test for <rdar://problem/10523721> Crash at WebCore::SubresourceLoader::releaseResources
1376
1377         Reviewed by Darin Adler.
1378
1379         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1380         * TestWebKitAPI/Tests/mac/SubresourceErrorCrash.mm: Added.
1381         (TestWebKitAPI::TEST):
1382
1383 2011-12-03  Philippe Normand  <pnormand@igalia.com>
1384
1385         Another GTK build fix after r101922.
1386
1387         Rubber-stamped by Martin James Robinson.
1388
1389         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1390         (LayoutTestController::pathToLocalResource): return value needs to
1391         be a URI.
1392
1393 2011-12-03  Martin Robinson  <mrobinson@igalia.com>
1394
1395         Small fix for my previous patch. Do not try to delete a member
1396         of a dictionary that does not exist.
1397
1398         Reviewed by Gustavo Noronha Silva.
1399
1400         * gtk/run-with-jhbuild:
1401         (install_and_run_jhbuild):
1402
1403 2011-12-03  Martin Robinson  <mrobinson@igalia.com>
1404
1405         Fix jhbuild installation for people who have unusual MAKE environment
1406         variables such as 'make -j4' as the jhbuild build does not handle it.
1407
1408         Reviewed by Gustavo Noronha Silva.
1409
1410         * gtk/run-with-jhbuild:
1411         (install_and_run_jhbuild): Remove any MAKE environment variable before building jhbuild.
1412
1413 2011-12-03  Martin Robinson  <mrobinson@igalia.com>
1414
1415         Fix the update-webkitgtk-libs script to properly trigger a jhbuild installation
1416         when run for the first time.
1417
1418         Reviewed by Gustavo Noronha Silva.
1419
1420         * Scripts/update-webkitgtk-libs: Fix first run.
1421
1422 2011-12-03  Philippe Normand  <pnormand@igalia.com>
1423
1424         Unreviewed, GTK build fix after r101922.
1425
1426         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1427         (LayoutTestController::pathToLocalResource): Added sentinel in
1428         g_build_filename call.
1429
1430 2011-12-03  Philippe Normand  <pnormand@igalia.com>
1431
1432         GTK API tests build fix. Run the tests inside jhbuild.
1433
1434         Rubber-stamped by Gustavo Noronha Silva.
1435
1436         * Scripts/run-gtk-tests:
1437         (TestRunner.__init__):
1438         (TestRunner.run._error_handler):
1439         (TestRunner):
1440         (TestRunner.run):
1441
1442 2011-12-03  Martin Robinson  <mrobinson@igalia.com>
1443
1444         Small build fix. Properly specify the path to the locally installed jhbuild.
1445
1446         * Scripts/update-webkitgtk-libs: Fix jhbuild path.
1447
1448 2011-12-03  Martin Robinson  <mrobinson@igalia.com>
1449
1450         [GTK][DRT] Normalize file:///tmp/LayoutTests in LayoutTestController::pathToLocalResource()
1451         https://bugs.webkit.org/show_bug.cgi?id=67256
1452
1453         Reviewed by Philippe Normand.
1454
1455         Implement LayoutTestController::pathToLocalResource for GTK+. Instead of passing 
1456         WEBKIT_TEST_FONTS to the GTK+ test harnesses, pass a more generic WEBKIT_TOP_LEVEL,
1457         which points to the more generic top-level path of the WebKit checkout. This code is
1458         duplicated between WK1 and WK2 harnesses because we do not currently have a way to
1459         share code here.
1460
1461         If WEBKIT_TOP_LEVEL is not provided, we search for the top level based on the binary
1462         location. This will cause the fallback to fail if you build into a non-typical location
1463         or even fake it with a symlink. In this case it's important to use the environment variable.
1464
1465         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1466         (getTopLevelPath): Added.
1467         (initializeFonts): Use the new helper to get the font path.
1468         * DumpRenderTree/gtk/DumpRenderTreeGtk.h: Expose the new helper.
1469         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1470         (LayoutTestController::pathToLocalResource):Use the new helper to normalize
1471         test paths.
1472         * Scripts/webkitpy/layout_tests/port/gtk.py:
1473         (GtkPort.setup_environ_for_server):Pass the top-level path of the
1474         checkout instead of passing the font path.
1475         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp:
1476         (WTR::getTopLevelPath): Added this helper.
1477         (WTR::inititializeFontConfigSetting): Use the helper to find the font path.
1478
1479 2011-12-02  David Levin  <levin@chromium.org>
1480
1481         Rename WTF class from TemporarilyChange to TemporaryChange.
1482         https://bugs.webkit.org/show_bug.cgi?id=73479
1483
1484         Reviewed by Eric Seidel.
1485
1486         * DumpRenderTree/ForwardingHeaders/wtf/TemporarilyChange.h: Removed.
1487         * DumpRenderTree/ForwardingHeaders/wtf/TemporaryChange.h: Added.
1488         * TestWebKitAPI/TestWebKitAPI.gypi:
1489         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1490         * TestWebKitAPI/Tests/WTF/TemporaryChange.cpp: Renamed from Tools/TestWebKitAPI/Tests/WTF/TemporarilyChange.cpp.
1491         (TestWebKitAPI::TEST):
1492         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
1493
1494 2011-12-02  Tony Chang  <tony@chromium.org>
1495
1496         [NRWT] reftest asserts intermittently on the Qt-WK2 bot
1497         https://bugs.webkit.org/show_bug.cgi?id=73453
1498
1499         Reviewed by Ryosuke Niwa.
1500
1501         We used to assert if a ref test didn't generate pixel results. Instead, just claim it is a failing test.
1502         The verbose output will include the failure reason.
1503
1504         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py: Don't assert, return FailureReftestNoImagesGenerated instead.
1505         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
1506         (write_test_result):
1507         * Scripts/webkitpy/layout_tests/controllers/test_result_writer_unittest.py: Remove stray letter t.
1508         * Scripts/webkitpy/layout_tests/models/test_failures.py: Add a new failure type: FailureReftestNoImagesGenerated
1509         (determine_result_type):
1510         (FailureReftestNoImagesGenerated.__init__):
1511         (FailureReftestNoImagesGenerated):
1512         (FailureReftestNoImagesGenerated.message):
1513         * Scripts/webkitpy/layout_tests/port/test.py: Add a test case for integration tests.
1514         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py: Update test counts.
1515         (MainTest.test_run_singly_actually_runs_tests):
1516         (MainTest.test_unexpected_failures):
1517         (EndToEndTest.test_end_to_end):
1518
1519 2011-12-02  Ryosuke Niwa  <rniwa@webkit.org>
1520
1521         ChangeLog should be able to parse annotated changelogs
1522         https://bugs.webkit.org/show_bug.cgi?id=72703
1523
1524         Reviewed by Eric Seidel.
1525
1526         Make parse_entries_from_file aware of svn blame annotation such as "100000 ossy@webkit.org ",
1527         and extract the revision number for each entry.
1528
1529         Also add a support to extract "svn blame" result from svn/git checkout to scm classes.
1530
1531         * Scripts/webkitpy/common/checkout/changelog.py:
1532         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
1533         * Scripts/webkitpy/common/checkout/scm/git.py:
1534         * Scripts/webkitpy/common/checkout/scm/scm.py:
1535         * Scripts/webkitpy/common/checkout/scm/svn.py:
1536
1537 2011-12-02  Eric Seidel  <eric@webkit.org>
1538
1539         Enable parallel testing for Win and Gtk bots
1540         after confirming with Lucas and Martin the bots should be OK.
1541
1542         Unreviewed. (I ran the idea by Lucas and Martin.)
1543
1544         * Scripts/run-webkit-tests:
1545         (platformIsReadyForParallelTesting):
1546
1547 2011-12-02  Martin Robinson  <mrobinson@igalia.com>
1548
1549         Try to fix the GTK+ build.
1550
1551         * gtk/jhbuild.modules: Properly specify the glib-networking
1552         dependency for soup.
1553
1554 2011-12-02  Martin Robinson  <mrobinson@igalia.com>
1555
1556         Fix a syntax error in the committers.py file.
1557
1558         * Scripts/webkitpy/common/config/committers.py: Add a missing comma.
1559
1560 2011-12-02  Ryosuke Niwa  <rniwa@webkit.org>
1561
1562         Add Alan Stearns and Peter Linss to the contributors list.
1563
1564         * Scripts/webkitpy/common/config/committers.py:
1565
1566 2011-12-02  Kent Tamura  <tkent@chromium.org>
1567
1568         [Chromium] Fix compilation warnings for ASSERT()
1569         https://bugs.webkit.org/show_bug.cgi?id=73623
1570
1571         Reviewed by Tony Chang.
1572
1573         * DumpRenderTree/chromium/ImageDiff.cpp:
1574          - Fix fprintf format: %s -> %d for __LINE__
1575          - Do nothing for ASSERT() if NDEBUG.
1576
1577 2011-12-02  Eric Seidel  <eric@webkit.org>
1578
1579         Enable parallel testing for all Mac bots.
1580
1581         Unreviewed.
1582
1583         * Scripts/run-webkit-tests:
1584         (platformIsReadyForParallelTesting):
1585
1586 2011-12-02  Eric Seidel  <eric@webkit.org>
1587
1588         Enable parallel testing for run-webkit-tests on Mac Lion.
1589         I plan to move all the machines to parallel testing shortly
1590         but we're starting with Lion.
1591
1592         Unreviewed.
1593
1594         * Scripts/run-webkit-tests:
1595         (platformIsReadyForParallelTesting):
1596
1597 2011-12-02  Kausalya Madhusudhanan  <kmadhusu@chromium.org>
1598
1599         Added helper method to identify whether the page has custom page size style.
1600         https://bugs.webkit.org/show_bug.cgi?id=73585
1601
1602         Reviewed by Darin Fisher.
1603
1604         * DumpRenderTree/chromium/LayoutTestController.cpp:
1605         (LayoutTestController::LayoutTestController):
1606         (LayoutTestController::hasCustomPageSizeStyle):
1607         * DumpRenderTree/chromium/LayoutTestController.h:
1608
1609 2011-12-01  Ryosuke Niwa  <rniwa@webkit.org>
1610
1611         [NRWT] reftest should support having multiple references per test
1612         https://bugs.webkit.org/show_bug.cgi?id=73613
1613
1614         Reviewed by Dirk Pranke.
1615
1616         Add a support for having multiple reference files for a single test.
1617
1618         Because a reftest succeeds when it matches at least one of expected matches and fails when it matches
1619         at least one of expected mismatches, we compare expected mismatches first in order to minimize
1620         the number of reference files to open on DRT.
1621
1622         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1623         (interpret_test_failures): Remove checks no longer applicable.
1624         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1625         (ResultSummaryTest.test_interpret_test_failures): Ditto.
1626         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
1627         (SingleTestRunner.__init__): Remove a bunch of code and just call port.reference_files.
1628         (SingleTestRunner._driver_input):
1629         (SingleTestRunner.run):
1630         (SingleTestRunner._run_reftest): Compare the output of the test to each reference file.
1631         * Scripts/webkitpy/layout_tests/models/test_input.py:
1632         (TestInput.__init__): Remove ref_file and is_mismatch_reftest because they are no longer used.
1633         * Scripts/webkitpy/layout_tests/port/base.py:
1634         (Port.reference_files): Renamed from _reference_file_for. Returns a list of expectation, filename pairs.
1635         (_parse_reftest_list): Now supports parsing multiple entries for a single test.
1636         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1637         (PortTest.test_parse_reftest_list):
1638         * Scripts/webkitpy/layout_tests/port/test.py:
1639         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1640         (MainTest.test_unexpected_failures):
1641         (MainTest.test_reftest_skipped_if_unlisted): Renamed from test_missing_and_unexpected_results.
1642         (EndToEndTest.test_end_to_end):
1643         (EndToEndTest.test_reftest_with_two_notrefs): Added.
1644
1645 2011-12-02  Gustavo Noronha Silva  <gns@gnome.org>
1646
1647         Build libsoup without gnome dependencies (like keyring).
1648
1649         * gtk/jhbuild.modules:
1650
1651 2011-12-02  Gustavo Noronha Silva  <gns@gnome.org>
1652
1653         Also build gcrypt and p11-kit, and make them dependencies of
1654         gnutls.
1655
1656         * gtk/jhbuild.modules:
1657
1658 2011-12-02  Kevin Ollivier  <kevino@theolliviers.com>
1659
1660         [wx] Unreviewed build fixes for Windows build.
1661
1662         * waf/build/wxpresets.py:
1663
1664 2011-12-01  Eric Seidel  <eric@webkit.org>
1665
1666         Reviewed by Adam Barth.
1667
1668         webkit-patch post, post-commits, upload should warn when posting to a closed bug, and offer to reopen it
1669         https://bugs.webkit.org/show_bug.cgi?id=32006
1670
1671         I decided not to make it warn, and just have it re-open the bug.
1672         That's not that different from today's behavior which will
1673         just silently attach the patch.
1674
1675         This patch makes behavior between upload and land-safely consistent
1676         (previously one would assign patches and the other would not)
1677         as well as adds the ability for both to ensure that the bug is open.
1678
1679         To test this I had to add a few more methods to MockBugzilla which
1680         (positively) affected a few other test results.
1681
1682         I also made AbstractStep keep a cached copy of the Bug object
1683         and used the cached copy where appropriate (including for 'bug_title').
1684         This should reduce the number of bug fetches we perform.
1685
1686         * Scripts/webkitpy/tool/commands/download_unittest.py:
1687         * Scripts/webkitpy/tool/commands/upload.py:
1688         * Scripts/webkitpy/tool/commands/upload_unittest.py:
1689         * Scripts/webkitpy/tool/mocktool.py:
1690         * Scripts/webkitpy/tool/steps/__init__.py:
1691         * Scripts/webkitpy/tool/steps/abstractstep.py:
1692         * Scripts/webkitpy/tool/steps/closebug.py:
1693         * Scripts/webkitpy/tool/steps/ensurebugisopenandassigned.py: Added.
1694         * Scripts/webkitpy/tool/steps/postdiff.py:
1695         * Scripts/webkitpy/tool/steps/postdiffforcommit.py:
1696         * Scripts/webkitpy/tool/steps/preparechangelog.py:
1697         * Scripts/webkitpy/tool/steps/steps_unittest.py:
1698         * Scripts/webkitpy/tool/steps/updatechangelogswithreviewer.py:
1699
1700 2011-12-02  Gustavo Noronha Silva  <gns@gnome.org>
1701
1702         Also pass --no-interact to jhbuild when updating dependencies.
1703
1704         * Scripts/update-webkitgtk-libs:
1705
1706 2011-12-02  Alice Boxhall  <aboxhall@chromium.org>
1707
1708         Make _build_path check more thoroughly that build paths exist.
1709         https://bugs.webkit.org/show_bug.cgi?id=73601
1710
1711         Reviewed by Ojan Vafai.
1712
1713         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1714         (ChromiumMacPort._build_path):
1715
1716 2011-12-02  Gustavo Noronha Silva  <gns@gnome.org>
1717
1718         Add libffi to the jhbuild modules list as a glib dependency.
1719
1720         Rubber-stamped by Martin Robinson.
1721
1722         * gtk/jhbuild.modules:
1723
1724 2011-12-02  Gustavo Noronha Silva  <gns@gnome.org>
1725
1726         Build fix. Use internal pcre on glib, make gnutls a dependency of
1727         glib-networking, unset AR_FLAGS for jhbuild builds, and run jhbuild
1728         in non-interactive mode.
1729
1730         * Scripts/update-webkitgtk-libs:
1731         * gtk/jhbuild.modules:
1732         * gtk/run-with-jhbuild:
1733
1734 2011-12-02  Adam Roben  <aroben@apple.com>
1735
1736         Fix for Python 2.6
1737
1738         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1739         (RunUnitTests.countFailures): Removed the flags parameter to re.split, which was only added
1740         in Python 2.7, and used whitespace instead of start-of-string/end-of-string markers when
1741         searching for "Tests that timed out:" etc. This makes our check less strict, but it
1742         shouldn't matter given the kinds of content we parse.
1743
1744 2011-12-02  Adam Roben  <aroben@apple.com>
1745
1746         Fix names of failed/timed out tests in run-api-tests output
1747
1748         Fixes <http://webkit.org/b/73663> run-api-tests prints suite names twice for failed or timed
1749         out tests
1750
1751         We were ending up with the suite name doubled (e.g., "WebKit2.WebKit2.MouseMoveAfterCrash").
1752
1753         Reviewed by Simon Fraser.
1754
1755         * Scripts/run-api-tests:
1756         (runTest): Don't prepend the suite name to $test. It's already been prepended at the start
1757         of this function.
1758
1759 2011-11-29  Martin Robinson  <mrobinson@igalia.com>
1760
1761         [GTK] generate-gtkdoc should respect build type
1762         https://bugs.webkit.org/show_bug.cgi?id=73296
1763
1764         Reviewed by Philippe Normand.
1765
1766         * Scripts/webkitdirs.pm:
1767         (buildAutotoolsProject): Pass the --debug to the generate-gtkdoc script.
1768         * gtk/common.py: If --build is passed give preference to the Debug build directory.
1769
1770 2011-12-02  Tony Chang  <tony@chromium.org>
1771
1772         [nrwt] fix wdiff output
1773         https://bugs.webkit.org/show_bug.cgi?id=73604
1774
1775         Reviewed by Adam Barth.
1776
1777         * Scripts/webkitpy/layout_tests/port/base.py:
1778         (Port.wdiff_text):
1779         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1780         (PortTest.test_wdiff_text):
1781
1782 2011-12-02  Adam Roben  <aroben@apple.com>
1783
1784         Teach build.webkit.org to display how many unit tests failed or timed out
1785
1786         Fixes <http://webkit.org/b/73659> It's hard to tell how many unit tests are failing on
1787         build.webkit.org
1788
1789         We now display something like "5 unit tests failed or timed out". Eventually we might want
1790         to split out how many failure vs. timeouts there were.
1791
1792         Reviewed by Darin Adler.
1793
1794         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1795         (TestWithFailureCount): Moved this class up to the top of the file so it is before any other
1796         classes that may want to subclass it.
1797         (RunUnitTests): Changed to inherit from TestWithFailureCount.
1798         (RunUnitTests.countFailures): Added. Counts the number of tests following the "Tests that
1799         timed out:" and "Tests that failed:" lines, if present.
1800
1801         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py:
1802         (StubStdio):
1803         (StubRemoteCommand):
1804         Added these two stub classes to mimic buildbot's RemoteCommand and log classes.
1805
1806         (RunUnitTestsTest.assertFailures): Helper method to check that we interpreted the results of
1807         the test run correctly.
1808
1809         (RunUnitTestsTest.test_no_failures_or_timeouts):
1810         (RunUnitTestsTest.test_one_failure):
1811         (RunUnitTestsTest.test_multiple_failures):
1812         (RunUnitTestsTest.test_one_timeout):
1813         (RunUnitTestsTest.test_multiple_timeouts):
1814         (RunUnitTestsTest.test_multiple_failures_and_timeouts):
1815         Test various cases.
1816
1817 2011-12-02  Adam Roben  <aroben@apple.com>
1818
1819         Teach prepare-ChangeLog to treat master.cfg as a Python file
1820
1821         Fixes <http://webkit.org/b/73658> prepare-ChangeLog doesn't show modified classes/functions
1822         in buildbot's master.cfg
1823
1824         Reviewed by Darin Adler.
1825
1826         * Scripts/prepare-ChangeLog:
1827         (get_function_line_ranges): Use get_function_line_ranges_for_python if the filename is
1828         master.cfg.
1829
1830 2011-12-02  Martin Robinson  <mrobinson@igalia.com>
1831
1832         [GTK] Create a wrapper script that knows how to install jhbuild and run commands with it
1833         https://bugs.webkit.org/show_bug.cgi?id=73669
1834
1835         Reviewed by Gustavo Noronha Silva.
1836
1837         Add a wrapper script that knows how to install jhbuild and also to wrap commands
1838         in the jhbuild shell.
1839
1840         * Scripts/webkitdirs.pm:
1841         (runAutogenForAutotoolsProject): Use the new wrapper script.
1842         (mustReRunAutogen): Ditto.
1843         (buildAutotoolsProject): Ditto.
1844         * Scripts/webkitpy/layout_tests/port/gtk.py:
1845         (GtkDriver.cmd_line): Ditto.
1846         * gtk/run-with-jhbuild: Added.
1847
1848 2011-12-01  Adam Roben  <aroben@apple.com>
1849
1850         Teach check-webkit-style how to check the syntax of JSON files
1851
1852         Fixes <http://webkit.org/b/73590> check-webkit-style doesn't flag JSON syntax errors
1853
1854         I previously landed this patch as r101711, but the new tests failed on Snow Leopard. This
1855         patch is identical to that one except for two change: assertIn(a, b) (which is new to Python
1856         2.7) has been replaced with assertTrue(a in b), and the test_missing_closing_brace test has
1857         been removed, since its output differs by platform.
1858
1859         Reviewed by Darin Adler.
1860
1861         * Scripts/webkitpy/style/checker.py:
1862         (_all_categories): Added JSONChecker's categories to the set of all categories.
1863         (FileType): Added a JSON type. Incremented other types.
1864         (CheckerDispatcher._file_type): Use the JSON file type for .json files.
1865         (CheckerDispatcher._create_checker): Use a JSONChecker for JSON files.
1866
1867         * Scripts/webkitpy/style/checker_unittest.py:
1868         (CheckerDispatcherDispatchTest.assert_checker_json): Added this helper method.
1869         (CheckerDispatcherDispatchTest.test_json_paths): Added. Based on test_python_paths.
1870
1871         * Scripts/webkitpy/style/checkers/jsonchecker.py: Added. (I didn't name this just "json",
1872         which would have matched our other checkers, because I couldn't figure out how to call
1873         "json.loads" without hitting namespace conflicts.)
1874         (JSONChecker.__init__): Turn of line filtering so that we always check the whole file, not
1875         just the modified lines from a patch.
1876         (JSONChecker.check): Try to parse the lines as JSON. Mark an error if there was an
1877         exception.
1878         (JSONChecker.line_number_from_json_exception): Parse the json modules exception message to
1879         try to extract a line number.
1880
1881         * Scripts/webkitpy/style/checkers/jsonchecker_unittest.py: Added.
1882         (MockErrorHandler.__init__):
1883         (MockErrorHandler.turn_off_line_filtering):
1884         (MockErrorHandler.__call__):
1885         Helper class. Copied from xml_unittest.py.
1886
1887         (JSONCheckerTest.test_line_number_from_json_exception): Test the
1888         line_number_from_json_exception helper method.
1889
1890         (JSONCheckerTest.assert_no_error):
1891         (JSONCheckerTest.assert_error):
1892         Helper methods to assert that we did or didn't get an error.
1893
1894         (JSONCheckerTest.mock_handle_style_error): Helper method.
1895
1896         (JSONCheckerTest.test_conflict_marker):
1897         (JSONCheckerTest.test_single_quote):
1898         (JSONCheckerTest.test_init):
1899         (JSONCheckerTest.test_no_error):
1900         Test various cases.
1901
1902 2011-12-02  Gustavo Noronha Silva  <gns@gnome.org>
1903
1904         [GTK] Add glib, glib-networking, gnutls and libsoup to jhbuild
1905         https://bugs.webkit.org/show_bug.cgi?id=73664
1906
1907         Reviewed by Martin Robinson.
1908
1909         This patch adds our HTTP library and its main dependencies; this
1910         should not affect layout or rendering.
1911
1912         * gtk/jhbuild.modules:
1913
1914 2011-12-02  Gustavo Noronha Silva  <gns@gnome.org>
1915
1916         [GTK] Also run DumpRenderTree under jhbuild
1917         https://bugs.webkit.org/show_bug.cgi?id=73646
1918
1919         Reviewed by Martin Robinson.
1920
1921         * Scripts/webkitpy/layout_tests/port/gtk.py:
1922         (GtkDriver.cmd_line):
1923
1924 2011-12-02  Hajime Morrita  <morrita@chromium.org>
1925
1926         Unreviewed, rolling out r101751 and r101775.
1927         http://trac.webkit.org/changeset/101751
1928         http://trac.webkit.org/changeset/101775
1929         https://bugs.webkit.org/show_bug.cgi?id=73191
1930
1931         breaks Windows build
1932
1933         * DumpRenderTree/chromium/config.h:
1934         * DumpRenderTree/config.h:
1935         * TestWebKitAPI/config.h:
1936         * WebKitTestRunner/config.h:
1937
1938 2011-12-02  Kentaro Hara  <haraken@chromium.org>
1939
1940         Move run-bindings-tests implementation to webkitpy/bindings/main.py
1941         https://bugs.webkit.org/show_bug.cgi?id=73619
1942
1943         Reviewed by Adam Barth.
1944
1945         As run-bindings-tests has been growing, this patch just moves its implementation
1946         to webkitpy/bindings/main.py to integrate with the rest of webkitpy.
1947
1948         No change in behavior.
1949
1950         * Scripts/run-bindings-tests:
1951         (main):
1952         * Scripts/webkitpy/bindings/__init__.py: Added.
1953         * Scripts/webkitpy/bindings/main.py: Added.
1954         (BindingsTests.__init__):
1955         (BindingsTests.generate_from_idl):
1956         (BindingsTests.generate_supplemental_dependency):
1957         (BindingsTests.detect_changes):
1958         (BindingsTests.run_tests):
1959
1960 2011-12-01  Ilya Tikhonovsky  <loislo@chromium.org>
1961
1962         Tools/Scripts: eliminate find_test_files from Port class.
1963         https://bugs.webkit.org/show_bug.cgi?id=73553
1964
1965         rebaseline.py and chromium_gpu.py use Port.find_test_files() the same way as it is used in Port.tests()
1966         I'd like to replace all calls to find_test_files with tests and eliminate find_tests_files.
1967         Also I'll move _is_test_file() and related functions close to Port.tests().
1968
1969         Reviewed by Dirk Pranke.
1970
1971         * Scripts/webkitpy/layout_tests/port/base.py:
1972         (Port.tests):
1973         (Port.is_reference_html_file):
1974         (Port._has_supported_extension):
1975         (Port._is_test_file):
1976         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1977         (PortTest.test_find_no_paths_specified):
1978         (PortTest.test_find_one_test):
1979         (PortTest.test_find_glob):
1980         (PortTest.test_find_with_skipped_directories):
1981         (PortTest.test_find_with_skipped_directories_2):
1982         (PortTest.test_is_test_file):
1983         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
1984         (_default_tests_paths):
1985         (ChromiumGpuLinuxPort.tests):
1986         (ChromiumGpuCgMacPort.tests):
1987         (ChromiumGpuMacPort.tests):
1988         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
1989         (ChromiumGpuTest.test_default_tests_paths.test_paths):
1990         (ChromiumGpuTest.test_default_tests_paths):
1991         (ChromiumGpuTest.test_test_files.test_paths):
1992         (ChromiumGpuTest):
1993         (ChromiumGpuTest.test_test_files):
1994         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1995         (get_tests_run.RecordingTestDriver.run_test):
1996         * Scripts/webkitpy/tool/commands/rebaseline.py:
1997         (OptimizeBaselines._optimize_baseline):
1998         (OptimizeBaselines.execute):
1999         (AnalyzeBaselines._analyze_baseline):
2000         (AnalyzeBaselines.execute):
2001
2002 2011-12-01  Hajime Morrita  <morrita@chromium.org>
2003
2004         JS_INLINE and WTF_INLINE should be visible from WebCore
2005         https://bugs.webkit.org/show_bug.cgi?id=73191
2006
2007         Reviewed by Kevin Ollivier.
2008
2009         Removed macro definitions which is now provided by Platform.h
2010
2011         * DumpRenderTree/chromium/config.h:
2012         * DumpRenderTree/config.h:
2013         * TestWebKitAPI/config.h:
2014         * WebKitTestRunner/config.h:
2015
2016 2011-12-01  Hayato Ito  <hayato@chromium.org>
2017
2018         Explicitly pass tolerance=0 to port.diff_image in case of RefTestMismatch failure.
2019         https://bugs.webkit.org/show_bug.cgi?id=73406
2020
2021         Reviewed by Ryosuke Niwa.
2022
2023         WebKitPort's image_diff uses tolerance='0.1' in default.
2024         When reftests fail, we should use tolerace=0 when diff-ing images.
2025
2026         Since ImageDiff on chromium port doesn't use tolerance value as of now,
2027         this change doesn't affect chromium port's behavior.
2028
2029         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
2030         (write_test_result):
2031         * Scripts/webkitpy/layout_tests/port/base.py:
2032         (Port.diff_image):
2033         * Scripts/webkitpy/layout_tests/port/chromium.py:
2034         (ChromiumPort.diff_image):
2035         * Scripts/webkitpy/layout_tests/port/test.py:
2036         (TestPort.diff_image):
2037         * Scripts/webkitpy/layout_tests/port/webkit.py:
2038         (WebKitPort.diff_image):
2039         (WebKitPort._start_image_diff_process):
2040         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2041         (MainTest.test_tolerance.ImageDiffTestPort.diff_image):
2042
2043 2011-12-01  Adam Klein  <adamk@chromium.org>
2044
2045         Add Chromium ToT GTest build bots (and group selection support) to flakiness dashboard
2046         https://bugs.webkit.org/show_bug.cgi?id=73599
2047
2048         Reviewed by Ojan Vafai.
2049
2050         * TestResultServer/static-dashboards/builders.js:
2051         * TestResultServer/static-dashboards/dashboard_base.js:
2052         ():
2053         (htmlForTestTypeSwitcher):
2054         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
2055         (testHtmlForTestTypeSwitcherGroup):
2056
2057 2011-12-01  Ryosuke Niwa  <rniwa@webkit.org>
2058
2059         Parse reftest.list and extract types of ref tests
2060         https://bugs.webkit.org/show_bug.cgi?id=66837
2061
2062         Reviewed by Dirk Pranke.
2063
2064         Add support for reftest.list to base port.
2065
2066         * Scripts/webkitpy/common/find_files.py:
2067         (find):
2068         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
2069         (SingleTestRunner.__init__):
2070         * Scripts/webkitpy/layout_tests/port/base.py:
2071         (Port.__init__): Initialize self._reftest_list. It's a dictionary mapping from a test directory
2072         to a dictionary of {test path: ("==" or "!=", reference file path)}
2073         (Port._get_reftest_list): Added; calls test_file.parse_reftest_list to fill self._reftest_list.
2074         (Port._reference_file_for): Added; obtains the reference file name given a test name.
2075         (Port.is_reftest): Added; Calls _reference_file_for.
2076         (Port.reftest_expected_filename): Calls _reference_file_for.
2077         (Port.reftest_expected_mismatch_filename): Ditto.
2078         (Port.find_test_files):
2079         (is_reference_html_file): Treat any file that starts with ref- or notref- or ends with
2080         -expected, -expected-mismach, -ref, or -notref as a reference file.
2081         (_is_test_file):
2082         (_parse_reftest_list): Added.
2083         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2084         (PortTest.test_is_test_file):
2085         (PortTest.test_parse_reftest_list):
2086         * Scripts/webkitpy/layout_tests/port/dryrun.py:
2087         (DryrunDriver.run_test):
2088         * Scripts/webkitpy/layout_tests/port/test.py:
2089         (unit_test_filesystem.add_test_file):
2090         (unit_test_filesystem.add_file):
2091         (unit_test_filesystem):
2092         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2093         (get_tests_run.RecordingTestDriver.run_test):
2094         (MainTest.test_unexpected_failures):
2095         (MainTest.test_missing_and_unexpected_results):
2096         (EndToEndTest.test_end_to_end):
2097         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
2098         (Rebaseliner._compile_rebaselining_tests):
2099
2100 2011-12-01  Sheriff Bot  <webkit.review.bot@gmail.com>
2101
2102         Unreviewed, rolling out r101711.
2103         http://trac.webkit.org/changeset/101711
2104         https://bugs.webkit.org/show_bug.cgi?id=73605
2105
2106         Broke 3 webkitpy tests (Requested by rniwa on #webkit).
2107
2108         * Scripts/webkitpy/style/checker.py:
2109         (_all_categories):
2110         (FileType):
2111         (CheckerDispatcher._file_type):
2112         (CheckerDispatcher._create_checker):
2113         * Scripts/webkitpy/style/checker_unittest.py:
2114         (CheckerDispatcherDispatchTest.assert_checker_cpp):
2115         (CheckerDispatcherDispatchTest.test_cpp_paths):
2116         * Scripts/webkitpy/style/checkers/jsonchecker.py: Removed.
2117         * Scripts/webkitpy/style/checkers/jsonchecker_unittest.py: Removed.
2118
2119 2011-12-01  Gustavo Noronha Silva  <gns@gnome.org>
2120
2121         Unreviewed. Add myself to GStreamer and WebKit2GTK+ watchlists.
2122
2123         * Scripts/webkitpy/common/config/watchlist:
2124
2125 2011-12-01  Martin Robinson  <mrobinson@igalia.com>
2126
2127         [GTK] Read fonts from the jhbuild root
2128         https://bugs.webkit.org/show_bug.cgi?id=73487
2129
2130         Reviewed by Gustavo Noronha Silva.
2131
2132         Read fonts from the jhbuild root instead of from the system. This will ensure
2133         that all testers use the same fonts instead of leaving this up to luck.
2134
2135         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2136         (initializeFonts): Load fonts from jhbuild root.
2137         * Scripts/webkitpy/layout_tests/port/gtk.py:
2138         (GtkPort.setup_environ_for_server): Pass an environment variable containing
2139         the path to the jhbuild root.
2140         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp:
2141         (WTR::inititializeFontConfigSetting): Load fonts from the jhbuild root.
2142
2143 2011-12-01  Adam Roben  <aroben@apple.com>
2144
2145         Teach check-webkit-style how to check the syntax of JSON files
2146
2147         Fixes <http://webkit.org/b/73590> check-webkit-style doesn't flag JSON syntax errors
2148
2149         Reviewed by Darin Adler.
2150
2151         * Scripts/webkitpy/style/checker.py:
2152         (_all_categories): Added JSONChecker's categories to the set of all categories.
2153         (FileType): Added a JSON type. Incremented other types.
2154         (CheckerDispatcher._file_type): Use the JSON file type for .json files.
2155         (CheckerDispatcher._create_checker): Use a JSONChecker for JSON files.
2156
2157         * Scripts/webkitpy/style/checker_unittest.py:
2158         (CheckerDispatcherDispatchTest.assert_checker_json): Added this helper method.
2159         (CheckerDispatcherDispatchTest.test_json_paths): Added. Based on test_python_paths.
2160
2161         * Scripts/webkitpy/style/checkers/jsonchecker.py: Added. (I didn't name this just "json",
2162         which would have matched our other checkers, because I couldn't figure out how to call
2163         "json.loads" without hitting namespace conflicts.)
2164         (JSONChecker.__init__): Turn of line filtering so that we always check the whole file, not
2165         just the modified lines from a patch.
2166         (JSONChecker.check): Try to parse the lines as JSON. Mark an error if there was an
2167         exception.
2168         (JSONChecker.line_number_from_json_exception): Parse the json modules exception message to
2169         try to extract a line number.
2170
2171         * Scripts/webkitpy/style/checkers/jsonchecker_unittest.py: Added.
2172         (MockErrorHandler.__init__):
2173         (MockErrorHandler.turn_off_line_filtering):
2174         (MockErrorHandler.__call__):
2175         Helper class. Copied from xml_unittest.py.
2176
2177         (JSONCheckerTest.test_line_number_from_json_exception): Test the
2178         line_number_from_json_exception helper method.
2179
2180         (JSONCheckerTest.assert_no_error):
2181         (JSONCheckerTest.assert_error):
2182         Helper methods to assert that we did or didn't get an error.
2183
2184         (JSONCheckerTest.mock_handle_style_error): Helper method.
2185
2186         (JSONCheckerTest.test_conflict_marker):
2187         (JSONCheckerTest.test_single_quote):
2188         (JSONCheckerTest.test_init):
2189         (JSONCheckerTest.test_missing_closing_brace):
2190         (JSONCheckerTest.test_no_error):
2191         Test various cases.
2192
2193 2011-12-01  Martin Robinson  <mrobinson@igalia.com>
2194
2195         [GTK] Add a helper function to find the current executable's path
2196         https://bugs.webkit.org/show_bug.cgi?id=73473
2197
2198         Reviewed by Gustavo Noronha Silva.
2199
2200         Update MiniBrowser to not pass the binary directory as an environment variable.
2201         This means that you can move the binaries around without it breaking.
2202
2203         * MiniBrowser/gtk/main.c:
2204         (main): No longer set the environment variable.
2205         * WebKitTestRunner/GNUmakefile.am: No longer add the directory path define.
2206         * WebKitTestRunner/gtk/main.cpp:  Ditto.
2207
2208 2011-12-01  Gustavo Noronha Silva  <gns@gnome.org>
2209
2210         Unreviewed. Fix bad file =(.
2211
2212         * gtk/jhbuild.modules:
2213         * Scripts/webkitdirs.pm:
2214         (runAutogenForAutotoolsProject):
2215
2216 2011-12-01  Gustavo Noronha Silva  <gns@gnome.org>
2217
2218         [GTK] Add freetype to our jhbuild setup
2219         https://bugs.webkit.org/show_bug.cgi?id=73488
2220
2221         Reviewed by Martin Robinson.
2222
2223         This adds the first library to our jhbuild setup, and makes sure
2224         build-webkit calls autogen.sh and make with jhbuild, so that the
2225         environment is properly set.
2226
2227         * Scripts/webkitdirs.pm:
2228         (saveSum):
2229         (hashFile):
2230         (runAutogenForAutotoolsProject): save md5sum of jhbuild-related files, and
2231         call autogen under jhbuild run;
2232         (mustRunAutogen): generalized the arguments change checking to also force
2233         running autogen when jhbuild files change;
2234         (buildAutotoolsProject): run make under jhbuild;
2235         * gtk/jhbuild.modules: add freetype.
2236
2237 2011-12-01  Dominic Mazzoni  <dmazzoni@google.com>
2238
2239         Fix WebKitTestRunner compile warnings with XCode 3.2
2240         https://bugs.webkit.org/show_bug.cgi?id=73378
2241
2242         Reviewed by Chris Fleizach.
2243
2244         * WebKitTestRunner/InjectedBundle/AccessibilityTextMarker.cpp:
2245         (WTR::AccessibilityTextMarker::AccessibilityTextMarker):
2246         * WebKitTestRunner/InjectedBundle/AccessibilityTextMarkerRange.cpp:
2247         (WTR::AccessibilityTextMarkerRange::AccessibilityTextMarkerRange):
2248         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
2249         (WTR::AccessibilityUIElement::AccessibilityUIElement):
2250         (WTR::AccessibilityUIElement::verticalScrollbar):
2251
2252 2011-12-01  Adam Roben  <aroben@apple.com>
2253
2254         Allow committers to use their Trac credentials to force builds on the buildbots
2255
2256         Fixes <http://webkit.org/b/73353>
2257
2258         A new class, CommitterAuth, handles authentication of WebKit committers. CommitterAuth uses
2259         three files to do its job: a config file that contains a list of WebKit committer usernames,
2260         an htdigest file that contains Trac credentials, and JSON file that gives the paths for
2261         those two files.
2262
2263         Reviewed by Darin Adler.
2264
2265         * BuildSlaveSupport/build.webkit.org-config/committer_auth.py: Added.
2266         (Error): Basic wrapper around Exception that we use for cases where we couldn't even check
2267         whether credentials were valid or not.
2268         (CommitterAuth.__init__): Just store the path to auth.json.
2269         (CommitterAuth.auth_json): Load, parse, and return auth.json.
2270         (CommitterAuth.auth_json_filename): Return the path to auth.json.
2271         (CommitterAuth.authenticate): Return true if the user is a WebKit committer and their
2272         credentials are valid Trac credentials. Return false otherwise or if an error occurred while
2273         checking those conditions.
2274         (CommitterAuth.is_webkit_committer): Return true if the user is a WebKit committer. Return
2275         false otherwise or if an exception was thrown.
2276         (CommitterAuth.is_webkit_trac_user): Return true if the username/password are present in the
2277         Trac credentials htdigest file. Return false otherwise or if an exception was thrown.
2278
2279         (CommitterAuth.open_auth_json_file):
2280         (CommitterAuth.open_trac_credentials_file):
2281         (CommitterAuth.open_webkit_committers_file):
2282         Open the specified file. These are mostly useful for testing purposes.
2283
2284         (CommitterAuth.trac_credentials_filename):
2285         (CommitterAuth.webkit_committers_filename):
2286         Return the path to the specified file by retrieving it from auth.json.
2287
2288         (CommitterAuth.webkit_committers): Load and parse the committers file and extract the list
2289         of WebKit committers from it.
2290
2291         * BuildSlaveSupport/build.webkit.org-config/committer_auth_unittest.py: Added.
2292         (CMStringIO.__enter__):
2293         (CMStringIO.__exit__):
2294         Helper class that makes it possible to use StringIO with the "with" statement.
2295
2296         (open_override): Helper context manager for overriding the global "open" function
2297         temporarily.
2298
2299         (CommitterAuthTest.setUp): Set up a somewhat-mocked CommitterAuth that is used by most
2300         tests.
2301         (CommitterAuthTest.fake_open_function): Returns a function that can be used in place of
2302         "open" to test that the expected path was opened.
2303         (CommitterAuthTest.test_authentication_success): Test that committers can authenticate
2304         successfully.
2305         (CommitterAuthTest.test_committer_without_trac_credentials_fails): Test that committers who
2306         somehow have no Trac account can't authenticate.
2307
2308         (CommitterAuthTest.test_fail_to_open_auth_json_file):
2309         (CommitterAuthTest.test_fail_to_open_trac_credentials_file):
2310         (CommitterAuthTest.test_fail_to_open_webkit_committers_file):
2311         Test what happens when we can't open the three files we depend upon.
2312
2313         (CommitterAuthTest.test_implements_IAuth): Test that we fulfill buildbot's expectations for
2314         an authentication class.
2315
2316         (CommitterAuthTest.test_invalid_auth_json_file):
2317         (CommitterAuthTest.test_invalid_committers_file):
2318         (CommitterAuthTest.test_invalid_trac_credentials_file):
2319         (CommitterAuthTest.test_missing_auth_json_keys):
2320         Test what happens when the three files we depend upon are invalid in some way.
2321
2322         (CommitterAuthTest.test_open_auth_json_file):
2323         (CommitterAuthTest.test_open_trac_credentials_file):
2324         (CommitterAuthTest.test_open_webkit_committers_file):
2325         Test that we open the expected paths.
2326
2327         (CommitterAuthTest.test_trac_credentials_filename):
2328         (CommitterAuthTest.test_webkit_committers_filename):
2329         Test that we extract filenames out of auth.json correctly.
2330
2331         (CommitterAuthTest.test_non_committer_fails):
2332         (CommitterAuthTest.test_unknown_user_fails):
2333         (CommitterAuthTest.test_username_is_prefix_of_valid_user):
2334         (CommitterAuthTest.test_wrong_password_fails):
2335         Test various failed authentication attempts.
2336
2337         (CommitterAuthTest.test_webkit_committers): Test that we can parse the list of WebKit
2338         committers out of the committers file correctly.
2339
2340         (CommitterAuthTest.fake_auth_json_file):
2341         (CommitterAuthTest.invalid_auth_json_file):
2342         (CommitterAuthTest.fake_committers_file):
2343         (CommitterAuthTest.invalid_committers_file):
2344         (CommitterAuthTest.fake_htdigest_file):
2345         (CommitterAuthTest.invalid_htdigest_file):
2346         Return various fake files for testing.
2347
2348         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Specify an instance of CommitterAuth
2349         to be used for authentication in the web interface, and specify that only authenticated
2350         users may force builds.
2351
2352 2011-12-01  Adam Roben  <aroben@apple.com>
2353
2354         Add an HTDigestParser class to webkitpy
2355
2356         Fixes <http://webkit.org/b/73575> webkitpy doesn't provide a way to parse htdigest files
2357
2358         This class can be used to parse Apache's htdigest files and check whether a given
2359         username/realm/password tuple is present in the file. Eventually this will be used for
2360         authenticating users on build.webkit.org (<http://webkit.org/b/73353>).
2361
2362         Reviewed by Eric Seidel.
2363
2364         * Scripts/webkitpy/common/net/htdigestparser.py: Added.
2365         (HTDigestParser.__init__): Stores the parsed representation of the file.
2366         (HTDigestParser.authenticate): Hashes the username/realm/password tuple to generate a hashed
2367         password and returns whether the resulting tuple is present in the file.
2368         (HTDigestParser.entries): Just returns the parsed representation of the file.
2369         (HTDigestParser.parse_file): Splits each line on colons and checks that each line has the
2370         expected syntax ('username:realm:hashed_password'). If any line is invalid, we treat the
2371         whole file as invalid and all authentication attempts will fail.
2372
2373         * Scripts/webkitpy/common/net/htdigestparser_unittest.py: Added.
2374         (HTDigestParserTest.assertEntriesEqual): Helper method to assert that fake_htdigest_file,
2375         optionally appended with some extra data, generates the expected entries.
2376         (HTDigestParserTest.test_authenticate): Tests that tuples present in the file can
2377         authenticate, and tuples not present cannot.
2378         (HTDigestParserTest.test_entries): Tests that we get the expected entries from
2379         fake_htdigest_file.
2380
2381         (HTDigestParserTest.test_empty_file):
2382         (HTDigestParserTest.test_too_few_colons):
2383         (HTDigestParserTest.test_too_many_colons):
2384         (HTDigestParserTest.test_invalid_hash):
2385         Test various forms of invalid files.
2386
2387         (HTDigestParserTest.fake_htdigest_file): Returns a fake valid htdigest file for testing.
2388
2389 2011-12-01  Philippe Normand  <pnormand@igalia.com>
2390
2391         [GTK] Make the new 64-bit Release bot part of the core set
2392         https://bugs.webkit.org/show_bug.cgi?id=73570
2393
2394         Reviewed by Adam Roben.
2395
2396         The 64-bit Release bot re-introduced in r101676 replaces the 64-bit
2397         Debug bot which was already part of the core set. The GTK
2398         gardeners are willing to actively maintain it.
2399
2400         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
2401         (BuildBot.__init__):
2402         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2403         (test_builder_name_regexps):
2404
2405 2011-12-01  Philippe Normand  <pnormand@igalia.com>
2406
2407         [GTK] Buildslave switch from 32-bit Debug to 64-bit Release
2408         https://bugs.webkit.org/show_bug.cgi?id=73547
2409
2410         Reviewed by Adam Roben.
2411
2412         * BuildSlaveSupport/build.webkit.org-config/config.json: Rename
2413         the 32-bit Debug configuration to 64-bit Release. The new slave is
2414         a 64-bit machine already. The 32-bit slave will be shut down.
2415
2416 2011-12-01  Martin Robinson  <mrobinson@igalia.com>
2417
2418         [GTK] build-jsc should not trigger gtkdoc generation
2419         https://bugs.webkit.org/show_bug.cgi?id=73552
2420
2421         Reviewed by Philippe Normand.
2422
2423         Do not generate gtkdoc when building projects other than WebKit. In particular, this ensures
2424         that build-jsc does not try to generate gtkdoc.
2425
2426         * Scripts/webkitdirs.pm:
2427         (buildAutotoolsProject): Only generate gtkdoc if the project is "WebKit".
2428
2429 2011-12-01  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
2430
2431         tests_run0.txt gets clobbered when re-running failing tests
2432         https://bugs.webkit.org/show_bug.cgi?id=63844
2433
2434         Reviewed by Dirk Pranke.
2435
2436         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2437         (Manager.results_directory):
2438         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
2439         (AbstractWorker.__init__):
2440         (_ManagerConnection.start_worker):
2441         (_InlineManager.start_worker):
2442         (_MultiProcessManager.start_worker):
2443         (_WorkerConnection.__init__):
2444         (_InlineWorkerConnection.__init__):
2445         (_MultiProcessWorkerConnection.__init__):
2446         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
2447         (_TestWorker.__init__):
2448         (_TestsMixin.test_cancel):
2449         (_TestsMixin.test_done):
2450         (_TestsMixin.test_unknown_message):
2451         (InterfaceTest.test_managerconnection_is_abstract):
2452         (InterfaceTest.test_workerconnection_is_abstract):
2453         * Scripts/webkitpy/layout_tests/controllers/worker.py:
2454         (Worker.__init__):
2455         (Worker.safe_init):
2456         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2457         (MainTest.test_retries_directory):
2458
2459 2011-12-01  Adam Roben  <aroben@apple.com>
2460
2461         Buildbot fix
2462
2463         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Moved all of the initialization of
2464         BuildmasterConfig not performed by loadBuilderConfig up to the top of the file. This makes
2465         it possible for build steps to reference things like the buildbotURL property of the config.
2466
2467 2011-11-30  Adam Roben  <aroben@apple.com>
2468
2469         Use curl for downloading builds to test slaves
2470
2471         Buildbot's built-in file transfer capabilities are transferring files much more slowly than
2472         curl does. This seems to be due to a combination of buildbot being single-threaded (and thus
2473         often busy servicing web requests, etc.) and using an IPC mechanism with fairly high
2474         overhead (courtesy of the Twisted framework).
2475
2476         Eventually we'd like to make Buildbot's built-in transferring more efficient. In the
2477         meantime we'll try using curl for downloads and keep thinking about a solution for uploads.
2478         This should make the test slaves quite a bit faster, and should ease the CPU load on
2479         build.webkit.org a bit.
2480
2481         Part of <http://webkit.org/b/73484> Mac slaves take 7+ minutes to upload or download a build
2482
2483         Reviewed by Mark Rowe.
2484
2485         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Added the CreateWebKitBuildDirectory
2486         to Test slaves. It's run just before DownloadBuiltProduct.
2487         (DownloadBuiltProduct): Changed to be a normal ShellCommand that invokes curl to perform the
2488         download.
2489         (CreateWebKitBuildDirectory): Added. Creates the WebKitBuild directory on a slave.
2490         FileDownload used to do this for us automatically.
2491
2492 2011-11-21  Philippe Normand  <pnormand@igalia.com>
2493
2494         [GTK] enable-webaudio websetting
2495         https://bugs.webkit.org/show_bug.cgi?id=69836
2496
2497         Reviewed by Martin Robinson.
2498
2499         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2500         (LayoutTestController::overridePreference): Update WebKitGTK
2501         enable-webaudio setting when a layout test requests it.
2502
2503 2011-11-30  Ilya Tikhonovsky  <loislo@chromium.org>
2504
2505         Web Inspector: chromium: I'd like to add a script for running perf tests for WebInspector.
2506         https://bugs.webkit.org/show_bug.cgi?id=73079
2507
2508         The idea is to have performance tests for WebInspector.
2509         I was suggested to put these tests into PerformanceTests/inspector.
2510         They produce output in a format that is suitable for chromium perf bot drawing scripts.
2511
2512         I'd like to reuse code for scanning folders for test files.
2513         Because of the fact that original test_files.py has common and layout specific parts
2514         I decided to extract the generic part and put it into webkitpy/common
2515         I think the two packages with name test_files look strange but I have no idea about a better name.
2516
2517         Reviewed by Dirk Pranke.
2518
2519         * Scripts/run-inspector-perf-tests.py: Added.
2520         * Scripts/webkitpy/common/find_files.py: Renamed from Tools/Scripts/webkitpy/layout_tests/port/test_files.py.
2521         (find):
2522         (_normalize):
2523         * Scripts/webkitpy/common/find_files_unittest.py: Added.
2524         (MockWinFileSystem.join):
2525         (MockWinFileSystem):
2526         (MockWinFileSystem.normpath):
2527         (TestWinNormalize.assert_filesystem_normalizes):
2528         (TestWinNormalize.test_mocked_win):
2529         (TestWinNormalize):
2530         (TestWinNormalize.test_win):
2531         * Scripts/webkitpy/layout_tests/port/base.py:
2532         (Port.tests):
2533         (Port.find_test_files):
2534         (Port._driver_class):
2535         (is_reference_html_file):
2536         (_has_supported_extension):
2537         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2538         (PortTest.test_find_no_paths_specified):
2539         (PortTest.test_find_one_test):
2540         (PortTest.test_find_glob):
2541         (PortTest.test_find_with_skipped_directories):
2542         (PortTest.test_find_with_skipped_directories_2):
2543         (PortTest):
2544         (PortTest.test_is_test_file):
2545         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2546         (_tests):
2547         * Scripts/webkitpy/layout_tests/port/test.py:
2548         * Scripts/webkitpy/layout_tests/port/test_files_unittest.py: Removed.
2549         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2550         * Scripts/webkitpy/performance_tests/__init__.py: Added.
2551         * Scripts/webkitpy/performance_tests/perftestsrunner.py: Added.
2552         (PerfTestsRunner.__init__):
2553         (PerfTestsRunner._parse_args):
2554         (PerfTestsRunner._collect_tests._is_test_file):
2555         (PerfTestsRunner._collect_tests):
2556         (PerfTestsRunner.run):
2557         (PerfTestsRunner._run_tests_set):
2558         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py: Added.
2559         (MainTest.TestPort):
2560         (MainTest.TestPort.create_driver):
2561         (MainTest.TestDriver.run_test):
2562         (MainTest.TestDriver):
2563         (MainTest.TestDriver.stop):
2564         (MainTest.create_runner):
2565         (MainTest.run_test):
2566         (MainTest.test_run_passing_test):
2567         (MainTest.test_run_silent_test):
2568         (MainTest.test_run_failed_test):
2569         (MainTest.test_run_tonguey_test):
2570         (MainTest.test_run_timeout_test):
2571         (MainTest.test_run_crash_test):
2572         (MainTest.test_run_test_set):
2573         (MainTest.test_collect_tests):
2574         (MainTest):
2575         (MainTest.test_parse_args):
2576         * Scripts/webkitpy/tool/commands/rebaseline.py:
2577         (OptimizeBaselines.execute):
2578         (AnalyzeBaselines.execute):
2579
2580 2011-11-30  nduca@chromium.org  <nduca@chromium.org@268f45cc-cd09-0410-ab3c-d52691b4dbfc>
2581
2582         [chromium] Enable threaded compositing via CCThreadProxy::hasThread only
2583         https://bugs.webkit.org/show_bug.cgi?id=70838
2584
2585         Reviewed by James Robinson.
2586
2587         * DumpRenderTree/chromium/TestShell.cpp:
2588         (TestShell::resetWebSettings):
2589         * DumpRenderTree/chromium/TestShell.h:
2590         (TestShell::threadedCompositingEnabled):
2591         * DumpRenderTree/chromium/WebPreferences.cpp:
2592         (WebPreferences::reset):
2593         (WebPreferences::applyTo):
2594         * DumpRenderTree/chromium/WebPreferences.h:
2595         * DumpRenderTree/chromium/WebViewHost.cpp:
2596         (WebViewHost::WebViewHost):
2597         (WebViewHost::~WebViewHost):
2598
2599 2011-11-30  Chris Fleizach  <cfleizach@apple.com>
2600
2601         AX: Nodes are reporting that focus can be set when they really can't
2602         https://bugs.webkit.org/show_bug.cgi?id=72791
2603
2604         Reviewed by Beth Dakin.
2605
2606         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2607         (AccessibilityUIElement::isFocusable):
2608         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityUIElement.idl:
2609         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm:
2610         (WTR::AccessibilityUIElement::isFocusable):
2611
2612 2011-11-30  Alejandro G. Castro  <alex@igalia.com>
2613
2614         Moved myself to the reviewers list.
2615
2616         * Scripts/webkitpy/common/config/committers.py:
2617
2618 2011-11-30  David Levin  <levin@chromium.org>
2619
2620         [chromium] Move didUpdateLayout from WebFrameClient to WebViewClient.
2621         https://bugs.webkit.org/show_bug.cgi?id=73415
2622
2623         Reviewed by Darin Fisher.
2624
2625         * DumpRenderTree/chromium/WebViewHost.cpp:
2626         (WebViewHost::didUpdateLayout): Move didUpdateLayout to account for
2627         it being moved to a new interface and having the parameter removed.
2628         * DumpRenderTree/chromium/WebViewHost.h: Ditto.
2629
2630 2011-11-30  Adam Roben  <aroben@apple.com>
2631
2632         Make NRWT show DRT/WTR build logs when the build fails or --verbose is passed
2633
2634         This will make it a lot easier to investigate build failures on the bots.
2635
2636         Fixes <http://webkit.org/b/71160> NRWT doesn't show build output when building DRT
2637
2638         Reviewed by Eric Seidel.
2639
2640         * Scripts/webkitpy/common/system/executive_mock.py:
2641         (MockExecutive.run_command): Include the mock output in the ScriptError we raise since it
2642         makes the expected output in WebKitPortTest.test_build_driver more closely mimic the output
2643         seen in practice.
2644
2645         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2646         (ManagerTest.test_http_locking): Pass False for the configure_logging parameter to Printer.
2647         The Printer was modifying the root logger in way that would cause my new tests in
2648         WebKitPort.test_build_driver to fail. The value of the configure_logging parameter doesn't
2649         seem important to the functioning of this test, and passing False makes Printer not modify
2650         the root logger.
2651
2652         * Scripts/webkitpy/layout_tests/port/webkit.py:
2653         (WebKitPort._run_script): Log the output of the command at the DEBUG level so it will show
2654         up when --verbose is passed but not otherwise.
2655         (WebKitPort._build_driver): When an error occurs, log the error message and the script's
2656         output at the ERROR level so it will always show up.
2657
2658         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2659         (WebKitPortTest.test_build_driver): Test that nothing is logged in the normal, successful
2660         cases. Test that the build scripts' output is logged when --verbose is passed or the build
2661         fails.
2662
2663 2011-11-30  Adam Roben  <aroben@apple.com>
2664
2665         Add a way to set OutputCapture's log level
2666
2667         This will be used by some unit tests I'm planning to add.
2668
2669         Fixes <http://webkit.org/b/73469> OutputCapture can't be used to capture DEBUG messages
2670
2671         Reviewed by Eric Seidel.
2672
2673         * Scripts/webkitpy/common/system/outputcapture.py:
2674         (OutputCapture.__init__): Initialize our internal logging level to INFO, which was the only
2675         level we ever used before this patch.
2676         (OutputCapture.set_log_level): Added. This is useful for testing what log level code that is
2677         being tested uses. Sets the log level for any current or future logging handlers.
2678         (OutputCapture.capture_output): Use our internal logging level instead of always using INFO.
2679
2680         * Scripts/webkitpy/common/system/outputcapture_unittest.py: Added.
2681         (OutputCaptureTest.setUp): Create an OutputCapture to test.
2682         (OutputCaptureTest.log_all_levels): Helper method to log a message at each log level.
2683         (OutputCaptureTest.assertLogged): Helper method to assert that a particular set of strings
2684         was logged.
2685         (OutputCaptureTest.test_initial_log_level): Test that we start out at the INFO level.
2686         (OutputCaptureTest.test_set_log_level): Test that set_log_level correctly modifies the log
2687         level.
2688
2689 2011-11-29  Beth Dakin  <bdakin@apple.com>
2690
2691         https://bugs.webkit.org/show_bug.cgi?id=72751
2692         WebKit2.MouseMoveAfterCrash API test is failing 
2693
2694         Reviewed by Sam Weinig.
2695
2696         Like DumpRenderTree, the NSWindow should act as an active window.
2697         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
2698         (-[ActiveOffscreenWindow isKeyWindow]):
2699         (TestWebKitAPI::PlatformWebView::PlatformWebView):
2700
2701 2011-11-30  Gustavo Noronha Silva  <gns@gnome.org>
2702
2703         Add step to update gtk dependencies to the buildbot
2704         https://bugs.webkit.org/show_bug.cgi?id=73455
2705
2706         Reviewed by Martin Robinson.
2707
2708         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2709
2710 2011-11-30  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
2711
2712         [Qt] Add a fullscreen option to MiniBrowser.
2713         https://bugs.webkit.org/show_bug.cgi?id=73445
2714
2715         Reviewed by Noam Rosenthal.
2716
2717         This bypasses the compositor on the N9 to increase performances
2718         and remove the huge useless gray bar part of the window frame
2719         at the bottom.
2720
2721         * MiniBrowser/qt/BrowserWindow.cpp:
2722         (BrowserWindow::BrowserWindow):
2723         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2724         (printHelp):
2725         (MiniBrowserApplication::handleUserOptions):
2726         * MiniBrowser/qt/MiniBrowserApplication.h:
2727         (WindowOptions::setStartFullScreen):
2728         (WindowOptions::startFullScreen):
2729
2730 2011-11-30  Martin Robinson  <mrobinson@igalia.com>
2731
2732         [GTK] Add an initial jhbuild setup which installs fonts into the WebKitBuild
2733         https://bugs.webkit.org/show_bug.cgi?id=73425
2734
2735         Reviewed by Gustavo Noronha Silva.
2736
2737         Add initial jhbuild support to the GTK+ port. update-webkitgtk-libs will
2738         ensure that the jhbuild root is up-to-date. Currently the only module is
2739         the font module.
2740
2741         * Scripts/update-webkitgtk-libs: Added.
2742         * gtk/common.py: Added a helper to get the number of CPUs.
2743         * gtk/jhbuild.modules: Added.
2744         * gtk/jhbuildrc: Added.
2745
2746 2011-11-30  Simon Hausmann  <simon.hausmann@nokia.com>
2747
2748         [Qt] Remove dependency of Qt DRT on JavaScriptCore.
2749
2750         Reviewed by Csaba Osztrogonác.
2751
2752         * DumpRenderTree/WorkQueue.cpp: Include the correct header file we're using anyway,
2753         instead of one that pulls in JavaScriptCore headers.
2754         * DumpRenderTree/qt/DumpRenderTree.pro: There's no need to pull in the JSC headers
2755         with load(javascriptcore).
2756
2757 2011-11-30  Sheriff Bot  <webkit.review.bot@gmail.com>
2758
2759         Unreviewed, rolling out r101440 and r101442.
2760         http://trac.webkit.org/changeset/101440
2761         http://trac.webkit.org/changeset/101442
2762         https://bugs.webkit.org/show_bug.cgi?id=73429
2763
2764         multiple crashes on layout tests (Requested by hayato on
2765         #webkit).
2766
2767         * DumpRenderTree/chromium/TestShell.cpp:
2768         (TestShell::resetWebSettings):
2769         * DumpRenderTree/chromium/TestShell.h:
2770         * DumpRenderTree/chromium/WebPreferences.cpp:
2771         (WebPreferences::reset):
2772         (WebPreferences::applyTo):
2773         * DumpRenderTree/chromium/WebPreferences.h:
2774         * DumpRenderTree/chromium/WebViewHost.cpp:
2775         (WebViewHost::WebViewHost):
2776         (WebViewHost::~WebViewHost):
2777
2778 2011-11-30  Fady Samuel  <fsamuel@chromium.org>
2779
2780         [Chromium] Fix broken DRT build for Aura Linux
2781         https://bugs.webkit.org/show_bug.cgi?id=72667
2782
2783         Reviewed by Tony Chang.
2784
2785         DRT now builds for Aura Linux. This, more or less, shares the same code as Android.
2786
2787         The following changes have been made:
2788
2789           1. Renamed TestShellAndroid to TestShellLinux.
2790           2. Stub code moved added in TestShellStub that will only compile on non-GTK Linux builds.
2791           3. Code common to Gtk, Aura, and Android moved to TestShellLinux.
2792
2793         * DumpRenderTree/DumpRenderTree.gypi:
2794         * DumpRenderTree/chromium/EventSender.cpp:
2795         (EventSender::keyDown):
2796         * DumpRenderTree/chromium/TestShellAndroid.cpp: Removed.
2797         * DumpRenderTree/chromium/TestShellGtk.cpp:
2798         * DumpRenderTree/chromium/TestShellLinux.cpp: Copied from Tools/DumpRenderTree/chromium/TestShellGtk.cpp.
2799         (AlarmHandler):
2800         (TestShell::waitTestFinished):
2801         (setupFontconfig):
2802         (platformInit):
2803         * DumpRenderTree/chromium/TestShellStub.cpp: Added.
2804         (checkLayoutTestSystemDependencies):
2805         (openStartupDialog):
2806
2807 2011-11-30  Fady Samuel  <fsamuel@chromium.org>
2808
2809         [Chromium] Set Result Before Early Exit for Fixed Layout Methods in LayoutTestController
2810         https://bugs.webkit.org/show_bug.cgi?id=73328
2811
2812         Reviewed by Kent Tamura.
2813
2814         * DumpRenderTree/chromium/LayoutTestController.cpp:
2815         (LayoutTestController::enableFixedLayoutMode):
2816         (LayoutTestController::setFixedLayoutSize):
2817
2818 2011-11-29  Hayato Ito  <hayato@chromium.org>
2819
2820         Skip writing the result of diff_image since non-chromium ports don't implement diff_image.
2821         https://bugs.webkit.org/show_bug.cgi?id=73381
2822
2823         Reviewed by Ryosuke Niwa.
2824
2825         This is a quick fix for the breakage of tests on non-chromium port.
2826         We should implement diff_image later on non-chromium ports.
2827
2828         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
2829         (write_test_result):
2830
2831 2011-11-29  Balazs Ankes  <Ankes.Balazs@stud.u-szeged.hu>
2832
2833         [NRWT] Fix --platform=qt-5.0 --new-baseline combo
2834         https://bugs.webkit.org/show_bug.cgi?id=72489
2835
2836         Reviewed by Eric Seidel.
2837
2838         Before the this fix always added "qt" option for the option list
2839         * Scripts/run-webkit-tests:
2840         * Scripts/webkitpy/layout_tests/port/qt.py:
2841
2842 2011-11-29  David Levin  <levin@chromium.org>
2843
2844         Add a way to revert a variable to its previous value after leaving a scope.
2845         https://bugs.webkit.org/show_bug.cgi?id=73371
2846
2847         Reviewed by Adam Barth.
2848
2849         * DumpRenderTree/ForwardingHeaders/wtf/TemporarilyChange.h: Added.
2850         * TestWebKitAPI/TestWebKitAPI.gypi: Added test file to the build.
2851         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Ditto.
2852         * TestWebKitAPI/win/TestWebKitAPI.vcproj: Ditto.
2853         * TestWebKitAPI/Tests/WTF/TemporarilyChange.cpp: Added.
2854         (TestWebKitAPI::TEST): Added a test for TemporarilyChange.
2855
2856 2011-11-28  Kentaro Hara  <haraken@chromium.org>
2857
2858         Implement [Supplemental] IDL and support it in run-bindings-tests
2859         https://bugs.webkit.org/show_bug.cgi?id=73162
2860
2861         Reviewed by Adam Barth.
2862
2863         run-bindings-tests supports the [Supplemental] IDL in the following way.
2864             (1) It creates a file |x| listing all IDL files.
2865             (2) It generates a supplemental dependency file |y| using resolve-supplemental.pl and |x|.
2866             (3) For each IDL file |z|, it runs generate-bindings.pl with |y|.
2867                 Then, generate-bindings.pl generates .h and .cpp files for the IDL file |z|,
2868                 including all attributes in IDL files that are supplementing the IDL file |z|.
2869
2870         * Scripts/run-bindings-tests:
2871         (generate_from_idl):
2872         (generate_supplemental_dependency):
2873         (detect_changes):
2874         (run_tests):
2875         (main):
2876
2877 2011-11-18  Nat Duca  <nduca@chromium.org>
2878
2879         [chromium] Enable threaded compositing via CCThreadProxy::hasThread only
2880         https://bugs.webkit.org/show_bug.cgi?id=70838
2881
2882         Reviewed by James Robinson.
2883
2884         * DumpRenderTree/chromium/TestShell.cpp:
2885         (TestShell::resetWebSettings):
2886         * DumpRenderTree/chromium/TestShell.h:
2887         (TestShell::threadedCompositingEnabled):
2888         * DumpRenderTree/chromium/WebPreferences.cpp:
2889         (WebPreferences::reset):
2890         (WebPreferences::applyTo):
2891         * DumpRenderTree/chromium/WebPreferences.h:
2892         * DumpRenderTree/chromium/WebViewHost.cpp:
2893         (WebViewHost::WebViewHost):
2894         (WebViewHost::~WebViewHost):
2895
2896 2011-11-29  Ojan Vafai  <ojan@chromium.org>
2897
2898         List of builders in the flakiness dashboard is out of date
2899         https://bugs.webkit.org/show_bug.cgi?id=73347
2900
2901         Reviewed by Adam Barth.
2902
2903         The chromium gpu bots no longer run layout tests, so kill that
2904         group entirely.
2905
2906         * TestResultServer/static-dashboards/builders.js:
2907         * TestResultServer/static-dashboards/dashboard_base.js:
2908
2909 2011-11-29  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2910
2911         [Qt] Don't hard-code the list of WebKit2 generated sources
2912
2913         The generated sources are... wait for it... generated. So
2914         use the generator itself to figure out which sources we need
2915         to compile.
2916
2917         Reviewed by Simon Hausmann.
2918
2919         * qmake/mkspecs/features/default_post.prf:
2920
2921 2011-11-29  Martin Robinson  <mrobinson@igalia.com>
2922
2923         [GTK] Add a method to detect 'make dist' errors without running 'make dist'
2924         https://bugs.webkit.org/show_bug.cgi?id=73216
2925
2926         Reviewed by Philippe Normand.
2927
2928         Add a script that tries to sniff out 'make dist' problems without running
2929         'make dist.' 'make distcheck' takes a very long time to run and this should
2930         reduce the amount of times it needs to be run consecutively.
2931
2932         * gtk/common.py:
2933         (get_build_path.is_valid_build_directory): Guess the source directory
2934         by the existence of the GNUmakefile instead of the .libs directory. This
2935         allows one to run the script after running autogen.sh but before fully
2936         building.
2937         * gtk/find-make-dist-errors: Added.
2938
2939 2011-11-29  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2940
2941         [Qt] Remove use of internal headers in the MiniBrowser
2942
2943         Reviewed by Simon Hausmann.
2944
2945         * MiniBrowser/qt/MiniBrowser.pro:
2946
2947 2011-11-29  Simon Hausmann  <simon.hausmann@nokia.com>
2948
2949         [Qt] Fix debug-shlib build without webkit2
2950
2951         Reviewed by Tor Arne Vestbø.
2952
2953         * qmake/mkspecs/features/qtwebkit.prf: Respect no_webkit2.
2954
2955 2011-11-28  Csaba Osztrogonác  <ossy@webkit.org>
2956
2957         [Qt][WK2] Unreviewed buildfix after r101307.
2958
2959         * qmake/mkspecs/features/webkit2.prf: Add missing includepath.
2960
2961 2011-11-29  Roland Steiner  <rolandsteiner@chromium.org>
2962
2963         <style scoped>: add ENABLE(STYLE_SCOPED) flag to WebKit
2964         https://bugs.webkit.org/show_bug.cgi?id=72848
2965
2966         Add option to build-webkit.
2967
2968         Reviewed by Dimitri Glazkov.
2969
2970         * Scripts/build-webkit:
2971
2972 2011-11-28  Ojan Vafai  <ojan@chromium.org>
2973
2974         gtest normalization in the test results server is sometimes wrong
2975         https://bugs.webkit.org/show_bug.cgi?id=73262
2976
2977         Reviewed by Tony Chang.
2978
2979         If a gtest has it's modified changed (e.g. add FLAKY_) then it will have two
2980         entries in the incremental JSON, one of the entries will have the correct data
2981         from the run and the other will have dummy no-data values. Make sure to
2982         always pick the one with real data.
2983
2984         * TestResultServer/model/jsonresults.py:
2985         (JsonResults._remove_gtest_modifiers):
2986         * TestResultServer/model/jsonresults_unittest.py:
2987         (JsonResultsTest.test_remove_gtest_modifiers):
2988
2989 2011-11-28  Ojan Vafai  <ojan@chromium.org>
2990
2991         Some of the results.json files have results/times entries at the directory level
2992         https://bugs.webkit.org/show_bug.cgi?id=73261
2993
2994         Reviewed by Tony Chang.
2995
2996         This is just a bug that got introduced in a temporary push of the results server.
2997         This patch repairs the broken files. After all the bots have cycled, we can simplify
2998         this code to just assert that results/times are not at the directory level.
2999
3000         Also, when catching exceptions, log the full stacktrace.
3001
3002         * TestResultServer/model/jsonresults.py:
3003         (_is_directory):
3004         (JsonResults._load_json):
3005         (JsonResults._merge_tests):
3006         (JsonResults.merge):
3007         * TestResultServer/model/jsonresults_unittest.py:
3008         (JsonResultsTest.test_merge_directory_hierarchy_extra_results_and_times):
3009
3010 2011-11-28  Tony Chang  <tony@chromium.org>
3011
3012         ews bots should pass --force to update-webkit-chromium
3013         https://bugs.webkit.org/show_bug.cgi?id=73230
3014
3015         Reviewed by Adam Barth.
3016
3017         This makes it less likely for gclient changes to break the bots (e.g.,
3018         if a DEPS repository moves or is switched to a branch).
3019
3020         Also pass --force when using |build-webkit --update-chromium| since this
3021         appears to only be used by the bots.
3022
3023         Take 2: Add Options.non_interactive to update.py's options() method.
3024
3025         * Scripts/update-webkit:
3026         * Scripts/webkitdirs.pm:
3027         (determineIsChromium): Add --force-update to update-webkit --chromium.
3028         (forceChromiumUpdate):
3029         (buildChromium): Pass --force to update-webkit-chromium.
3030         * Scripts/webkitpy/tool/steps/update.py:
3031         (Update.run): Add --force-update if non-interactive (i.e., bots).
3032         * Scripts/webkitpy/tool/steps/update_unittest.py:
3033
3034 2011-11-28  Tony Chang  <tony@chromium.org>
3035
3036         Revert r101279, broke the ews and cq bots.
3037
3038         * Scripts/update-webkit:
3039         * Scripts/webkitdirs.pm:
3040         (determineIsChromium):
3041         (buildChromium):
3042         * Scripts/webkitpy/common/config/ports.py:
3043         (WebKitPort.update_webkit_command):
3044         (ChromiumPort.update_webkit_command):
3045         * Scripts/webkitpy/common/config/ports_mock.py:
3046         (MockPort.update_webkit_command):
3047         * Scripts/webkitpy/tool/steps/update.py:
3048         * Scripts/webkitpy/tool/steps/update_unittest.py: Removed.
3049
3050 2011-11-28  Ojan Vafai  <ojan@chromium.org>
3051
3052         testlistjson on the test results server doesn't understand hierarchical results format
3053         https://bugs.webkit.org/show_bug.cgi?id=73246
3054
3055         Reviewed by Tony Chang.
3056
3057         * TestResultServer/model/jsonresults.py:
3058         (JsonResults._delete_results_and_times):
3059         Instead of just getting the top-level keys, we now walk the tests tree and
3060         delete the results and times values.
3061
3062         * TestResultServer/model/jsonresults_unittest.py:
3063         Simplify the test harness to take in the same JSON format for the tests
3064         that the actual results.json uses instead of something very similar but
3065         needlessly different. This also allows for testing the broken case of
3066         results and times values being at non-leaf level.
3067
3068 2011-11-28  Tony Chang  <tony@chromium.org>
3069
3070         ews bots should pass --force to update-webkit-chromium
3071         https://bugs.webkit.org/show_bug.cgi?id=73230
3072
3073         Reviewed by Adam Barth.
3074
3075         This makes it less likely for gclient changes to break the bots (e.g.,
3076         if a DEPS repository moves or is switched to a branch).
3077
3078         Also pass --force when using |build-webkit --update-chromium| since this
3079         appears to only be used by the bots.
3080
3081         * Scripts/update-webkit:
3082         * Scripts/webkitdirs.pm:
3083         (determineIsChromium): Add --force-update to update-webkit --chromium.
3084         (forceChromiumUpdate):
3085         (buildChromium): Pass --force to update-webkit-chromium.
3086         * Scripts/webkitpy/tool/steps/update.py:
3087         (Update.run): Add --force-update if non-interactive (i.e., bots).
3088         * Scripts/webkitpy/tool/steps/update_unittest.py:
3089
3090 2011-11-28  Michael Saboff  <msaboff@apple.com>
3091
3092         Fixed help message for --exclude-kraken to say that
3093         it excludes Kraken and not SunSpider.
3094
3095         Rubber-stamped by Filip Pizlo.
3096
3097         * Scripts/bencher:
3098
3099 2011-11-28  Sheriff Bot  <webkit.review.bot@gmail.com>
3100
3101         Unreviewed, rolling out r101273.
3102         http://trac.webkit.org/changeset/101273
3103         https://bugs.webkit.org/show_bug.cgi?id=73244
3104
3105         multiple test failures and timeouts (Requested by tony^work on
3106         #webkit).
3107
3108         * DumpRenderTree/DumpRenderTree.gypi:
3109         * DumpRenderTree/chromium/EventSender.cpp:
3110         (EventSender::keyDown):
3111         * DumpRenderTree/chromium/TestShellAndroid.cpp: Copied from Tools/DumpRenderTree/chromium/TestShellGtk.cpp.
3112         (AlarmHandler):
3113         (TestShell::waitTestFinished):
3114         (platformInit):
3115         (openStartupDialog):
3116         (checkLayoutTestSystemDependencies):
3117         * DumpRenderTree/chromium/TestShellGtk.cpp:
3118         (AlarmHandler):
3119         (setupFontconfig):
3120         (TestShell::waitTestFinished):
3121         (platformInit):
3122         * DumpRenderTree/chromium/TestShellLinux.cpp: Removed.
3123         * DumpRenderTree/chromium/TestShellStub.cpp: Removed.
3124
3125 2011-11-28  John Yani  <vanuan@gmail.com>
3126
3127         new-run-webkit-tests is locale dependent
3128         https://bugs.webkit.org/show_bug.cgi?id=68691
3129
3130         Reviewed and modified by Eric Seidel.
3131
3132         This is a simpler version of John's patch which I'm landing.
3133         This just engages the hacks from the Host constructor on a global
3134         basis.
3135
3136         * Scripts/webkitpy/common/host.py:
3137         (Host.__init__):
3138         (Host._engage_awesome_locale_hacks):
3139
3140 2011-11-28  Fady Samuel  <fsamuel@chromium.org>
3141
3142         [Chromium] Fix broken DRT build for Aura Linux
3143         https://bugs.webkit.org/show_bug.cgi?id=72667
3144
3145         Reviewed by Tony Chang.
3146
3147         DRT now builds for Aura Linux. This, more or less, shares the same code as Android.
3148
3149         The following changes have been made:
3150
3151           1. Renamed TestShellAndroid to TestShellLinux.
3152           2. Stub code moved added in TestShellStub that will only compile on non-GTK Linux builds.
3153           3. Code common to Gtk, Aura, and Android moved to TestShellLinux.
3154
3155         * DumpRenderTree/DumpRenderTree.gypi:
3156         * DumpRenderTree/chromium/EventSender.cpp:
3157         (EventSender::keyDown):
3158         * DumpRenderTree/chromium/TestShellAndroid.cpp: Removed.
3159         * DumpRenderTree/chromium/TestShellGtk.cpp:
3160         * DumpRenderTree/chromium/TestShellLinux.cpp: Copied from Tools/DumpRenderTree/chromium/TestShellGtk.cpp.
3161         (AlarmHandler):
3162         (TestShell::waitTestFinished):
3163         (setupFontconfig):
3164         (platformInit):
3165         * DumpRenderTree/chromium/TestShellStub.cpp: Added.
3166         (checkLayoutTestSystemDependencies):
3167         (openStartupDialog):
3168
3169 2011-11-28  Eric Seidel  <eric@webkit.org>
3170
3171         Add Environment object to Host and fix the GCC smartquotes trouble seen on the commit-queue
3172         https://bugs.webkit.org/show_bug.cgi?id=71983
3173
3174         Reviewed by Adam Barth.
3175
3176         We'll add more code to Environment overtime,
3177         allowing us to mock out more of our direct interactions with os.environ.
3178
3179         This patch also makes run_command print the passed in environment.
3180
3181         * Scripts/webkitpy/common/host_mock.py:
3182         * Scripts/webkitpy/common/system/environment.py: Copied from Tools/Scripts/webkitpy/tool/steps/build.py.
3183         * Scripts/webkitpy/common/system/environment_mock.py: Copied from Tools/Scripts/webkitpy/tool/steps/build.py.
3184         * Scripts/webkitpy/common/system/environment_unittest.py: Copied from Tools/Scripts/webkitpy/tool/steps/build.py.
3185         * Scripts/webkitpy/common/system/executive.py:
3186         * Scripts/webkitpy/common/system/executive_mock.py:
3187         * Scripts/webkitpy/layout_tests/port/webkit.py:
3188         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3189         * Scripts/webkitpy/tool/commands/download_unittest.py:
3190         * Scripts/webkitpy/tool/steps/build.py:
3191
3192 2011-11-28  Simon Hausmann  <simon.hausmann@nokia.com>
3193
3194         [Qt] Build system fixes against V8.
3195
3196         Reviewed by Tor Arne Vestbø.
3197
3198         * qmake/mkspecs/features/webcore.prf: Get rid of old v8 cruft.
3199         * qmake/mkspecs/features/wtf.prf: If requested, pull in v8 from Qt and configure
3200         WTF accordingly to use v8. (WTF in the implementation needs it as well as users of WTF)
3201
3202 2011-11-28  Kenneth Rohde Christiansen  <kenneth@webkit.org>
3203
3204         Make sure the useFixedLayout feature is consistently handled
3205         https://bugs.webkit.org/show_bug.cgi?id=73212
3206
3207         Reviewed by Simon Hausmann.
3208
3209         The web page parameters are set to our Qt defaults and are now altered
3210         in the WebKitTestRunner.
3211
3212         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
3213         (WTR::WrapperWindow::WrapperWindow):
3214
3215 2011-11-28  Simon Hausmann  <simon.hausmann@nokia.com>
3216
3217         [Qt] WTF should be built as separate static library
3218         https://bugs.webkit.org/show_bug.cgi?id=73201
3219
3220         Reviewed by Tor Arne Vestbø.
3221
3222         * DumpRenderTree/qt/DumpRenderTree.pro: Require wtf.
3223         * MiniBrowser/qt/BrowserWindow.cpp: Remove unnecessary config.h inclusion and
3224         thus wtf dependency.
3225         * WebKitTestRunner/InjectedBundle/Target.pri: Require wtf.
3226         * WebKitTestRunner/Target.pri: Ditto.
3227         * qmake/mkspecs/features/default_post.prf: When re-ordering the static libs
3228         consider wtf, too.
3229         * qmake/mkspecs/features/javascriptcore.prf: Remove WTF specific build stuff,
3230         moved into wtf.prf.
3231         * qmake/mkspecs/features/qtwebkit.prf: For debug builds with shared libs, also
3232         link in libwtf.
3233         * qmake/mkspecs/features/wtf.prf: Added. Stuff needed for building and using WTF.
3234
3235 2011-11-28  Kentaro Hara  <haraken@chromium.org>
3236
3237         run-bindings-tests shows diff error for a newly added binding test
3238         https://bugs.webkit.org/show_bug.cgi?id=73179
3239
3240         Reviewed by Adam Barth.
3241
3242         Currently, when we add a new binding test, say bindings/scripts/test/TestSupplemental.idl,
3243         run-bindings-tests just shows diff error because reference files
3244         (i.e. {V8,JS,...}TestSupplemental.h and {V8,JS,...}TestSupplemental.cpp) do not yet exist:
3245
3246             FAIL: (V8) V8TestSupplemental.h
3247             diff: WebCore/bindings/scripts/test/V8/V8TestSupplemental.h: No such file or directory
3248
3249         Instead, this patch treats a non-existent reference file as an empty file and
3250         shows the result of the newly added binding test, like this:
3251
3252             FAIL: (V8) V8TestSupplemental.h
3253             --- WebCore/bindings/scripts/test/V8/V8TestSupplemental.h       1970-01-01 09:00:00.000000000 +0900
3254             +++ /tmp/tmpmX2i2F/V8TestSupplemental.h 2011-11-28 11:08:34.150707677 +0900
3255             @@ -0,0 +1,84 @@
3256             +/*
3257             +    This file is part of the WebKit open source project.
3258             +    This file has been generated by generate-bindings.pl. DO NOT MODIFY!
3259             +
3260             (omitted)
3261
3262         * Scripts/run-bindings-tests:
3263         (detect_changes): Just added an -N option to the diff command.
3264
3265 2011-11-28  Martin Robinson  <mrobinson@igalia.com>
3266
3267         [Gtk][gtkdoc] Syntax error in Tools/gtk/generate-gtkdoc
3268         https://bugs.webkit.org/show_bug.cgi?id=73165
3269
3270         Reviewed by Philippe Normand.
3271
3272         Fix some syntax errors in this script and remove an unused variable.
3273
3274         * gtk/generate-gtkdoc:
3275         (get_webkit2_options.src_path): 
3276         (get_webkit2_options):
3277         (get_webkit1_options.src_path):
3278
3279 2011-11-26  Kevin Ollivier  <kevino@theolliviers.com>
3280
3281         [wx] Unreviewed build fix. Enable the inspector.
3282         
3283         * waf/build/settings.py:
3284
3285 2011-11-26  Martin Robinson  <mrobinson@igalia.com>
3286
3287         Try to fix the build after r101174.
3288
3289         * gtk/generate-gtkdoc: Properly handle empty PKG_CONFIG_PATHs.
3290
3291 2011-11-17  Martin Robinson  <mrobinson@igalia.com>
3292
3293         [GTK] Integrate build-gtkdoc into build-webkit and make
3294         https://bugs.webkit.org/show_bug.cgi?id=72626
3295
3296         Reviewed by Philippe Normand.
3297
3298         * GNUmakefile.am: Added support for new scripts and "make docs".
3299         * Scripts/webkitdirs.pm: Call generate-gtkdoc when building.
3300         (buildAutotoolsProject):
3301         * gtk/generate-gtkdoc: Added.
3302         * gtk/gtkdoc.py: Renamed from Source/WebKit2/UIProcess/API/gtk/docs/gtkdoc.py.
3303
3304 2011-11-25  Ádám Kallai  <Kallai.Adam@stud.u-szeged.hu>
3305
3306         [Qt] Add support for using OWRT with Qt5
3307         https://bugs.webkit.org/show_bug.cgi?id=72947
3308
3309         Add qt-5.0 platform and don't pass -graphicssystem to QtTestBrowser if using Qt5.
3310
3311         Reviewed by Csaba Osztrogonác.
3312
3313         * Scripts/old-run-webkit-tests:
3314
3315 2011-11-25  Yury Semikhatsky  <yurys@chromium.org>
3316
3317         [Chromium] Web Inspector: get rid of WebDevToolsFrontendClient::sendFrontendLoaded method
3318         https://bugs.webkit.org/show_bug.cgi?id=73126
3319
3320         No need to have a separate client call for this. The front-end can store messages
3321         which arrive before front-end loaded and dispatch them later. This happens in layout
3322         tests only by the way as all other messages should be initiated by the front-end code.
3323
3324         Reviewed by Pavel Feldman.
3325
3326         * DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
3327         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
3328         * DumpRenderTree/chromium/DRTDevToolsClient.cpp:
3329         * DumpRenderTree/chromium/DRTDevToolsClient.h:
3330
3331 2011-11-24  Kentaro Hara  <haraken@chromium.org>
3332
3333         Replace subprocess.call() with subprocess.Popen()
3334         https://bugs.webkit.org/show_bug.cgi?id=73105
3335
3336         Reviewed by Adam Barth.
3337
3338         Currently, the stdout and stderr of the child processes of run-bindings-tests
3339         are mixed (or even gone away) with the stdout and stderr of run-bindings-tests itself.
3340         This is the problem of subprocess.call(), which does not synchronize the stdout and stderr
3341         between a parent process and a child process. This patch replaces subprocess.call() with
3342         subprocess.Popen() and synchronizes the stdout and stderr between them.
3343         This patch also makes output messages of run-bindings-tests more readable.
3344
3345         With this patch, the output messages of run-bindings-tests look like as follows:
3346
3347             PASS: (CPP) WebDOMTestObj.h
3348             PASS: (CPP) WebDOMTestObj.cpp
3349             FAIL: (CPP) WebDOMTestInterface.h
3350             --- Source/WebCore/bindings/scripts/test/CPP/WebDOMTestInterface.h      2011-11-25 13:34:09.313516268 +0900
3351             +++ /tmp/tmpDAV87G/WebDOMTestInterface.h        2011-11-25 13:44:42.712946812 +0900
3352             @@ -23,7 +23,7 @@
3353              #ifndef WebDOMTestInterface_h
3354              #define WebDOMTestInterface_h
3355
3356             -#if ENABLE(Condition1) || ENABLE(Condition2)
3357             +#if ENABLE(Condition1) || ENABLE(Condition3)
3358
3359              #include <WebDOMObject.h>
3360              #include <WebDOMString.h>
3361              @@ -53,5 +53,5 @@
3362              WebDOMTestInterface toWebKit(WebCore::TestInterface*);
3363
3364              #endif
3365             -#endif // ENABLE(Condition1) || ENABLE(Condition2)
3366             +#endif // ENABLE(Condition1) || ENABLE(Condition3)
3367             PASS: (CPP) WebDOMFloat64Array.cpp
3368             PASS: (CPP) WebDOMFloat64Array.h
3369
3370             Some tests FAIL! (To update the reference files, execute "run-bindings-tests --reset-results")
3371
3372         * Scripts/run-bindings-tests:
3373         (generate_from_idl):
3374         (detect_changes):
3375         (run_tests):
3376         (main):
3377
3378 2011-11-24  Kentaro Hara  <haraken@chromium.org>
3379
3380         Fix the current working directory of run-bindings-tests
3381         https://bugs.webkit.org/show_bug.cgi?id=73106
3382
3383         Reviewed by Adam Barth.
3384
3385         Currently, run-bindings-tests outputs the following error
3386         (although it outputs "all tests passed!" in the end):
3387
3388             Testing the ObjC generator on TestObj.idl
3389             gcc: WebCore/bindings/objc/PublicDOMInterfaces.h: No such file or directory
3390             gcc: warning: '-x objective-c' after last input file has no effect
3391             gcc: no input files
3392
3393         This is because CodeGeneratorObjC.pm assumes that the current working directory
3394         is WebKit/Source/, but run-bindings-tests sets the current working directory to WebKit/.
3395         Thus, this patch changes it to WebKit/Source/.
3396
3397         * Scripts/run-bindings-tests:
3398         (generate_from_idl):
3399         (main):
3400
3401 2011-11-24  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3402
3403         [Qt] Get rid of the buildDirForSource() function in the build system
3404
3405         At some point the plan was to allow for running qmake on sub-trees
3406         of the sources, into the top level build directory, but this is no
3407         longer possible, so no reason to keep the convenience function around.
3408
3409         Reviewed by Simon Hausmann.
3410
3411         * WebKitTestRunner/DerivedSources.pri:
3412         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
3413         * qmake/mkspecs/features/default_post.prf:
3414         * qmake/mkspecs/features/default_pre.prf:
3415         * qmake/mkspecs/features/functions.prf:
3416         * qmake/mkspecs/features/javascriptcore.prf:
3417         * qmake/mkspecs/features/webcore.prf:
3418         * qmake/mkspecs/features/webkit2.prf:
3419
3420 2011-11-23  Sheriff Bot  <webkit.review.bot@gmail.com>
3421
3422         Unreviewed, rolling out r101107.
3423         http://trac.webkit.org/changeset/101107
3424         https://bugs.webkit.org/show_bug.cgi?id=73062
3425
3426         Breaks the commit-queue (Requested by abarth on #webkit).
3427
3428         * Scripts/webkitpy/common/host.py:
3429         (Host._engage_awesome_windows_hacks):
3430         * Scripts/webkitpy/common/host_mock.py:
3431         (MockHost.__init__):
3432         * Scripts/webkitpy/common/system/environment.py: Removed.
3433         * Scripts/webkitpy/common/system/environment_unittest.py: Removed.
3434         * Scripts/webkitpy/common/system/executive.py:
3435         (Executive._run_command_with_teed_output):
3436         (Executive.run_and_throw_if_fail):
3437         * Scripts/webkitpy/common/system/executive_mock.py:
3438         (MockExecutive.run_and_throw_if_fail):
3439         (MockExecutive.run_command):
3440         * Scripts/webkitpy/layout_tests/port/webkit.py:
3441         (WebKitPort._driver_build_script_name):
3442         (WebKitPort._run_script):
3443         (WebKitPort._build_driver):
3444         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3445         (test_build_driver):
3446         * Scripts/webkitpy/tool/commands/download_unittest.py:
3447         * Scripts/webkitpy/tool/steps/build.py:
3448         (Build.build):
3449
3450 2011-11-23  Ojan Vafai  <ojan@chromium.org>
3451
3452         Unreviewed, rolling out r101110.
3453         http://trac.webkit.org/changeset/101110
3454         https://bugs.webkit.org/show_bug.cgi?id=73057
3455
3456         The bots have now cycled and this patch is no longer needed.
3457
3458         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3459         (Manager._upload_json_files):
3460
3461 2011-11-23  Ojan Vafai  <ojan@chromium.org>
3462
3463         Temporarily save the incremental results json in order to identify bug
3464         https://bugs.webkit.org/show_bug.cgi?id=73057
3465
3466         Reviewed by Dean Jackson.
3467
3468         The test results server needed to be rolled back from tip of tree
3469         due to a JSON merging bug. It's hard to identify what the bug is without
3470         the raw data.
3471
3472         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3473         (Manager._upload_json_files):
3474
3475 2011-11-23  Ojan Vafai  <ojan@chromium.org>
3476
3477         Merging two version 4 results hits an error in the test results server
3478         https://bugs.webkit.org/show_bug.cgi?id=73042
3479
3480         Reviewed by Adam Barth.
3481
3482         The old code assumed that if the aggregate results value was not a leaf
3483         value that the incremental results would exist, which is just totally
3484         wrong. Added a test for that case as well.
3485
3486         * TestResultServer/model/jsonresults.py:
3487         (JsonResults._merge_tests):
3488         * TestResultServer/model/jsonresults_unittest.py:
3489         (JsonResultsTest.test_merge_build_directory_hierarchy_old_version):
3490         (JsonResultsTest.test_merge_build_directory_hierarchy):
3491
3492 2011-11-23  Eric Seidel  <eric@webkit.org>
3493
3494         Add Environment object to Host and fix the GCC smartquotes trouble seen on the commit-queue
3495         https://bugs.webkit.org/show_bug.cgi?id=71983
3496
3497         Reviewed by Adam Barth.
3498
3499         We'll add more code to Environment overtime,
3500         allowing us to mock out more of our direct interactions with os.environ.
3501
3502         This patch also makes run_command print the passed in environment.
3503
3504         * Scripts/webkitpy/common/host_mock.py:
3505         * Scripts/webkitpy/common/system/environment.py: Copied from Tools/Scripts/webkitpy/tool/steps/build.py.
3506         * Scripts/webkitpy/common/system/environment_mock.py: Copied from Tools/Scripts/webkitpy/tool/steps/build.py.
3507         * Scripts/webkitpy/common/system/environment_unittest.py: Copied from Tools/Scripts/webkitpy/tool/steps/build.py.
3508         * Scripts/webkitpy/common/system/executive.py:
3509         * Scripts/webkitpy/common/system/executive_mock.py:
3510         * Scripts/webkitpy/layout_tests/port/webkit.py:
3511         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3512         * Scripts/webkitpy/tool/commands/download_unittest.py:
3513         * Scripts/webkitpy/tool/steps/build.py:
3514
3515 2011-11-23  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3516
3517         [Qt] Re-generate QtWebKit API forwarding headers when API changes
3518
3519         The forwarding headers are generated by syncqt, and syncqt is run by qmake,
3520         so we need sync.profile to live in the same directory as the project file
3521         that will be touched when adding/removing API (api.pri in our case).
3522
3523         Since the WebKit2 APIs live in a separate static library, we have to
3524         add the project file for WebKit2 as a dependency of the api.pri file
3525         as well, so that any changes to the WebKit2 API will still trigger a
3526         run of qmake (and then syncqt).
3527
3528         Lastly, we use the new QMAKE_SYNCQT_OUTDIR variable in Qt 5 to ensure
3529         that the forwarding headers are still generated in the root build dir,
3530         not in the Source dir along with the internal forwarding headers.
3531
3532         Reviewed by Andreas Kling.
3533
3534         * qmake/mkspecs/features/default_pre.prf:
3535         * qmake/mkspecs/features/functions.prf:
3536
3537 2011-11-23  Johnny Ding  <jnd@chromium.org>
3538
3539         Set right default value to baseURL in LayoutTestController::queueLoadHTMLString.
3540         https://bugs.webkit.org/show_bug.cgi?id=72950
3541
3542         Reviewed by Tony Chang.
3543
3544         * DumpRenderTree/chromium/LayoutTestController.cpp:
3545         (LayoutTestController::queueLoadHTMLString):
3546
3547 2011-11-23  Adam Barth  <abarth@webkit.org>
3548
3549         garden-o-matic should display a message if it can't find results to display
3550         https://bugs.webkit.org/show_bug.cgi?id=72991
3551
3552         Reviewed by Dimitri Glazkov.
3553
3554         Not all failures (e.g., TIMEOUT) have results that we can display.
3555         This patch adds a short message so that the user can see that we're
3556         done trying to load the results.
3557
3558         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
3559         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
3560         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
3561
3562 2011-11-23  Adam Barth  <abarth@webkit.org>
3563
3564         garden-o-matic should explain more clearly what the builder lists at the top of the page mean
3565         https://bugs.webkit.org/show_bug.cgi?id=72992
3566
3567         Reviewed by Dimitri Glazkov.
3568
3569         This patch also prepares us to list other sorts of failures, like
3570         performance and unit test.
3571
3572         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
3573         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
3574         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
3575         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
3576
3577 2011-11-23  Ryan Sleevi  <rsleevi@chromium.org>
3578
3579         Add new Chromium tests to the flakiness dashboard
3580         https://bugs.webkit.org/show_bug.cgi?id=72997
3581
3582         Reviewed by Ojan Vafai.
3583
3584         * TestResultServer/static-dashboards/dashboard_base.js:
3585         ():
3586
3587 2011-11-23  Adam Barth  <abarth@webkit.org>
3588
3589         garden-o-matic should support AUDIO failures
3590         https://bugs.webkit.org/show_bug.cgi?id=69477
3591
3592         Reviewed by Dimitri Glazkov.
3593
3594         This is a first iteration at supporting AUDIO failures in
3595         garden-o-matic.  We'll probably need to iterate a bit when we have a
3596         real audio failure in the tree to play with.  For example, we'll
3597         probably want to add and <audio> element to the "Examine" view to hear
3598         the differences, but that all can be done in a subsequent patch.
3599
3600         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
3601
3602 2011-11-23  Scott Graham  <scottmg@chromium.org>
3603
3604         Adding gamepad support
3605         https://bugs.webkit.org/show_bug.cgi?id=69451
3606
3607         Reviewed by Darin Fisher.
3608
3609         Add a 'gamepadController' to Chromium DRT. Allows for setting of mock
3610         gamepad data that will be retrieved when JS code pulls via API access.
3611
3612         * DumpRenderTree/DumpRenderTree.gypi:
3613         * DumpRenderTree/chromium/GamepadController.cpp: Added.
3614         (GamepadController::GamepadController):
3615         (GamepadController::bindToJavascript):
3616         (GamepadController::reset):
3617         (GamepadController::connect):
3618         (GamepadController::disconnect):
3619         (GamepadController::setId):
3620         (GamepadController::setButtonCount):
3621         (GamepadController::setButtonData):
3622         (GamepadController::setAxisCount):
3623         (GamepadController::setAxisData):
3624         (GamepadController::fallbackCallback):
3625         * DumpRenderTree/chromium/GamepadController.h: Added.
3626         * DumpRenderTree/chromium/TestShell.cpp:
3627         (TestShell::TestShell):
3628         (TestShell::resetTestController):
3629         (TestShell::bindJSObjectsToWindow):
3630         * DumpRenderTree/chromium/TestShell.h:
3631         (TestShell::gamepadController):
3632
3633 2011-11-23  Raphael Kubo da Costa  <kubo@profusion.mobi>
3634
3635         [CMake] Move the top-level logic to the top-level directory.
3636         https://bugs.webkit.org/show_bug.cgi?id=72685
3637
3638         Reviewed by Brent Fulgham.
3639
3640         * CMakeLists.txt: Added.
3641         * DumpRenderTree/efl/CMakeLists.txt: Keep building into Programs/.
3642         * EWebLauncher/CMakeLists.txt: Renamed from Tools/CMakeListsEfl.txt.
3643         * Scripts/webkitdirs.pm:
3644         (generateBuildSystemFromCMakeProject):
3645         * WinCELauncher/CMakeLists.txt: Renamed from Tools/CMakeListsWinCE.txt.
3646
3647 2011-11-17  Vincent Scheib  <scheib@chromium.org>
3648
3649         Pointer Lock: Initial Tests for navigator.webkitPonter
3650         https://bugs.webkit.org/show_bug.cgi?id=72659
3651
3652         Reviewed by Darin Fisher.
3653
3654         * DumpRenderTree/chromium/TestShell.cpp:
3655         (TestShell::TestShell):
3656             Enabling PointerLock in TestShell
3657
3658 2011-11-22  Ryan Sleevi  <rsleevi@chromium.org>
3659
3660         Update flakiness dashboard for new Chromium builders and tests
3661         https://bugs.webkit.org/show_bug.cgi?id=72931
3662
3663         Reviewed by Ojan Vafai.
3664
3665         * TestResultServer/static-dashboards/builders.js:
3666         * TestResultServer/static-dashboards/dashboard_base.js:
3667
3668 2011-11-22  Adam Roben  <aroben@apple.com>
3669
3670         Remove some debugging code I accidentally left in r101005
3671
3672         * Scripts/prepare-ChangeLog:
3673         (get_function_line_ranges_for_python):
3674
3675 2011-11-22  Adam Roben  <aroben@apple.com>
3676
3677         Teach prepare-ChangeLog how to find changed classes/methods/functions in Python files
3678
3679         Fixes <http://webkit.org/b/57008> prepare-ChangeLog doesn't find names of modified
3680         classes/methods in Python source files
3681
3682         Reviewed by Dan Bates.
3683
3684         * Scripts/prepare-ChangeLog:
3685         (get_function_line_ranges): Call get_function_line_ranges_for_python for files ending in .py
3686         or that use python as their interpreter. Modified the code that extracts the interpreter to
3687         ignore a leading "/usr/bin/env " in the interpreter line, as is common for our Python
3688         scripts.
3689         (get_function_line_ranges_for_python): Added. Does simple parsing of Python files to look
3690         for class/def lines and generate ranges based on them.
3691
3692 2011-11-21  Yuta Kitamura  <yutak@chromium.org>
3693
3694         [GTK] Enable WebSocket hybi tests
3695         https://bugs.webkit.org/show_bug.cgi?id=72870
3696
3697         Reviewed by Xan Lopez.
3698
3699         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3700         (resetDefaultsToConsistentValues):
3701         Reset the setting value to the default value (true).
3702         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
3703         (LayoutTestController::overridePreference):
3704         Change the setting value through a function in DumpRenderTreeSupportGtk.
3705
3706 2011-11-21  Ryosuke Niwa  <rniwa@webkit.org>
3707
3708         ChangeLogEntry should be able to parse entries with multiple authors
3709         https://bugs.webkit.org/show_bug.cgi?id=72690
3710
3711         Reviewed by Eric Seidel.
3712
3713         Support multi-author change log entries in ChangeLogEntry. Split author names using
3714         the same logic as splitting reviewer names. _authors now stores list of dictionaries
3715         with "name" and "email" as keys. Unlike reviewers, we can't always resolve authors to
3716         accounts because some contributors are never listed in committers.py. 
3717
3718         * Scripts/webkitpy/common/checkout/changelog.py:
3719         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
3720         * Scripts/webkitpy/common/config/committers.py: Fixed a bug that *_by_{name|email}
3721         throws an exception when the argument is None.
3722
3723 2011-11-21  Simon Hausmann  <simon.hausmann@nokia.com>
3724
3725         [Qt] Speed up debug builds.
3726         https://bugs.webkit.org/show_bug.cgi?id=72882
3727
3728         Reviewed by Tor Arne Vestbø.
3729
3730         Introduce a special type of build that replaces the static jsc/wc/wk2
3731         static libraries with shared libraries, that individually are much faster
3732         to link and don't require relinking into the big library when changing just
3733         one file in one module. It can be enabled with
3734         --qmakearg="CONFIG+=force_static_libs_as_shared"
3735
3736         * qmake/mkspecs/features/force_static_libs_as_shared.prf: Added. With this configuration,
3737         disable symbol visibility, static linkage and enable rpath.
3738         * qmake/mkspecs/features/functions.prf: Add a helper function to determine if
3739         we need to link the convenience libraries together or not and renamed linkStaticLibrary
3740         to just linkLibrary (because linkage is not necessarily static).
3741         * qmake/mkspecs/features/qtwebkit.prf: When using the trick, make sure to also
3742         link against jsc/wc/wk2 when linking app code against QtWebKit.
3743         * qmake/mkspecs/features/javascriptcore.prf: Use new +needToLinkLibrary helper
3744         function.
3745         * qmake/mkspecs/features/webcore.prf: Ditto.
3746         * qmake/mkspecs/features/webkit2.prf: Ditto.
3747
3748 2011-11-21  Adam Klein  <adamk@chromium.org>
3749
3750         Update flakiness dashboard linkification regexps for crbug.com/100000 and above
3751         https://bugs.webkit.org/show_bug.cgi?id=72898
3752
3753         Reviewed by Ojan Vafai.
3754
3755         Also removed a comment that was going to be more trouble to keep up to
3756         date than seemed worthwhile (the code is now the documentation).
3757
3758         * TestResultServer/static-dashboards/flakiness_dashboard.html:
3759
3760 2011-11-21  Dominic Mazzoni  <dmazzoni@google.com>
3761
3762         [Chromium] WebAccessibilityNotification should use AssertMatchingEnums.
3763         https://bugs.webkit.org/show_bug.cgi?id=72895
3764
3765         Reviewed by Chris Fleizach.
3766
3767         * DumpRenderTree/chromium/WebViewHost.cpp:
3768         (WebViewHost::postAccessibilityNotification):
3769
3770 2011-11-21  Philippe Normand  <pnormand@igalia.com>
3771
3772         Unreviewed, run-gtk-tests temporary workaround after r98500.
3773
3774         * Scripts/run-gtk-tests: Try to find the build directory like in
3775         build-gtkdoc, until webkit-build-directory --configuration gets fixed.
3776
3777 2011-11-21  Carlos Garcia Campos  <cgarcia@igalia.com>
3778
3779         [GTK] Add a way to skip unit tests in the bots
3780         https://bugs.webkit.org/show_bug.cgi?id=72879
3781
3782         Reviewed by Philippe Normand.
3783
3784         Rewrite the script in python to make it easier for everybody to
3785         change it and add support for skipping tests.
3786
3787         * Scripts/run-gtk-tests:
3788
3789 2011-11-21  Raphael Kubo da Costa  <kubo@profusion.mobi>
3790
3791         [EFL] Add more appcache functions to ewk_settings.
3792         https://bugs.webkit.org/show_bug.cgi?id=72143
3793
3794         Reviewed by Kenneth Rohde Christiansen.
3795
3796         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
3797         (LayoutTestController::clearAllApplicationCaches):
3798         (LayoutTestController::setAppCacheMaximumSize): Use the right
3799         function, this is not related to the memory cache, but rather to the
3800         appcache.
3801
3802 2011-11-18  Adam Roben  <aroben@apple.com>
3803
3804         Make NRWT consider all lines of output from parse-malloc-history, not just the first
3805
3806         Fixes <http://webkit.org/b/72652> REGRESSION (NRWT): NRWT always reports "1 unique leaks
3807         found" even when there are more than 1 unique leaks
3808
3809         Reviewed by Darin Adler.
3810
3811         * Scripts/webkitpy/layout_tests/port/leakdetector.py:
3812         (LeakDetector.count_total_bytes_and_unique_leaks): Pass re.MULTILINE to re.findall so that
3813         "^" will be considered to match the beginning of any line, not only the beginning of the
3814         string.
3815
3816         * Scripts/webkitpy/layout_tests/port/leakdetector_unittest.py:
3817         (LeakDetectorTest.test_count_total_bytes_and_unique_leaks): Added a second callstack to the
3818         fake output and updated expectations to match.
3819
3820 2011-11-21  Andras Becsi  <andras.becsi@nokia.com>
3821
3822         [Qt] Clean up MiniBrowser project file after recent changes
3823         https://bugs.webkit.org/show_bug.cgi?id=72869
3824
3825         Reviewed by Simon Hausmann.
3826
3827         Remove nonexistent files from OTHER_FILES after our touch and desktop webviews were merged.
3828
3829         * MiniBrowser/qt/MiniBrowser.pro:
3830
3831 2011-11-21  Dominic Mazzoni  <dmazzoni@google.com>
3832
3833         Accessibility: Multiselect list boxes need to report the active option in addition to which items are selected.
3834         https://bugs.webkit.org/show_bug.cgi?id=72479
3835
3836         Reviewed by Chris Fleizach.
3837
3838         * DumpRenderTree/AccessibilityUIElement.cpp:
3839         (getIsSelectedOptionActiveCallback):
3840         (AccessibilityUIElement::getJSClass):
3841         * DumpRenderTree/AccessibilityUIElement.h:
3842         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
3843         (AccessibilityUIElement::AccessibilityUIElement):
3844         (AccessibilityUIElement::isSelectedOptionActiveGetterCallback):
3845         * DumpRenderTree/chromium/AccessibilityUIElement.h:
3846         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3847         (AccessibilityUIElement::isSelectedOptionActive):
3848         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
3849         (AccessibilityUIElement::isSelectedOptionActive):
3850         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
3851         (AccessibilityUIElement::isSelectedOptionActive):
3852
3853 2011-11-21  Takashi Toyoshima  <toyoshim@chromium.org>
3854
3855         Use Host().port_factory instead of webkitpy.layout_tests.port.factory.
3856         https://bugs.webkit.org/show_bug.cgi?id=72859
3857
3858         Reviewed by Eric Seidel.
3859         
3860         * Scripts/new-run-webkit-websocketserver:
3861
3862 2011-11-20  Shinya Kawanaka  <shinyak@google.com>
3863
3864         [Chromium] Layout test does not return correct misspelling positions if there are multiple misspelled words.
3865         https://bugs.webkit.org/show_bug.cgi?id=72655
3866
3867         Reviewed by Hajime Morita.
3868
3869         Fixed the calculation of offset.
3870
3871         * DumpRenderTree/chromium/WebViewHost.cpp:
3872         (WebViewHost::finishLastTextCheck):
3873
3874 2011-11-20  Ojan Vafai  <ojan@chromium.org>
3875
3876         Change the final place where we use version 3 of the results json output
3877         https://bugs.webkit.org/show_bug.cgi?id=72838
3878
3879         Reviewed by Adam Barth.
3880
3881         This converts the json from being a flat map of test name --> results
3882         to being hierarchical by directory. This will make the json files 
3883         considerably smaller.
3884
3885         Also cleaned up some functions that were returning/checking boolean values
3886         that were always True.
3887
3888         * TestResultServer/model/jsonresults.py:
3889         * TestResultServer/model/jsonresults_unittest.py:
3890
3891 2011-11-20  Ojan Vafai  <ojan@chromium.org>
3892
3893         Include the master name when querying the test results server.
3894         This will soon be required since leaving the master name out mean
3895         that multiple files could match the query.
3896
3897         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
3898
3899 2011-11-19  Ojan Vafai  <ojan@chromium.org>
3900
3901         Fix jsonp callback code to add the callback if json content
3902         is not jsonp.
3903         * TestResultServer/handlers/testfilehandler.py:
3904
3905 2011-11-19  Alice Boxhall  <aboxhall@chromium.org>
3906
3907         Modify dashboard_base.js to optionally accept version 4 results json.
3908
3909         Reviewed by Ojan Vafai.
3910
3911         * TestResultServer/static-dashboards/dashboard_base.js:
3912         (ADD_RESULTS):
3913         (flattenTrie):
3914         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
3915         (testFlattenTrie):
3916
3917 2011-11-19  Ojan Vafai  <ojan@chromium.org>
3918
3919         Stop storing results files as jsonp in the test results server
3920         https://bugs.webkit.org/show_bug.cgi?id=72814
3921
3922         Reviewed by Adam Barth.
3923
3924         * TestResultServer/model/jsonresults.py:
3925         * TestResultServer/model/jsonresults_unittest.py:
3926
3927 2011-11-19  Ojan Vafai  <ojan@chromium.org>
3928
3929         Remove the dependence on jsonp from more of new-run-webkit-tests and the test results server
3930         https://bugs.webkit.org/show_bug.cgi?id=72813
3931
3932         Reviewed by Adam Barth.
3933
3934         Once this lands, we can start storing pure json in the test results server and then
3935         we can delete the code with all the FIXMEs added here.
3936
3937         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3938         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
3939         Only add jsonp for full_results.json.
3940         * TestResultServer/model/jsonresults.py:
3941         * TestResultServer/model/jsonresults_unittest.py:
3942         Accept pure json uploads.
3943         * TestResultServer/static-dashboards/dashboard_base.js:
3944         (appendJSONScriptElementFor):
3945         Use the callback parameter so that the server can start returning pure json if it's left out.
3946
3947 2011-11-19  Ojan Vafai  <ojan@chromium.org>
3948
3949         Allow json NRWT downloads to be pure json and not jsonp
3950         https://bugs.webkit.org/show_bug.cgi?id=72809
3951
3952         Reviewed by Adam Barth.
3953
3954         I'm moving the server to storing and serving up raw json instead of jsonp.
3955         You can still get the jsonp by passing a "callback" parameter, but there's no
3956         need for run-webkit-tests to get jsonp when all it wants is the raw json.
3957
3958         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
3959         Make the wrapper stripping only happen if the json is actually wrapped.
3960         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
3961
3962 2011-11-18  Daniel Bates  <dbates@rim.com>
3963
3964         Add CMake build infrastructure for the BlackBerry port
3965         https://bugs.webkit.org/show_bug.cgi?id=72768
3966
3967         Reviewed by Antonio Gomes.
3968
3969         Add optional --blackberry command line argument to build-webkit to build the
3970         BlackBerry port.
3971
3972         * Scripts/build-webkit:
3973         * Scripts/webkitdirs.pm:
3974         (determineBaseProductDir):
3975         (argumentsForConfiguration):
3976         (builtDylibPathForName):
3977         (determineIsQt):
3978         (isBlackBerry): Added.
3979         (determineIsBlackBerry): Added.
3980         (blackberryTargetArchitecture): Added.
3981         (isAppleWebKit):
3982
3983 2011-11-18  James Robinson  <jamesr@chromium.org>
3984
3985         Unreviewed, rolling out r100835.
3986         http://trac.webkit.org/changeset/100835
3987         https://bugs.webkit.org/show_bug.cgi?id=72778
3988
3989         Broke some page scale tests
3990
3991         * DumpRenderTree/chromium/EventSender.cpp:
3992         (EventSender::scalePageBy):
3993
3994 2011-11-18  Adam Barth  <abarth@webkit.org>
3995
3996         Prepare to move manual-tests out of WebCore
3997         https://bugs.webkit.org/show_bug.cgi?id=72782
3998
3999         Reviewed by Eric Seidel.
4000
4001         Update this file to refer to the new location of ManualTests.
4002
4003         * Scripts/webkitpy/common/config/build.py:
4004
4005 2011-11-18  James Robinson  <jamesr@chromium.org>
4006
4007         [chromium] Set min/max page scale factors when calling eventSender.scalePageBy to avoid clamping
4008         https://bugs.webkit.org/show_bug.cgi?id=72778
4009
4010         Reviewed by Kenneth Russell.
4011
4012         Set the min/max page scale clamps when a test calls eventSender.scalePageBy() so that the scale isn't clamped.
4013
4014         * DumpRenderTree/chromium/EventSender.cpp:
4015         (EventSender::scalePageBy):
4016
4017 2011-11-18  Scott Graham  <scottmg@chromium.org>
4018
4019         IDL changes for gamepad support
4020         https://bugs.webkit.org/show_bug.cgi?id=71753
4021
4022         Reviewed by Adam Barth.
4023
4024         Runtime enable gamepad in chromium test shell.
4025
4026         * DumpRenderTree/chromium/TestShell.cpp:
4027         (TestShell::TestShell):
4028
4029 2011-10-27  Darin Fisher  <darin@chromium.org>
4030
4031         [chromium] Improve fullscreen API
4032         https://bugs.webkit.org/show_bug.cgi?id=70477
4033
4034         Reviewed by Adam Barth.
4035
4036         Implement enter/exitFullScreen.  The implementation is asynchronous,
4037         via postDelayedTask, to simulate how the browser actually behaves.
4038
4039         * DumpRenderTree/chromium/WebViewHost.cpp:
4040         (WebViewHost::enterFullScreen):
4041         (WebViewHost::exitFullScreen):
4042         (WebViewHost::setAddressBarURL):
4043         (WebViewHost::enterFullScreenNow):
4044         (WebViewHost::exitFullScreenNow):
4045         * DumpRenderTree/chromium/WebViewHost.h:
4046
4047 2011-11-18  Adam Roben  <aroben@apple.com>
4048
4049         Ignore an ANGLE leak that is not WebKit's fault
4050
4051         * Scripts/old-run-webkit-tests:
4052         (countAndPrintLeaks):
4053         * Scripts/webkitpy/layout_tests/port/leakdetector.py:
4054         (LeakDetector._callstacks_to_exclude_from_leaks):
4055         Added ScanFromString to the call stacks to exclude on all OS versions.
4056
4057 2011-11-18  Adam Roben  <aroben@apple.com>
4058
4059         Ignore another Lion-specific leak that is not WebKit's fault
4060
4061         * Scripts/old-run-webkit-tests:
4062         (countAndPrintLeaks):
4063         * Scripts/webkitpy/layout_tests/port/leakdetector.py:
4064         (LeakDetector._callstacks_to_exclude_from_leaks):
4065         Added SecTransformExecute to the call stacks to exclude.
4066
4067 2011-11-18  Yuta Kitamura  <yutak@chromium.org>
4068
4069         [Qt] Enable WebSocket hybi tests
4070         https://bugs.webkit.org/show_bug.cgi?id=72687
4071
4072         Reviewed by Simon Hausmann.
4073
4074         Let LayoutTestControllerQt be able to switch WebSocket protocols via overridePreference().
4075
4076         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
4077         (WebCore::WebPage::resetSettings):
4078         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
4079         (LayoutTestController::overridePreference):
4080
4081 2011-11-18  Sergio Villar Senin  <svillar@igalia.com>
4082
4083         [GTK][WK2] Invalid TestNetscapePlugin path when using new-run-webkit-tests
4084         https://bugs.webkit.org/show_bug.cgi?id=72707
4085
4086         Reviewed by Philippe Normand.
4087
4088         Make new-run-webkit-tests use the right path to locate the test
4089         plugin used by WebKitTestRunner.
4090
4091         * Scripts/webkitpy/layout_tests/port/gtk.py:
4092
4093 2011-11-18  Simon Hausmann  <simon.hausmann@nokia.com>
4094
4095         [Qt] WTR and InjectedBundle should not link statically against JSC/WTF
4096         https://bugs.webkit.org/show_bug.cgi?id=72697
4097
4098         Reviewed by Kenneth Rohde Christiansen.
4099
4100         * Tools.pro: Make WTR linux only for the moment.
4101         * WebKitTestRunner/InjectedBundle/Target.pri: Remove static jsc/wtf linkage.
4102         * WebKitTestRunner/Target.pri: Ditto.
4103         * qmake/mkspecs/features/default_post.prf: Don't build the main webkit sources
4104         with ELF visibility on Linux. Instead we're temporarily using an LD version script.
4105
4106 2011-11-18  Kenneth Rohde Christiansen  <kenneth@webkit.org>
4107
4108         [Qt] Change default size of the MiniBrowser
4109
4110         Reviewed by Simon Hausmann.
4111
4112         As we lay out desktop pages using the width of 980, change our default
4113         size to 980x735 so that we are not scaling down by default.
4114
4115         * MiniBrowser/qt/MiniBrowserApplication.h:
4116         (WindowOptions::WindowOptions):
4117
4118 2011-11-17  Raphael Kubo da Costa  <kubo@profusion.mobi>
4119
4120         [EFL] Clean up the use of DATA_DIR in the buildsystem
4121         https://bugs.webkit.org/show_bug.cgi?id=72681
4122
4123         Reviewed by Daniel Bates.
4124
4125         Instead of adding -DDATA_DIR in all scopes, only define it in
4126         WebCore/PlatformEFL.cmake, as WebCore is the only place that uses
4127         it.
4128
4129         We can then define DATA_DIR to the generated theme directory for
4130         EWebLauncher, since it's what we want here.
4131
4132         * CMakeListsEfl.txt:
4133
4134 2011-11-17  Raphael Kubo da Costa  <kubo@profusion.mobi>
4135
4136         [EFL] Move platform includes last in DRT's include directories.
4137         https://bugs.webkit.org/show_bug.cgi?id=72684
4138
4139         Reviewed by Daniel Bates.
4140
4141         This is the first step in fixing the build when a platform include (in
4142         this case, Valgrind 3.7.0 ships its config.h) and the wrong config.h
4143         is included instead of WebKit's.
4144
4145         * DumpRenderTree/efl/CMakeLists.txt:
4146
4147 2011-11-17  Ryosuke Niwa  <rniwa@webkit.org>
4148
4149         fuzzy_match doesn't recognize "Dan B" or "hyatt"
4150         https://bugs.webkit.org/show_bug.cgi?id=72636
4151
4152         Reviewed by Eric Seidel.
4153
4154         Make contributors_by_fuzz_match search string in shorthand names of contributors such as "Ryosuke N"
4155         and user names of email addreses such as rniwa in rniwa@webkit.org.
4156
4157         Also import legacy contributor names and mispell lists from Eric's script on the bug 26533 as a test.
4158
4159         * Scripts/webkitpy/common/config/committers.py:
4160         * Scripts/webkitpy/common/config/committers_unittest.py:
4161
4162 2011-11-17  Eric Seidel  <eric@webkit.org>
4163
4164         Remove support for running check-webkit-style w/o a webkit checkout
4165         https://bugs.webkit.org/show_bug.cgi?id=72678
4166
4167         Reviewed by Adam Barth.
4168
4169         After talking to Dave Levin we've decided to remove this
4170         code.  I know of no users of check-webkit-style w/o a webkit checkout
4171         and this needlessly adds complexity.
4172
4173         * Scripts/webkitpy/style/main.py:
4174
4175 2011-11-17  Eric Seidel  <eric@webkit.org>
4176
4177         Remove more direct uses of os.path
4178         https://bugs.webkit.org/show_bug.cgi?id=72677
4179
4180         Reviewed by Adam Barth.
4181
4182         Remove more uses of os.path, including refactoring
4183         CommitterValidator to use a Host object instead of
4184         manual hacks to emulate what SCM and FileSystem provide.
4185
4186         * Scripts/webkitpy/common/checkout/changelog.py:
4187         * Scripts/webkitpy/common/config/committervalidator.py:
4188         * Scripts/webkitpy/common/config/committervalidator_unittest.py:
4189         * Scripts/webkitpy/common/system/filesystem_mock.py:
4190          - Turns out none of our unittests actually looked at this path.
4191            now I've made it be a valid path.
4192         * Scripts/webkitpy/tool/bot/feeders.py:
4193         * Scripts/webkitpy/tool/commands/queues.py:
4194
4195 2011-11-17  Eric Seidel  <eric@webkit.org>
4196
4197         Teach TextFileReader about FileSystem
4198         https://bugs.webkit.org/show_bug.cgi?id=72673
4199
4200         Reviewed by Adam Barth.
4201
4202         Unfortunately TextFileReader doesn't use FileSystem
4203         everywhere yet, so we can't move the unittests to
4204         using MockFileSystem, but we're close.
4205
4206         * Scripts/webkitpy/style/filereader.py:
4207         * Scripts/webkitpy/style/filereader_unittest.py:
4208         * Scripts/webkitpy/style/main.py:
4209
4210 2011-11-17  Eric Seidel  <eric@webkit.org>
4211
4212         Give check-webkit-style a Host
4213         https://bugs.webkit.org/show_bug.cgi?id=72670
4214
4215         Reviewed by Adam Barth.
4216
4217         * Scripts/webkitpy/common/checkout/scm/detection.py:
4218         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
4219          - We don't run these unittests, but this one was looking for the wrong exception.
4220         * Scripts/webkitpy/style/main.py:
4221         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
4222          - This was create a new scm object even though it already had one?
4223
4224 2011-11-17  Eric Seidel  <eric@webkit.org>
4225
4226         Move check-webkit-style change_directory code to more modern mocks
4227         https://bugs.webkit.org/show_bug.cgi?id=72664
4228
4229         Reviewed by Adam Barth.
4230
4231         * Scripts/check-webkit-style:
4232         * Scripts/webkitpy/style/main.py:
4233         * Scripts/webkitpy/style/main_unittest.py:
4234
4235 2011-11-17  Zan Dobersek  <zandobersek@gmail.com>
4236
4237         [GTK] enable ref tests
4238         https://bugs.webkit.org/show_bug.cgi?id=72601
4239
4240         Re-enable ref tests. Run the driver under evenly numbered display if performing
4241         pixel tests or under oddly numbered display otherwise. This prevents the ref tests
4242         being run in an already used Xvfb instance if pixel tests are disabled.
4243
4244         Reviewed by Tony Chang.
4245
4246         * Scripts/webkitpy/layout_tests/port/gtk.py: Also set Xvfb variable to None
4247         after the process is killed
4248
4249 2011-11-17  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
4250
4251         Make check-webkit-style accept xxx_p.h as a primary header for xxx.cpp for Qt's sake
4252         https://bugs.webkit.org/show_bug.cgi?id=72620
4253
4254         Reviewed by David Levin.
4255
4256         Qt's convention of keeping private (but exposed) API suffixed by "_p.h" but not
4257         adding the suffix the cpp file. One example of false positive
4258         https://bugs.webkit.org/show_bug.cgi?id=72319#c12.
4259
4260         * Scripts/webkitpy/style/checkers/cpp.py: accept headers ending with _p as primary if its
4261         basename contains the cpp's basename.
4262         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
4263
4264 2011-11-17  Eric Seidel  <eric@webkit.org>
4265
4266         Move check-webkit-style's guts into webkitpy/style/main.py
4267         https://bugs.webkit.org/show_bug.cgi?id=72657
4268
4269         Reviewed by Adam Barth.
4270
4271         * Scripts/check-webkit-style:
4272         * Scripts/webkitpy/style/main.py:
4273
4274 2011-11-17  Eric Seidel  <eric@webkit.org>
4275
4276         new-run-webkit-tests is locale dependent
4277         https://bugs.webkit.org/show_bug.cgi?id=68691
4278
4279         Reviewed by Adam Barth.
4280
4281         Make NRWT use a clean environment, just like ORWT did.
4282
4283         * Scripts/webkitpy/layout_tests/port/base.py:
4284         * Scripts/webkitpy/layout_tests/port/chromium.py:
4285         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
4286         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
4287         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
4288         * Scripts/webkitpy/layout_tests/port/efl.py:
4289         * Scripts/webkitpy/layout_tests/port/google_chrome_unittest.py:
4290         * Scripts/webkitpy/layout_tests/port/mac.py:
4291         * Scripts/webkitpy/layout_tests/port/qt.py:
4292