1dc6ef2f81c7493d4d23b8797352832facde86de
[WebKit-https.git] / Tools / ChangeLog
1 2012-09-12  Lucas Forschler  <lforschler@apple.com>
2
3         Remove unused bot, apple-xserve-1.
4         https://bugs.webkit.org/show_bug.cgi?id=96570
5         
6         Reviewed by Stephanie Lewis.
7
8         * BuildSlaveSupport/build.webkit.org-config/config.json:
9
10 2012-09-12  Xan Lopez  <xlopez@igalia.com>
11
12         [GTK] We attempt to rebase documentation even if it's not present
13         https://bugs.webkit.org/show_bug.cgi?id=96553
14
15         Reviewed by Martin Robinson.
16
17         Do not make the documentation rebase step fatal. This allows make
18         install to succeed when there's no documentation generated.
19
20         * gtk/generate-gtkdoc:
21
22 2012-09-12  Kenneth Rohde Christiansen  <kenneth@webkit.org>
23
24         Respect WEBKITOUTPUTDIR when running EFL tests
25         https://bugs.webkit.org/show_bug.cgi?id=96528
26
27         Reviewed by Dirk Pranke.
28
29         Expose user set WEBKITOUTPUTDIR to the web process.
30
31         * Scripts/webkitpy/layout_tests/port/driver.py:
32         (Driver._start): Add WEBKITOUTPUTDIR to the environment
33         of the web process and its potential jhbuild wrapper.
34
35 2012-09-12  Dirk Pranke  <dpranke@chromium.org>
36
37         webkitdirs: fix uname version handling for cygwin
38         https://bugs.webkit.org/show_bug.cgi?id=96436
39
40         Reviewed by Jon Honeycutt.
41
42         Newer versions of cygwin embed an additional version string
43         inside parentheses, so you get "1.7.16(0.249/5/3)" instead of "1.7.16".
44         Update the code to handle that.
45
46         * Scripts/webkitdirs.pm:
47         (setupAppleWinEnv):
48
49 2012-09-12  Christophe Dumez  <christophe.dumez@intel.com>
50
51         Regression(r128338): Broke Windows build
52         https://bugs.webkit.org/show_bug.cgi?id=96537
53
54         Unreviewed build fix.
55
56         Fix Apple-Win build by allocating array dynamically
57         since its size is not constant.
58
59         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
60         (WTR::stringArrayToJS):
61
62 2012-09-12  Alex Sakhartchouk  <alexst@chromium.org>
63
64         [chromium] Add a virtual test suite for fast/hidpi
65         https://bugs.webkit.org/show_bug.cgi?id=90192
66
67         Reviewed by Dirk Pranke.
68
69         Add a virtual test suite to make sure the pixel tests in fast/hidpi give the same result
70         on the hardware accelerated path as the software path.
71
72         * DumpRenderTree/chromium/WebPreferences.cpp:
73         (WebPreferences::applyTo):
74         * Scripts/webkitpy/layout_tests/port/chromium.py:
75         (ChromiumPort.virtual_test_suites):
76
77 2012-09-12  Adam Klein  <adamk@chromium.org>
78
79         [chromium] Add content_browsertests to the flakiness dashboard
80         https://bugs.webkit.org/show_bug.cgi?id=96535
81
82         Reviewed by Ojan Vafai.
83
84         * TestResultServer/static-dashboards/dashboard_base.js:
85
86 2012-09-12  Christophe Dumez  <christophe.dumez@intel.com>
87
88         [WK2][WKTR] TestRunner needs to implement originsWithApplicationCache
89         https://bugs.webkit.org/show_bug.cgi?id=96496
90
91         Reviewed by Kenneth Rohde Christiansen.
92
93         Add implementation for originsWithApplicationCache to
94         WebKitTestRunner.
95
96         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
97         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
98         (WTR::stringArrayToJS):
99         (WTR):
100         (WTR::TestRunner::originsWithApplicationCache):
101         * WebKitTestRunner/InjectedBundle/TestRunner.h:
102         (TestRunner):
103
104 2012-09-12  Simon Hausmann  <simon.hausmann@nokia.com>
105
106         Fix Qt/Windows build with Python3
107         https://bugs.webkit.org/show_bug.cgi?id=96473
108
109         Reviewed by Csaba Osztrogonác.
110
111         In Python 3 print is a real function, so we must use parentheses around
112         the function parameters. This is backwards compatible with Python 2.
113
114         * Scripts/generate-win32-export-forwards:
115
116 2012-09-12  Simon Hausmann  <simon.hausmann@nokia.com>
117
118         Unreviewed trivial build fix for Qt/Windows after bug #96358.
119
120         As pointed out by Kevin, we should inject the "set PATH=..." statement
121         to add the GnuWin32 directory only if that directory actually exists.
122
123         * qmake/mkspecs/features/default_post.prf:
124
125 2012-09-12  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
126
127         [EFL] [WK2] Memory leaks in TestControllerEfl
128         https://bugs.webkit.org/show_bug.cgi?id=96525
129
130         Reviewed by Kenneth Rohde Christiansen.
131
132         Fix memory leaks in EFL's TestRunner code by adopting an allocation
133         of WKString created with WKStringCreateWithUTF8CString().
134
135         * WebKitTestRunner/efl/TestControllerEfl.cpp:
136         (WTR::TestController::initializeInjectedBundlePath):
137         (WTR::TestController::initializeTestPluginDirectory):
138
139 2012-09-12  Christophe Dumez  <christophe.dumez@intel.com>
140
141         [WK2][WKTR] TestRunner needs to implement dumpApplicationCacheDelegateCallbacks
142         https://bugs.webkit.org/show_bug.cgi?id=96374
143
144         Reviewed by Kenneth Rohde Christiansen.
145
146         Implement support for dumpApplicationCacheDelegateCallbacks
147         and disallowIncreaseForApplicationCacheQuota in
148         WebKitTestRunner and properly dump the information
149         expected by the tests.
150
151         If the application cache quota is reached for a given
152         security origin, WebKitTestRunner will reset the quota
153         to its default value, unless intructed not to via
154         disallowIncreaseForApplicationCacheQuota().
155
156         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
157         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
158         (WTR::InjectedBundlePage::InjectedBundlePage):
159         (WTR::InjectedBundlePage::didReachApplicationCacheOriginQuota):
160         (WTR):
161         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
162         (InjectedBundlePage):
163         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
164         (WTR::TestRunner::TestRunner):
165         (WTR::TestRunner::disallowIncreaseForApplicationCacheQuota):
166         (WTR):
167         * WebKitTestRunner/InjectedBundle/TestRunner.h:
168         (WTR::TestRunner::dumpApplicationCacheDelegateCallbacks):
169         (TestRunner):
170         (WTR::TestRunner::shouldDisallowIncreaseForApplicationCacheQuota):
171         (WTR::TestRunner::shouldDumpApplicationCacheDelegateCallbacks):
172
173 2012-09-12  Jochen Eisinger  <jochen@chromium.org>
174
175         [chromium] remove deprecated and unused sets import from chromium_android driver
176         https://bugs.webkit.org/show_bug.cgi?id=96485
177
178         Reviewed by Dirk Pranke.
179
180         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
181
182 2012-09-12  Zan Dobersek  <zandobersek@gmail.com>
183
184         Flakiness dashboard doesn't recognize new Chromium Android test builder
185         https://bugs.webkit.org/show_bug.cgi?id=96523
186
187         Reviewed by Ojan Vafai.
188
189         Properly return 'ANDROID' as the Chromium platform for Android builders.
190
191         * TestResultServer/static-dashboards/flakiness_dashboard.js:
192         (chromiumPlatform):
193         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
194         (test):
195
196 2012-09-12  Simon Hausmann  <simon.hausmann@nokia.com>
197
198         [Qt] Build on Windows requires bison/flex in PATH
199         https://bugs.webkit.org/show_bug.cgi?id=96358
200
201         Reviewed by Tor Arne Vestbø.
202
203         The build requires flex, bison, etc. and they need to be in the PATH when building. On Mac OS X
204         and Linux that is rarely a problem given how easily available the tools are. On Windows however
205         a separate installation of various GNU tools is required as the operating system doesn't come with
206         them. To make the development more convenient, Qt 5 provides a copy of the most essential tools in
207         the gnuwin32 directory of the qt5.git top-level repository.
208
209         This patch tries to detect the presence of those tools and prepends them to the PATH if found.
210
211         This is required in preparation for the elimination of qt5/qtwebkit.pri, which currently expands
212         PATH before calling build-webkit. It it also required for the upcoming introduction of win_flex
213         as dependency over flex, which can be done with less hassle when qt5's gnuwin32 directory has been
214         updated with the new tool.
215
216         * Scripts/webkitdirs.pm:
217         (checkRequiredSystemConfig):
218         * qmake/mkspecs/features/default_post.prf:
219
220 2012-09-12  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
221
222         [Qt] Update build-jsc after r128174
223
224         Reviewed by Ossy.
225
226         * Scripts/build-jsc:
227
228 2012-09-12  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
229
230         [Qt] Teach addStrictSubdirOrderBetween to handle more than two targets
231
232         By hard-coding the names of the targets we defined we ended up just
233         redefining the previous target when using addStrictSubdirOrderBetween
234         more than once in a single project file.
235
236         We now embed the two base targets into the target names.
237
238         Reviewed by Simon Hausmann.
239
240         * qmake/mkspecs/features/functions.prf:
241
242 2012-09-12  Philippe Liard  <pliard@google.com>
243
244         Depend on {base,net} GYP targets rather than {base,net}_java.
245         https://bugs.webkit.org/show_bug.cgi?id=95690
246
247         Reviewed by Adam Barth.
248
249         {base,net}_java became 'private' targets on the Chromium side that
250         clients should not depend on (see
251         https://chromiumcodereview.appspot.com/10913083/).
252         This removes a bunch of trailing white spaces additionally in
253         TestWebkitAPI.gyp (which includes non-trailing white space changes).
254
255         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
256         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
257
258 2012-09-12  Simon Hausmann  <simon.hausmann@nokia.com>
259
260         [Qt] Fix the build with ENABLE_NETSCAPE_PLUGIN_API=0
261         https://bugs.webkit.org/show_bug.cgi?id=96494
262
263         Reviewed by Tor Arne Vestbø.
264
265         WK2's ENABLE_PLUGIN_PROCESS uses the NPAPI functions unconditionally, so disable the
266         plugin process feature if we don't have NPAPI.
267
268         * qmake/mkspecs/features/features.prf:
269
270 2012-09-12  Sheriff Bot  <webkit.review.bot@gmail.com>
271
272         Unreviewed, rolling out r128280.
273         http://trac.webkit.org/changeset/128280
274         https://bugs.webkit.org/show_bug.cgi?id=96498
275
276         it broke compilation on windows debug bot (Requested by loislo
277         on #webkit).
278
279         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
280         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
281         (WTR::InjectedBundlePage::InjectedBundlePage):
282         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
283         (InjectedBundlePage):
284         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
285         (WTR::TestRunner::TestRunner):
286         * WebKitTestRunner/InjectedBundle/TestRunner.h:
287         (TestRunner):
288
289 2012-09-12  Simon Hausmann  <simon.hausmann@nokia.com>
290
291         [Qt] Make it possible to build with "make release" and "make debug" on Windows
292         https://bugs.webkit.org/show_bug.cgi?id=96488
293
294         Reviewed by Tor Arne Vestbø.
295
296         A "make debug" is passed through recursively and currently it aborts at Makefile.DerivedSources
297         because there are no such targets. We want the generated sources to be independent from release
298         or debug build configurations, so it is sufficient to provide fake debug and release targets that
299         redirect to the same general-purpose target (first) of creating the derived sources.
300
301         * qmake/mkspecs/features/default_post.prf:
302
303 2012-09-12  Kenneth Rohde Christiansen  <kenneth@webkit.org>
304
305         [EFL] Make DumpRenderTree smarter at finding the fonts
306         http://webkit.org/b/96281
307
308         Reviewed by Gyuyoung Kim.
309
310         Respect WEBKITOUTPUTDIR and expand the font dir from it.
311         Use CString consistently.
312
313         * DumpRenderTree/efl/FontManagement.cpp:
314         (buildPath):
315         (getCoreFontFiles):
316         (addFontDirectory):
317         (addFontFiles):
318         (getCustomBuildDir):
319         (getPlatformFontsPath):
320         (addFontsToEnvironment):
321
322 2012-09-12  Christophe Dumez  <christophe.dumez@intel.com>
323
324         [WK2][WKTR] TestRunner needs to implement dumpApplicationCacheDelegateCallbacks
325         https://bugs.webkit.org/show_bug.cgi?id=96374
326
327         Reviewed by Kenneth Rohde Christiansen.
328
329         Implement support for dumpApplicationCacheDelegateCallbacks
330         and disallowIncreaseForApplicationCacheQuota in
331         WebKitTestRunner and properly dump the information
332         expected by the tests.
333
334         If the application cache quota is reached for a given
335         security origin, WebKitTestRunner will reset the quota
336         to its default value, unless intructed not to via
337         disallowIncreaseForApplicationCacheQuota().
338
339         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
340         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
341         (WTR::InjectedBundlePage::InjectedBundlePage):
342         (WTR::InjectedBundlePage::didReachApplicationCacheOriginQuota):
343         (WTR):
344         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
345         (InjectedBundlePage):
346         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
347         (WTR::TestRunner::TestRunner):
348         (WTR::TestRunner::disallowIncreaseForApplicationCacheQuota):
349         (WTR):
350         * WebKitTestRunner/InjectedBundle/TestRunner.h:
351         (WTR::TestRunner::dumpApplicationCacheDelegateCallbacks):
352         (TestRunner):
353         (WTR::TestRunner::shouldDisallowIncreaseForApplicationCacheQuota):
354         (WTR::TestRunner::shouldDumpApplicationCacheDelegateCallbacks):
355
356 2012-09-12  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
357
358         [WK2] [WTR] WebKitTestRunner needs TestRunner.workerThreadCount
359         https://bugs.webkit.org/show_bug.cgi?id=96388
360
361         Reviewed by Kenneth Rohde Christiansen.
362
363         Exported TestRunner.workerThreadCount as readonly attribute.
364
365         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
366         * WebKitTestRunner/InjectedBundle/TestRunner.cpp: Added workerThreadCount() method implementation.
367         (WTR::TestRunner::workerThreadCount): Returns count of worker threads.
368         (WTR):
369         * WebKitTestRunner/InjectedBundle/TestRunner.h: Added workerThreadCount() method. 
370         (TestRunner):
371
372 2012-09-12  Jochen Eisinger  <jochen@chromium.org>
373
374         [chromium] Consume a user gesture when creating a new view.
375         https://bugs.webkit.org/show_bug.cgi?id=96373
376
377         Reviewed by Adam Barth.
378
379         * DumpRenderTree/chromium/WebViewHost.cpp:
380         (WebViewHost::createView):
381
382 2012-09-12  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
383
384         [WK2][WTR] Some of TestRunner special options are not reset before testing
385         https://bugs.webkit.org/show_bug.cgi?id=96384
386
387         Reviewed by Kenneth Rohde Christiansen.
388
389         Now values of the following special options are reset:
390         void setAcceptsEditing(in boolean value);
391         void setCloseRemainingWindowsWhenComplete(in boolean value);
392         void setXSSAuditorEnabled(in boolean value);
393         void setAllowFileAccessFromFileURLs(in boolean value);
394         void setPluginsEnabled(in boolean value);
395         void setPopupBlockingEnabled(in boolean value);
396
397         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
398         (WTR::InjectedBundle::beginTesting):
399
400 2012-09-11  Ryuan Choi  <ryuan.choi@samsung.com>
401
402         [CMAKE] Supply feature defines to CodeGeneratorTestRunner.
403         https://bugs.webkit.org/show_bug.cgi?id=96273
404
405         Reviewed by Gyuyoung Kim.
406
407         Passed FEATURE_DEFINES_WITH_SPACE_SEPARATOR to CodeGeneratorTestRunner
408         in order to distinguish compile time options such as ENABLE(TOUCH_EVENTS).
409
410         * WebKitTestRunner/CMakeLists.txt:
411
412 2012-09-11  Dirk Pranke  <dpranke@chromium.org>
413
414         [chromium] rename webkit_resources
415         https://bugs.webkit.org/show_bug.cgi?id=95874
416
417         Reviewed by James Robinson.
418
419         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
420
421 2012-09-11  Jochen Eisinger  <jochen@chromium.org>
422
423         [chromium] don't include content shell builders in the garden-o-matic view
424         https://bugs.webkit.org/show_bug.cgi?id=96418
425
426         Reviewed by James Robinson.
427
428         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
429
430 2012-09-11  Adam Klein  <adamk@chromium.org>
431
432         Garden-o-matic should ignore a wider variety of warnings in buildbot json
433         https://bugs.webkit.org/show_bug.cgi?id=96411
434
435         Reviewed by Adam Barth.
436
437         Previously only the exact string "warning" was treated as a warning
438         result. This patch treats any string with "warning" as a substring
439         as a warning (e.g., "warnings", as seen on the cr-win buildbots).
440
441         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
442         (.):
443         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
444
445 2012-09-11  Marcelo Lira  <marcelo.lira@openbossa.org>
446
447         Restore original value of mock scrollbars enabled in InternalSettings
448         https://bugs.webkit.org/show_bug.cgi?id=87680
449
450         Reviewed by Adam Barth.
451
452         Added backup support for mock scrollbars state in
453         InternalSettings::Backup, and removed support for these
454         from Qt's DumpRenderTree.
455
456         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
457         (WebCore::WebPage::resetSettings):
458         * DumpRenderTree/qt/TestRunnerQt.cpp:
459         (TestRunner::setFrameFlatteningEnabled):
460         * DumpRenderTree/qt/TestRunnerQt.h:
461         (TestRunner):
462
463 2012-09-11  Dan Bernstein  <mitz@apple.com>
464
465         Fixed a bug where kill-old-processes failed to kill clang++.
466
467         Rubber-stamped by Alexey Proskuryakov.
468
469         * BuildSlaveSupport/kill-old-processes: Escape the + signs in "clang++"
470         where it is passed as a regular expression to killall.
471
472 2012-09-11  Philippe Liard  <pliard@google.com>
473
474         Make sure that md5sum is not setup on non-Chromium Android platforms.
475         https://bugs.webkit.org/show_bug.cgi?id=96393
476
477         Reviewed by Adam Barth.
478
479         Previously md5sum was setup in ChromiumAndroidDriver's constructor
480         which is also invoked on non-Chromium Android platforms. The following patch
481         moves this logic to _startup_test() which is guaranteed to be invoked
482         only on Android.
483
484         This also updates the prompt unit test in chromium_android_unittest.py.
485
486         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
487         (ChromiumAndroidDriver.__init__):
488         (ChromiumAndroidDriver._setup_md5sum_and_push_data_if_needed):
489         (ChromiumAndroidDriver._setup_test):
490
491 2012-09-11  Raphael Kubo da Costa  <rakuco@webkit.org>
492
493         [EFL] Rewrite the EFL-related Find modules
494         https://bugs.webkit.org/show_bug.cgi?id=95237
495
496         Reviewed by Kenneth Rohde Christiansen.
497
498         * DumpRenderTree/efl/CMakeLists.txt: Stop setting the LINK_FLAGS
499         property and include all the Enlightenment Foundation Libraries
500         required by the target.
501         * EWebLauncher/CMakeLists.txt: Ditto.
502         * MiniBrowser/efl/CMakeLists.txt: Ditto.
503         * TestWebKitAPI/PlatformEfl.cmake: Add missing include directories
504         now that they are not added implicitly.
505         * WebKitTestRunner/CMakeLists.txt: Stop setting the LINK_FLAGS
506         property.
507         * WebKitTestRunner/PlatformEfl.cmake: Stop setting the LINK_FLAGS
508         property and include all the Enlightenment Foundation Libraries
509         required by the target.
510
511 2012-09-11  Christophe Dumez  <christophe.dumez@intel.com>
512
513         [WK2][WKTR] TestRunner needs to implement setApplicationCacheOriginQuota
514         https://bugs.webkit.org/show_bug.cgi?id=96379
515
516         Reviewed by Kenneth Rohde Christiansen.
517
518         Add implementation for setApplicationCacheOriginQuota
519         to WebKitTestRunner.
520
521         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
522         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
523         (WTR::TestRunner::setApplicationCacheOriginQuota):
524         (WTR):
525         * WebKitTestRunner/InjectedBundle/TestRunner.h:
526         (TestRunner):
527
528 2012-09-11  Luciano Wolf  <luciano.wolf@openbossa.org>
529
530         [Qt] [WK2] editing/inserting/typing-tab-designmode tests are failing
531         https://bugs.webkit.org/show_bug.cgi?id=91988
532
533         Reviewed by Kenneth Rohde Christiansen.
534
535         Regular key events generated by Qt for the tab key fill both
536         keyCode and keyText, so we should do the same for our fake events.
537
538         * WebKitTestRunner/qt/EventSenderProxyQt.cpp:
539         (WTR::EventSenderProxy::keyDown):
540
541 2012-09-11  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
542
543         [WK2][WTR] WebKitTestRunner needs testRunner.setSpatialNavigationEnabled
544         https://bugs.webkit.org/show_bug.cgi?id=96269
545
546         Reviewed by Kenneth Rohde Christiansen.
547
548         Exported missing testRunner.setSpatialNavigationEnabled() method.
549
550         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
551         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
552         (WTR::InjectedBundle::beginTesting):
553         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
554         (WTR::TestRunner::setSpatialNavigationEnabled):
555         (WTR):
556         * WebKitTestRunner/InjectedBundle/TestRunner.h:
557         (TestRunner):
558
559 2012-09-11  Philippe Liard  <pliard@chromium.org>
560
561         Support LayoutTests on non-rooted devices for Chromium Android.
562         https://bugs.webkit.org/show_bug.cgi?id=95346
563
564         Reviewed by Dirk Pranke.
565
566         We were requiring root permissions to sync resource files with the
567         device by comparing host and device files' timestamp after we
568         synchronized the device and host clocks.
569         We now rely on md5sum provided by Chromium to handle data files
570         synchronization.
571
572         * Scripts/webkitpy/layout_tests/port/chromium.py:
573         (ChromiumPort._check_driver_build_up_to_date):
574         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
575         (ChromiumAndroidDriver.__init__):
576         (ChromiumAndroidDriver._setup_test):
577         (ChromiumAndroidDriver._push_executable):
578         (ChromiumAndroidDriver._push_test_resources):
579         (ChromiumAndroidDriver._read_prompt):
580
581 2012-09-11  Simon Hausmann  <simon.hausmann@nokia.com>
582
583         [Qt] Fix output paths on Windows
584
585         Reviewed by Tor Arne Vestbø.
586
587         BUILD_ROOT_RELATIVE_TO_OUT_PWD was calculated incorrectly Windows, resulting in an incorrect
588         DESTDIR transformation at the end of default_post.prf. The calculation should be baased no qmake
589         paths, which can use drive letters but always use forward slashes. So the use of QMAKE_DIR_SEP is
590         incorrect here and instead for the calculation we should use forward slashes.
591
592         * qmake/mkspecs/features/default_pre.prf:
593
594 2012-09-11  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
595
596         [Qt] Fix passing of defines from the build system
597
598         We explicitly have to pass -DENABLE_FOO=0 for every single feature
599         that's not explicitly enabled in WEBKIT_CONFIG, since Platform.h
600         will add it's own defaults if the features are not defined.
601
602         At some point we might want to run a configure-test to inspect what
603         Platform.h will do, so we can sync up the build system's view of
604         the feature-defines with the compile-time situation.
605
606         Reviewed by Simon Hausmann.
607
608         * qmake/mkspecs/features/functions.prf:
609
610 2012-09-11  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
611
612         [Qt] Fix --qt option to build-webkit
613
614         Reviewed by Ossy.
615
616         * Scripts/webkitdirs.pm:
617         (buildQMakeProjects):
618
619 2012-09-11  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
620
621         [Qt] Add a configure step to the Qt build system
622
623         This allows building the Qt port using just 'qmake WebKit.pro'. Using
624         the build-webkit script is still supported, and will add slightly more
625         logic to the build, such as detecting the need for clean or incremental
626         builds.
627
628         Internally, the build system now uses a new variable, WEBKIT_CONFIG, for
629         all things related to configuring the build, such as use/have/enable flags,
630         and these are translated to defines in default_post. Project files should
631         no longer check contains(DEFINES,...) to detect features, but use the new
632         enable?(), use?(), and have?() functions. The no_webkit1 and no_webkit2
633         options have been translated into WEBKIT_CONFIG options as well, and can
634         be checked using build?().
635
636         Reviewed by Simon Hausmann.
637
638         * DumpRenderTree/qt/DumpRenderTree.pro:
639         * MiniBrowser/qt/MiniBrowser.pro:
640         * QtTestBrowser/QtTestBrowser.pro:
641         * Scripts/build-webkit:
642         * Scripts/webkitdirs.pm:
643         (qtFeatureDefaults):
644         (buildQMakeProjects):
645         * Tools.pro:
646         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
647         * WebKitTestRunner/InjectedBundle/Target.pri:
648         * WebKitTestRunner/Target.pri:
649         * qmake/.qmake.cache: Added.
650         * qmake/configure.pri: Removed.
651         * qmake/configure.pro: Removed.
652         * qmake/dump-features: Added.
653         * qmake/mkspecs/features/configure.prf: Added.
654         * qmake/mkspecs/features/default_post.prf:
655         * qmake/mkspecs/features/default_pre.prf:
656         * qmake/mkspecs/features/features.prf:
657         * qmake/mkspecs/features/features.pri:
658         * qmake/mkspecs/features/functions.prf:
659         * qmake/mkspecs/features/gprof.prf:
660         * qmake/mkspecs/features/mac/default_pre.prf:
661         * qmake/mkspecs/features/win32/default_pre.prf:
662
663 2012-09-11  Christophe Dumez  <christophe.dumez@intel.com>
664
665         [WK2][WKTR] TestRunner needs to implement clearApplicationCacheForOrigin
666         https://bugs.webkit.org/show_bug.cgi?id=96372
667
668         Reviewed by Kenneth Rohde Christiansen.
669
670         Implement clearApplicationCacheForOrigin() in WebKitTestRunner.
671
672         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
673         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
674         (WTR::TestRunner::clearApplicationCacheForOrigin):
675         (WTR):
676         * WebKitTestRunner/InjectedBundle/TestRunner.h:
677         (TestRunner):
678
679 2012-09-11  Peter Beverloo  <peter@chromium.org>
680
681         Don't ignore .pak files when making an archive for Android
682         https://bugs.webkit.org/show_bug.cgi?id=96375
683
684         Reviewed by Jochen Eisinger.
685
686         The .pak files do need to be pushed to the device, which is currently
687         preventing the layout test runner from correctly pushing all files.
688
689         * BuildSlaveSupport/built-product-archive:
690         (archiveBuiltProduct):
691
692 2012-09-11  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
693
694         [WK2][WTR] WebKitTestRunner needs layoutTestController.setMinimumTimerInterval
695         https://bugs.webkit.org/show_bug.cgi?id=96256
696
697         Reviewed by Kenneth Rohde Christiansen.
698
699         Exported TestRunner::setMinimumTimerInterval() method.
700
701         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
702         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
703         (WTR::InjectedBundle::beginTesting):
704         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
705         (WTR::TestRunner::setMinimumTimerInterval):
706         (WTR):
707         * WebKitTestRunner/InjectedBundle/TestRunner.h:
708         (TestRunner):
709
710 2012-09-11  Tommy Widenflycht  <tommyw@google.com>
711
712         MediaStream API: add RTCPeerConnection::onnegotiationneeded
713         https://bugs.webkit.org/show_bug.cgi?id=96097
714
715         Reviewed by Adam Barth.
716
717         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
718         (MockWebRTCPeerConnectionHandler::addStream):
719         (MockWebRTCPeerConnectionHandler::removeStream):
720
721 2012-09-11  Peter Beverloo  <peter@chromium.org>
722
723         Recognize the "--chromium-android" argument in run-webkit-tests
724         https://bugs.webkit.org/show_bug.cgi?id=96369
725
726         Reviewed by Jochen Eisinger.
727
728         Chromium for the Android platform cannot determine the platform based on
729         the host itself, so we'll use the --chromium-android argument as the build
730         master passes to run-webkit-tests. Make sure the layout test runner
731         can pick the right port when doing so.
732
733         * Scripts/run-webkit-tests:
734         * Scripts/webkitpy/layout_tests/port/factory.py:
735         (port_options):
736
737 2012-09-11  Simon Hausmann  <simon.hausmann@nokia.com>
738
739         [QT] Fix build with newer qmake from Qt 5
740
741         Rubber-stamped by Tor Arne Vestbø.
742
743         Support for "qmake -query QMAKE_MKSPECS" was removed from qmake. Replace
744         it with the use of QT_HOST_DATA/mkspecs.
745         Based on patch by Oswald Buddenhagen <oswald.buddenhagen@nokia.com>.
746
747         * Scripts/webkitdirs.pm:
748         (builtDylibPathForName):
749         (qtMakeCommand):
750
751 2012-09-11  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
752
753         [Qt] Remove Qt 4 syncqt, no longer needed
754
755         Reviewed by Simon Hausmann.
756
757         * qmake/syncqt-4.8: Removed.
758
759 2012-09-11  Andras Becsi  <andras.becsi@nokia.com>
760
761         [Qt][WK2] Fix the build with recent Qt5
762         https://bugs.webkit.org/show_bug.cgi?id=96283
763
764         Reviewed by Simon Hausmann.
765
766         We should not use deprecated Qt API because doing that results
767         in build failure since all the deprecated API was disabled
768         in qtbase (https://codereview.qt-project.org/#change,24890).
769
770         * DumpRenderTree/qt/main.cpp:
771         (main):
772         * MiniBrowser/qt/MiniBrowserApplication.cpp:
773         (MiniBrowserApplication::notify):
774         * MiniBrowser/qt/raw/View.cpp:
775         (View::exposeEvent):
776         (main):
777
778 2012-09-11  Christophe Dumez  <christophe.dumez@intel.com>
779
780         WTR needs an implementation of applicationCacheDiskUsageForOrigin
781         https://bugs.webkit.org/show_bug.cgi?id=87188
782
783         Reviewed by Kenneth Rohde Christiansen.
784
785         Add implementation for applicationCacheDiskUsageForOrigin()
786         in WebKitTestRunner.
787
788         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
789         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
790         (WTR::TestRunner::applicationCacheDiskUsageForOrigin):
791         (WTR):
792         * WebKitTestRunner/InjectedBundle/TestRunner.h:
793         (TestRunner):
794
795 2012-09-10  Joanmarie Diggs  <jdiggs@igalia.com>
796
797         Properly expose <legend> elements to ATs
798         https://bugs.webkit.org/show_bug.cgi?id=84137
799
800         Reviewed by Chris Fleizach.
801
802         Created a new WebCore Accessibility Role, LegendRole. Used it to map to
803         the expected platform role, ATK_ROLE_LABEL. Also established the needed
804         AtkRelation pair, label-for/labelled-by between the legend and fieldset.
805
806         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
807         (roleToString): added "Legend" string for WebAccessibilityRoleLegend
808
809 2012-09-10  Jer Noble  <jer.noble@apple.com>
810
811         Unreviewed; rolling out r128081.
812         http://trac.webkit.org/changeset/128081
813         https://bugs.webkit.org/show_bug.cgi?id=96134
814
815         New test http/tests/media/video-donottrack.html fails to run on Apple test bots.
816
817         * DumpRenderTree/TestRunner.cpp:
818         (TestRunner::TestRunner):
819         (TestRunner::staticFunctions):
820         * DumpRenderTree/TestRunner.h:
821         (TestRunner):
822         * DumpRenderTree/mac/FrameLoadDelegate.mm:
823         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
824         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
825         (WTR::InjectedBundlePage::InjectedBundlePage):
826         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
827         (InjectedBundlePage):
828         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
829         (WTR::TestRunner::TestRunner):
830         * WebKitTestRunner/InjectedBundle/TestRunner.h:
831         (TestRunner):
832
833 2012-09-10  Christophe Dumez  <christophe.dumez@intel.com>
834
835         TestExpectationsChecker._determine_port_from_expectations_path() does not support cascaded TestExpectations
836         https://bugs.webkit.org/show_bug.cgi?id=96205
837
838         Reviewed by Tony Chang.
839
840         _determine_port_from_expectations_path() was calling port.path_to_test_expectations_file()
841         internally, which means that it would support only 1 TestExpectations file per port. This
842         is an issue for ports such as EFL that support cascased TestExpectations (efl-wk2 -> efl
843         and efl-wk1 -> efl).
844
845         This patch makes _determine_port_from_expectations_path() call port.expectations_files()
846         instead so that all the ports TestExpectations are recognized. The ports are also
847         constructed twice, with "webkit_test_runner" option set to True and False so that we
848         retrieve the TestExpectations paths for both WebKit1 and WebKit2.
849
850         * Scripts/webkitpy/style/checkers/test_expectations.py:
851         (TestExpectationsChecker._determine_port_from_expectations_path):
852         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Add corresponding unit test.
853         (TestExpectationsTestCase.test_determine_port_from_expectations_path):
854
855 2012-09-10  Ryuan Choi  <ryuan.choi@samsung.com>
856
857         [WTR] Generated source files should include config.h
858         https://bugs.webkit.org/show_bug.cgi?id=96254
859
860         Reviewed by Benjamin Poulain.
861
862         Current coding style has encouraged to include config.h at the first position,
863         But derived sources, generated by CodeGeneratorTestRunner.pm, don't follow
864         this rule.
865         Therefore, this patch added it.
866
867         * WebKitTestRunner/InjectedBundle/Bindings/CodeGeneratorTestRunner.pm:
868         (_generateImplementationFile):
869
870 2012-09-10  Rick Byers  <rbyers@chromium.org>
871
872         Add handling of new GestureTapCancel in DRT
873         
874         https://bugs.webkit.org/show_bug.cgi?id=96183
875
876         Reviewed by Adam Barth.
877
878         * DumpRenderTree/chromium/TestWebPlugin.cpp:
879         (TestWebPlugin::handleInputEvent):
880
881 2012-09-10  Brady Eidson  <beidson@apple.com>
882
883         Javascript in foreground tabs should not wait synchronously for plug-ins to load
884         <rdar://problem/12067415> and https://bugs.webkit.org/show_bug.cgi?id=96167
885
886         Reviewed by Geoff Garen.
887
888         Enhance the "Slow NPP_New" plug-in to also be able to return properties to javascript.
889
890         * DumpRenderTree/TestNetscapePlugIn/Tests/SlowNPPNew.cpp:
891         (PluginObject):
892         (SlowNPPNew::PluginObject::PluginObject):
893         (SlowNPPNew::PluginObject::~PluginObject):
894         (SlowNPPNew::PluginObject::hasProperty):
895         (SlowNPPNew::PluginObject::getProperty):
896         (SlowNPPNew::NPP_GetValue):
897         (SlowNPPNew):
898
899 2012-09-10  Tim Horton  <timothy_horton@apple.com>
900
901         WKTR often reports an unresponsive WebProcess on Mac bots
902         https://bugs.webkit.org/show_bug.cgi?id=95906
903
904         Reviewed by Dan Bernstein.
905
906         Attempt to work around resource contention issues by increasing the NRWT
907         timeout when running against WKTR, as well as WKTR's own WebProcess timeout.
908
909         * Scripts/webkitpy/layout_tests/port/base.py:
910         (Port.default_timeout_ms): Default timeout when using WKTR -> 80 seconds.
911         * WebKitTestRunner/TestController.cpp:
912         (WTR): LongTimeout (used when waiting for WebProcess to finish) -> 60 seconds.
913
914 2012-09-07  Jer Noble  <jer.noble@apple.com>
915
916         <audio> and <video> should send Do Not Track when appropriate
917         https://bugs.webkit.org/show_bug.cgi?id=96134
918
919         Reviewed by Eric Carlson.
920
921         Add a testRunner JS API for calling setShouldSendDoNotTrackHTTPHeader to DRT and 
922         WTR. Add FrameLoaderClient client methods for retrieving that value.
923
924         * DumpRenderTree/TestRunner.cpp:
925         (setShouldSendDoNotTrackHTTPHeaderCallback): Pass through to the TestRunner.
926         (TestRunner::staticFunctions): Add the setShouldSendDoNotTrackHTTPHeader function.
927         * DumpRenderTree/TestRunner.h:
928         (TestRunner::setShouldSendDoNotTrackHTTPHeader): Simple setter.
929         (TestRunner::shouldSendDoNotTrackHTTPHeader): Simple getter.
930         * DumpRenderTree/mac/FrameLoadDelegate.mm:
931         (-[FrameLoadDelegate webViewShouldSendDoNotTrackHTTPHeader:]): 
932             Pass the question on to the TestRunner.
933         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
934         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
935         (WTR::InjectedBundlePage::InjectedBundlePage):
936             Add the setShouldSendDoNotTrackHTTPHeader client function.
937         (WTR::InjectedBundlePage::shouldSendDoNotTrackHTTPHeader):
938             Pass the question on to the TestRunner.
939         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
940         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
941         * WebKitTestRunner/InjectedBundle/TestRunner.h:
942         (WTR::TestRunner::shouldSendDoNotTrackHTTPHeader): Simple getter.
943         (WTR::TestRunner::setShouldSendDoNotTrackHTTPHeader): Simple Setter.
944
945 2012-09-10  Ojan Vafai  <ojan@chromium.org>
946
947         Make EFL bots work on the flakiness dashboard
948         https://bugs.webkit.org/show_bug.cgi?id=96164
949
950         Reviewed by Tony Chang.
951
952         Also, update outdated unittests that had started failing.
953
954         * TestResultServer/static-dashboards/builders.js:
955         (isWebkitTestRunner):
956         * TestResultServer/static-dashboards/flakiness_dashboard.js:
957         (nonChromiumPlatform):
958         * TestResultServer/static-dashboards/flakiness_dashboard_unittests.js:
959         (test):
960
961 2012-09-10  Tommy Widenflycht  <tommyw@google.com>
962
963         [chromium] MediaStream API: Remove the Descriptor postfix
964         https://bugs.webkit.org/show_bug.cgi?id=96268
965
966         Reviewed by Adam Barth.
967
968         After consideration I realized that there is no need whatsoever in using the Descriptor prefix
969         in the WekKit embedder API. Removed from WebRTCSessionDescription and WebRTCICECandidate to start with.
970
971         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
972         (RTCSessionDescriptionRequestSuccededTask::RTCSessionDescriptionRequestSuccededTask):
973         (MockWebRTCPeerConnectionHandler::createOffer):
974         (MockWebRTCPeerConnectionHandler::createAnswer):
975         (MockWebRTCPeerConnectionHandler::setLocalDescription):
976         (MockWebRTCPeerConnectionHandler::setRemoteDescription):
977         (MockWebRTCPeerConnectionHandler::localDescription):
978         (MockWebRTCPeerConnectionHandler::remoteDescription):
979         (MockWebRTCPeerConnectionHandler::addICECandidate):
980         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
981         (MockWebRTCPeerConnectionHandler):
982
983 2012-09-10  Arnaud Renevier  <a.renevier@sisa.samsung.com>
984
985         [Gtk] failure to build jhbuild dependencies with cups 1.6
986         https://bugs.webkit.org/show_bug.cgi?id=95991
987
988         Reviewed by Gustavo Noronha Silva.
989
990         Bump gtk+ and gnome-themes-standard dependencies to 3.4.2 to allow
991         building with cups >= 1.6 [GNOME #670373].
992
993         * gtk/jhbuild.modules:
994
995 2012-09-10  Christophe Dumez  <christophe.dumez@intel.com>
996
997         [EFL][DRT] Encoding issue in printed console messages
998         https://bugs.webkit.org/show_bug.cgi?id=96263
999
1000         Reviewed by Kenneth Rohde Christiansen.
1001
1002         Correctly handle CONSOLE messages that contain non-ascii
1003         characters in EFL's DRT. We need to use String::fromUTF8()
1004         to construct a WTF::String from a raw string.
1005
1006         * DumpRenderTree/efl/DumpRenderTreeView.cpp:
1007         (onConsoleMessage):
1008
1009 2012-09-10  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1010
1011         [WK2][WTR] User style sheet location should be reset before each test
1012         https://bugs.webkit.org/show_bug.cgi?id=96262
1013
1014         Reviewed by Kenneth Rohde Christiansen.
1015
1016         User style sheet location is reset before each test now.
1017
1018         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1019         (WTR::InjectedBundle::beginTesting):
1020
1021 2012-09-10  Simon Hausmann  <simon.hausmann@nokia.com>
1022
1023         Fix build of WTR with Qt on Windows
1024         https://bugs.webkit.org/show_bug.cgi?id=96249
1025
1026         Reviewed by Kenneth Rohde Christiansen.
1027
1028         Windows loves to define min and max as macros, which beautifully
1029         clashes in many places where min and max are used as variable or method
1030         names. It can be convinced not to do this when min and max are already
1031         define as macros (expanding to themselves). We have this kind of
1032         workaround in config.h files all over WebKit and as it turns out when
1033         building WTR with Qt on Windows we do end up with such a clash. Therefore
1034         we need the same workaround.
1035
1036         * WebKitTestRunner/config.h:
1037
1038 2012-09-10  Christophe Dumez  <christophe.dumez@intel.com>
1039
1040         [WK2] New fast/forms/number/number-interactive-validation-required.html fails
1041         https://bugs.webkit.org/show_bug.cgi?id=95936
1042
1043         Reviewed by Kenneth Rohde Christiansen.
1044
1045         Enable interactive form validation setting in WebKitTestRunner as
1046         we are currently doing in DumpRenderTree. This is needed for some
1047         tests to pass.
1048
1049         * WebKitTestRunner/TestController.cpp:
1050         (WTR::TestController::resetStateToConsistentValues):
1051
1052 2012-09-09  Zan Dobersek  <zandobersek@gmail.com>
1053
1054         Clean up the code in XvfbDriver
1055         https://bugs.webkit.org/show_bug.cgi?id=96202
1056
1057         Reviewed by Philippe Normand.
1058
1059         Changes:
1060         - Remove unneeded imports,
1061         - assign the _lock_file member variable when the display number is determined,
1062         - no need for parentheses around a single string format argument,
1063         - use FileSystem.exists method when checking if lock file is present as it
1064         makes the intention much more clearer.
1065
1066         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
1067         (XvfbDriver._start):
1068         (XvfbDriver.stop):
1069
1070 2012-09-09  Patrick Gansterer  <paroga@webkit.org>
1071
1072         Make the String initialization on the function side of String::number()
1073         https://bugs.webkit.org/show_bug.cgi?id=95940
1074
1075         Reviewed by Benjamin Poulain.
1076
1077         * TestWebKitAPI/Tests/WTF/IntegerToStringConversion.cpp:
1078         (testBoundaries):
1079         (testNumbers):
1080
1081 2012-09-09  Pierre Rossi  <pierre.rossi@gmail.com>
1082
1083         [WK2] expose element rect for color input type
1084         https://bugs.webkit.org/show_bug.cgi?id=91664
1085
1086         Reviewed by Simon Hausmann.
1087
1088         Update the QML logic in MiniBrowser to take advantage of this.
1089         The picker size is now hardcoded and we try to place it below the
1090         input element if possible.
1091
1092         * MiniBrowser/qt/qml/ColorChooser.qml:
1093
1094 2012-09-08  Christophe Dumez  <christophe.dumez@intel.com>
1095
1096         [WK2] New fast/events/tab-focus-link-in-canvas fails from r126908
1097         https://bugs.webkit.org/show_bug.cgi?id=95329
1098
1099         Reviewed by Kenneth Rohde Christiansen.
1100
1101         Explicitely disable TabToLinks setting in WebKitTestRunner.
1102         Tests override the "WebKitTabToLinksPreferenceKey" preference
1103         to enable it.
1104
1105         * WebKitTestRunner/TestController.cpp:
1106         (WTR::TestController::resetStateToConsistentValues):
1107
1108 2012-09-08  Kenichi Ishibashi  <bashi@chromium.org>
1109
1110         webkit.py gdb pretty printer can't print CString
1111         https://bugs.webkit.org/show_bug.cgi?id=96068
1112
1113         Reviewed by Tony Chang.
1114
1115         Update WTFCStringPrinter to follow r126191.
1116
1117         * gdb/webkit.py:
1118         (WTFCStringPrinter.to_string):
1119
1120 2012-09-07  Benjamin Poulain  <bpoulain@apple.com>
1121
1122         Rename the ustring() accessor to string()
1123         https://bugs.webkit.org/show_bug.cgi?id=95919
1124
1125         Reviewed by Geoffrey Garen.
1126
1127         Update EFL's DRT.
1128
1129         Fix some string conversion that have been made useless
1130         with recent changes.
1131
1132         * DumpRenderTree/efl/EventSender.cpp:
1133         (keyPadNameFromJSValue):
1134         (keyNameFromJSValue):
1135         * DumpRenderTree/efl/TestRunnerEfl.cpp:
1136         (TestRunner::queueLoad):
1137         (TestRunner::addOriginAccessWhitelistEntry):
1138         (TestRunner::removeOriginAccessWhitelistEntry):
1139         (TestRunner::setUserStyleSheetLocation):
1140         (TestRunner::setValueForUser):
1141         (TestRunner::elementDoesAutoCompleteForElementWithId):
1142         (TestRunner::execCommand):
1143         (TestRunner::findString):
1144         (TestRunner::isCommandEnabled):
1145         (TestRunner::clearApplicationCacheForOrigin):
1146         (TestRunner::setDomainRelaxationForbiddenForURLScheme):
1147         (TestRunner::pauseAnimationAtTimeOnElementWithId):
1148         (TestRunner::pauseTransitionAtTimeOnElementWithId):
1149         (toInt):
1150         (TestRunner::overridePreference):
1151         (TestRunner::addUserScript):
1152         (TestRunner::addUserStyleSheet):
1153         (TestRunner::evaluateInWebInspector):
1154         (TestRunner::evaluateScriptInIsolatedWorld):
1155         (TestRunner::setTextDirection):
1156         * DumpRenderTree/efl/WorkQueueItemEfl.cpp:
1157         (LoadItem::invoke):
1158         (LoadHTMLStringItem::invoke):
1159         (ScriptItem::invoke):
1160
1161 2012-09-06  Jeffrey Pfau  <jpfau@apple.com>
1162
1163         Extend third-party storage blocking API to optionally allow blocking all storage
1164         https://bugs.webkit.org/show_bug.cgi?id=95915
1165
1166         Reviewed by Brady Eidson.
1167
1168         Modified API test to use new revised API.
1169
1170         * TestWebKitAPI/Tests/WebKit2/WKPreferences.cpp:
1171         (TestWebKitAPI::TEST):
1172
1173 2012-09-07  Dominic Mazzoni  <dmazzoni@google.com>
1174
1175         AX: WebCore accessibility roles should be cross-platform
1176         https://bugs.webkit.org/show_bug.cgi?id=94870
1177
1178         Reviewed by Chris Fleizach.
1179
1180         Add debug strings to Chromium for new accessibility roles.
1181
1182         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
1183         (roleToString):
1184
1185 2012-09-07  Dominic Mazzoni  <dmazzoni@google.com>
1186
1187         AX: Chromium needs access to canvasHasFallbackContent
1188         https://bugs.webkit.org/show_bug.cgi?id=96124
1189
1190         Reviewed by Chris Fleizach.
1191
1192         Distinguish between canvas roles with and without fallback
1193         content on Chromium.
1194
1195         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
1196         (getRole):
1197
1198 2012-09-07  Ojan Vafai  <ojan@chromium.org>
1199
1200         Handle non-existant TestExpectations files gracefully
1201         https://bugs.webkit.org/show_bug.cgi?id=96149
1202
1203         Reviewed by Eric Seidel.
1204
1205         Even when we get rid of Skipped files, we shouldn't have each port
1206         require a TestExpectations file for each platform if it doesn't
1207         need one.
1208
1209         * Scripts/webkitpy/layout_tests/port/base.py:
1210         (Port.expectations_dict):
1211         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1212         (PortTest.test_nonexistant_expectations):
1213
1214 2012-09-07  Tony Chang  <tony@chromium.org>
1215
1216         Regression: check-webkit-style no longer checks TestExpectations for syntax errors
1217         https://bugs.webkit.org/show_bug.cgi?id=96136
1218
1219         Reviewed by Ojan Vafai.
1220
1221         When we renamed the file to TestExpectations, it was no longer considered a text file so
1222         it was no longer being checked. Since then, test_expectations_overrides() has been removed
1223         from the port object, so I've removed that from the check. I'm not sure what it was
1224         supposed to do.
1225
1226         * Scripts/webkitpy/style/checker.py:
1227         (CheckerDispatcher._file_type): Mark TestExpectations as a text file.
1228         * Scripts/webkitpy/style/checkers/test_expectations.py:
1229         (TestExpectationsChecker.check_test_expectations): Remove overrides param.
1230         (TestExpectationsChecker.check): Remove overrides param.
1231         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
1232         (TestExpectationsTestCase.assert_lines_lint): Remove overrides param.
1233
1234 2012-09-07  Christophe Dumez  <christophe.dumez@intel.com>
1235
1236         [EFL] Use same default minimum logical font size in DRT and WTR
1237         https://bugs.webkit.org/show_bug.cgi?id=96116
1238
1239         Reviewed by Martin Robinson.
1240
1241         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1242         (DumpRenderTreeChrome::resetDefaultsToConsistentValues): Set minimum
1243         logical font size to 9 to match the value used in WTR.
1244         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1245         (WTR::InjectedBundle::beginTesting): Explicitly set the minimum
1246         logical font size to 9 (default value in WK2)
1247
1248 2012-09-07  Tommy Widenflycht  <tommyw@google.com>
1249
1250         MediaStream API: add RTCPeerConnection::createAnswer
1251         https://bugs.webkit.org/show_bug.cgi?id=96092
1252
1253         Reviewed by Adam Barth.
1254
1255         Expanding and enhancing MockWebRTCPeerConnectionHandler.
1256
1257         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
1258         (MockWebRTCPeerConnectionHandler::createOffer):
1259         (MockWebRTCPeerConnectionHandler::createAnswer):
1260         (MockWebRTCPeerConnectionHandler::setLocalDescription):
1261         (MockWebRTCPeerConnectionHandler::setRemoteDescription):
1262         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
1263         (MockWebRTCPeerConnectionHandler):
1264
1265 2012-09-07  Dean Jackson  <dino@apple.com>
1266
1267         Unreviewed changes to watchlist.
1268
1269         - Add new WebGL and Filters definitions
1270         - Add myself to OpenGL bugs
1271
1272         * Scripts/webkitpy/common/config/watchlist:
1273
1274 2012-09-07  Peter Beverloo  <peter@chromium.org>
1275
1276         Actually trigger the Android Tester
1277         https://bugs.webkit.org/show_bug.cgi?id=96118
1278
1279         Reviewed by Adam Barth.
1280
1281         * BuildSlaveSupport/build.webkit.org-config/config.json:
1282
1283 2012-09-07  Dominic Mazzoni <dmazzoni@google.com>
1284
1285         New time input needs accessibility
1286         https://bugs.webkit.org/show_bug.cgi?id=96032
1287
1288         Reviewed by Chris Fleizach.
1289
1290         Add support for valueDescription for testing.
1291
1292         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
1293         (getValueDescription):
1294         (AccessibilityUIElement::AccessibilityUIElement):
1295         (AccessibilityUIElement::valueDescriptionGetterCallback):
1296         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.h:
1297         (AccessibilityUIElement):
1298
1299 2012-09-07  Peter Beverloo  <peter@chromium.org>
1300
1301         Add the google-android-tests build slave to the master's config
1302         https://bugs.webkit.org/show_bug.cgi?id=96113
1303
1304         Reviewed by Csaba Osztrogonác.
1305
1306         * BuildSlaveSupport/build.webkit.org-config/config.json:
1307
1308 2012-09-07  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1309
1310         [WK2] [EFL] Set theme for WebKit2 API test runner
1311         https://bugs.webkit.org/show_bug.cgi?id=96100
1312
1313         Reviewed by Kenneth Rohde Christiansen.
1314
1315         Set the theme generated by the build for the WebKit2 API test runner.
1316
1317         * TestWebKitAPI/PlatformEfl.cmake:
1318         * TestWebKitAPI/efl/PlatformWebView.cpp:
1319         (TestWebKitAPI::PlatformWebView::PlatformWebView):
1320
1321 2012-09-07  Csaba Osztrogonác  <ossy@webkit.org>
1322
1323         Unreviewed. Add Gábor Ballabás as contributor.
1324
1325         * Scripts/webkitpy/common/config/committers.py:
1326
1327 2012-09-07  Zoltan Arvai  <zarvai@inf.u-szeged.hu>
1328
1329         [Qt] Fix ROOT_BUILD_DIR on Windows MSVC build.
1330         https://bugs.webkit.org/show_bug.cgi?id=96078
1331
1332         Reviewed by Tor Arne Vestbø.
1333
1334         Correct JSC, DumpRenderTree, ImageDiff, QtTestBrowser binaries location by adding $$toSystemPath to ROOT_BUILD_DIR generation.
1335
1336         * qmake/mkspecs/features/default_pre.prf:
1337
1338 2012-09-07  Christophe Dumez  <christophe.dumez@intel.com>
1339
1340         [EFL] Add TestExpectations file for efl-wk1
1341         https://bugs.webkit.org/show_bug.cgi?id=96019
1342
1343         Reviewed by Gyuyoung Kim.
1344
1345         Update TestExpectations cascading for EFL port
1346         so that we use efl-wk1 -> efl search paths
1347         for WebKit1 EFL.
1348
1349         * Scripts/webkitpy/layout_tests/port/efl.py:
1350         (EflPort._search_paths):
1351
1352 2012-09-07  Zan Dobersek  <zandobersek@gmail.com>
1353
1354         [NRWT] XvfbDriver should stop the Xvfb more aggressively
1355         https://bugs.webkit.org/show_bug.cgi?id=95097
1356
1357         Reviewed by Dirk Pranke.
1358
1359         The Xvfb process is now killed when the XvfbDriver is stopping. Simply
1360         terminating the process does not always work as expected, leading to timeouts
1361         in run-webkit-tests on the buildbot. The lock file Xvfb creates is also removed
1362         as it's not removed by the killed Xvfb process.
1363
1364         Also included is a thorough unittest for XvfbDriver. Proper starting of the driver
1365         is tested in different circumstances - whether pixel tests are enabled or not,
1366         with an arbitrary worker number or with an Xorg process already running on the system.
1367         The stopping of the driver is tested by checking that the Xvfb process is truly killed
1368         and then set to None.
1369
1370         * BuildSlaveSupport/kill-old-processes:
1371         Kill any stale Xvfb processes at the start of the build cycle so no Xvfb
1372         is left running, potentially clogging up a display at a certain position.
1373         * Scripts/webkitpy/common/system/executive_mock.py:
1374         (MockExecutive.__init__):
1375         The _running_pids is now a dictionary with process names as keys and pids
1376         as values.
1377         (MockExecutive.check_running_pid):
1378         Check whether the passed-in pid is in the _running_pids dictionary's values.
1379         (MockExecutive):
1380         (MockExecutive.running_pids):
1381         This method returns the list of all the process pids of which the name passes
1382         through the process_name_filter.
1383         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
1384         Use the Executive object of the port's host to check for running pids and open
1385         new subprocesses or kill them throughout the class.
1386         (XvfbDriver._start):
1387         Use the _server_process_constructor to make testing the XvfbDriver possible. Also,
1388         start the server process after it's created.
1389         (XvfbDriver.stop):
1390         Now kills the Xvfb process instead of terminating it and waiting for it to close.
1391         * Scripts/webkitpy/layout_tests/port/xvfbdriver_unittest.py: Added.
1392         (XvfbDriverTest):
1393         (XvfbDriverTest.make_driver):
1394         (XvfbDriverTest.assertDriverStartSuccessful):
1395         (XvfbDriverTest.test_start_no_pixel_tests):
1396         (XvfbDriverTest.test_start_pixel_tests):
1397         (XvfbDriverTest.test_start_arbitrary_worker_number):
1398         (XvfbDriverTest.test_start_existing_xorg_process):
1399         (XvfbDriverTest.test_stop):
1400         (XvfbDriverTest.test_stop.FakeXvfbProcess):
1401
1402 2012-09-06  Peter Beverloo  <peter@chromium.org>
1403
1404         Introduce the Chromium Android Release (Tests) bot
1405         https://bugs.webkit.org/show_bug.cgi?id=95980
1406
1407         Reviewed by Adam Barth.
1408
1409         This adds the Chromium Android Release (Tests) bot to the WebKit waterfall,
1410         which will be running layout tests on actual Android devices. Right now,
1411         four Galaxy Nexus phones are attached.
1412
1413         * BuildSlaveSupport/build.webkit.org-config/config.json:
1414
1415 2012-09-06  James Robinson  <jamesr@chromium.org>
1416
1417         [chromium] Use WebCompositorSupport functions instead of WebCompositor statics
1418         https://bugs.webkit.org/show_bug.cgi?id=96007
1419
1420         Reviewed by Adrienne Walker.
1421
1422         Fixes initialization / settings code to use WebCompositorSupport.
1423
1424         * DumpRenderTree/chromium/DumpRenderTree.cpp:
1425         (main):
1426         * DumpRenderTree/chromium/TestShell.cpp:
1427         (TestShell::initialize):
1428         (TestShell::setPerTilePaintingEnabled):
1429
1430 2012-09-06  Christophe Dumez  <christophe.dumez@intel.com>
1431
1432         [EFL][WK2] Provide implementation for PlatformWebView::windowFrame()
1433         https://bugs.webkit.org/show_bug.cgi?id=95995
1434
1435         Reviewed by Kenneth Rohde Christiansen.
1436
1437         Provide proper implementation for PlatformWebView::windowFrame()
1438         in EFL's WebKitTestRunner.
1439
1440         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp:
1441         (WTR::PlatformWebView::windowFrame):
1442
1443 2012-09-06  Fady Samuel  <fsamuel@chromium.org>
1444
1445         Fix WTF.Lrint on 32-bit platforms
1446         https://bugs.webkit.org/show_bug.cgi?id=96003
1447
1448         Reviewed by Benjamin Poulain.
1449
1450         Fixed failing WTF.Lrint test on 32-bit platforms.
1451
1452         * TestWebKitAPI/Tests/WTF/MathExtras.cpp:
1453         (TestWebKitAPI::TEST):
1454
1455 2012-09-06  Tommy Widenflycht  <tommyw@google.com>
1456
1457         MediaStream API: Add the local and remote description functionality to RTCPeerConnection
1458         https://bugs.webkit.org/show_bug.cgi?id=95839
1459
1460         Reviewed by Adam Barth.
1461
1462         In addition to adding the new methods to MockWebRTCPeerConnectionHandler I cleaned
1463         up the task definitions a bit. There was no need to have them in the header file.
1464
1465         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
1466         (RTCSessionDescriptionRequestSuccededTask):
1467         (RTCSessionDescriptionRequestSuccededTask::RTCSessionDescriptionRequestSuccededTask):
1468         (RTCSessionDescriptionRequestFailedTask):
1469         (RTCSessionDescriptionRequestFailedTask::RTCSessionDescriptionRequestFailedTask):
1470         (RTCVoidRequestTask):
1471         (RTCVoidRequestTask::RTCVoidRequestTask):
1472         (MockWebRTCPeerConnectionHandler::createOffer):
1473         (MockWebRTCPeerConnectionHandler::setLocalDescription):
1474         (MockWebRTCPeerConnectionHandler::setRemoteDescription):
1475         (MockWebRTCPeerConnectionHandler::localDescription):
1476         (MockWebRTCPeerConnectionHandler::remoteDescription):
1477         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
1478         (MockWebRTCPeerConnectionHandler):
1479
1480 2012-09-06  Simon Hausmann  <simon.hausmann@nokia.com>
1481
1482         [Qt] Fix build on Mac OS X 10.6
1483         https://bugs.webkit.org/show_bug.cgi?id=95200
1484
1485         Reviewed by Tor Arne Vestbø.
1486
1487         The build on Mac OS X 10.6 is broken when using QuickTime. Enable it only for
1488         10.7 and newer.
1489
1490         * qmake/mkspecs/features/features.prf:
1491
1492 2012-09-06  Christophe Dumez  <christophe.dumez@intel.com>
1493
1494         [WKTR] addSelectionKey should be mapped to "Control" modifiers on other OSes than Mac
1495         https://bugs.webkit.org/show_bug.cgi?id=95973
1496
1497         Reviewed by Kenneth Rohde Christiansen.
1498
1499         Correctly map "addSelectionKey" to "Control" modifier
1500         in WKTR's EventSender on other OSes than Mac OS X,
1501         instead of mapping it to the "Meta" modifier.
1502
1503         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
1504         (WTR::parseModifier):
1505
1506 2012-09-06  Balazs Kelemen  <kbalazs@webkit.org>
1507
1508         [nrwt] Harden against empty stdout/stderr after stopping a crashed/timed out driver
1509         https://bugs.webkit.org/show_bug.cgi?id=95710
1510
1511         Reviewed by Dirk Pranke.
1512
1513         The stdout / stderr output can be None so make sure
1514         we won't raise an exception in this case.
1515
1516         * Scripts/webkitpy/layout_tests/port/driver.py:
1517         (Driver.run_test):
1518
1519 2012-09-06  Christophe Dumez  <christophe.dumez@intel.com>
1520
1521         [EFL] Sanitize URLs in EWebLauncher / Minibrowser
1522         https://bugs.webkit.org/show_bug.cgi?id=95849
1523
1524         Reviewed by Kenneth Rohde Christiansen.
1525
1526         Sanitize URLs passed by the user as program
1527         argument or typed in the URL bar in order to
1528         be more tolerant.
1529
1530         * EWebLauncher/CMakeLists.txt:
1531         * EWebLauncher/main.c:
1532         (main):
1533         * EWebLauncher/url_bar.c:
1534         (_url_bar_url_get_with_protocol):
1535         (on_urlbar_key_down):
1536         * EWebLauncher/url_utils.c: Added.
1537         (has_scheme):
1538         (url_from_user_input): Add utility function to construct a
1539         URL from user input.
1540         * EWebLauncher/url_utils.h: Added.
1541         * MiniBrowser/efl/CMakeLists.txt:
1542         * MiniBrowser/efl/main.c:
1543         (main):
1544
1545 2012-09-06  Simon Hausmann  <simon.hausmann@nokia.com>
1546
1547         [Qt] Various small Windows / MSVC build fixes
1548         https://bugs.webkit.org/show_bug.cgi?id=95934
1549
1550         Reviewed by Kenneth Rohde Christiansen.
1551
1552         setFocus takes a boolean as argument: Don't try to implicitly convert
1553         an enum to a boolean, MSVC doesn't like that. The intent here is to
1554         enable focus, so just pass true.
1555
1556         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
1557         (WTR::PlatformWebView::focus):
1558
1559 2012-09-06  Simon Hausmann  <simon.hausmann@nokia.com>
1560
1561         [Qt] Fix build with ANGLE on platforms that provide EGL/GLESv2
1562         https://bugs.webkit.org/show_bug.cgi?id=95965
1563
1564         Reviewed by Tor Arne Vestbø.
1565
1566         Add ANGLE as a static library module to the Qt WebKit build system.
1567
1568         * qmake/mkspecs/features/default_post.prf:
1569         * qmake/mkspecs/modules/angle.prf: Added.
1570
1571 2012-09-06  Andras Becsi  <andras.becsi@nokia.com>
1572
1573         [WK2] Make [Qt]ViewportHandler cross platform
1574         https://bugs.webkit.org/show_bug.cgi?id=91257
1575
1576         Reviewed by Kenneth Rohde Christiansen.
1577
1578         Remove initial scale from viewport info item since it is
1579         set to -1 after applied thus it does not hold any valuable
1580         information.
1581
1582         * MiniBrowser/qt/qml/ViewportInfoItem.qml:
1583
1584 2012-09-05  Tony Chang  <tony@chromium.org>
1585
1586         [chromium] Remove dependency on webkit_user_agent
1587         https://bugs.webkit.org/show_bug.cgi?id=95886
1588
1589         Reviewed by Adam Barth.
1590
1591         The code isn't used by DumpRenderTree.
1592
1593         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
1594
1595 2012-09-05  Rick Byers  <rbyers@chromium.org>
1596
1597         Update chromium DumpRenderTree for new WebGestureEvent fields
1598         https://bugs.webkit.org/show_bug.cgi?id=95573
1599
1600         Reviewed by James Robinson.
1601
1602         Update EventSender to write gesture details into the appropriate
1603         type-specific WebGestureEvent fields.
1604         * DumpRenderTree/chromium/TestRunner/EventSender.cpp:
1605         (EventSender::gestureEvent):
1606         (EventSender::gestureFlingStart):
1607
1608 2012-09-05  Dirk Pranke  <dpranke@chromium.org>
1609
1610         NRWT lies about the number of tests that it's skipping when used with --repeat-each
1611         https://bugs.webkit.org/show_bug.cgi?id=95789
1612
1613         Reviewed by Ojan Vafai.
1614
1615         Fix miscalculation in log message.
1616
1617         * Scripts/webkitpy/layout_tests/views/printing.py:
1618         (Printer.print_found):
1619
1620 2012-09-05  Kenichi Ishibashi  <bashi@chromium.org>
1621
1622         Unreviewed, rolling out r127612, r127660, and r127664.
1623         http://trac.webkit.org/changeset/127612
1624         http://trac.webkit.org/changeset/127660
1625         http://trac.webkit.org/changeset/127664
1626         https://bugs.webkit.org/show_bug.cgi?id=95920
1627
1628         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
1629         (MockWebRTCPeerConnectionHandler::SuccessCallbackTask::SuccessCallbackTask):
1630         (MockWebRTCPeerConnectionHandler::SuccessCallbackTask::runIfValid):
1631         (MockWebRTCPeerConnectionHandler::FailureCallbackTask::FailureCallbackTask):
1632         (MockWebRTCPeerConnectionHandler::FailureCallbackTask::runIfValid):
1633         (MockWebRTCPeerConnectionHandler::createOffer):
1634         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
1635         (MockWebRTCPeerConnectionHandler):
1636         (SuccessCallbackTask):
1637         (FailureCallbackTask):
1638
1639 2012-09-05  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
1640
1641         Ignore-metrics includes scroll offsets.
1642         https://bugs.webkit.org/show_bug.cgi?id=95845
1643
1644         Reviewed by Dirk Pranke.
1645
1646         Strip scroll-offsets the same way scroll-height and scroll-width is stripped.
1647
1648         * Scripts/old-run-webkit-tests:
1649         (stripMetrics):
1650         * Scripts/webkitpy/layout_tests/port/driver.py:
1651         (DriverOutput):
1652
1653 2012-09-05  Zan Dobersek  <zandobersek@gmail.com>
1654
1655         Unreviewed test fix.
1656
1657         Including a plugin source file into the build that was added in r127595.
1658
1659         * GNUmakefile.am:
1660
1661 2012-09-05  Tommy Widenflycht  <tommyw@google.com>
1662
1663         MediaStream API: Add the local and remote description functionality to RTCPeerConnection
1664         https://bugs.webkit.org/show_bug.cgi?id=95839
1665
1666         Reviewed by Adam Barth.
1667
1668         In addition to adding the new methods to MockWebRTCPeerConnectionHandler I cleaned
1669         up the task definitions a bit. There was no need to have them in the header file.
1670
1671         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
1672         (RTCSessionDescriptionRequestSuccededTask):
1673         (RTCSessionDescriptionRequestSuccededTask::RTCSessionDescriptionRequestSuccededTask):
1674         (RTCSessionDescriptionRequestFailedTask):
1675         (RTCSessionDescriptionRequestFailedTask::RTCSessionDescriptionRequestFailedTask):
1676         (RTCVoidRequestTask):
1677         (RTCVoidRequestTask::RTCVoidRequestTask):
1678         (MockWebRTCPeerConnectionHandler::createOffer):
1679         (MockWebRTCPeerConnectionHandler::setLocalDescription):
1680         (MockWebRTCPeerConnectionHandler::setRemoteDescription):
1681         (MockWebRTCPeerConnectionHandler::localDescription):
1682         (MockWebRTCPeerConnectionHandler::remoteDescription):
1683         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
1684         (MockWebRTCPeerConnectionHandler):
1685
1686 2012-09-05  Kangil Han  <kangil.han@samsung.com>
1687
1688         [EFL] Remove '+=' usage in String
1689         https://bugs.webkit.org/show_bug.cgi?id=95808
1690
1691         Reviewed by Benjamin Poulain.
1692
1693         Replaced use of String operator '+=' with StringBuilder.append*() in order to adopt efficient String concatenation.
1694
1695         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1696         (descriptionSuitableForTestResult):
1697
1698 2012-09-05  Christophe Dumez  <christophe.dumez@intel.com>
1699
1700         [EFL][WK2] Provide implementation for TestRunner::pathToLocalResource()
1701         https://bugs.webkit.org/show_bug.cgi?id=95842
1702
1703         Reviewed by Kenneth Rohde Christiansen.
1704
1705         Provide proper implementation for TestRunner::pathToLocalResource()
1706         in EFL WKTR, identical to the one for DumpRenderTree.
1707
1708         Map /tmp to ${DUMPRENDERTREE_TEMP} environment variable and
1709         /tmp/LayoutTests to ${LOCAL_RESOURCE_ROOT} so that local resources
1710         are found my WebKitTestRunner.
1711
1712         * WebKitTestRunner/InjectedBundle/efl/TestRunnerEfl.cpp:
1713         (WTR::TestRunner::pathToLocalResource):
1714
1715 2012-09-05  Brady Eidson  <beidson@apple.com>
1716
1717         Frequent crashes in PluginView::scriptObject under runtimeObjectCustomGetOwnPropertySlot
1718         <rdar://problem/12142226> and https://bugs.webkit.org/show_bug.cgi?id=95026
1719
1720         Reviewed by Andy Estes.
1721
1722         Add a plug-in that always fails to initialize:
1723         * DumpRenderTree/TestNetscapePlugIn/Tests/NPPNewFails.cpp: Added.
1724         (NPPNewFails):
1725         (NPPNewFails::NPPNewFails):
1726         (NPPNewFails::NPP_New):
1727
1728         Add it to all the project files:
1729         * DumpRenderTree/DumpRenderTree.gypi:
1730         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
1731         * DumpRenderTree/TestNetscapePlugIn/CMakeLists.txt:
1732         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
1733         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
1734
1735 2012-09-05  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1736
1737         [Qt] Fix makefile race condition between derived sources and target
1738
1739         r121966 removed too much of the old qmake logic, including the rule
1740         that ensured that qmake_all would not trigger dual runs of qmake --
1741         one of the derived sources -qmake-all rule, and one for making the
1742         derived sources.
1743
1744         We now follow a similar pattern as before, by introducing a custom
1745         target that explicitly depends on the derived sources' qmake-all before
1746         doing a manual run of make. That ensures a single qmake run for the derived
1747         sources.
1748
1749         https://bugs.webkit.org/show_bug.cgi?id=93847
1750
1751         Reviewed by Csaba Osztrogonác.
1752
1753         * qmake/mkspecs/features/functions.prf:
1754
1755 2012-09-05  Christophe Dumez  <christophe.dumez@intel.com>
1756
1757         WKTR doesn't implement dumpWillCacheResponse()
1758         https://bugs.webkit.org/show_bug.cgi?id=94607
1759
1760         Reviewed by Kenneth Rohde Christiansen.
1761
1762         Define shouldCacheResponse() callback in resource
1763         load client and provide implementation for
1764         dumpWillCacheResponse() in WebKitTestRunner.
1765
1766         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
1767         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1768         (WTR::InjectedBundlePage::InjectedBundlePage):
1769         (WTR::InjectedBundlePage::shouldCacheResponse):
1770         (WTR):
1771         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
1772         (InjectedBundlePage):
1773         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
1774         (WTR::TestRunner::TestRunner):
1775         * WebKitTestRunner/InjectedBundle/TestRunner.h:
1776         (WTR::TestRunner::dumpWillCacheResponse):
1777         (WTR::TestRunner::shouldDumpWillCacheResponse):
1778         (TestRunner):
1779
1780 2012-09-05  Kaustubh Atrawalkar  <kaustubh@motorola.com>
1781
1782         [DRT] LTC:: Move printing related APIs from LayoutTestController to Internals
1783         https://bugs.webkit.org/show_bug.cgi?id=92735
1784
1785         Reviewed by Hajime Morita.
1786
1787         Move numberOfPages, pageProperty & pageSizeAndMarginsInPixels in Internals and remove duplicated code from DumprenderTree & WebkitTestRunner.
1788
1789         * DumpRenderTree/TestRunner.cpp:
1790         (TestRunner::staticFunctions):
1791         * DumpRenderTree/TestRunner.h:
1792         (TestRunner):
1793         * DumpRenderTree/blackberry/TestRunnerBlackBerry.cpp:
1794         (TestRunner::addUserStyleSheet):
1795         * DumpRenderTree/chromium/DRTTestRunner.cpp:
1796         (DRTTestRunner::DRTTestRunner):
1797         (DRTTestRunner::hasCustomPageSizeStyle):
1798         * DumpRenderTree/chromium/DRTTestRunner.h:
1799         (DRTTestRunner):
1800         * DumpRenderTree/efl/TestRunnerEfl.cpp:
1801         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
1802         * DumpRenderTree/mac/TestRunnerMac.mm:
1803         * DumpRenderTree/qt/TestRunnerQt.cpp:
1804         (TestRunner::evaluateScriptInIsolatedWorld):
1805         * DumpRenderTree/qt/TestRunnerQt.h:
1806         (TestRunner):
1807         * DumpRenderTree/win/TestRunnerWin.cpp:
1808         * DumpRenderTree/wx/TestRunnerWx.cpp:
1809         (TestRunner::abortModal):
1810         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
1811         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
1812         (WTR):
1813         * WebKitTestRunner/InjectedBundle/TestRunner.h:
1814         (TestRunner):
1815
1816 2012-09-05  Simon Hausmann  <simon.hausmann@nokia.com>
1817
1818         [Qt] Automate the use of AllInOne files in the Qt build
1819         https://bugs.webkit.org/show_bug.cgi?id=95837
1820
1821         Reviewed by Tor Arne Vestbø.
1822
1823         New feature file that implements the all-in-one file automation by
1824         adding the requested all-in-one files to SOURCES and removing the files
1825         it includes from SOURCES.
1826
1827         * qmake/mkspecs/features/use_all_in_one_files.prf: Added.
1828
1829 2012-09-04  Tim Horton  <timothy_horton@apple.com>
1830
1831         Have NRWT log more output when crash logs are missing
1832         https://bugs.webkit.org/show_bug.cgi?id=95814
1833
1834         Reviewed by Dan Bernstein.
1835
1836         Log the stderr and stdout of the test, in order to provide slightly more
1837         context when the WebProcess times out while running tests.
1838
1839         * Scripts/webkitpy/layout_tests/port/driver.py:
1840         (Driver.run_test):
1841
1842 2012-09-04  Xianzhu Wang  <wangxianzhu@chromium.org>
1843
1844         [Chromium-Android] Include device serial number in ChromiumAndroidDriver log messages
1845         https://bugs.webkit.org/show_bug.cgi?id=95770
1846
1847         Reviewed by Adam Barth.
1848
1849         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1850         (ChromiumAndroidDriver.__init__):
1851         (ChromiumAndroidDriver._log_error): Added to replace original _log.error().
1852         (ChromiumAndroidDriver._log_debug): Added to replace original _log.debug().
1853         (ChromiumAndroidDriver._abort): Added to replace original raise AssertionError().
1854         (ChromiumAndroidDriver._push_executable):
1855         (ChromiumAndroidDriver._push_fonts):
1856         (ChromiumAndroidDriver._push_test_resources):
1857         (ChromiumAndroidDriver._run_adb_command):
1858         (ChromiumAndroidDriver._get_last_stacktrace):
1859         (ChromiumAndroidDriver._get_crash_log):
1860         (ChromiumAndroidDriver._start):
1861         (ChromiumAndroidDriver._start_once):
1862         (ChromiumAndroidDriver._start_once.deadlock_detector):
1863         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
1864         (ChromiumAndroidDriverTest.test_get_crash_log):
1865
1866 2012-09-04  Satish Sampath  <satish@chromium.org>
1867
1868         Run git commands within the checkout directory
1869         https://bugs.webkit.org/show_bug.cgi?id=95742
1870
1871         Reviewed by Eric Seidel.
1872
1873         git commands run from scm/git.py are sometimes run with cwd set to the
1874         checkout_root and other times with whatever directory was the default cwd.
1875         Instead all these should run with the checkout root set as the cwd
1876         (except where we explicitly need a different path)
1877
1878         * Scripts/webkitpy/common/checkout/scm/git.py:
1879         (Git._run_git):
1880         (Git.find_checkout_root):
1881         (Git.discard_local_commits):
1882         (Git.local_commits):
1883         (Git.working_directory_is_clean):
1884         (Git.clean_working_directory):
1885         (Git.add_list):
1886         (Git.delete_list):
1887         (Git.exists):
1888         (Git._current_branch):
1889         (Git._changes_files_for_commit):
1890         (Git.revisions_changing_file):
1891         (Git.svn_revision):
1892         (Git._run_git_svn_find_rev):
1893         (Git.contents_at_revision):
1894         (Git.diff_for_file):
1895         (Git.show_head):
1896         (Git.committer_email_for_revision):
1897         (Git.apply_reverse_diff):
1898         (Git.revert_files):
1899         (Git.commit_with_message):
1900         (Git._commit_on_branch):
1901         (Git.svn_commit_log):
1902         (Git.last_svn_commit_log):
1903         (Git.svn_blame):
1904         (Git._branch_ref_exists):
1905         (Git.delete_branch):
1906         (Git.remote_merge_base):
1907         (Git.commit_locally_with_message):
1908         (Git.push_local_commits_to_server):
1909         (Git.commit_ids_from_commitish_arguments):
1910         (Git.commit_message_for_local_commit):
1911         (Git.files_changed_summary_for_commit):
1912
1913 2012-09-04  Tommy Widenflycht  <tommyw@google.com>
1914
1915         MediaStream API: Add the async createOffer functionality to RTCPeerConnection
1916         https://bugs.webkit.org/show_bug.cgi?id=95734
1917
1918         Reviewed by Adam Barth.
1919
1920         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
1921         (MockWebRTCPeerConnectionHandler::SuccessCallbackTask::SuccessCallbackTask):
1922         (MockWebRTCPeerConnectionHandler::SuccessCallbackTask::runIfValid):
1923         (MockWebRTCPeerConnectionHandler::FailureCallbackTask::FailureCallbackTask):
1924         (MockWebRTCPeerConnectionHandler::FailureCallbackTask::runIfValid):
1925         (MockWebRTCPeerConnectionHandler::createOffer):
1926         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
1927         (MockWebRTCPeerConnectionHandler):
1928         (MockWebRTCPeerConnectionHandler::taskList):
1929         (SuccessCallbackTask):
1930         (FailureCallbackTask):
1931
1932 2012-09-04  Xianzhu Wang  <wangxianzhu@chromium.org>
1933
1934         [Chromium-Android] Output time and thread info in crash log
1935         https://bugs.webkit.org/show_bug.cgi?id=95768
1936
1937         Reviewed by Adam Barth.
1938
1939         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1940         (ChromiumAndroidDriver._get_logcat): Add '-v threadtime' to adb logcat command line to get time and thread info.
1941
1942 2012-09-04  Tommy Widenflycht  <tommyw@google.com>
1943
1944         MediaStream API: Change the MediaStreamTrackList track added/removed signaling
1945         https://bugs.webkit.org/show_bug.cgi?id=95721
1946
1947         Reviewed by Adam Barth.
1948
1949         * DumpRenderTree/chromium/MockWebMediaStreamCenter.cpp:
1950         (MockWebMediaStreamCenter::didAddMediaStreamTrack):
1951         (MockWebMediaStreamCenter::didRemoveMediaStreamTrack):
1952         * DumpRenderTree/chromium/MockWebMediaStreamCenter.h:
1953         (MockWebMediaStreamCenter):
1954
1955 2012-09-04  Tony Chang  <tony@chromium.org>
1956
1957         Elide test names when running new-run-webkit-tests
1958         https://bugs.webkit.org/show_bug.cgi?id=95626
1959
1960         Reviewed by Dirk Pranke.
1961
1962         When the test name exceeds the number of terminal columns, it leaves
1963         text on the screen. Avoid this by eliding the test name when running
1964         tests. This is similar to what ninja does when compiling.
1965
1966         * Scripts/webkitpy/layout_tests/views/metered_stream.py:
1967         (MeteredStream):
1968         (MeteredStream.number_of_columns): Helper method to get number of
1969         terminal columns. Won't work on Win32 python, but meh.
1970         * Scripts/webkitpy/layout_tests/views/printing.py:
1971         (Printer._test_status_line): New method for eliding the text. Doesn't use \u2026
1972         because stderr appears to be ascii only.
1973         (Printer.print_started_test): Route through _test_status_line.
1974         (Printer.print_finished_test): Route through _test_status_line.
1975         (Printer._print_test_trace): Route through _test_status_line.
1976         (Printer._print_baseline): Route through _test_status_line.
1977         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
1978         (test_test_status_line): Test for _test_status_line.
1979
1980 2012-09-04  Martin Robinson  <mrobinson@igalia.com>
1981
1982         [GTK] Print API missing documentation when generating gtkdoc
1983         https://bugs.webkit.org/show_bug.cgi?id=95703
1984
1985         Reviewed by Carlos Garcia Campos.
1986
1987         Print out API missing documentation when generating gtkdoc. This makes it
1988         easier to fix documentation errors.
1989
1990         * gtk/generate-gtkdoc:
1991         (print_missing_api): Added.
1992         (generate_doc): Inline the rebase step, since it now reports an error
1993         when you try to rebase without generating documentation first.
1994         * gtk/gtkdoc.py:
1995         (GTKDoc.rebase_installed_docs): Raise an error when rebasing without
1996         generating documentation first to enable a small cleanup at the caller.
1997         (GTKDoc.api_missing_documentation): Added.
1998
1999 2012-09-04  Vincent Scheib  <scheib@chromium.org>
2000
2001         [Chromium] Allow asyncronous response of pointer lock requests in layout tests.
2002         https://bugs.webkit.org/show_bug.cgi?id=95628
2003
2004         Reviewed by James Robinson.
2005
2006         Future tests need the ability to respond asynchronously that
2007         pointer lock requests have been accepted or rejected. This patch
2008         adds the following:
2009             testRunner.setPointerLockWillRespondAsynchronously();
2010             testRunner.didAcquirePointerLock();
2011             testRunner.didNotAcquirePointerLock();
2012
2013         Tested in existing test:
2014             pointerlockchange-pointerlockerror-events.html
2015
2016         * DumpRenderTree/chromium/DRTTestRunner.cpp:
2017         (DRTTestRunner::DRTTestRunner):
2018         (DRTTestRunner::didAcquirePointerLock):
2019         (DRTTestRunner::didNotAcquirePointerLock):
2020         (DRTTestRunner::setPointerLockWillRespondAsynchronously):
2021         * DumpRenderTree/chromium/DRTTestRunner.h:
2022         (DRTTestRunner):
2023         * DumpRenderTree/chromium/WebViewHost.cpp:
2024         (WebViewHost::requestPointerLock):
2025         (WebViewHost::didAcquirePointerLock):
2026         (WebViewHost::didNotAcquirePointerLock):
2027         * DumpRenderTree/chromium/WebViewHost.h:
2028         (WebViewHost):
2029         (WebViewHost::setPointerLockWillRespondAsynchronously):
2030
2031 2012-09-04  Ojan Vafai  <ojan@chromium.org>
2032
2033         Move content shell bots to their own builder group
2034         https://bugs.webkit.org/show_bug.cgi?id=95765
2035
2036         Reviewed by Adam Barth.
2037
2038         In their current state of instability, they add too much noise. Once they get more stable,
2039         we can add them back to the regular group or just have them replace the other bots.
2040
2041         Also, fix the filter for Apple Windows bots since some of them are now "Apple Win ...".
2042
2043         * TestResultServer/static-dashboards/builders.js:
2044         (isWebkitTestRunner):
2045         (isChromiumContentShellTestRunner):
2046         (isChromiumWebkitTipOfTreeTestRunner):
2047         (loadBuildersList):
2048
2049 2012-09-03  Ryuan Choi  <ryuan.choi@samsung.com>
2050
2051         [EFL] Change the focus when the web view is clicked in MiniBrowser and EWebLauncher
2052         https://bugs.webkit.org/show_bug.cgi?id=95689
2053
2054         Reviewed by Kenneth Rohde Christiansen.
2055
2056         In Minibrowser & EWebLauncher, the web view loses the focus when url bar is clicked.
2057         So, there are bugs that key events are always generated in url bar although
2058         the user clicked the web view.
2059
2060         This patch recovers the focus of the web view when left button is pressed and
2061         adds for MiniBrowser to toggle the focus when middle button is pressed
2062         like EWebLauncher already did.
2063
2064         * EWebLauncher/main.c:
2065         (on_mouse_down):
2066         * MiniBrowser/efl/main.c:
2067         (on_mouse_down):
2068         (browserCreate):
2069
2070 2012-09-03  Tao Bai  <michaelbai@chromium.org>
2071
2072         Added ANDROID_GDBSERVER definition for Ant
2073         https://bugs.webkit.org/show_bug.cgi?id=95265
2074
2075         Reviewed by Adam Barth.
2076
2077         ANDROID_GDBSERVER will replace ANDROID_TOOLCHAIN, because the ANDROID_TOOLCHAIN
2078         is only used for finding the gdbserver.
2079
2080         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2081         * TestWebKitAPI/TestWebKitAPI.gyp/TestWebKitAPI.gyp:
2082
2083 2012-09-03  Tommy Widenflycht  <tommyw@google.com>
2084
2085         MediaStream API: Add Ice-related functionality to RTCPeerConnection
2086         https://bugs.webkit.org/show_bug.cgi?id=95565
2087
2088         Reviewed by Adam Barth.
2089
2090         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
2091         (MockWebRTCPeerConnectionHandler::MockWebRTCPeerConnectionHandler):
2092         (MockWebRTCPeerConnectionHandler::updateICE):
2093         (MockWebRTCPeerConnectionHandler::addICECandidate):
2094         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
2095         (MockWebRTCPeerConnectionHandler):
2096
2097 2012-09-03  Christophe Dumez  <christophe.dumez@intel.com>
2098
2099         [EFL] Fix resizing issues related to URL bar in Minibrowser / EWebLauncher
2100         https://bugs.webkit.org/show_bug.cgi?id=95670
2101
2102         Reviewed by Kenneth Rohde Christiansen.
2103
2104         Fix layout issues related to the URL bar in Minibrowser
2105         and EWebLauncher when the window is resized.
2106
2107         The URL bar was not resized accordingly and the Web
2108         page was moved behing the URL bar, hiding part of it.
2109
2110         * EWebLauncher/main.c:
2111         (on_ecore_evas_resize):
2112         (closeWindow):
2113         * EWebLauncher/url_bar.c:
2114         (url_bar_width_set):
2115         (url_bar_add):
2116         * EWebLauncher/url_bar.h:
2117         * MiniBrowser/efl/main.c:
2118         (on_ecore_evas_resize):
2119         (main):
2120
2121 2012-09-01  Xianzhu Wang  <wangxianzhu@chromium.org>
2122
2123         [Chromium-Android] Remove hard-coded skipped directories in chromium_android.py
2124         https://bugs.webkit.org/show_bug.cgi?id=95605
2125
2126         Reviewed by Dirk Pranke.
2127
2128         Now skipping rules are put in LayoutTests/platform/chromium-android/TestExpectations.
2129
2130         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2131         (ChromiumAndroidPort.check_sys_deps):
2132         (ChromiumAndroidPort.expectations_files):
2133         (ChromiumAndroidPort.clean_up_test_run):
2134
2135 2012-09-01  Joanmarie Diggs  <jdiggs@igalia.com>
2136
2137         [Gtk] Incorrect/unexpected characters in the text of certain accessibles
2138         https://bugs.webkit.org/show_bug.cgi?id=95180
2139
2140         Reviewed by Chris Fleizach.
2141
2142         The bug that was fixed stood in the way of fully implementing stringValue().
2143         Testing that the bug is fixed requires stringValue() to be fully implemented
2144         and object replacement characters to be printable.
2145
2146         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
2147         (replaceCharactersForResults): New method which turns object replacement
2148         characters into "<obj>" so that the characters can be properly shown in
2149         Layout Test results. Also turns "\n" into "<\\n>" so that printing the
2150         accessible text of a single object in the accessible tree doesn't mess up
2151         the readibility of the results.
2152         (AccessibilityUIElement::stringValue): Remove the code that immediately
2153         returned upon encountering an object of ATK_ROLE_PANEL and call the new
2154         replaceCharactersForResults() prior to returning the accessible string
2155         value.
2156
2157 2012-09-01  Tommy Widenflycht  <tommyw@google.com>
2158
2159         MediaStream API: Add MediaStream management to RTCPeerConnection
2160         https://bugs.webkit.org/show_bug.cgi?id=95543
2161
2162         Reviewed by Adam Barth.
2163
2164         Extending MockWebRTCPeerConnectionHandler to handle addStream/removeStream.
2165
2166         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
2167         (MockWebRTCPeerConnectionHandler::MockWebRTCPeerConnectionHandler):
2168         (MockWebRTCPeerConnectionHandler::addStream):
2169         (MockWebRTCPeerConnectionHandler::removeStream):
2170         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
2171         (MockWebRTCPeerConnectionHandler):
2172
2173 2012-08-31  Tony Chang  <tony@chromium.org>
2174
2175         Remove ENABLE_CSS3_FLEXBOX compile time flag
2176         https://bugs.webkit.org/show_bug.cgi?id=95382
2177
2178         Reviewed by Ojan Vafai.
2179
2180         Everyone is already enabling this by default and the spec has stablized.
2181
2182         * Scripts/webkitperl/FeatureList.pm:
2183         * qmake/mkspecs/features/features.pri:
2184
2185 2012-08-31  Pratik Solanki  <psolanki@apple.com>
2186
2187         objc_msgSend and IMP should be cast appropriately before using
2188         https://bugs.webkit.org/show_bug.cgi?id=95242
2189
2190         Reviewed by Benjamin Poulain.
2191
2192         Fix for older compilers. Pass id as the return type to the template
2193         instead of relying on default type.
2194
2195         * DumpRenderTree/mac/DumpRenderTree.mm:
2196         (drt_NSFontManager_availableFontFamilies):
2197         * WebKitTestRunner/InjectedBundle/mac/ActivateFonts.mm:
2198         (WTR::wtr_NSFontManager_availableFontFamilies):
2199
2200 2012-08-31  Zan Dobersek  <zandobersek@gmail.com>
2201
2202         nrwt: use scm instead of calling svn directly to get the revision in json results generator
2203         https://bugs.webkit.org/show_bug.cgi?id=89616
2204
2205         Reviewed by Eric Seidel.
2206
2207         Add the new 'executive' optional function argument to the SCM.in_working_directory
2208         class method. This way we get to reuse an Executive instance which is necessary in testing
2209         when we're operating with a mock version of the object.
2210
2211         In JSONResultsGeneratorBase, it's now possible to get the current SVN revision by using
2212         the SCMDetector with the generator's FileSystem and Executive instances.
2213
2214         * Scripts/webkitpy/common/checkout/scm/detection.py:
2215         (SCMDetector.detect_scm_system):
2216         * Scripts/webkitpy/common/checkout/scm/detection_unittest.py:
2217         (SCMDetectorTest.test_detect_scm_system): Update the unit test to capture the mock
2218         Executive output and test it against expected output.
2219         * Scripts/webkitpy/common/checkout/scm/git.py:
2220         (Git.in_working_directory):
2221         * Scripts/webkitpy/common/checkout/scm/scm.py:
2222         (SCM.in_working_directory):
2223         * Scripts/webkitpy/common/checkout/scm/svn.py:
2224         (SVN.in_working_directory):
2225         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2226         (JSONResultsGeneratorBase.__init__):
2227         (JSONResultsGeneratorBase._get_svn_revision):
2228
2229 2012-08-31  Jon Lee  <jonlee@apple.com>
2230
2231         [Tests] Add basic tests to http/tests/notifications
2232         https://bugs.webkit.org/show_bug.cgi?id=95493
2233         <rdar://problem/12209303>
2234
2235         Reviewed by Alexey Proskuryakov.
2236
2237         Add permission request support for WTR.
2238
2239         * WebKitTestRunner/TestController.cpp:
2240         (WTR::TestController::initialize):
2241         (WTR::TestController::decidePolicyForNotificationPermissionRequest): Always allow.
2242         * WebKitTestRunner/TestController.h:
2243         (TestController):
2244
2245 2012-08-31  Jon Lee  <jonlee@apple.com>
2246
2247         [DRT] Make simulating a web click on a notification a queued task
2248         https://bugs.webkit.org/show_bug.cgi?id=95546
2249         <rdar://problem/12214170>
2250
2251         Reviewed by Alexey Proskuryakov.
2252
2253         Making the web click happen asynchronously better mimics user interaction with the platform.
2254
2255         * DumpRenderTree/TestRunner.h: Add a flag that determines whether we have a pending notification click.
2256         (TestRunner::hasPendingWebNotificationClick): Exposed so that dump() can check that the flag is
2257         not set.
2258         * DumpRenderTree/mac/DumpRenderTree.mm:
2259         (dump): Assert that the flag is not set.
2260         * DumpRenderTree/mac/TestRunnerMac.mm:
2261         (TestRunner::simulateWebNotificationClick): Set the flag, then add a task to click the
2262         notification. In the case where the block gets executed after the flag has been unset, we avoid
2263         performing the click.
2264
2265 2012-08-31  Balazs Kelemen  <kbalazs@webkit.org>
2266
2267         [Qt] ImageDiff output is not in the expected form if image dimensions differ
2268         https://bugs.webkit.org/show_bug.cgi?id=95575
2269
2270         Reviewed by Csaba Osztrogonác.
2271
2272         Output 100% difference in the case of differing image dimensions so
2273         the test harness knows that we won't produce any output further.
2274         This matches the behavior of other ports.
2275
2276         * DumpRenderTree/qt/ImageDiff.cpp:
2277         (main):
2278
2279 2012-08-31  Sheriff Bot  <webkit.review.bot@gmail.com>
2280
2281         Unreviewed, rolling out r127261.
2282         http://trac.webkit.org/changeset/127261
2283         https://bugs.webkit.org/show_bug.cgi?id=95572
2284
2285         This patch depends on other build fixes (see blockers of bug
2286         95466) which should land first (Requested by milian on
2287         #webkit).
2288
2289         * qmake/mkspecs/features/features.prf:
2290
2291 2012-08-31  Milian Wolff  <milian.wolff@kdab.com>
2292
2293         [Qt] Support WebKit2/QML on QNX
2294         https://bugs.webkit.org/show_bug.cgi?id=95466
2295
2296         Reviewed by Noam Rosenthal.
2297
2298         Enable USE(3D_GRAPHICS) on QNX again, paired with the other bug fixes
2299         this should finally work now.
2300
2301         * qmake/mkspecs/features/features.prf:
2302
2303 2012-08-30  Xianzhu Wang  <wangxianzhu@chromium.org>
2304
2305         [Chromium-Android] Skip compositing/webgl and platform/chromium/virtual/threaded/compositing/webgl tests
2306         https://bugs.webkit.org/show_bug.cgi?id=95400
2307
2308         Reviewed by Dirk Pranke.
2309
2310         WebGL is not enabled and should be skipped on Android (http://crbug.com/135877).
2311
2312         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2313         (ChromiumAndroidPort.skipped_layout_tests):
2314
2315 2012-08-30  Gustavo Noronha Silva  <gns@gnome.org>
2316
2317         [GTK] Tries to run empty string when calling generate-gtkdoc when not using jhbuild
2318         https://bugs.webkit.org/show_bug.cgi?id=95499
2319
2320         Unreviewed. One more fix to avoid trying to run the empty string.
2321
2322         * Scripts/webkitdirs.pm:
2323         (buildAutotoolsProject):
2324
2325 2012-08-30  Sheriff Bot  <webkit.review.bot@gmail.com>
2326
2327         Unreviewed, rolling out r127171.
2328         http://trac.webkit.org/changeset/127171
2329         https://bugs.webkit.org/show_bug.cgi?id=95505
2330
2331         testRunner does not need dumpWebNotificationCallbacks().
2332         (Requested by jonlee on #webkit).
2333
2334         * DumpRenderTree/TestRunner.cpp:
2335         (TestRunner::TestRunner):
2336         (TestRunner::staticFunctions):
2337         * DumpRenderTree/TestRunner.h:
2338         (TestRunner):
2339         * DumpRenderTree/mac/MockWebNotificationProvider.mm:
2340         (-[MockWebNotificationProvider webView:didShowNotification:]):
2341         (-[MockWebNotificationProvider webView:didClickNotification:]):
2342         (-[MockWebNotificationProvider webView:didCloseNotifications:]):
2343         * DumpRenderTree/mac/UIDelegate.mm:
2344         (-[UIDelegate webView:decidePolicyForNotificationRequestFromOrigin:listener:]):
2345
2346 2012-08-30  Pratik Solanki  <psolanki@apple.com>
2347
2348         objc_msgSend and IMP should be cast appropriately before using
2349         https://bugs.webkit.org/show_bug.cgi?id=95242
2350
2351         Reviewed by Benjamin Poulain.
2352
2353         Use wtfObjcMsgSend and wtfCallIMP templates which do appropriate casts
2354         to a function pointer with right types when calling objc_msgSend and an
2355         IMP method directly.
2356
2357         * DumpRenderTree/mac/DumpRenderTree.mm:
2358         (drt_NSFontManager_availableFontFamilies):
2359         * WebKitTestRunner/InjectedBundle/mac/ActivateFonts.mm:
2360         (WTR::wtr_NSFontManager_availableFontFamilies):
2361
2362 2012-08-30  Benjamin Poulain  <bpoulain@apple.com>
2363
2364         Replace JSC::UString by WTF::String
2365         https://bugs.webkit.org/show_bug.cgi?id=95271
2366
2367         Reviewed by Geoffrey Garen.
2368
2369         Get rid of UString.
2370
2371         * DumpRenderTree/efl/WorkQueueItemEfl.cpp:
2372         * gdb/webkit.py:
2373         (WTFStringPrinter.to_string):
2374         (JSCIdentifierPrinter.to_string):
2375         (JSCJSStringPrinter.to_string):
2376         (add_pretty_printers):
2377
2378 2012-08-30  Dirk Pranke  <dpranke@chromium.org>
2379
2380         nrwt should have TestExpectations everywhere we have Skipped files for apple ports
2381         https://bugs.webkit.org/show_bug.cgi?id=95495
2382
2383         Reviewed by Ojan Vafai.
2384
2385         This is a follow-on to bug 95370 that adds in support for
2386         platform/wk2/TestExpectations and
2387         platform/{mac,win}-$version/TestExpectations.
2388
2389         * Scripts/webkitpy/layout_tests/port/apple.py:
2390         (ApplePort.expectations_files):
2391         * Scripts/webkitpy/layout_tests/port/base.py:
2392         (Port._skipped_file_search_paths):
2393         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2394         (MacTest.test_expectations_files):
2395         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
2396         (MockDRTPortTest.make_port):
2397         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
2398         (WinPortTest.test_expectations_files):
2399
2400 2012-08-30  Dirk Pranke  <dpranke@chromium.org>
2401
2402         executive.run_in_parallel() hangs if given nothing to do
2403         https://bugs.webkit.org/show_bug.cgi?id=95387
2404
2405         Reviewed by Ojan Vafai.
2406
2407         Check to make sure that Executive.run_in_parallel() requires
2408         a non-empty list of commands to execute; passing an empty list
2409         seems surely like a programming error.
2410
2411         * Scripts/webkitpy/common/system/executive.py:
2412         (Executive.run_in_parallel):
2413         * Scripts/webkitpy/common/system/executive_unittest.py:
2414         (ExecutiveTest.test_run_in_parallel_assert_nonempty):
2415
2416 2012-08-30  Dirk Pranke  <dpranke@chromium.org>
2417
2418         decide what (and how) we should set the tolerance for ref test pixel compares and test for that
2419         https://bugs.webkit.org/show_bug.cgi?id=94746
2420
2421         Reviewed by Ojan Vafai.
2422
2423         Add an assertion to diff_image() to check that we are passing
2424         tolerance=0 explicitly when diffing ref test results.
2425
2426         * Scripts/webkitpy/layout_tests/port/test.py:
2427         (TestPort.diff_image):
2428
2429 2012-08-30  Dirk Pranke  <dpranke@chromium.org>
2430
2431         NRWT should look in mac-wk2 for a TestExpecations file
2432         https://bugs.webkit.org/show_bug.cgi?id=95370
2433
2434         Reviewed by Ojan Vafai.
2435
2436         Adds support for mac-wk2 and win-wk2 to expectations_files()
2437         for the apple mac and apple win ports.
2438         
2439         * Scripts/webkitpy/layout_tests/port/apple.py:
2440         (ApplePort.expectations_files):
2441         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2442         (MacTest.test_expectations_files):
2443         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
2444         (WinTest.test_expectations_files):
2445         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2446         (PortTestCase.test_expectations_files):
2447
2448 2012-08-30  Jon Lee  <jonlee@apple.com>
2449
2450         [Mac] Add testRunner.dumpWebNotificationCallbacks() to DRT
2451         https://bugs.webkit.org/show_bug.cgi?id=95232
2452         <rdar://problem/12190776>
2453
2454         Reviewed by Alexey Proskuryakov.
2455
2456         Add additional output when running test, if testRunner.dumpWebNotificationCallbacks() is called.
2457
2458         * DumpRenderTree/TestRunner.cpp:
2459         (TestRunner::TestRunner):
2460         (dumpWebNotificationCallbacksCallback):
2461         (TestRunner::staticFunctions):
2462         * DumpRenderTree/TestRunner.h:
2463         (TestRunner):
2464         (TestRunner::dumpWebNotificationCallbacks):
2465         (TestRunner::setDumpWebNotificationCallbacks):
2466         * DumpRenderTree/mac/MockWebNotificationProvider.mm:
2467         (-[WebNotification _drt_descriptionSuitableForTestResult]):
2468         (-[MockWebNotificationProvider webView:didShowNotification:]):
2469         (-[MockWebNotificationProvider webView:didClickNotification:]):
2470         (-[MockWebNotificationProvider webView:didCloseNotifications:]):
2471         * DumpRenderTree/mac/UIDelegate.mm:
2472         (-[UIDelegate webView:decidePolicyForNotificationRequestFromOrigin:listener:]):
2473
2474 2012-08-30  Tommy Widenflycht  <tommyw@google.com>
2475
2476         MediaStream API: Introduce MediaConstraints
2477         https://bugs.webkit.org/show_bug.cgi?id=95198
2478
2479         Reviewed by Adam Barth.
2480
2481         Extending the MockWebRTCPeerConnectionHandler with MediaConstraints functionality.
2482
2483         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
2484         (MockWebRTCPeerConnectionHandler::MockWebRTCPeerConnectionHandler):
2485         (isSupportedConstraint):
2486         (isValidConstraint):
2487         (MockWebRTCPeerConnectionHandler::initialize):
2488         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
2489         (MockWebRTCPeerConnectionHandler):
2490
2491 2012-08-30  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
2492
2493         [WK2][WTR] Add didReceiveServerRedirectForProvisionalLoadForFrame dumping
2494         https://bugs.webkit.org/show_bug.cgi?id=95454
2495
2496         Reviewed by Kenneth Rohde Christiansen.
2497
2498         Added didReceiveServerRedirectForProvisionalLoadForFrame dumping to WTR.
2499
2500         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2501         (WTR::InjectedBundlePage::didReceiveServerRedirectForProvisionalLoadForFrame):
2502
2503 2012-08-30  Leo Yang  <leoyang@rim.com>
2504
2505         Update my email address.
2506
2507         Not reviewed.
2508
2509         * Scripts/webkitpy/common/config/committers.py:
2510
2511 2012-08-30  Jochen Eisinger  <jochen@chromium.org>
2512
2513         [NRWT] Add support for recognizing arbitrary process names in crash lines
2514         https://bugs.webkit.org/show_bug.cgi?id=95435
2515
2516         Reviewed by Adam Barth.
2517
2518         When running layout tests in the chromium port using the content shell,
2519         we want to be able to report sub-process crashes as well.
2520
2521         * Scripts/webkitpy/layout_tests/port/driver.py:
2522         (Driver._check_for_driver_crash):
2523         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
2524         (DriverTest.test_check_for_driver_crash):
2525
2526 2012-08-29  Zan Dobersek  <zandobersek@gmail.com>
2527
2528         Random test-webkitpy failures on the buildbot
2529         https://bugs.webkit.org/show_bug.cgi?id=95096
2530
2531         Reviewed by Dirk Pranke.
2532
2533         Suppress occasional errors when running test-webkitpy on GTK builders
2534         by running these tests serially. The 64-bit Release builder is especially
2535         prone to these as it can run up to 24 tests in parallel.
2536
2537         Despite the tests not being run in parallel, the testing only lasts a handful
2538         of seconds more, so this is a worthy trade-off to avoid unnecessary
2539         false-alarm redness on the GTK builders.
2540
2541         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2542         (RunPythonTests.start):
2543
2544 2012-08-29  Benjamin Poulain  <bpoulain@apple.com>
2545
2546         REGRESSION(r126780): Crash using StringImpl::is8Bit before checking if there is an impl
2547         https://bugs.webkit.org/show_bug.cgi?id=95380
2548
2549         Reviewed by Michael Saboff.
2550
2551         * TestWebKitAPI/Tests/WTF/WTFString.cpp:
2552         Add very basic tests for String::ascii(). This covers the case of null strings that caused
2553         the crash.
2554
2555 2012-08-29  Dominic Mazzoni  <dmazzoni@google.com>
2556
2557         AX: Canvas should have a distinct role
2558         https://bugs.webkit.org/show_bug.cgi?id=95248
2559
2560         Reviewed by Chris Fleizach.
2561
2562         Add support for canvas accessibility role.
2563
2564         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
2565         (roleToString):
2566
2567 2012-08-29  José Dapena Paz  <jdapena@igalia.com>
2568
2569         [Gtk] Process Gtk 3.4 smooth scroll events properly
2570         https://bugs.webkit.org/show_bug.cgi?id=88070
2571
2572         Added continousMouseScrollBy support in WebKitTestRunner, and added
2573         implementation for gtk, and stub for Qt, mac and EFL.
2574
2575         Added layout tests support for smooth scroll in Gtk 3.4, and use smooth
2576         scroll for emulating multi-tick mouseScrollBy events.
2577
2578         Reviewed by Martin Robinson.
2579
2580         * DumpRenderTree/gtk/EventSender.cpp:
2581         (mouseScrollByCallback):
2582         (continuousMouseScrollByCallback):
2583         * WebKitTestRunner/EventSenderProxy.h:
2584         (EventSenderProxy):
2585         * WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
2586         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
2587         (WTR::EventSendingController::mouseScrollBy):
2588         (WTR):
2589         (WTR::EventSendingController::continuousMouseScrollBy):
2590         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
2591         (EventSendingController):
2592         * WebKitTestRunner/TestController.cpp:
2593         (WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
2594         * WebKitTestRunner/gtk/EventSenderProxyGtk.cpp:
2595         (WTR):
2596         (WTR::EventSenderProxy::mouseScrollBy):
2597         (WTR::EventSenderProxy::continuousMouseScrollBy):
2598         * WebKitTestRunner/efl/EventSenderProxyEfl.cpp:
2599         (WTR):
2600         (WTR::EventSenderProxy::continuousMouseScrollBy):
2601         * WebKitTestRunner/mac/EventSenderProxy.mm:
2602         (WTR::EventSenderProxy::continuousScrollBy):
2603         * WebKitTestRunner/qt/EventSenderProxyQt.cpp:
2604         (WTR::EventSenderProxy::continuousMouseScrollBy):
2605         (WTR):
2606
2607 2012-08-29  Alex Sakhartchouk  <alexst@chromium.org>
2608
2609         [chromium] Support high DIP pixel tests with DumpRenderTree
2610         https://bugs.webkit.org/show_bug.cgi?id=94935
2611
2612         Reviewed by James Robinson.
2613
2614         Adding device scale factor into the drt image dumping code
2615
2616         * DumpRenderTree/chromium/DRTTestRunner.cpp:
2617         (DRTTestRunner::reset):
2618         (DRTTestRunner::setBackingScaleFactor):
2619         * DumpRenderTree/chromium/WebViewHost.cpp:
2620         (WebViewHost::setDeviceScaleFactor):
2621         (WebViewHost::paintInvalidatedRegion):
2622         (WebViewHost::canvas):
2623         * DumpRenderTree/chromium/WebViewHost.h:
2624         (WebViewHost):
2625
2626 2012-08-29  Kevin Funk  <kevin.funk@kdab.com>
2627
2628         Fix check-webkit-style (and probably others) when WebKit is in a git submodule
2629         https://bugs.webkit.org/show_bug.cgi?id=95177
2630
2631         Reviewed by Dirk Pranke.
2632
2633         Find the real checkout root by using 'git rev-parse --show-toplevel' instead of '--git-dir'
2634
2635         * Scripts/webkitpy/common/checkout/scm/git.py:
2636
2637 2012-08-29  Dirk Pranke  <dpranke@chromium.org>
2638
2639         webkit-patch rebaseline-expectations wrongly touches other platforms' TestExpectations
2640         https://bugs.webkit.org/show_bug.cgi?id=95222
2641
2642         Reviewed by Adam Barth.
2643
2644         The code we had to remove REBASELINE lines from the
2645         TestExpectations files did not handle multiple files (in a
2646         cascade) correctly; we weren't limiting lines to a particular
2647         file correctly in without_rebaseline_modifier(). This patch
2648         fixes that and corrects the tests (which weren't correct and
2649         didn't cover things very well).
2650
2651         Note also that the webkit-patch rebaseline-commands are still
2652         using the include_overrides=False option in a couple places;
2653         this is never correct or needed at this point. This patch fixes
2654         the usage for webkit-patch rebaseline-expectations, but I still
2655         need to fix rebaseline-test-internal, which I will do in a
2656         separate patch (see bug 95268).
2657
2658         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2659         (TestExpectations.remove_rebaselined_tests.without_rebaseline_modifier):
2660         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
2661         (RebaseliningTest.assertRemove):
2662         (RebaseliningTest.test_remove):
2663         * Scripts/webkitpy/tool/commands/rebaseline.py:
2664         (RebaselineExpectations._update_expectations_files):
2665
2666 2012-08-29  Dirk Pranke  <dpranke@chromium.org>
2667
2668         webkit-patch rebaseline-expectations hangs
2669         https://bugs.webkit.org/show_bug.cgi?id=95243
2670
2671         Reviewed by Tony Chang.
2672
2673         Don't try to run no commands in parallel; report an error
2674         instead that we didn't find any tests to rebaseline.
2675
2676         Filed bug 95387 as well to make sure run_in_parallel() doesn't hang forever.
2677
2678         * Scripts/webkitpy/tool/commands/rebaseline.py:
2679         (RebaselineExpectations.execute):
2680         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
2681         (_assert_command):
2682         (test_rebaseline_expectations_noop):
2683
2684 2012-08-29  Brady Eidson  <beidson@apple.com>
2685
2686         REGRESSION: Not sending NPP_SetWindow is causing Flash to not throttle itself
2687         <rdar://problem/12133021> and https://bugs.webkit.org/show_bug.cgi?id=95274
2688
2689         Reviewed by Sam Weinig.
2690
2691         Add a test plug-in that calls back into the page with info on the NPWindow passed
2692         in to NPP_SetWindow.
2693
2694         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
2695         * DumpRenderTree/TestNetscapePlugIn/Tests/LogNPPSetWindow.cpp: Added.
2696         (LogNPPSetWindow):
2697         (LogNPPSetWindow::LogNPPSetWindow):
2698         (LogNPPSetWindow::NPP_SetWindow):
2699
2700 2012-08-29  Jon Lee  <jonlee@apple.com>
2701
2702         [Mac] Basic DRT support for web notifications
2703         https://bugs.webkit.org/show_bug.cgi?id=79492
2704         <rdar://problem/10357639>
2705
2706         Reviewed by Alexey Proskuryakov.
2707
2708         This patch implements the stubs for web notification support on DRT. Assume that when
2709         Notification.requestPermission() is called, the user always chooses to allow it. This can be easily
2710         overridden in layout tests with testRunner.denyWebNotificationPermission() when needed.
2711
2712         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Add MockWebNotificationProvider.
2713         * DumpRenderTree/mac/DumpRenderTree.mm:
2714         (createWebViewAndOffscreenWindow): Set MockWebNotificationProvider singleton as the provider for the
2715         web view.
2716         (resetWebViewToConsistentStateBeforeTesting): Make sure to reset the provider's state between each test.
2717
2718         * DumpRenderTree/mac/TestRunnerMac.mm:
2719         (TestRunner::grantWebNotificationPermission): Set permission to granted for specified origin.
2720         (TestRunner::denyWebNotificationPermission): Set permission to denied for specified origin.
2721         (TestRunner::removeAllWebNotificationPermissions):
2722         (TestRunner::simulateWebNotificationClick): Retrieve the notification ID from the JS notification object, and
2723         simulate the click. This is similar to how WTR performs the click.
2724         * DumpRenderTree/mac/UIDelegate.mm:
2725         (-[UIDelegate webView:decidePolicyForNotificationRequestFromOrigin:listener:]): Assume when asked, the
2726         user allows web notifications.
2727
2728         * DumpRenderTree/mac/MockWebNotificationProvider.h: Added. Maintains a list of registered web views, known permissions,
2729         a map of notification IDs to WebNotification instances, and a map of the WebViews from which the notifications were dispatched.
2730         * DumpRenderTree/mac/MockWebNotificationProvider.mm: Added.
2731         (+[MockWebNotificationProvider shared]): The provider is a singleton.
2732         (-[MockWebNotificationProvider init]): Instantiate the permissions map.
2733         (-[MockWebNotificationProvider registerWebView:]):
2734         (-[MockWebNotificationProvider unregisterWebView:]):
2735         (-[MockWebNotificationProvider showNotification:fromWebView:]): Add the notification to the maps. Tell the web view that the
2736         notification did show.
2737         (-[MockWebNotificationProvider cancelNotification:]): Tell the web view the notification did close.
2738         (-[MockWebNotificationProvider notificationDestroyed:]): Remove the notification from the maps. Here we don't make a callback.
2739         (-[MockWebNotificationProvider clearNotifications:]): Remove the specified notifications from the maps. Here we don't make a callback.
2740         (-[MockWebNotificationProvider webView:didShowNotification:]): Delegate callback. Dispatch the show event.
2741         (-[MockWebNotificationProvider webView:didClickNotification:]): Delegate callback. Dispatch the click event.
2742         (-[MockWebNotificationProvider webView:didCloseNotifications:]): Delegate callback. Dispatch the close event.
2743         (-[MockWebNotificationProvider simulateWebNotificationClick:]): Tell the web view we clicked the notification.
2744         (-[MockWebNotificationProvider policyForOrigin:]): Look in the permission dictionary.
2745         (-[MockWebNotificationProvider setWebNotificationOrigin:permission:]): Set the permission for the origin in the dictionary.
2746         (-[MockWebNotificationProvider removeAllWebNotificationPermissions]): Clear the permission map.
2747         (-[MockWebNotificationProvider reset]): Remove all mappings and permissions.
2748
2749 2012-08-29  Rob Buis  <rbuis@rim.com>
2750
2751         [BlackBerry] Adjust wtf include header style
2752         https://bugs.webkit.org/show_bug.cgi?id=95368
2753
2754         Reviewed by Yong Li.
2755
2756         Switch to #include <wtf/...> like the other ports.
2757
2758         * DumpRenderTree/blackberry/DumpRenderTree.cpp:
2759         * DumpRenderTree/blackberry/PNGImageEncoder.cpp:
2760         * DumpRenderTree/blackberry/PNGImageEncoder.h:
2761         * DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.cpp:
2762         * DumpRenderTree/blackberry/PixelDumpSupportBlackBerry.h:
2763         * DumpRenderTree/blackberry/WorkQueueItemBlackBerry.cpp:
2764
2765 2012-08-29  Roger Fong  <roger_fong@apple.com>
2766
2767         Update cygwin-downloader.zip to match modified downloader script.
2768         https://bugs.webkit.org/show_bug.cgi?id=76936
2769
2770         Rubber Stamped by Jon Honeycutt.
2771
2772         cygwin-downloader.py was modified in http://trac.webkit.org/changeset/126948.
2773         Updating zip to reflect changes.
2774
2775         * CygwinDownloader/cygwin-downloader.zip:
2776
2777 2012-08-29  Dominik Röttsches  <dominik.rottsches@intel.com>
2778
2779         Stylechecker warns about comparison to zero when comparing to 0.5
2780         https://bugs.webkit.org/show_bug.cgi?id=94913
2781
2782         Reviewed by Dirk Pranke.
2783
2784         According to the python documentation, \W is the character group with all
2785         non-alphanumeric characters, equivalent to [^a-zA-Z0-9_], which is equal to
2786         to [^\w]. We need to exclude the dot "." as well, so that floating point
2787         values do not trigger this warning. Adding a unit test that shows the problem.
2788
2789         * Scripts/webkitpy/style/checkers/cpp.py:
2790         (check_for_comparisons_to_zero): Modifying the regex to not get triggered by comparing to floats.
2791         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
2792         (WebKitStyleTest.test_null_false_zero): Unit test exposing this issue.
2793
2794 2012-08-29  Jon Lee  <jonlee@apple.com>
2795
2796         WTR build fixes.
2797
2798         * WebKitTestRunner/CMakeLists.txt:
2799         * WebKitTestRunner/GNUmakefile.am:
2800         * WebKitTestRunner/Target.pri:
2801         * WebKitTestRunner/WebNotificationProvider.cpp:
2802         (WTR::WebNotificationProvider::showWebNotification): Make ML happier.
2803         * WebKitTestRunner/win/WebKitTestRunner.vcproj:
2804
2805 2012-08-29  Jon Lee  <jonlee@apple.com>
2806
2807         [WK2] Basic WTR support for web notifications
2808         https://bugs.webkit.org/show_bug.cgi?id=95154
2809         <rdar://problem/12184492>
2810
2811         Reviewed by Alexey Proskuryakov.
2812
2813         Implement WTR support, without platform event output (bug 95233).
2814
2815         Legacy APIs will not be supported, since many of the tests are skipped
2816         on ports that do test notifications, and the issue with using file:// as the origin means a lot of the permissions
2817         tests won't work. Bugs 81048 and 81697 will track migrating the existing tests to use the new testRunner API, and move
2818         the tests to http/tests/notifications.
2819
2820         Permissions will be handled only in the injected bundle. This allows tests to set permission synchronously, and not
2821         require a chain of setTimeout()'s when writing a notification test. Clicking a notification, however, should
2822         be invoked from the UIProcess.
2823
2824         * WebKitTestRunner/InjectedBundle/Bindings/TestRunner.idl:
2825         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
2826         (WTR::InjectedBundle::didReceiveMessage): When told to "Reset", clear out all of the permissions.
2827         (WTR::InjectedBundle::postSimulateWebNotificationClick): Tell the bundle client to simulate a click based on the
2828         notification's internal ID.
2829         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
2830         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
2831         (WTR::TestRunner::grantWebNotificationPermission): Manually set the permission using WKBundle SPI.
2832         (WTR::TestRunner::denyWebNotificationPermission): Manually set the permission using WKBundle SPI.
2833         (WTR::TestRunner::removeAllWebNotificationPermissions):
2834         (WTR::TestRunner::simulateWebNotificationClick): Post a message through the injected bundle.
2835         * WebKitTestRunner/InjectedBundle/TestRunner.h:
2836
2837         Because supporting web notifications goes beyond a couple functions, move all of the logic into a WebNotificationProvider
2838         class.
2839         * WebKitTestRunner/TestInvocation.cpp:
2840         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): When the injected bundle sends a message to simulate a click,
2841         TestInvocation forwards that request to the TestController.
2842         * WebKitTestRunner/TestController.cpp:
2843         (WTR::TestController::initialize): Set the provider.
2844         (WTR::TestController::resetStateToConsistentValues): Tell the provider to reset.
2845         (WTR::TestController::simulateWebNotificationClick): Tell the provider to simulate a user clicking on a platform notification.
2846         * WebKitTestRunner/TestController.h: Holds an instance to the provider.
2847
2848         The provider maintains a pointer to one notification manager (presumably the web process associated with WTR), and a set
2849         of notification IDs that have been "shown" by the platform. This allows us to simulate a click on a notification.
2850         * WebKitTestRunner/WebNotificationProvider.h: Added.
2851         * WebKitTestRunner/WebNotificationProvider.cpp: Added.
2852         (WTR::WebNotificationProvider::showWebNotification): Adds the notification to set of shown notifications. Notifies provider that
2853         the notification got shown.
2854         (WTR::WebNotificationProvider::closeWebNotification): Removes the notification from the set. Notifies provider that the notification
2855         got closed.
2856         (WTR::WebNotificationProvider::addNotificationManager): Maintains one manager, which is fine for testing purposes.
2857         (WTR::WebNotificationProvider::removeNotificationManager): Maintains one manager.
2858         (WTR::WebNotificationProvider::notificationPermissions): When the web process starts, it maintains an initial set of known
2859         permissions. For testing purposes, we always want this set to be empty.
2860         (WTR::WebNotificationProvider::simulateWebNotificationClick): Notifies provider that the notification got "clicked".
2861         (WTR::WebNotificationProvider::reset): To reset the state, we pretend that we closed all of the platform notifications.
2862
2863         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj: Add WebNotificationProvider.
2864
2865 2012-08-29  Kamil Blank  <k.blank@samsung.com>
2866
2867         [EFL] Add setting API for allow universal/file access from file URLs.
2868         https://bugs.webkit.org/show_bug.cgi?id=83121
2869
2870         Reviewed by Eric Seidel.
2871
2872         Implementation of setAllowUniversalAccessFromFileURLs and setAllowFileAccessFromFileURLs.
2873
2874         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2875         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
2876         * DumpRenderTree/efl/TestRunnerEfl.cpp:
2877         (TestRunner::setAllowUniversalAccessFromFileURLs):
2878         (TestRunner::setAllowFileAccessFromFileURLs):
2879
2880 2012-08-29  Jon Lee  <jonlee@apple.com>
2881
2882         Update TestRunner API for web notifications
2883         https://bugs.webkit.org/show_bug.cgi?id=95093
2884         <rdar://problem/12179649>
2885
2886         Reviewed by Alexey Proskuryakov.
2887
2888         As it turns out the patch for this accidentally got squashed into the commit for
2889         bug 95099 (r126909). This amendment patch adds some comments about the renaming
2890         of the older TestRunner API for web notifications.
2891
2892         * DumpRenderTree/TestRunner.h:
2893         (TestRunner):
2894
2895 2012-08-29  Sheriff Bot  <webkit.review.bot@gmail.com>
2896
2897         Unreviewed, rolling out r126972.
2898         http://trac.webkit.org/changeset/126972
2899         https://bugs.webkit.org/show_bug.cgi?id=95349
2900
2901         accessibility/canvas-description-and-role.html has been
2902         failing consistently on several bots and Dominic needs some
2903         time to investigate why (Requested by jchaffraix on #webkit).
2904
2905         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
2906         (roleToString):
2907
2908 2012-08-29  Raphael Kubo da Costa  <rakuco@webkit.org>
2909
2910         [EFL] Resolve CMake warnings on overlapping search paths for EFL jhbuild-enabled build
2911         https://bugs.webkit.org/show_bug.cgi?id=84707
2912
2913         Reviewed by Gustavo Noronha Silva.
2914
2915         Most of the warnings have been fixed now that FindCairo.cmake and
2916         FindGLIB.cmake have been rewritten. The remaining one, related to
2917         FindFontconfig.cmake, shows up when building on a 64-bit
2918         Debian-based systems.
2919
2920         jhbuild installs libraries into lib64/ by default on 64-bit Linux
2921         installations, while CMake does not look for libraries in lib64/
2922         when /etc/debian_version exists on the system. The FIND_LIBRARY()
2923         would then sometimes end up using the system installation instead of
2924         the jhbuild one, causing mismatches and, when pkg-config is not used
2925         at all, failing to find libraries altogether.
2926
2927         * efl/jhbuildrc: Set the CMAKE_LIBRARY_PATH environment variable
2928         when use_lib64 is set to force CMake to look into lib64/ regardless
2929         of the presence of /etc/debian_version.
2930
2931 2012-08-29  Dominik Röttsches  <dominik.rottsches@intel.com>
2932
2933         The 2d.imageData.object.round canvas test is failing
2934         https://bugs.webkit.org/show_bug.cgi?id=40272
2935
2936         Reviewed by Benjamin Poulain.
2937
2938         Updating patch to address Benjamin's review comments.
2939         Adding a WTF test to test lrint implementation.
2940
2941         * TestWebKitAPI/CMakeLists.txt: Added MathExtras.cpp test file.
2942         * TestWebKitAPI/GNUmakefile.am: Added MathExtras.cpp test file.
2943         * TestWebKitAPI/TestWebKitAPI.gypi: Added MathExtras.cpp test file.
2944         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added MathExtras.cpp test file.
2945         * TestWebKitAPI/Tests/WTF/MathExtras.cpp: Added this test file containing a test for lrint().
2946         (TestWebKitAPI):
2947         (TestWebKitAPI::TEST):
2948         * TestWebKitAPI/win/TestWebKitAPI.vcproj: Added MathExtras.cpp test file.
2949
2950 2012-08-29  Florin Malita  <fmalita@chromium.org>
2951
2952         Unreviewed, updating Kelly Norton's email address at his request.
2953
2954         * Scripts/webkitpy/common/config/committers.py:
2955
2956 2012-08-29  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
2957
2958         [NRWT] The nrwt should check the contents of the skipped files with --lint-test-files
2959         https://bugs.webkit.org/show_bug.cgi?id=93723
2960
2961         Reviewed by Dirk Pranke.
2962
2963         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2964         (TestExpectationParser.expectation_for_skipped_test):
2965         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
2966         (SkippedTests.test_skipped_entry_dont_exist):
2967
2968 2012-08-29  Sheriff Bot  <webkit.review.bot@gmail.com>
2969
2970         Unreviewed, rolling out r120113.
2971         http://trac.webkit.org/changeset/120113
2972         https://bugs.webkit.org/show_bug.cgi?id=95320
2973
2974         Wrong fix for the problem, experimentally rolling it out for
2975         bug 95237. (Requested by rakuco on #webkit).
2976
2977         * efl/jhbuildrc:
2978
2979 2012-08-29  Dominic Mazzoni  <dmazzoni@google.com>
2980
2981         AX: Canvas should have a distinct role
2982         https://bugs.webkit.org/show_bug.cgi?id=95248
2983
2984         Reviewed by Chris Fleizach.
2985
2986         Add support for canvas accessibility role.
2987
2988         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
2989         (roleToString):
2990
2991 2012-08-28  Eric Seidel  <eric@webkit.org>
2992
2993         cr-ews bot doesn't set CWD correctly when zipping layout test results
2994         https://bugs.webkit.org/show_bug.cgi?id=91265
2995
2996         Reviewed by Kenneth Russell.
2997
2998         This should make the zips slightly less cumbersome to deal with.
2999
3000         * Scripts/webkitpy/common/system/workspace.py:
3001         (Workspace.create_zip):
3002         * Scripts/webkitpy/common/system/workspace_unittest.py:
3003         (WorkspaceTest.test_create_zip):
3004         (WorkspaceTest.test_create_zip_exception):
3005
3006 2012-08-28  Benjamin Poulain  <benjamin@webkit.org>
3007
3008         Fix the Commiters script
3009
3010         Unreviewed.
3011
3012         Remove Roger Fong from the Contributor list, he appears in the Commiter list since r126949.
3013
3014         * Scripts/webkitpy/common/config/committers.py:
3015
3016 2012-08-28  Roger Fong  <roger_fong@apple.com>
3017
3018         Unreviewed. Adding myself to committers list.
3019
3020         * Scripts/webkitpy/common/config/committers.py:
3021
3022 2012-08-28  Roger Fong  <roger_fong@apple.com>
3023
3024         Add minires to list of required cygwin install packages.
3025         https://bugs.webkit.org/show_bug.cgi?id=76936
3026
3027         Reviewed by Tim Horton.
3028
3029         * CygwinDownloader/cygwin-downloader.py:
3030
3031 2012-08-28  Xianzhu Wang  <wangxianzhu@chromium.org>
3032
3033         [Chromium-Android] Upstream layout test expectations (Part 2)
3034         https://bugs.webkit.org/show_bug.cgi?id=95212
3035
3036         Reviewed by Adam Barth.
3037
3038         - Now run both gpu path and non-gpu path of 2d canvas layout tests
3039         because both are used in chromium-android.
3040
3041         - Exclude all webgl tests because webgl is not enabled yet.
3042
3043         - Removed 'win' from the baseline fallback list. This was a mistake.
3044
3045         - Moved some small directories to TestExpectations from skipped_layout_tests().
3046
3047         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3048         (ChromiumAndroidPort):
3049         (ChromiumAndroidPort.skipped_layout_tests):
3050
3051 2012-08-28  Joanmarie Diggs  <jdiggs@igalia.com>
3052
3053         Implement AccessibilityUIElement::titleUIElement() and AccessibilityUIElement::stringValue()
3054         https://bugs.webkit.org/show_bug.cgi?id=95185
3055
3056         Reviewed by Chris Fleizach.
3057
3058         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3059         (AccessibilityUIElement::titleUIElement): Implemented
3060         Gets the ATK_RELATION_LABELLED_BY target.
3061         (AccessibilityUIElement::stringValue): Implemented for all roles but ATK_ROLE_PANEL.
3062         Gets the string from atk_text_get_text(). Will implement for ATK_ROLE_PANEL after
3063         bug 95180 is fixed.
3064
3065 2012-08-28  Dominic Mazzoni  <dmazzoni@google.com>
3066
3067         AX: Crash due to object getting deleted inside updateBackingStore
3068         https://bugs.webkit.org/show_bug.cgi?id=94619
3069
3070         Reviewed by Chris Fleizach.
3071
3072         Change isValid to !isDetached after deleting isValid from
3073         chromium WebAccessibilityObject as being redundant.
3074
3075         * DumpRenderTree/chromium/TestRunner/AccessibilityUIElementChromium.cpp:
3076         (AccessibilityUIElement::isValidGetterCallback):
3077
3078 2012-08-28  Sheriff Bot  <webkit.review.bot@gmail.com>
3079
3080         Unreviewed, rolling out r126914.
3081         http://trac.webkit.org/changeset/126914
3082         https://bugs.webkit.org/show_bug.cgi?id=95239
3083
3084         it breaks everything and fixes nothing (Requested by pizlo on
3085         #webkit).
3086
3087         * DumpRenderTree/chromium/MockWebSpeechInputController.cpp:
3088         (MockWebSpeechInputController::addMockRecognitionResult):
3089         * DumpRenderTree/chromium/NotificationPresenter.cpp:
3090         (NotificationPresenter::simulateClick):
3091         (NotificationPresenter::show):
3092         * DumpRenderTree/chromium/TestRunner/CppBoundClass.cpp:
3093         (CppBoundClass::~CppBoundClass):
3094         (CppBoundClass::invoke):
3095         (CppBoundClass::getProperty):
3096         (CppBoundClass::setProperty):
3097         (CppBoundClass::bindCallback):
3098         (CppBoundClass::bindProperty):
3099         * DumpRenderTree/chromium/WebPreferences.cpp:
3100         (applyFontMap):
3101         * DumpRenderTree/chromium/WebViewHost.cpp:
3102         (WebViewHost::printResourceDescription):
3103         * DumpRenderTree/mac/TestRunnerMac.mm:
3104         (worldIDForWorld):
3105         (TestRunner::evaluateScriptInIsolatedWorld):
3106         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
3107         (AccessibilityController::~AccessibilityController):
3108         (AccessibilityController::winNotificationReceived):
3109         * DumpRenderTree/win/ResourceLoadDelegate.cpp:
3110         (ResourceLoadDelegate::descriptionSuitableForTestResult):
3111         * DumpRenderTree/win/TestRunnerWin.cpp:
3112         (worldIDForWorld):
3113         (TestRunner::evaluateScriptInIsolatedWorld):
3114         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp:
3115         (TestWebKitAPI::DOMWindowExtensionBasic::willDestroyPage):
3116         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionNoCache_Bundle.cpp:
3117         (TestWebKitAPI::DOMWindowExtensionNoCache::willDestroyPage):
3118         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3119         (WTR::TestRunner::worldIDForWorld):
3120         (WTR::TestRunner::evaluateScriptInIsolatedWorld):
3121
3122 2012-08-28  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
3123
3124         Rename first/second to key/value in HashMap iterators
3125         https://bugs.webkit.org/show_bug.cgi?id=82784
3126
3127         Reviewed by Eric Seidel.
3128
3129         * DumpRenderTree/chromium/MockWebSpeechInputController.cpp:
3130         (MockWebSpeechInputController::addMockRecognitionResult):
3131         * DumpRenderTree/chromium/NotificationPresenter.cpp:
3132         (NotificationPresenter::simulateClick):
3133         (NotificationPresenter::show):
3134         * DumpRenderTree/chromium/TestRunner/CppBoundClass.cpp:
3135         (CppBoundClass::~CppBoundClass):
3136         (CppBoundClass::invoke):
3137         (CppBoundClass::getProperty):
3138         (CppBoundClass::setProperty):
3139         (CppBoundClass::bindCallback):
3140         (CppBoundClass::bindProperty):
3141         * DumpRenderTree/chromium/WebPreferences.cpp:
3142         (applyFontMap):
3143         * DumpRenderTree/chromium/WebViewHost.cpp:
3144         (WebViewHost::printResourceDescription):
3145         * DumpRenderTree/mac/TestRunnerMac.mm:
3146         (worldIDForWorld):
3147         (TestRunner::evaluateScriptInIsolatedWorld):
3148         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
3149         (AccessibilityController::~AccessibilityController):
3150         (AccessibilityController::winNotificationReceived):
3151         * DumpRenderTree/win/ResourceLoadDelegate.cpp:
3152         (ResourceLoadDelegate::descriptionSuitableForTestResult):
3153         * DumpRenderTree/win/TestRunnerWin.cpp:
3154         (worldIDForWorld):
3155         (TestRunner::evaluateScriptInIsolatedWorld):
3156         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp:
3157         (TestWebKitAPI::DOMWindowExtensionBasic::willDestroyPage):
3158         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionNoCache_Bundle.cpp:
3159         (TestWebKitAPI::DOMWindowExtensionNoCache::willDestroyPage):
3160         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3161         (WTR::TestRunner::worldIDForWorld):
3162         (WTR::TestRunner::evaluateScriptInIsolatedWorld):
3163
3164 2012-08-27  Jon Lee  <jonlee@apple.com>
3165
3166         Update TestRunner API for web notifications
3167         https://bugs.webkit.org/show_bug.cgi?id=95093
3168         <rdar://problem/12179649>
3169
3170         Reviewed by Jessie Berlin.
3171
3172         This is work toward providing Mac support for web notifications in DRT and WTR (77969).
3173
3174         Some of the legacy APIs are less than ideal, and not tenable with the WTR infrastructure.
3175         This first patch renames the TestRunner calls to use the term "web notifications" instead of
3176         "desktop notifications", deprecates a couple API calls that are not used by anyone, and adds
3177         additional calls that will be used in the tests that test the standard API.
3178
3179         * DumpRenderTree/TestRunner.h:
3180         (TestRunner): For consistency, rename a couple member variables. Remove unused checkDesktopNotificationPermission()
3181         and areDesktopNotificationPermissionRequestsIgnored(). Remove origin mapping since each port implements its own solution.
3182         * DumpRenderTree/TestRunner.cpp: Push grantWebNotificationPermission() to individual ports.
3183         (TestRunner::TestRunner):
3184         (ignoreLegacyWebNotificationPermissionRequestsCallback): Renamed.
3185         (simulateLegacyWebNotificationClickCallback): Renamed.
3186         (grantWebNotificationPermissionCallback): Renamed.
3187         (denyWebNotificationPermissionCallback): Added.
3188         (removeAllWebNotificationPermissionsCallback): Added.
3189         (simulateWebNotificationClickCallback): Added.
3190         (TestRunner::staticFunctions):
3191         (TestRunner::ignoreLegacyWebNotificationPermissionRequests):
3192
3193         * DumpRenderTree/chromium/DRTTestRunner.cpp:
3194         (DRTTestRunner::DRTTestRunner): Added bindings for new APIs.
3195         (DRTTestRunner::grantWebNotificationPermission):
3196         (DRTTestRunner::denyWebNotificationPermission): Stub.
3197         (DRTTestRunner::removeAllWebNotificationPermissions): Stub.
3198         (DRTTestRunner::simulateWebNotificationClick): Stub.
3199         (DRTTestRunner::simulateLegacyWebNotificationClick):
3200         * DumpRenderTree/chromium/DRTTestRunner.h:
3201         (DRTTestRunner):
3202
3203         Added stubs.
3204         * DumpRenderTree/blackberry/TestRunnerBlackBerry.cpp:
3205         * DumpRenderTree/efl/TestRunnerEfl.cpp:
3206         * DumpRenderTree/gtk/TestRunnerGtk.cpp:
3207         * DumpRenderTree/mac/TestRunnerMac.mm: Later patch will contain implementation of this API.
3208         * DumpRenderTree/qt/TestRunnerQt.cpp: Added Qt-based stubs.
3209         * DumpRenderTree/win/TestRunnerWin.cpp:
3210
3211 2012-08-28  Szilard Ledan  <szledan@inf.u-szeged.hu>
3212
3213         [webkit-patch] gets stuck into an infinite loop if bugzilla doesn't respond in time.
3214         https://bugs.webkit.org/show_bug.cgi?id=94700
3215
3216         Reviewed by Eric Seidel.
3217
3218         Interrupting download with a timeout would be a possible solution.
3219         I imported the socket module and used the setdefaulttimeout() method.
3220         Fixing it is important, because EWS bots need to be restarted regularly
3221         because of this bug.
3222
3223         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
3224         (Bugzilla._get_browser):
3225         (Bugzilla.setdefaulttimeout):
3226
3227 2012-08-28  Gabor Rapcsanyi  <rgabor@webkit.org>
3228
3229         [Qt] Keep QT_QPA_PLATFORM_PLUGIN_PATH environment variable in NRWT
3230         https://bugs.webkit.org/show_bug.cgi?id=95194
3231
3232         Reviewed by Simon Hausmann.
3233
3234         * Scripts/webkitpy/layout_tests/port/qt.py:
3235         (QtPort.setup_environ_for_server):
3236
3237 2012-08-28  Andras Becsi  <andras.becsi@nokia.com>
3238
3239         [watchlist] Unreviewed, subscribe to Qt bugs.
3240
3241         * Scripts/webkitpy/common/config/watchlist:
3242
3243 2012-08-28  Simon Hausmann  <simon.hausmann@nokia.com>
3244
3245         [Qt] New test introduced in r126611 fails: fast/css/image-set-setting.html
3246         https://bugs.webkit.org/show_bug.cgi?id=95054
3247
3248         Reviewed by Eric Seidel.
3249
3250         Enable CSS_IMAGE_SET for the Qt build.
3251
3252         * qmake/mkspecs/features/features.pri:
3253
3254 2012-08-27  Csaba Osztrogonác  <ossy@webkit.org>
3255
3256         [Qt] Enable CSS sticky position
3257         https://bugs.webkit.org/show_bug.cgi?id=95172
3258
3259         Reviewed by Simon Hausmann.
3260
3261         * qmake/mkspecs/features/features.pri:
3262
3263 2012-08-27  Zan Dobersek  <zandobersek@gmail.com>
3264
3265         [GTK] Memory cache should be cleared in between test runs
3266         https://bugs.webkit.org/show_bug.cgi?id=95105
3267
3268         Reviewed by Martin Robinson.
3269
3270         Call the new DumpRenderTreeSupportGtk helper method after every test
3271         to clear the memory cache.
3272
3273         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3274         (runTest):
3275
3276 2012-08-27  Sheriff Bot  <webkit.review.bot@gmail.com>
3277
3278         Unreviewed, rolling out r126836.
3279         http://trac.webkit.org/changeset/126836
3280         https://bugs.webkit.org/show_bug.cgi?id=95163
3281
3282         Broke all Apple ports, EFL, and Qt. (Requested by tkent on
3283         #webkit).
3284
3285         * DumpRenderTree/chromium/MockWebSpeechInputController.cpp:
3286         (MockWebSpeechInputController::addMockRecognitionResult):
3287         * DumpRenderTree/chromium/NotificationPresenter.cpp:
3288         (NotificationPresenter::simulateClick):
3289         (NotificationPresenter::show):
3290         * DumpRenderTree/chromium/TestRunner/CppBoundClass.cpp:
3291         (CppBoundClass::~CppBoundClass):
3292         (CppBoundClass::invoke):
3293         (CppBoundClass::getProperty):
3294         (CppBoundClass::setProperty):
3295         (CppBoundClass::bindCallback):
3296         (CppBoundClass::bindProperty):
3297         * DumpRenderTree/chromium/WebPreferences.cpp:
3298         (applyFontMap):
3299         * DumpRenderTree/chromium/WebViewHost.cpp:
3300         (WebViewHost::printResourceDescription):
3301         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
3302         (AccessibilityController::~AccessibilityController):
3303         (AccessibilityController::winNotificationReceived):
3304         * DumpRenderTree/win/ResourceLoadDelegate.cpp:
3305         (ResourceLoadDelegate::descriptionSuitableForTestResult):
3306         * DumpRenderTree/win/TestRunnerWin.cpp:
3307         (worldIDForWorld):
3308         (TestRunner::evaluateScriptInIsolatedWorld):
3309         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp:
3310         (TestWebKitAPI::DOMWindowExtensionBasic::willDestroyPage):
3311         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionNoCache_Bundle.cpp:
3312         (TestWebKitAPI::DOMWindowExtensionNoCache::willDestroyPage):
3313         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3314         (WTR::TestRunner::worldIDForWorld):
3315         (WTR::TestRunner::evaluateScriptInIsolatedWorld):
3316
3317 2012-08-27  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
3318
3319         Rename first/second to key/value in HashMap iterators
3320         https://bugs.webkit.org/show_bug.cgi?id=82784
3321
3322         Reviewed by Eric Seidel.
3323
3324         * DumpRenderTree/chromium/MockWebSpeechInputController.cpp:
3325         (MockWebSpeechInputController::addMockRecognitionResult):
3326         * DumpRenderTree/chromium/NotificationPresenter.cpp:
3327         (NotificationPresenter::simulateClick):
3328         (NotificationPresenter::show):
3329         * DumpRenderTree/chromium/TestRunner/CppBoundClass.cpp:
3330         (CppBoundClass::~CppBoundClass):
3331         (CppBoundClass::invoke):
3332         (CppBoundClass::getProperty):
3333         (CppBoundClass::setProperty):
3334         (CppBoundClass::bindCallback):
3335         (CppBoundClass::bindProperty):
3336         * DumpRenderTree/chromium/WebPreferences.cpp:
3337         (applyFontMap):
3338         * DumpRenderTree/chromium/WebViewHost.cpp:
3339         (WebViewHost::printResourceDescription):
3340         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
3341         (AccessibilityController::~AccessibilityController):
3342         (AccessibilityController::winNotificationReceived):
3343         * DumpRenderTree/win/ResourceLoadDelegate.cpp:
3344         (ResourceLoadDelegate::descriptionSuitableForTestResult):
3345         * DumpRenderTree/win/TestRunnerWin.cpp:
3346         (worldIDForWorld):
3347         (TestRunner::evaluateScriptInIsolatedWorld):
3348         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionBasic_Bundle.cpp:
3349         (TestWebKitAPI::DOMWindowExtensionBasic::willDestroyPage):
3350         * TestWebKitAPI/Tests/WebKit2/DOMWindowExtensionNoCache_Bundle.cpp:
3351         (TestWebKitAPI::DOMWindowExtensionNoCache::willDestroyPage):
3352         * WebKitTestRunner/InjectedBundle/TestRunner.cpp:
3353         (WTR::TestRunner::worldIDForWorld):
3354         (WTR::TestRunner::evaluateScriptInIsolatedWorld):
3355
3356 2012-08-24  James Robinson  <jamesr@chromium.org>
3357
3358         WebGL should not flip textures on presentation if contents are unchanged
3359         https://bugs.webkit.org/show_bug.cgi?id=94961
3360
3361         Reviewed by Kenneth Russell.
3362
3363         Run some compositing webgl tests in threaded mode to catch regressions specific to that mode.
3364
3365         * Scripts/webkitpy/layout_tests/port/chromium.py:
3366         (ChromiumPort.virtual_test_suites):
3367
3368 2012-08-27  Gavin Peters  <gavinp@chromium.org>
3369
3370         [webkit-patch] Don't crash chrome-channels command when a previously unknown platform shows up.
3371         https://bugs.webkit.org/show_bug.cgi?id=95104
3372
3373         Reviewed by Adam Barth.
3374
3375         Today I found out this command was broken by the new iOS platform showing up on omahaproxy. Now the webkit-patch chrome-channels command gracefully handles new, previously unknown platforms. Also, it prints iOS in the proper mixed case.
3376
3377         * Scripts/webkitpy/common/net/omahaproxy.py:
3378         (OmahaProxy):
3379         (OmahaProxy.get_revisions):
3380         * Scripts/webkitpy/common/net/omahaproxy_unittest.py:
3381         (OmahaProxyTest):
3382         (OmahaProxyTest.test_get_revisions):
3383
3384 2012-08-27  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
3385
3386         [EFL][WK2] Free Url_Bar on program exit
3387         https://bugs.webkit.org/show_bug.cgi?id=95086
3388
3389         Reviewed by Eric Seidel.
3390
3391         Fix a memory leak by freeing the Url_Bar on program exit.
3392
3393         * MiniBrowser/efl/main.c:
3394         (main): Call url_bar_del() to free the Url_Bar.
3395
3396 2012-08-27  James Robinson  <jamesr@chromium.org>
3397
3398         [chromium] Fix exit code status logic in run-chromium-webkit-unit-tests
3399         https://bugs.webkit.org/show_bug.cgi?id=95115
3400
3401         Reviewed by Adam Barth.
3402
3403         Use VCSUtils utility function to check if webkit_unit_tests terminates normally so we generate a non-zero exit
3404         code if the unit tests crash.
3405
3406         * Scripts/run-chromium-webkit-unit-tests:
3407
3408 2012-08-27  Benjamin Poulain  <benjamin@webkit.org>
3409
3410         Add ECMAScript Number to String conversion to WTF::String
3411         https://bugs.webkit.org/show_bug.cgi?id=95016
3412
3413         Reviewed by Geoffrey Garen.
3414
3415         * TestWebKitAPI/Tests/WTF/WTFString.cpp:
3416         (TestWebKitAPI::testNumberToStringECMAScript):
3417         (TestWebKitAPI): Add tests for String's ECMAString number conversion.
3418
3419 2012-08-27  Philippe Normand  <pnormand@igalia.com>
3420
3421         [GStreamer][Qt] WebAudio support
3422         https://bugs.webkit.org/show_bug.cgi?id=94806
3423
3424         Reviewed by Simon Hausmann.
3425
3426         * MiniBrowser/qt/qml/BrowserWindow.qml: Enable WebAudio support.
3427
3428 2012-08-27  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
3429
3430         Rename RegisterProtocolHandler API to NavigatorContentUtils
3431         https://bugs.webkit.org/show_bug.cgi?id=94920
3432
3433         Reviewed by Adam Barth.
3434
3435         Renamed 'register-protocol-handler' feature to 'navigator-content-utils'.
3436
3437         * Scripts/webkitperl/FeatureList.pm:
3438         * qmake/mkspecs/features/features.pri:
3439
3440 2012-08-24  Benjamin Poulain  <bpoulain@apple.com>
3441
3442         Unify Number to StringImpl conversion
3443         https://bugs.webkit.org/show_bug.cgi?id=94879
3444
3445         Reviewed by Geoffrey Garen.
3446
3447         Add testing for the new IntegerToStringConversion.
3448
3449         * TestWebKitAPI/CMakeLists.txt:
3450         * TestWebKitAPI/GNUmakefile.am:
3451         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
3452         * TestWebKitAPI/Tests/WTF/IntegerToStringConversion.cpp: Added.
3453
3454 2012-08-23  James Robinson  <jamesr@chromium.org>
3455
3456         [chromium] Convert WebLayerTreeView interface into pure virtual
3457         https://bugs.webkit.org/show_bug.cgi?id=94866
3458
3459         Reviewed by Adrienne Walker.
3460
3461         Updates DumpRenderTree's WebViewHost to vend a WebCompositorOutputSurface so DumpRenderTree can run compositing
3462         tests.
3463
3464         * DumpRenderTree/DumpRenderTree.gypi:
3465         * DumpRenderTree/chromium/WebViewHost.cpp:
3466         (WebViewHost::createOutputSurface):
3467         * DumpRenderTree/chromium/WebViewHost.h:
3468         (WebViewHost):
3469         * DumpRenderTree/chromium/WebViewHostOutputSurface.cpp: Copied from Source/WebKit/chromium/src/WebLayerTreeViewImpl.h.
3470         (WebKit):
3471         (WebKit::WebViewHostOutputSurface::WebViewHostOutputSurface):
3472         (WebKit::WebViewHostOutputSurface::~WebViewHostOutputSurface):
3473         (WebKit::WebViewHostOutputSurface::bindToClient):
3474         (WebKit::WebViewHostOutputSurface::capabilities):
3475         (WebKit::WebViewHostOutputSurface::context3D):
3476         (WebKit::WebViewHostOutputSurface::sendFrameToParentCompositor):
3477         * DumpRenderTree/chromium/WebViewHostOutputSurface.h: Copied from Source/WebKit/chromium/src/WebLayerTreeViewImpl.h.
3478         (WebKit):
3479         (WebViewHostOutputSurface):
3480
3481 2012-08-24  Xianzhu Wang  <wangxianzhu@chromium.org>
3482
3483         [Chromium-Android] Many 'Unexpected EOF ...' when a test crashes
3484         https://bugs.webkit.org/show_bug.cgi?id=94865
3485
3486         Reviewed by Dirk Pranke.
3487
3488         Android detects crash of test by checking if the return value of read() is None.
3489         Added a parameter 'treat_no_data_as_crash' to let ServerProcess know the special requirement of Android.
3490         If set, ServerProcess will set _crashed unconditionaly when there is no data.
3491
3492         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3493         (ChromiumAndroidPort.__init__):
3494         (ChromiumAndroidPort):
3495         (ChromiumAndroidPort._android_server_process_constructor):
3496         (ChromiumAndroidDriver._start_once):
3497         * Scripts/webkitpy/layout_tests/port/server_process.py:
3498         (ServerProcess.__init__):
3499         (ServerProcess._wait_for_data_and_update_buffers_using_select):
3500
3501 2012-08-24  Arnaud Renevier  <a.renevier@sisa.samsung.com>
3502
3503         build-webkit --gtk fails with Can't exec ""
3504         https://bugs.webkit.org/show_bug.cgi?id=94944
3505
3506         Reviewed by Martin Robinson.
3507
3508         When jhbuild wrapper is not needed, do unshift a blank string to build
3509         arguments.
3510
3511         * Scripts/webkitdirs.pm:
3512         (runAutogenForAutotoolsProjectIfNecessary):
3513
3514 2012-08-24  Tommy Widenflycht  <tommyw@google.com>
3515
3516         MediaStream API: Add readyState functionality to RTCPeerConnection
3517         https://bugs.webkit.org/show_bug.cgi?id=94813
3518
3519         Reviewed by Adam Barth.
3520
3521         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.cpp:
3522         (MockWebRTCPeerConnectionHandler::MockWebRTCPeerConnectionHandler):
3523         (MockWebRTCPeerConnectionHandler::stop):
3524         * DumpRenderTree/chromium/MockWebRTCPeerConnectionHandler.h:
3525         (MockWebRTCPeerConnectionHandler):
3526
3527 2012-08-24  Michael Brüning  <michaelbruening@gmail.com>
3528
3529         Unreviewed. Adding myself to committers list.
3530
3531         Also updated email information.
3532
3533         * Scripts/webkitpy/common/config/committers.py:
3534
3535 2012-08-24  Michael Brüning  <michaelbruening@gmail.com>
3536
3537         [Qt][WK2] Make viewport related experimental.test properties encapsulated.
3538         https://bugs.webkit.org/show_bug.cgi?id=88320
3539
3540         Reviewed by Kenneth Rohde Christiansen.
3541
3542         Changed the access to the viewport attributes to use the viewport object that is introduced by this change.
3543
3544         * MiniBrowser/qt/qml/ViewportInfoItem.qml:
3545
3546 2012-08-23  Dirk Pranke  <dpranke@chromium.org>
3547
3548         leak detection broken after r126247 on mac
3549         https://bugs.webkit.org/show_bug.cgi?id=94893
3550
3551         Unreviewed, build fix.
3552
3553         Adds is_snowleopard(), is_lion back() in for the mac port. We'll
3554         probably need an is_mountainlion() or equivalent at some point,
3555         I'm guessing (but I've deleted the leopard code).
3556
3557         * Scripts/webkitpy/layout_tests/port/leakdetector.py:
3558         (LeakDetector._callstacks_to_exclude_from_leaks):
3559         * Scripts/webkitpy/layout_tests/port/mac.py:
3560         (MacPort):
3561         (MacPort.is_snowleopard):
3562         (MacPort.is_lion):
3563
3564 2012-08-23  Julien Chaffraix  <jchaffraix@webkit.org>
3565
3566         [Chromium] Enable AtomicString && CString && WTFString && StringImpl unit tests
3567         https://bugs.webkit.org/show_bug.cgi?id=94871
3568
3569         Reviewed by Benjamin Poulain.
3570
3571         * TestWebKitAPI/TestWebKitAPI.gypi:
3572         Added the missing files to Chromium's build system.
3573
3574 2012-08-23  Raphael Kubo da Costa  <rakuco@webkit.org>
3575
3576         [watchlist] Unreviewed, subscribe to WebKit2-EFL bugs.
3577
3578         * Scripts/webkitpy/common/config/watchlist:
3579
3580 2012-08-23  Ryosuke Niwa  <rniwa@webkit.org>
3581
3582         Replay tests take forever to load
3583         https://bugs.webkit.org/show_bug.cgi?id=94764
3584
3585         Reviewed by Adam Barth.
3586
3587         Don't load about:blank at the beginning. This appears to clog the DRT.
3588
3589         * Scripts/webkitpy/performance_tests/perftest.py:
3590         (ReplayPerfTest.run_single):
3591         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
3592         (TestReplayPerfTest.test_run_single):
3593         (TestReplayPerfTest.test_run_single_fails_when_output_has_error):
3594
3595 2012-08-23  Kenneth Russell  <kbr@google.com>
3596
3597         Unreviewed, rolling out r126497.
3598         http://trac.webkit.org/changeset/126497
3599         https://bugs.webkit.org/show_bug.cgi?id=94764
3600
3601         Caused webkitpy-test step to start failing.
3602
3603         * Scripts/webkitpy/performance_tests/perftest.py:
3604         (ReplayPerfTest.run_single):
3605
3606 2012-08-23  Emil A Eklund  <eae@chromium.org>
3607
3608         Add saturation arithmetic support to FractionalLayoutUnit
3609         https://bugs.webkit.org/show_bug.cgi?id=94364
3610
3611         Reviewed by Benjamin Poulain.
3612
3613         Add test for new saturatedArithmetic functions.
3614
3615         * TestWebKitAPI/CMakeLists.txt:
3616         * TestWebKitAPI/GNUmakefile.am:
3617         * TestWebKitAPI/TestWebKitAPI.gypi:
3618         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
3619         * TestWebKitAPI/Tests/WTF/SaturatedArithmeticOperations.cpp: Added.
3620         (TestWebKitAPI):
3621         (TestWebKitAPI::TEST):
3622         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
3623
3624 2012-08-23  Ryosuke Niwa  <rniwa@webkit.org>
3625
3626         Replay tests take forever to load
3627         https://bugs.webkit.org/show_bug.cgi?id=94764
3628
3629         Reviewed by Adam Barth.
3630
3631         Don't load about:blank at the beginning. This appears to clog the DRT.
3632
3633         * Scripts/webkitpy/performance_tests/perftest.py:
3634         (ReplayPerfTest.run_single):
3635
3636 2012-08-23  Wei Jia  <wjia@chromium.org>
3637
3638         create different WebKit::WebMediaPlayer based on URL
3639         https://bugs.webkit.org/show_bug.cgi?id=91301
3640
3641         Reviewed by Adam Barth.
3642
3643         Pass URL to createMediaPlayer. This corresponds to change in base
3644         class WebKit::WebFrameClient.
3645
3646         * DumpRenderTree/chromium/WebViewHost.cpp:
3647         (WebViewHost::createMediaPlayer): add URL as additional argument.
3648         * DumpRenderTree/chromium/WebViewHost.h:
3649         (WebViewHost):
3650
3651 2012-08-23  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
3652
3653         Unreviewed, upgrade gyuyoung to reviewer.
3654         http://lists.webkit.org/pipermail/webkit-dev/2012-August/022004.html
3655
3656         * Scripts/webkitpy/common/config/committers.py:
3657
3658 2012-08-23  Carlos Garcia Campos  <cgarcia@igalia.com>
3659
3660         REGRESSION(r126306): it broke the plugin process
3661         https://bugs.webkit.org/show_bug.cgi?id=94797
3662
3663         Reviewed by Xan Lopez.
3664
3665         * GNUmakefile.am:
3666
3667 2012-08-23  Ryuan Choi  <ryuan.choi@samsung.com>
3668
3669      &