0eaed75998732153240fba90ea4e1e6a892410e3
[WebKit-https.git] / Tools / ChangeLog
1 2012-07-31  Jochen Eisinger  <jochen@chromium.org>
2
3         [chromium] move EventSender into TestRunner.a
4         https://bugs.webkit.org/show_bug.cgi?id=92277
5
6         Reviewed by Adam Barth.
7
8         * DumpRenderTree/DumpRenderTree.gypi:
9         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
10         * DumpRenderTree/chromium/TestRunner/EventSender.cpp: Renamed from Tools/DumpRenderTree/chromium/EventSender.cpp.
11         * DumpRenderTree/chromium/TestRunner/EventSender.h: Renamed from Tools/DumpRenderTree/chromium/EventSender.h.
12         * DumpRenderTree/chromium/TestRunner/TestInterfaces.cpp:
13         (TestInterfaces::Internal::eventSender):
14         (TestInterfaces::Internal):
15         (TestInterfaces::Internal::Internal):
16         (TestInterfaces::Internal::~Internal):
17         (TestInterfaces::Internal::setWebView):
18         (TestInterfaces::Internal::setDelegate):
19         (TestInterfaces::Internal::bindTo):
20         (TestInterfaces::Internal::resetAll):
21         (TestInterfaces::eventSender):
22         * DumpRenderTree/chromium/TestRunner/TestInterfaces.h:
23         (TestInterfaces):
24         * DumpRenderTree/chromium/TestShell.cpp:
25         (TestShell::initialize):
26         (TestShell::createMainWindow):
27         (TestShell::~TestShell):
28         (TestShell::resetTestController):
29         (TestShell::bindJSObjectsToWindow):
30         * DumpRenderTree/chromium/TestShell.h:
31         (TestShell::eventSender):
32         (TestShell):
33         * DumpRenderTree/chromium/WebViewHost.cpp:
34
35 2012-07-31  Jochen Eisinger  <jochen@chromium.org>
36
37         Unreviewed. Update watchlist
38
39         * Scripts/webkitpy/common/config/watchlist: Turned out to be less useful than I had hoped
40
41 2012-07-31  Kent Tamura  <tkent@chromium.org>
42
43         Unreviewed, update watchlist
44
45         * Scripts/webkitpy/common/config/watchlist:
46         - Improve "Forms" filename pattern.
47         - tkent stops watching Chromium DRT. It was not so helpful.
48
49 2012-07-31  Martin Robinson  <mrobinson@igalia.com>
50
51         [jhbuild] Also try to load jhbuild Python module from the source checkout
52
53         Reviewed by Philippe Normand.
54
55         JHBuild also seems to have an installation mode where it prefers to load its
56         Python modules from the source checkout. This seems to be used particularly
57         on the GTK+ 64-bit release bot. We should fall back to this mode.
58
59         * jhbuild/jhbuildutils.py:
60         (enter_jhbuild_environment_if_available): Use the source checkout if there
61         is no installation path.
62
63 2012-07-30  Ryosuke Niwa  <rniwa@webkit.org>
64
65         Revert r124153 and land the "right" build fix per ap's comment on the bug 92676.
66
67         * MiniBrowser/win/BrowserView.cpp:
68         (BrowserView::create):
69
70 2012-07-30  Sam Weinig  <sam@webkit.org>
71
72         Rename WebKit2.h to WebKit2_C.h
73         https://bugs.webkit.org/show_bug.cgi?id=92704
74
75         Reviewed by Dan Bernstein.
76
77         * MiniBrowser/mac/MiniBrowser_Prefix.pch:
78         * MiniBrowser/win/BrowserView.h:
79         * TestWebKitAPI/config.h:
80         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
81         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
82         * WebKitTestRunner/WebKitTestRunnerPrefix.h:
83         * WebKitTestRunner/config.h:
84         Update for changed header file name.
85
86 2012-07-30  Dirk Pranke  <dpranke@chromium.org>
87
88         chromium win compile broken after removing webkit.py
89         https://bugs.webkit.org/show_bug.cgi?id=92549
90
91         Reviewed by Ryosuke Niwa.
92
93         Remove the stub file for webkit.py now that it's no longer
94         necessary.
95
96         * Scripts/webkitpy/layout_tests/port/webkit.py: Removed.
97
98 2012-07-30  Ryosuke Niwa  <rniwa@webkit.org>
99
100         Windows build fix attempt after r124092.
101
102         * MiniBrowser/win/BrowserView.cpp:
103         (BrowserView::create):
104
105 2012-07-30  Ryosuke Niwa  <rniwa@webkit.org>
106
107         Remove erroneously committed debugging print call from a test.
108
109         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
110         (test_run_generates_results_page):
111
112 2012-07-30  Arnaud Renevier  <a.renevier@sisa.samsung.com>
113
114         webkit-patch: system keyring is not used to read my password
115         https://bugs.webkit.org/show_bug.cgi?id=92532
116
117         Reviewed by Dirk Pranke.
118
119         In case no username can be fetched from environment, git or keychain,
120         prompt for username, and try to get password from keyring associated
121         with that username.
122
123         * Scripts/webkitpy/common/net/credentials.py:
124         (Credentials.read_credentials):
125         * Scripts/webkitpy/common/net/credentials_unittest.py:
126         (test_keyring_without_git_repo_nor_keychain):
127         (test_keyring_without_git_repo_nor_keychain.MockKeyring):
128         (test_keyring_without_git_repo_nor_keychain.MockKeyring.get_password):
129         (test_keyring_without_git_repo_nor_keychain.FakeCredentials):
130         (test_keyring_without_git_repo_nor_keychain.FakeCredentials._credentials_from_keychain):
131         (test_keyring_without_git_repo_nor_keychain.FakeCredentials._credentials_from_environment):
132         (test_keyring_without_git_repo_nor_keychain.FakeUser):
133         (test_keyring_without_git_repo_nor_keychain.FakeUser.prompt):
134         (test_keyring_without_git_repo_nor_keychain.FakeUser.prompt_password):
135
136 2012-07-30  Dirk Pranke  <dpranke@chromium.org>
137
138         nrwt: move the code that identifies the chunk of tests to run into finder
139         https://bugs.webkit.org/show_bug.cgi?id=92694
140
141         Reviewed by Ryosuke Niwa.
142
143         Another patch to make manager.py smaller and clearer; this patch moves
144         the handling of breaking the list of tests into chunks (--run-part,
145         --run-chunk) from manager.py to finder.py.
146
147         There are no functional changes and this is covered by existing tests.
148
149         * Scripts/webkitpy/layout_tests/controllers/finder.py:
150         (Finder.__init__):
151         (Finder._read_test_files):
152         (Finder):
153         (Finder.split_into_chunks_if_necessary):
154         * Scripts/webkitpy/layout_tests/controllers/manager.py:
155         (Manager.__init__):
156         (Manager._collect_tests):
157         (Manager._parse_expectations):
158         (Manager.prepare_lists_and_print_output):
159
160 2012-07-30  Dirk Pranke  <dpranke@chromium.org>
161
162         nrwt: split test-finding code out from manager.py
163         https://bugs.webkit.org/show_bug.cgi?id=92693
164
165         Reviewed by Ryosuke Niwa.
166
167         In the interest of making manager.py smaller, this patch moves
168         the code that actually takes the command line arguments and
169         --test-file lists of tests to run and expands them into an
170         actual list of tests out into a separate module.
171
172         * Scripts/webkitpy/layout_tests/controllers/finder.py: Added.
173         (LayoutTestFinder):
174         (LayoutTestFinder.__init__):
175         (LayoutTestFinder.find_tests):
176         (LayoutTestFinder._strip_test_dir_prefixes):
177         (LayoutTestFinder._strip_test_dir_prefix):
178         (LayoutTestFinder._read_test_names_from_file):
179         * Scripts/webkitpy/layout_tests/controllers/manager.py:
180         (Manager._collect_tests):
181         (Manager._handle_finished_test):
182         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
183         (strip_comments):
184
185 2012-07-30  Adam Barth  <abarth@webkit.org>
186
187         commit-queue is corrupting ChangeLogs
188         https://bugs.webkit.org/show_bug.cgi?id=92681
189
190         Reviewed by Tony Chang.
191
192         Some commits made with the commit-queue are ending up with duplicate
193         ChangeLog entries. I've hot-patched the servers to not use a special
194         merge driver for ChangeLogs, which should fix this issue. This patch
195         removes the merge driver from our cold-boot.sh script so that we won't
196         use it on new commit-queue instances.
197
198         * EWSTools/cold-boot.sh:
199
200 2012-07-28  Sam Weinig  <sam@webkit.org>
201
202         Add ability to load from a string to the ObjC WK API
203         https://bugs.webkit.org/show_bug.cgi?id=92590
204
205         Reviewed by Dan Bernstein.
206
207         Add tests for [WKBrowsingContextController loadHTMLString:baseURL:]. We will be able
208         to greatly improve these tests (to test more than just not crashing) when methods to
209         access page content are added (soon!).
210
211         Adds:
212             Test: WKBrowsingContextLoadDelegateTest_SimpleLoadOfHTMLString
213             Test: WKBrowsingContextLoadDelegateTest_SimpleLoadOfHTMLString_NilBaseURL
214             Test: WKBrowsingContextLoadDelegateTest_SimpleLoadOfHTMLString_NilHTMLStringAndBaseURL
215
216         * TestWebKitAPI/Tests/WebKit2ObjC/WKBrowsingContextLoadDelegateTest.mm:
217         Adds tests and refactors delegates to not use global state.
218
219 2012-07-30  Thiago Marcos P. Santos  <thiago.santos@intel.com>
220
221         [CMake] TestWebKitAPI bundle should link with WTF
222         https://bugs.webkit.org/show_bug.cgi?id=92616
223
224         Reviewed by Antonio Gomes.
225
226         Fixes build when shared core is disabled.
227
228         * TestWebKitAPI/CMakeLists.txt:
229
230 2012-07-30  Dirk Pranke  <dpranke@chromium.org>
231
232         nrwt: clean up handling of 'expected' stats
233         https://bugs.webkit.org/show_bug.cgi?id=92527
234
235         Reviewed by Tony Chang.
236
237         This patch alters the way we compute and log the "expected"
238         results and how we treat skipped tests; we will now log the
239         number of skipped tests separately from the categories, e.g.:
240
241         Found 31607 tests; running 24464.
242         Expect: 23496 passes   (23496 now,    0 wontfix)
243         Expect:   548 failures (  543 now,    5 wontfix)
244         Expect:   420 flaky    (  245 now,  175 wontfix)
245
246         (so that the "expect" totals add up to the "running" totals);
247         in addition, the totals in the one-line-progress reflect the
248         number of tests we will actually run. If --iterations or
249         --repeat-each are specified, the number of tests we run are
250         multiplied as appropriate, but the "expect" numbers are
251         unchanged, since we don't count multiple invocations of the same
252         test multiple times. In addition, if we are using --run-part or
253         --run-chunk, the tests we don't run are treated as skipped
254         for consistency. We will also log the values for --iterations
255         and --repeat each as part of the found/running line.
256
257         Previously the code had parsed and re-parsed the
258         TestExpectations files several times in an attempt to come up
259         with some sane statistics, but this was expensive and lead to
260         confusing layer; treating files as skipped in the way described
261         above is more consistent and cleaner.
262
263         * Scripts/webkitpy/layout_tests/controllers/manager.py:
264         (Manager._split_into_chunks_if_necessary):
265         (Manager.prepare_lists_and_print_output):
266         (Manager.run):
267         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
268         (ManagerTest.test_interrupt_if_at_failure_limits):
269         (ManagerTest.test_update_summary_with_result):
270         (ManagerTest.test_look_for_new_crash_logs):
271         (ResultSummaryTest.get_result_summary):
272         * Scripts/webkitpy/layout_tests/models/result_summary.py:
273         (ResultSummary.__init__):
274         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
275         (TestExpectationParser.expectation_for_skipped_test):
276         (TestExpectations.__init__):
277         (TestExpectations.add_skipped_tests):
278           Here we make add_skipped_tests() public, so that we can update
279           the expectations for tests that we are skipping due to
280           --run-part or --run-chunk; we use the wontfix flag so that
281           the tests that are intentionally skipped aren't considered
282           "fixable".
283         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
284         (SkippedTests.check):
285         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
286         (parse_args):
287         * Scripts/webkitpy/layout_tests/views/printing.py:
288         (Printer.print_found):
289         (Printer):
290         (Printer.print_expected):
291         (Printer._print_result_summary):
292         (Printer._print_result_summary_entry):
293           Here we split out printing the number of tests found and run
294           from the expected results, to be clearer and so that we don't
295           have to reparse the expectations to update the stats.
296         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
297         (Testprinter.get_result_summary):
298
299 2012-07-30  Sadrul Habib Chowdhury  <sadrul@chromium.org>
300
301         Propagate gesture events to plugins.
302         https://bugs.webkit.org/show_bug.cgi?id=92281
303
304         Reviewed by Adam Barth.
305
306         Update TestWebPlugin to receive events, and output logs for the events.
307
308         * DumpRenderTree/chromium/TestWebPlugin.cpp:
309         (TestWebPlugin::handleInputEvent):
310         * DumpRenderTree/chromium/TestWebPlugin.h:
311         (TestWebPlugin::acceptsInputEvents):
312         (TestWebPlugin):
313
314 2012-07-30  Raphael Kubo da Costa  <rakuco@webkit.org>
315
316         [jhbuild] gnutls 2.12.14 does not build with glibc 2.16.0
317         https://bugs.webkit.org/show_bug.cgi?id=90643
318
319         Reviewed by Gustavo Noronha Silva.
320
321         The 2.x series of gnutls has a version of Gnulib that is
322         incompatible with glibc 2.16.0, which now does not define gets()
323         by default.
324
325         After talking to kov, mrobinson and philn, it looks like it makes
326         sense to simply remove gnutls from jhbuild.modules and rely on the
327         version installed on the system being recent enough.
328
329         This means at least version 2.12.8 for PKCS11 support, or at the
330         very least 2.11.0 (these are the minimum versions enforced by
331         glib-networking's configure.ac).
332
333         * efl/jhbuild.modules: Do not build gnutls anymore, make
334         glib-networking depend on glib instead, and move the libgcrypt
335         dependency to eet, which is what currently needs it.
336         * gtk/jhbuild.modules: Do not build gnutls anymore.
337
338 2012-07-30  Dirk Pranke  <dpranke@chromium.org>
339
340         nrwt: --verbose logging does not work right on windows
341         https://bugs.webkit.org/show_bug.cgi?id=92673
342
343         Unreviewed, build fix.
344
345         The --verbose log level is not being propagated to the worker
346         processes on windows properly; this has been broken ever since I
347         refactored logging into the message pool directly :(.
348
349         Also, fixing this revealed that outputcapture wasn't saving and
350         restoring log levels correctly, so I had to fix that as well.
351
352         * Scripts/webkitpy/common/message_pool.py:
353         (_MessagePool._start_workers):
354         (_MessagePool._worker_log_level):
355         (_Worker.__init__):
356         (_Worker._set_up_logging):
357         (_WorkerLogHandler.__init__):
358         * Scripts/webkitpy/common/system/outputcapture.py:
359         (OutputCapture.capture_output):
360         (OutputCapture.restore_output):
361
362 2012-07-30  Dirk Pranke  <dpranke@chromium.org>
363
364         After r123895, new-run-webkit-tests can fail with TypeError when a test crashes
365         https://bugs.webkit.org/show_bug.cgi?id=92664
366
367         Reviewed by Michael Saboff.
368
369         Updated to handle crash_logs being returned as a tuple of (stderr, crash_log).
370
371         * Scripts/webkitpy/layout_tests/port/mac.py:
372         (MacPort.look_for_new_crash_logs):
373
374 2012-07-30  Stephen White  <senorblanco@chromium.org>
375
376         Added a SkiaGraphics definition to the watchlist file; cc'ed myself
377         to that group.  Unreviewed.
378
379         * Scripts/webkitpy/common/config/watchlist:
380
381 2012-07-30  Sheriff Bot  <webkit.review.bot@gmail.com>
382
383         Unreviewed, rolling out r123966 and r123967.
384         http://trac.webkit.org/changeset/123966
385         http://trac.webkit.org/changeset/123967
386         https://bugs.webkit.org/show_bug.cgi?id=92656
387
388         This patch is causing assertion failures on the debug bot
389         (also rolling out a dependent patch) (Requested by mrobinson
390         on #webkit).
391
392         * gtk/generate-gtkdoc:
393         (get_webkit2_options):
394
395 2012-07-30  Martin Robinson  <mrobinson@igalia.com>
396
397         [GTK] Add a non-subprocess jhbuild environment and use it for run-gtk-tests
398         https://bugs.webkit.org/show_bug.cgi?id=92626
399
400         Reviewed by Gustavo Noronha Silva.
401
402         Instead of always shelling out to enter a jhbuild environment, add a Python
403         hook for entering a jhbuild environment directly. This avoids requiring a
404         script to be wrapped in jhbuild.
405
406         * GNUmakefile.am: Remove gtk/run-api-tests from the distribution list.
407         * Scripts/run-gtk-tests: Move most of the logic from run-api-tests here
408         with additional code for entering a jhbuild environment directly.
409         * gtk/run-api-tests: Removed.
410         * jhbuild/jhbuild-wrapper: Use the new helper methods defined in jhbuildutils.py.
411         * jhbuild/jhbuildutils.py: Add some new helper methods here including
412         a method which can transform the current environment into a jhbuild-ified one.
413
414 2012-07-30  Alexander Pavlov  <apavlov@chromium.org>
415
416         Unreviewed, upgrade apavlov to reviewer.
417         http://lists.webkit.org/mailman/private/webkit-committers/2012-July/000184.html
418
419         * Scripts/webkitpy/common/config/committers.py:
420
421 2012-07-30  Robin Cao  <robin.cao@torchmobile.com.cn>
422
423         [BlackBerry] Adapt to changes in the SharedArray platform API
424         https://bugs.webkit.org/show_bug.cgi?id=92631
425
426         Reviewed by Rob Buis.
427
428         Adapt to changes in the SharedArray platform API. No behavioural change.
429
430         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
431         (LayoutTestController::webHistoryItemCount):
432
433 2012-07-30  Balazs Kelemen  <kbalazs@webkit.org>
434
435         [Qt][NRWT] REGRESSION(123729): Forcing pixel tests with -p doesn't work
436         https://bugs.webkit.org/show_bug.cgi?id=92627
437
438         Reviewed by Noam Rosenthal.
439
440         Remove the default behavior I added in r123729 so that we run all tests as pixel
441         test with -p. It was intentional but it seems like we don't have consensus on
442         that we want it.
443
444         * Scripts/webkitpy/layout_tests/port/qt.py:
445         (QtPort._supports_switching_pixel_tests_per_test):
446
447 2012-07-30  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
448
449         [EFL][WTR] Enforce SHIFT modifier to upper case keys in event sender.
450         https://bugs.webkit.org/show_bug.cgi?id=92366
451
452         Reviewed by Antonio Gomes.
453
454         Add SHIFT modifier for capital letters as VK_
455         codes do not distinguish between lows and caps and also to be consistent
456         with other ports which send caps as lowercase letters with the shift key down.
457
458         * WebKitTestRunner/efl/EventSenderProxyEfl.cpp:
459         (WTR::EventSenderProxy::keyDown):
460
461 2012-07-29  Raphael Kubo da Costa  <rakuco@webkit.org>
462
463         [EFL][jhbuild] Make run-with-jhbuild become the jhbuild process itself.
464         https://bugs.webkit.org/show_bug.cgi?id=92592
465
466         Reviewed by Eric Seidel.
467
468         Adapt to the changes made to GTK+'s run-with-jhbuild in r123979
469         and make EFL's run-with-jhbuild call os.execve() instead of
470         launching a subprocess as well.
471
472         This should make it possible to interrupt execution of jhbuild
473         cleanly.
474
475         * efl/run-with-jhbuild:
476
477 2012-07-28  Ryosuke Niwa  <rniwa@webkit.org>
478
479         run-perf-tests should generate a results page
480         https://bugs.webkit.org/show_bug.cgi?id=92575
481
482         Reviewed by Eric Seidel.
483
484         Added the ability to generate a results page that summarizes performance test results from
485         multiple runs of run-perf-tests when --output-json-path is specified and --test-results-server
486         is not specified. We cannot generate a results page when --test-results-server is specified
487         because perf-o-matic cannot parse the new JSON format.
488
489         The new JSON format is simply an array of the old JSON output. This JSON is then merged into
490         results-template.html along with jQuery and copied as a HTML file of the same filename as
491         the output JSON file with a .html extension.
492
493         We merge all scripts and the JSON output into the results page as opposed to including them
494         with the src content attribute to make it standalone so that we may post it on Bugzilla or
495         e-mail it to someone else without breaking it.
496
497         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
498         (PerfTestsRunner.run):
499         (PerfTestsRunner._generate_json): Merge "contents" with the existing JSON file when generating
500         a results page, and generate the results page from results-template.html merged with jQuery and
501         the generated JSON output.
502         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
503         (_test_run_with_json_output): Extracted from test_run_with_upload_json to be shared with
504         several test cases.
505         (_test_run_with_json_output.mock_upload_json): Ditto.
506         (test_run_with_json_output): Refactored to use _test_run_with_json_output.
507         (test_run_generates_results_page): Added. Tests the new feature.
508         (test_run_with_json_source): Refactored to use _test_run_with_json_output.
509         (test_run_with_multiple_repositories): Ditto.
510         (test_run_with_upload_json): Ditto.
511
512 2012-07-29  Rik Cabanier  <cabanier@adobe.com>
513
514         Add ENABLE_CSS_COMPOSITING flag
515         https://bugs.webkit.org/show_bug.cgi?id=92553
516
517         Reviewed by Dirk Schulze.
518
519         Adds compiler flag CSS_COMPOSITING to build systems to enable CSS blending and compositing. See spec https://dvcs.w3.org/hg/FXTF/rawfile/tip/compositing/index.html
520
521         * Scripts/webkitperl/FeatureList.pm:
522
523 2012-07-29  Gustavo Noronha Silva  <gns@gnome.org>
524
525         Reviewed by Martin Robinson.
526
527         Make the jhbuild wrapper scripts become the jhbuild process
528         instead of starting it as a child. This should make it possible to
529         interrupt execution of jhbuild cleanly.
530
531         * gtk/run-with-jhbuild:
532         * jhbuild/jhbuild-wrapper:
533
534 2012-07-29  Mike West  <mkwst@chromium.org>
535
536         Extend `application/x-webkit-test-netscape` plugins to better support multiple frames.
537         https://bugs.webkit.org/show_bug.cgi?id=92478
538
539         Reviewed by Adam Barth.
540
541         DumpRenderTree currently understands
542         `<object src="data:application/x-webkit-test-netscape,alertwhenloaded">`.
543         It would be useful if it understood `logifloaded`, which could give
544         more context regarding test expectations, and can be written in such
545         as way as to support plugins loaded into multiple iframes in a single
546         test.
547
548         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
549         (NPP_New):
550             Adding support for `logifloaded`.
551
552 2012-07-28  Peter Kasting  <pkasting@google.com>
553
554         Fix Chromium/Win compile.
555         https://bugs.webkit.org/show_bug.cgi?id=88787
556
557         Unreviewed, build fix.
558
559         Chromium/Win needs _USE_MATH_DEFINES but doesn't define PLATFORM(WIN).
560         The other files in WebKit that define this macro (both in
561         WebCore/platform/graphics/cg) do it by unconditionally #defining it to 1
562         above the #include of "config.h", so copy that pattern here.
563
564         * TestWebKitAPI/Tests/WTF/MediaTime.cpp:
565
566 2012-07-28  Mario Sanchez Prada  <msanchez@igalia.com>
567
568         [WK2][GTK] Implement a new spell checker API for WebKit2GTK+
569         https://bugs.webkit.org/show_bug.cgi?id=90268
570
571         Reviewed by Martin Robinson.
572
573         Ignore WebKitTextChecker.* private files for gtk-doc.
574
575         * gtk/generate-gtkdoc:
576         (get_webkit2_options):
577
578 2012-07-28  Philippe Normand  <pnormand@igalia.com>
579
580         [GTK] Transparent Media controls timeline and panel
581         https://bugs.webkit.org/show_bug.cgi?id=85279
582
583         Reviewed by Martin Robinson.
584
585         * gtk/jhbuild.modules: Provide gnome-themes-standard build
586         support, this is necessary so the media controls panel colors are
587         properly initialized from the GTK+ theme.
588
589 2012-07-27  Kihong Kwon  <kihong.kwon@samsung.com>
590
591         [EFL] Support for HTML media capture
592         https://bugs.webkit.org/show_bug.cgi?id=91842
593
594         Reviewed by Ryosuke Niwa.
595
596         Add HTML media capture feature to the feature list.
597         In addition, enable that on the Efl port.
598
599         * Scripts/webkitperl/FeatureList.pm:
600
601 2012-07-27  Dirk Pranke  <dpranke@chromium.org>
602
603         "webkit-patch rebaseline <test>" fails with exit_code 1
604         https://bugs.webkit.org/show_bug.cgi?id=91560
605
606         Reviewed by Adam Barth.
607
608         webkit-patch rebaseline-json crashes in an svn checkout if it
609         tries to add files and there are no files to add; arguably
610         svn.add() should be robust against this, but rebaseline-json
611         shouldn't be calling it, either. This patch fixes the latter and
612         adds better debugging info to garden-o-matic so we can see what
613         caused the crash.
614
615         * Scripts/webkitpy/tool/commands/rebaseline.py:
616         (AbstractParallelRebaselineCommand._rebaseline):
617         * Scripts/webkitpy/tool/servers/gardeningserver.py:
618         (GardeningHTTPRequestHandler.rebaselineall):
619         (GardeningHTTPRequestHandler.rebaselineall.error_handler):
620         (GardeningHTTPRequestHandler):
621
622 2012-07-27  Brady Eidson  <beidson@apple.com>
623
624         Plugins should not be allowed to override standard properties/attributes in non-standard worlds
625         <rdar://problem/11975252> and https://bugs.webkit.org/show_bug.cgi?id=92519
626
627         Reviewed by Anders Carlsson.
628
629         Add a test plugin which overrides all properties it is asked about:
630         * DumpRenderTree/TestNetscapePlugIn/Tests/PluginScriptableObjectOverridesAllProperties.cpp: Added.
631         (PluginScriptableObjectOverridesAllProperties):
632         (PluginScriptableObjectOverridesAllProperties::PluginScriptableObjectOverridesAllProperties):
633         (PluginObject):
634         (PluginScriptableObjectOverridesAllProperties::PluginObject::PluginObject):
635         (PluginScriptableObjectOverridesAllProperties::PluginObject::~PluginObject):
636         (PluginScriptableObjectOverridesAllProperties::PluginObject::hasProperty):
637         (PluginScriptableObjectOverridesAllProperties::PluginObject::getProperty):
638         (PluginScriptableObjectOverridesAllProperties::NPP_GetValue):
639
640         Expose NPN_MemAlloc to the test plugin:
641         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
642         (PluginTest::NPN_MemAlloc):
643         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
644         (PluginTest):
645
646         * DumpRenderTree/DumpRenderTree.gypi:
647         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
648         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
649         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
650
651 2012-07-27  Rafael Brandao  <rafael.lobo@openbossa.org>
652
653         "webkit-patch upload" should strip colors from patch if necessary
654         https://bugs.webkit.org/show_bug.cgi?id=92306
655
656         Reviewed by Adam Barth.
657
658         * Scripts/webkitpy/common/checkout/scm/git.py: Explicity remove colors from diff.
659
660 2012-07-27  Eric Seidel  <eric@webkit.org>
661
662         clean-pending-commit throws exception when encountering Ossy
663         https://bugs.webkit.org/show_bug.cgi?id=92534
664
665         Reviewed by Adam Barth.
666
667         Mark a couple strings as unicode so that reviewers with unicode names do not cause
668         clean-pending-commit to barf.
669
670         * Scripts/webkitpy/tool/commands/upload.py:
671         (CleanPendingCommit._flags_to_clear_on_patch):
672         (CleanPendingCommit.execute):
673         (AssignToCommitter._assign_bug_to_last_patch_attacher):
674
675 2012-07-27  Dirk Pranke  <dpranke@chromium.org>
676
677         chromium win compile broken after removing webkit.py
678         https://bugs.webkit.org/show_bug.cgi?id=92549
679
680         Unreviewed, build fix.
681
682         It turns out that this file is listed in the chromium downstream
683         'browser_tests.isolate' file, and so deleting it broke gyp.
684         Adding a stub file back in until we can remove the entry from
685         the file.
686
687         * Scripts/webkitpy/layout_tests/port/webkit.py: Added.
688
689 2012-07-27  Dirk Pranke  <dpranke@chromium.org>
690
691         nrwt: remove 'webkit.py' and the 'WebKitPort'
692         https://bugs.webkit.org/show_bug.cgi?id=92450
693
694         Reviewed by Ryosuke Niwa.
695
696         This patch finishes the work of merging WebKitPort into Port.
697         Now Port is way too big :).
698
699         This is all cutting and pasting; no new functionality and
700         no changes in the tests.
701
702         * Scripts/webkitpy/layout_tests/port/apple.py:
703         (ApplePort):
704         (ApplePort.__init__):
705         * Scripts/webkitpy/layout_tests/port/base.py:
706         (Port.skipped_layout_tests):
707         (Port):
708         (Port._port_flag_for_scripts):
709         (Port._arguments_for_configuration):
710         (Port._run_script):
711         (Port._build_driver):
712         (Port._build_driver_flags):
713         (Port._tests_for_other_platforms):
714         (Port._runtime_feature_list):
715         (Port.nm_command):
716         (Port._modules_to_search_for_symbols):
717         (Port._symbols_string):
718         (Port._missing_feature_to_skipped_tests):
719         (Port._missing_symbol_to_skipped_tests):
720         (Port._has_test_in_directories):
721         (Port._skipped_tests_for_unsupported_features):
722         (Port._wk2_port_name):
723         (Port._skipped_file_search_paths):
724         * Scripts/webkitpy/layout_tests/port/chromium.py:
725         (ChromiumPort):
726         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
727         * Scripts/webkitpy/layout_tests/port/chromium_port_testcase.py:
728         (ChromiumPortTestCase.test_check_build):
729         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
730         * Scripts/webkitpy/layout_tests/port/efl.py:
731         (EflPort):
732         (EflPort.__init__):
733         * Scripts/webkitpy/layout_tests/port/gtk.py:
734         (GtkPort):
735         (GtkPort.setup_environ_for_server):
736         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
737         (TestWebKitPort):
738         (TestWebKitPort.__init__):
739         (TestWebKitPort.all_test_configurations):
740         (TestWebKitPort._symbols_string):
741         (TestWebKitPort._tests_for_other_platforms):
742         (TestWebKitPort._tests_for_disabled_features):
743         (PortTestCase):
744         (PortTestCase.test_diff_image):
745         (PortTestCase.test_path_to_test_expectations_file):
746         (PortTestCase.test_skipped_directories_for_symbols):
747         (test_skipped_directories_for_features):
748         (test_skipped_directories_for_features_no_matching_tests_in_test_list):
749         (test_skipped_tests_for_unsupported_features_empty_test_list):
750         (test_skipped_layout_tests):
751         (test_skipped_file_search_paths):
752         (test_root_option):
753         (test_test_expectations):
754         (test_build_driver):
755         (_assert_config_file_for_platform):
756         (test_linux_distro_detection):
757         (test_apache_config_file_name_for_platform):
758         (test_path_to_apache_config_file):
759         * Scripts/webkitpy/layout_tests/port/qt.py:
760         (QtPort):
761         (QtPort.__init__):
762         (QtPort.setup_environ_for_server):
763         * Scripts/webkitpy/layout_tests/port/webkit.py: Removed.
764         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py: Removed.
765         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
766         (get_test_baselines.AllPlatformsPort):
767         (get_test_baselines.AllPlatformsPort.__init__):
768         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
769         (get_test_config.TestMacPort):
770
771 2012-07-27  Christophe Dumez  <christophe.dumez@intel.com>
772
773         [WK2][WTR] LayoutTestController.deliverWebIntent() needs to be implemented
774         https://bugs.webkit.org/show_bug.cgi?id=92228
775
776         Reviewed by Anders Carlsson.
777
778         Implement LayoutTestController.deliverWebIntent() to
779         allow test cases to deliver a Web intent and test this
780         functionality.
781
782         This functionality is needed by:
783         - webintents/web-intents-delivery.html
784
785         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
786         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
787         (WTR::LayoutTestController::deliverWebIntent):
788         (WTR):
789         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
790         (LayoutTestController):
791
792 2012-07-27  Jer Noble  <jer.noble@apple.com>
793
794         Unreviewed build fix.
795
796         Wrap MSVC-specific fixes is a COMPILER() check, not a PLATFORM() check.
797
798         * TestWebKitAPI/Tests/WTF/MediaTime.cpp:
799
800 2012-07-27  Xianzhu Wang  <wangxianzhu@chromium.org>
801
802         [Chromium-Android] NRWT better handling of DRT deadlocks and crashes
803         https://bugs.webkit.org/show_bug.cgi?id=92299
804
805         Reviewed by Dirk Pranke.
806
807         1. Changed the deadlock detector thread target from a method of driver to a function to avoid its reference to the
808            current driver (see the bug for the problem of the reference.
809         2. Let the deadlock detector thread exit early when the normal_startup_event is set to avoid it from blocking the main thread when the main thread is exiting.
810         3. Extracted common loop until timeout logic into ChromiumAndroidDriver._loop_with_timeout().
811         4. Use the "for i in range(n)" style suggested by dpranke in https://bugs.webkit.org/show_bug.cgi?id=89124 in several functions.
812
813         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
814         (ChromiumAndroidPort.__init__):
815         (ChromiumAndroidPort._run_adb_command): Reduced the verbose output.
816         (ChromiumAndroidDriver.__init__):
817         (ChromiumAndroidDriver._loop_with_timeout): Extracted common loop until timeout logic here.
818         (ChromiumAndroidDriver._all_pipes_created):
819         (ChromiumAndroidDriver._remove_all_pipes):
820         (ChromiumAndroidDriver._start): Changed the retry loop style.
821         (ChromiumAndroidDriver._start_once): Changed the loop style and code related to deadlock detector.
822         (ChromiumAndroidDriver._start_once.deadlock_detector): Moved the original ChromiumAndroidDriver._deadlock_detector() here.
823         (ChromiumAndroidDriver.stop): Changed the loop style.
824
825 2012-07-27  Dirk Pranke  <dpranke@chromium.org>
826
827         nrwt: move collect_tests(), parse_expectations() into manager
828         https://bugs.webkit.org/show_bug.cgi?id=92423
829
830         Reviewed by Ojan Vafai.
831
832         This is the first in a series of patches refactoring
833         run_webkit_tests.py and manager.py to make them more
834         maintainable; all this patch does is move a couple of functions
835         into manager.run_tests() so manager's public interface is
836         simple.
837
838         * Scripts/webkitpy/layout_tests/controllers/manager.py:
839         (Manager._collect_tests):
840         (Manager._parse_expectations):
841         (Manager._split_into_chunks_if_necessary):
842         (Manager.run):
843         (read_test_files):
844         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
845         (ManagerTest.test_http_locking):
846         (ManagerTest.test_perf_locking):
847         (ManagerTest.integration_test_needs_servers.get_manager_with_tests):
848         (ManagerTest.test_look_for_new_crash_logs.get_manager_with_tests):
849         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
850         (run):
851
852 2012-07-27  Dirk Pranke  <dpranke@chromium.org>
853
854         nrwt: print filtered stderr after a crash correctly
855         https://bugs.webkit.org/show_bug.cgi?id=92428
856
857         Reviewed by Adam Barth.
858
859         Some ports will run the stderr output from DRT/WTR through a
860         filter after a crash in order to print a more useful stack
861         trace; while that gets written to the crash log correctly, we
862         don't log it to the debug output properly. This change fixes
863         that and makes it clearer that we're logging stderr, not the
864         crash log.
865
866         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
867         (SingleTestRunner._handle_error):
868         * Scripts/webkitpy/layout_tests/port/base.py:
869         (Port._get_crash_log):
870         * Scripts/webkitpy/layout_tests/port/chromium.py:
871         (ChromiumPort._get_crash_log):
872         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
873         (ChromiumAndroidPortTest.test_get_crash_log):
874         * Scripts/webkitpy/layout_tests/port/driver.py:
875         (Driver.run_test):
876         * Scripts/webkitpy/layout_tests/port/gtk.py:
877         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
878         * Scripts/webkitpy/layout_tests/port/mac.py:
879         (MacPort._get_crash_log):
880         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
881         (PortTestCase.test_get_crash_log):
882
883 2012-07-27  Dirk Pranke  <dpranke@chromium.org>
884
885         nrwt: move image diffing code to a separate module
886         https://bugs.webkit.org/show_bug.cgi?id=92447
887
888         Reviewed by Ryosuke Niwa.
889
890         This patch moves the code to talk to ImageDiff into its own
891         module, and adds more tests for it. In addition, the patch
892         modifies diff_image() so that we don't automatically stop
893         ImageDiff after a single invocation, and thus subsequent
894         diffs may be slightly faster. (Note that the chromium ports
895         don't use any of this code; that is not changed by this patch).
896
897         The main motivation for this change is to move more "generic"
898         code out of the port/* classes, and in particular to move more
899         code out of webkit.py so that we can eventually eliminate it by
900         merging it into base.py.
901
902         This patch also splits MockServerProcess out from driver_unittest.py
903         so that it can be re-used.
904
905         * Scripts/webkitpy/layout_tests/port/base.py:
906         (Port.__init__):
907         (Port.diff_image):
908         (Port.clean_up_test_run):
909         * Scripts/webkitpy/layout_tests/port/driver.py:
910         (Driver.__init__):
911         (Driver._start):
912         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
913         (DriverTest.test_stop_cleans_up_properly):
914         (DriverTest.test_two_starts_cleans_up_properly):
915         (DriverTest.test_start_actually_starts):
916         * Scripts/webkitpy/layout_tests/port/efl.py:
917         (EflPort.clean_up_test_run):
918         * Scripts/webkitpy/layout_tests/port/gtk.py:
919         (GtkPort.clean_up_test_run):
920         * Scripts/webkitpy/layout_tests/port/image_diff.py: Added.
921         (ImageDiffer):
922         (ImageDiffer.__init__):
923         (ImageDiffer.diff_image):
924         (ImageDiffer._start):
925         (ImageDiffer._read):
926         (ImageDiffer.stop):
927         * Scripts/webkitpy/layout_tests/port/image_diff_unittest.py: Added.
928         (for):
929         (FakePort):
930         (FakePort.__init__):
931         (FakePort._path_to_image_diff):
932         (FakePort.setup_environ_for_server):
933         (TestImageDiffer):
934         (TestImageDiffer.test_diff_image):
935         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
936         (MockDRTPortTest.test_diff_image):
937         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
938         (PortTestCase.test_diff_image__missing_both):
939         (PortTestCase.test_diff_image):
940         (PortTestCase.test_diff_image.make_proc):
941         * Scripts/webkitpy/layout_tests/port/server_process.py:
942         (ServerProcess._start):
943         * Scripts/webkitpy/layout_tests/port/server_process_mock.py: Added.
944         (MockServerProcess):
945         (MockServerProcess.__init__):
946         (MockServerProcess.write):
947         (MockServerProcess.has_crashed):
948         (MockServerProcess.read_stdout_line):
949         (MockServerProcess.read_stdout):
950         (MockServerProcess.pop_all_buffered_stderr):
951         (MockServerProcess.read_either_stdout_or_stderr_line):
952         (MockServerProcess.start):
953         (MockServerProcess.stop):
954         (MockServerProcess.kill):
955         * Scripts/webkitpy/layout_tests/port/webkit.py:
956         (WebKitPort._build_driver_flags):
957         (WebKitPort._symbols_string):
958
959 2012-07-27  Tom Hudson  <hudson@google.com>
960
961         Activate committer bit granted April 6th so I can garden.
962         https://bugs.webkit.org/show_bug.cgi?id=92500
963
964         Reviewed by Jochen Eisinger.
965
966         * Scripts/webkitpy/common/config/committers.py:
967
968 2012-07-27  Jer Noble  <jer.noble@apple.com>
969
970         Support a rational time class for use by media elements.
971         https://bugs.webkit.org/show_bug.cgi?id=88787
972
973         Add unit tests for the WTF::MediaTime class.
974
975         Reviewed by Eric Carlson.
976
977         * TestWebKitAPI/GNUmakefile.am:
978         * TestWebKitAPI/TestWebKitAPI.gypi:
979         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
980         * TestWebKitAPI/Tests/WTF/MediaTime.cpp: Added.
981         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
982         (WTF::operator<<):
983         (TestWebKitAPI):
984         (TestWebKitAPI::TEST):
985
986 2012-07-27  Balazs Kelemen  <kbalazs@webkit.org>
987
988         [NRWT] should have a way to restrict pixel tests for individual directories
989         https://bugs.webkit.org/show_bug.cgi?id=91754
990
991         Unreviewed attempt to fix the chaos I introduced.
992
993         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
994         (WebCore::DumpRenderTree::processLine):
995
996 2012-07-27  YoungTaeck Song  <youngtaeck.song@samsung.com>
997
998         [WK2][EFL] Add an ACCELERATED_COMPOSITING implementation for Efl WebKit2
999         https://bugs.webkit.org/show_bug.cgi?id=91581
1000
1001         Reviewed by Noam Rosenthal.
1002
1003         Add a MiniBrowser's option for selecting evas engine.
1004
1005         * MiniBrowser/efl/main.c:
1006         (browserCreate):
1007         (main):
1008
1009 2012-07-27  Sheriff Bot  <webkit.review.bot@gmail.com>
1010
1011         Unreviewed, rolling out r123869.
1012         http://trac.webkit.org/changeset/123869
1013         https://bugs.webkit.org/show_bug.cgi?id=92501
1014
1015         "it did not fix anything but made things even worst"
1016         (Requested by kbalazs on #webkit).
1017
1018         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1019         (WebCore::DumpRenderTree::DumpRenderTree):
1020         (WebCore::DumpRenderTree::open):
1021         (WebCore::DumpRenderTree::processLine):
1022         (WebCore::DumpRenderTree::setDumpPixelsForAllTests):
1023         (WebCore::DumpRenderTree::dump):
1024         * DumpRenderTree/qt/DumpRenderTreeQt.h:
1025         (DumpRenderTree):
1026         * DumpRenderTree/qt/main.cpp:
1027         (main):
1028         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
1029         (SingleTestRunner._should_fetch_expected_checksum):
1030         (SingleTestRunner._overwrite_baselines):
1031         (SingleTestRunner._compare_output):
1032         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1033         (Worker._update_test_input):
1034         * Scripts/webkitpy/layout_tests/port/base.py:
1035         (Port.supports_switching_pixel_tests_per_test):
1036         (Port):
1037         (Port._supports_switching_pixel_tests_per_test):
1038         (Port.should_run_as_pixel_test):
1039         (Port._should_run_as_pixel_test):
1040         * Scripts/webkitpy/layout_tests/port/driver.py:
1041         (Driver.cmd_line):
1042         (Driver._command_from_driver_input):
1043         * Scripts/webkitpy/layout_tests/port/qt.py:
1044         (QtPort.check_sys_deps):
1045         (QtPort):
1046         (QtPort._supports_switching_pixel_tests_per_test):
1047         (QtPort._should_run_as_pixel_test):
1048         (QtPort._default_pixel_test_directories):
1049         * Scripts/webkitpy/layout_tests/port/test.py:
1050         (TestPort.supports_switching_pixel_tests_per_test):
1051         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1052         (_set_up_derived_options):
1053         (parse_args):
1054         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1055         (MainTest.test_run_singly_actually_runs_tests):
1056         (MainTest.test_pixel_test_directories):
1057         * WebKitTestRunner/TestController.cpp:
1058         (WTR::TestController::TestController):
1059         (WTR::TestController::initialize):
1060         (WTR::TestController::runTest):
1061         * WebKitTestRunner/TestController.h:
1062         (TestController):
1063
1064 2012-07-27  Balazs Kelemen  <kbalazs@webkit.org>
1065
1066         [NRWT] should have a way to restrict pixel tests for individual directories
1067         https://bugs.webkit.org/show_bug.cgi?id=91754
1068
1069         Reviewed by Zoltan Herczeg.
1070
1071         Rollout r123729 because it made Qt debug bots crasy.
1072
1073         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1074         (WebCore::DumpRenderTree::DumpRenderTree):
1075         (WebCore::DumpRenderTree::open):
1076         (WebCore::DumpRenderTree::processLine):
1077         (WebCore::DumpRenderTree::setDumpPixels):
1078         (WebCore::DumpRenderTree::dump):
1079         * DumpRenderTree/qt/DumpRenderTreeQt.h:
1080         (DumpRenderTree):
1081         * DumpRenderTree/qt/main.cpp:
1082         (main):
1083         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
1084         (SingleTestRunner._should_fetch_expected_checksum):
1085         (SingleTestRunner._overwrite_baselines):
1086         (SingleTestRunner._compare_output):
1087         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1088         (Worker._update_test_input):
1089         * Scripts/webkitpy/layout_tests/port/base.py:
1090         (Port.lookup_virtual_test_args):
1091         * Scripts/webkitpy/layout_tests/port/driver.py:
1092         (Driver.cmd_line):
1093         (Driver._command_from_driver_input):
1094         * Scripts/webkitpy/layout_tests/port/qt.py:
1095         (QtPort.check_sys_deps):
1096         * Scripts/webkitpy/layout_tests/port/test.py:
1097         (TestPort.virtual_test_suites):
1098         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1099         (_set_up_derived_options):
1100         (parse_args):
1101         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1102         (MainTest.test_run_singly_actually_runs_tests):
1103         (MainTest.test_missing_and_unexpected_results):
1104         * WebKitTestRunner/TestController.cpp:
1105         (WTR::TestController::TestController):
1106         (WTR::TestController::initialize):
1107         (WTR::TestController::runTest):
1108         * WebKitTestRunner/TestController.h:
1109         (TestController):
1110
1111 2012-07-27  Adam Barth  <abarth@webkit.org>
1112
1113         webkit.review.bot should run clean-review-queue and clean-pending-commit periodically
1114         https://bugs.webkit.org/show_bug.cgi?id=92472
1115
1116         Reviewed by Eric Seidel.
1117
1118         We need to run these command periodically to keep cruft from
1119         accumulating in bugs.webkit.org. Rather than running them manually, we
1120         should just have a bot run them. This patch has the style-queue run the
1121         commands, which admittedly is a bit odd but it doesn't seem worthwhile
1122         to create another bot specifically for this purposes.
1123
1124         * Scripts/webkitpy/tool/commands/queues.py:
1125         (AbstractReviewQueue):
1126         (StyleQueue.begin_work_queue):
1127         (StyleQueue):
1128         (StyleQueue.clean_bugzilla):
1129         * Scripts/webkitpy/tool/commands/queues_unittest.py:
1130         (StyleQueueTest.test_style_queue_with_style_exception):
1131         (test_style_queue_with_watch_list_exception):
1132
1133 2012-07-27  Dominik Röttsches  <dominik.rottsches@intel.com>
1134
1135         [Cairo] Add complex font drawing using HarfbuzzNG
1136         https://bugs.webkit.org/show_bug.cgi?id=91864
1137
1138         Reviewed by Simon Hausmann and Martin Robinson.
1139
1140         Bringing Harfbuzz support to EFL with this patch, so we need HarfBuzz in the pulled in jhbuild dependencies.
1141
1142         * WebKitTestRunner/CMakeLists.txt: Adding additional header include directories.
1143         * efl/jhbuild.modules: Adding source for HarfBuzz release version 0.9.0.
1144
1145 2012-07-27  KwangYong Choi  <ky0.choi@samsung.com>
1146
1147         [WK2][EFL][WTR] TestNetscapePlugin is required.
1148         https://bugs.webkit.org/show_bug.cgi?id=88756
1149
1150         Reviewed by Simon Hausmann.
1151
1152         libTestNetscapePlugin.so should be required for DRT/WTR/EFL.
1153
1154         * CMakeLists.txt:
1155         * DumpRenderTree/TestNetscapePlugin/CMakeList.txt:
1156
1157 2012-07-27  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1158
1159         [EFL] Memory leaks in EFL DRT
1160         https://bugs.webkit.org/show_bug.cgi?id=92347
1161
1162         Reviewed by Simon Hausmann.
1163
1164         Fix a memory leak in EFL's DRT code by adopting an allocation
1165         of JSString created with JSStringCreateWithUTF8CString.
1166
1167         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1168         (DumpRenderTreeChrome::onWindowObjectCleared):
1169
1170 2012-07-27  Adam Barth  <abarth@webkit.org>
1171
1172         Simplify ExpectedFailures
1173         https://bugs.webkit.org/show_bug.cgi?id=92216
1174
1175         Reviewed by Eric Seidel.
1176
1177         This patch simplifies the ExpectedFailures class we use to remember
1178         which tests are currently failing on the bots. When we wrote this code
1179         originally, we weren't entirely sure how it would work. Now that we
1180         understand it more clearly, we can write the code more clearly.
1181
1182         * Scripts/webkitpy/tool/bot/expectedfailures.py:
1183         (_has_failures):
1184         (_is_trustworthy):
1185         (ExpectedFailures.__init__):
1186         (ExpectedFailures.failures_were_expected):
1187         (ExpectedFailures.unexpected_failures_observed):
1188         (ExpectedFailures.update):
1189         * Scripts/webkitpy/tool/bot/expectedfailures_unittest.py:
1190         (ExpectedFailuresTest._assert_can_trust):
1191         (ExpectedFailuresTest.test_failures_were_expected):
1192         (ExpectedFailuresTest.test_unexpected_failures_observed):
1193         (ExpectedFailuresTest.test_unexpected_failures_observed_when_tree_is_hosed):
1194         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
1195         (PatchAnalysisTask._test):
1196         (PatchAnalysisTask._build_and_test_without_patch):
1197         (PatchAnalysisTask._test_patch):
1198
1199 2012-07-27  Csaba Osztrogonác  <ossy@webkit.org>
1200
1201         [Qt][WK2] REGRESSION(r119127): resetting window.internals settings between tests doesn't work properly
1202         https://bugs.webkit.org/show_bug.cgi?id=88064
1203
1204         Reviewed by Zoltan Herczeg.
1205
1206         Use resetInternalsObject() instead of injectInternalsObject().
1207
1208         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1209         (WTR::InjectedBundlePage::resetAfterTest):
1210
1211 2012-07-26  Ryuan Choi  <ryuan.choi@samsung.com>
1212
1213         [EFL] Remove forwarding headers from MiniBrowser/Efl.
1214         https://bugs.webkit.org/show_bug.cgi?id=91627
1215
1216         Reviewed by Kentaro Hara.
1217
1218         * MiniBrowser/efl/CMakeLists.txt: Remove unnecessary forwarding headers.
1219
1220 2012-07-26  Yoshifumi Inoue  <yosin@chromium.org>
1221
1222         [Forms] Introduce runtime feature flags for input type datetime, datetimelocal, month, time, week
1223         https://bugs.webkit.org/show_bug.cgi?id=92339
1224
1225         Reviewed by Kent Tamura.
1226
1227         This patch enables runtime features for input type datetime, datetime-local,
1228         month, time and week for Chromium DRT.
1229
1230         * DumpRenderTree/chromium/TestShell.cpp:
1231         (TestShell::TestShell): Call enableInputType{DateTime,DateTimeLocal,Month,Time,Week}.
1232
1233 2012-07-26  Jeffrey Pfau  <jpfau@apple.com>
1234
1235         Reloading substitute-data/alternate html string for unreachableURL will add an item to the back-forward-history for each reload
1236         https://bugs.webkit.org/show_bug.cgi?id=84041
1237
1238         Reviewed by Brady Eidson.
1239
1240         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1241         * TestWebKitAPI/Tests/mac/BackForwardList.mm: Added.
1242         (-[BackForwardListTest webView:didFinishLoadForFrame:]):
1243         (-[BackForwardListTest webView:didFailProvisionalLoadWithError:forFrame:]):
1244         (TestWebKitAPI):
1245         (TestWebKitAPI::TEST):
1246
1247 2012-07-26  Sheriff Bot  <webkit.review.bot@gmail.com>
1248
1249         Unreviewed, rolling out r123799.
1250         http://trac.webkit.org/changeset/123799
1251         https://bugs.webkit.org/show_bug.cgi?id=92431
1252
1253         broke the Apple Mac build (Requested by bweinstein on
1254         #webkit).
1255
1256         * DumpRenderTree/chromium/TestWebPlugin.cpp:
1257         * DumpRenderTree/chromium/TestWebPlugin.h:
1258         (TestWebPlugin::acceptsInputEvents):
1259         (TestWebPlugin::handleInputEvent):
1260
1261 2012-07-26  Dirk Pranke  <dpranke@chromium.org>
1262
1263         nrwt: pass --verbose twice on the bots to ensure getting the debug output for now
1264         https://bugs.webkit.org/show_bug.cgi?id=92172
1265
1266         Reviewed by Tony Chang.
1267
1268         Right now new-run-webkit-tests --verbose prints out too much
1269         information (it's like --verbose + debugging). I plan to fix
1270         this in bug 88702, and also clean up and simplify all of the
1271         --print options, but in preparation for these changes I need to
1272         pass --verbose twice to get the same level of output
1273         temporarily (for compatibility).
1274
1275         * Scripts/run-webkit-tests:
1276
1277 2012-07-26  Sadrul Habib Chowdhury  <sadrul@chromium.org>
1278
1279         Propagate gesture events to plugins.
1280         https://bugs.webkit.org/show_bug.cgi?id=92281
1281
1282         Reviewed by Adam Barth.
1283
1284         Update TestWebPlugin to receive events, and output logs for the events.
1285
1286         * DumpRenderTree/chromium/TestWebPlugin.cpp:
1287         (TestWebPlugin::handleInputEvent):
1288         * DumpRenderTree/chromium/TestWebPlugin.h:
1289         (TestWebPlugin::acceptsInputEvents):
1290         (TestWebPlugin):
1291
1292 2012-07-26  Balazs Kelemen  <kbalazs@webkit.org>
1293
1294         [nrwt] get rid of --skip-pixel-test-if-no-baseline
1295         https://bugs.webkit.org/show_bug.cgi?id=92377
1296
1297         Reviewed by Adam Barth.
1298
1299         Remove the option and all code paths that handled it.
1300
1301         * Scripts/webkitpy/layout_tests/port/base.py:
1302         (Port.should_run_as_pixel_test):
1303         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1304         (_set_up_derived_options):
1305         (parse_args):
1306         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1307         (MainTest.test_repeat_each):
1308
1309 2012-07-26  Olivier Blin  <olivier.blin@softathome.com>
1310
1311         Add FastMalloc statistics in window.internals
1312         https://bugs.webkit.org/show_bug.cgi?id=91274
1313
1314         Reviewed by Ryosuke Niwa.
1315
1316         * GNUmakefile.am:
1317
1318 2012-07-26  Olivier Blin  <olivier.blin@softathome.com>
1319
1320         Add FastMalloc statistics in window.internals
1321         https://bugs.webkit.org/show_bug.cgi?id=91274
1322
1323         Reviewed by Ryosuke Niwa.
1324
1325         * GNUmakefile.am:
1326
1327 2012-07-20  Holger Hans Peter Freyther  <holger@moiji-mobile.com>
1328
1329         buildbot: Remove the configuration of hfreyther*
1330         https://bugs.webkit.org/show_bug.cgi?id=91849
1331
1332         Reviewed by Adam Roben.
1333
1334         The machine running these configs was decommissioned and
1335         there is no replacement at this point. Remove them from the
1336         configuration.
1337
1338         * BuildSlaveSupport/build.webkit.org-config/config.json: Remove hfreyther* from
1339         the config. The Qt Linux SH4 and Qt Linux MIPS schedulers are orphaned, remove
1340         them too.
1341
1342 2012-07-26  Balazs Kelemen  <kbalazs@webkit.org>
1343
1344         Unreviewed speculative Mac buildfix after r123729 - 2nd attempt.
1345
1346         * WebKitTestRunner/TestController.cpp:
1347         (WTR::TestController::runTest):
1348
1349 2012-07-26  Balazs Kelemen  <kbalazs@webkit.org>
1350
1351         Unreviewed speculative buildfix after r123729.
1352
1353         * WebKitTestRunner/TestController.cpp:
1354         (WTR::TestController::runTest): Fix sign errors.
1355
1356 2012-07-26  Zoltan Nyul  <zoltan.nyul@intel.com>
1357
1358         [EFL] EFL port should use XDG paths
1359         https://bugs.webkit.org/show_bug.cgi?id=91719
1360
1361         Reviewed by Kenneth Rohde Christiansen.
1362
1363         The appcache/localstorage/webdatabase path should be set in the
1364         ewk_main.cpp instead of the DumpRenderTree, and the XDG_DATA_HOME and
1365         XDG_CACHE_HOME should be set to different directory for each processes
1366         for the layout-tests.
1367
1368         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1369         (DumpRenderTreeChrome::initialize):
1370         * Scripts/webkitpy/layout_tests/port/efl.py:
1371         (EflPort.setup_environ_for_server):
1372         * efl/jhbuild.modules:
1373
1374 2012-07-26  Balazs Kelemen  <kbalazs@webkit.org>
1375
1376         [NRWT] should have a way to restrict pixel tests for individual directories
1377         https://bugs.webkit.org/show_bug.cgi?id=91754
1378
1379         Reviewed by Dirk Pranke.
1380
1381         Added --pixel-test-directories option to nrwt. It implies --pixel-tests.
1382         If the platform supports this option only tests from these directories will
1383         be executed as pixel tests. Furthermore, platforms can have a default set of
1384         pixel test directories. To support the feature the test harness have to be
1385         able to inform the driver whether to dump pixels for each tests. For this
1386         purpose I changed WTR and the Qt DRT to handle a --pixel-test argument from
1387         the standard output and made NRWT use this argument with these drivers.
1388
1389         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1390         (WebCore::DumpRenderTree::DumpRenderTree):
1391         (WebCore::DumpRenderTree::open):
1392         (WebCore::DumpRenderTree::processLine):
1393         (WebCore::DumpRenderTree::setDumpPixelsForAllTests):
1394         (WebCore::DumpRenderTree::dump):
1395         * DumpRenderTree/qt/DumpRenderTreeQt.h:
1396         (DumpRenderTree):
1397         * DumpRenderTree/qt/main.cpp:
1398         (main):
1399         Handle the --pixel-test argument.
1400
1401         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
1402         (SingleTestRunner._should_fetch_expected_checksum):
1403         (SingleTestRunner._overwrite_baselines):
1404         (SingleTestRunner._compare_output):
1405         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1406         (Worker._update_test_input):
1407         * Scripts/webkitpy/layout_tests/port/base.py:
1408         (Port.supports_switching_pixel_tests_per_test): Ports can override this to inform
1409         the harness that --pixel-test argument is supported. True for WTR, otherwise calls
1410         private version that can be overridden by the actual platform. Only true with WTR
1411         and Qt DRT currently. This supposed to exist only temporary until there are ports
1412         that doesn't support the feature in their DRT's.
1413         (Port):
1414         (Port._supports_switching_pixel_tests_per_test):
1415         (Port.should_run_as_pixel_test):
1416         (Port._should_run_as_pixel_test): Ports can override this to define their
1417         default set of pixel tests directories.
1418         * Scripts/webkitpy/layout_tests/port/driver.py:
1419         (Driver.cmd_line): Don't pass --pixel-tests if the platform supports per test
1420         --pixel-test.
1421         (Driver._command_from_driver_input): Pass --pixel-test if it is supported and pixel
1422         tests are enabled.
1423         * Scripts/webkitpy/layout_tests/port/qt.py:
1424         (QtPort.check_sys_deps):
1425         (QtPort):
1426         (QtPort._supports_switching_pixel_tests_per_test):
1427         (QtPort._should_run_as_pixel_test):
1428         (QtPort._default_pixel_test_directories): Only compositing will be our default pixel
1429         test directory for the time being.
1430         * Scripts/webkitpy/layout_tests/port/test.py:
1431         (TestPort.supports_switching_pixel_tests_per_test):
1432         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1433         (_set_up_derived_options):
1434         (parse_args):
1435         Handle the --pixel-test-directory option. It expect a directory below LayoutTests
1436         and can be specified multiple times. Added fixme's that we should improve how we
1437         handle these arguments.
1438
1439         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1440         (MainTest.test_run_singly_actually_runs_tests):
1441         (MainTest.test_pixel_test_directories):
1442         Added an integration test for the feature.
1443
1444         * WebKitTestRunner/TestController.cpp:
1445         (WTR::TestController::TestController):
1446         (WTR::TestController::initialize):
1447         (WTR::TestController::runTest):
1448         * WebKitTestRunner/TestController.h:
1449         (TestController):
1450         Handle the --pixel-test argument.
1451
1452 2012-07-26  Christophe Dumez  <christophe.dumez@intel.com>
1453
1454         [EFL] Regression(r123604): webintents/web-intents-obj-constructor.html is crashing
1455         https://bugs.webkit.org/show_bug.cgi?id=92349
1456
1457         Reviewed by Kentaro Hara.
1458
1459         Fix crash occurring after r123604, which replaced strdup()
1460         usage with eina_stringshare_add(). Unfortunately, the DRT
1461         was not updated accordingly to use eina_stringshare_del()
1462         instead of free() on the values returned by Ewk_Intent
1463         getters and it was causing
1464         webintents/web-intents-obj-constructor.html to crash.
1465
1466         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1467         (DumpRenderTreeChrome::onFrameIntentNew):
1468
1469 2012-07-26  Adam Barth  <abarth@webkit.org>
1470
1471         Unreviewed attempt to fix the chromium-win build.
1472
1473         * DumpRenderTree/chromium/TestRunner/TestRunner.cpp:
1474
1475 2012-07-26  Kaustubh Atrawalkar  <kaustubh@motorola.com>
1476
1477         [DRT] LTC:: pageNumberForElementById() could be moved to Internals
1478         https://bugs.webkit.org/show_bug.cgi?id=92091
1479
1480         Reviewed by Adam Barth.
1481
1482         Move the pageNumberForElementById from LayoutTestCotroller to Internals and
1483         remove the old platform specific implementations as it exclusively tests WebCore functionality.
1484
1485         * DumpRenderTree/LayoutTestController.cpp:
1486         (LayoutTestController::staticFunctions):
1487         * DumpRenderTree/LayoutTestController.h:
1488         (LayoutTestController):
1489         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
1490         * DumpRenderTree/chromium/LayoutTestController.cpp:
1491         (LayoutTestController::LayoutTestController):
1492         (LayoutTestController::setPrinting):
1493         * DumpRenderTree/chromium/LayoutTestController.h:
1494         (LayoutTestController):
1495         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1496         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1497         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
1498         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1499         * DumpRenderTree/qt/LayoutTestControllerQt.h:
1500         (LayoutTestController):
1501         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1502         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
1503         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
1504         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1505         (WTR):
1506         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1507         (LayoutTestController):
1508
1509 2012-07-25  Adam Barth  <abarth@webkit.org>
1510
1511         TestRunner.a should contain a TestRunner object
1512         https://bugs.webkit.org/show_bug.cgi?id=92313
1513
1514         Reviewed by Kent Tamura.
1515
1516         This patch adds a stub implementation of TestRunner, which is the new
1517         home for our implementation of window.testRunner. A future patch will
1518         migrate functions from LayoutTestController to TestRunner.
1519
1520         * DumpRenderTree/DumpRenderTree.gypi:
1521         * DumpRenderTree/chromium/LayoutTestController.h:
1522         * DumpRenderTree/chromium/TestRunner/TestRunner.cpp: Added.
1523         * DumpRenderTree/chromium/TestRunner/TestRunner.h: Added.
1524
1525 2012-07-25  Wei James  <james.wei@intel.com>
1526
1527         [Chromium] Fix infinite loop issue for chromium android layout test
1528         https://bugs.webkit.org/show_bug.cgi?id=92197
1529
1530         Reviewed by Adam Barth.
1531
1532         The change brought in by r123530 will cause infinite loop when running
1533         Chromium Android Layout Test.
1534         Revert the change for this line only.
1535
1536         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1537         (ChromiumAndroidPort._build_path):
1538
1539 2012-07-25  Adam Barth  <abarth@webkit.org>
1540
1541         Jochen is our a newest reviewer!
1542
1543         * Scripts/webkitpy/common/config/committers.py:
1544
1545 2012-07-25  Jochen Eisinger  <jochen@chromium.org>
1546
1547         [chromium] don't use webkit_support methods directly from test runners but through the test delegate
1548         https://bugs.webkit.org/show_bug.cgi?id=92256
1549
1550         Reviewed by Adam Barth.
1551
1552         This will allow for the content_shell to provide its own implementation
1553         for setting the Gamepad data instead.
1554
1555         * DumpRenderTree/chromium/TestRunner/GamepadController.cpp:
1556         (GamepadController::setDelegate):
1557         (GamepadController::reset):
1558         (GamepadController::connect):
1559         (GamepadController::disconnect):
1560         (GamepadController::setId):
1561         (GamepadController::setButtonCount):
1562         (GamepadController::setButtonData):
1563         (GamepadController::setAxisCount):
1564         (GamepadController::setAxisData):
1565         * DumpRenderTree/chromium/TestRunner/GamepadController.h:
1566         (GamepadController):
1567         * DumpRenderTree/chromium/TestRunner/TestDelegate.h:
1568         (WebKit):
1569         (TestDelegate):
1570         * DumpRenderTree/chromium/TestRunner/TestInterfaces.cpp:
1571         (TestInterfaces::Internal):
1572         (TestInterfaces::Internal::~Internal):
1573         (TestInterfaces::Internal::setDelegate):
1574         (TestInterfaces::setDelegate):
1575         * DumpRenderTree/chromium/TestRunner/TestInterfaces.h:
1576         (TestInterfaces):
1577         * DumpRenderTree/chromium/TestShell.cpp:
1578         (TestShell::createMainWindow):
1579         (TestShell::~TestShell):
1580         * DumpRenderTree/chromium/WebViewHost.cpp:
1581         (WebViewHost::setGamepadData):
1582         * DumpRenderTree/chromium/WebViewHost.h:
1583         (WebViewHost):
1584
1585 2012-07-25  Christophe Dumez  <christophe.dumez@intel.com>
1586
1587         [WK2][WTR] LayoutTestController.sendWebIntentResponse() needs to be implemented
1588         https://bugs.webkit.org/show_bug.cgi?id=92227
1589
1590         Reviewed by Kenneth Rohde Christiansen.
1591
1592         Implement LayoutTestController.sendWebIntentResponse() so
1593         that a test case can reply to a Web intent request and
1594         test this functionality.
1595
1596         In order to support this functionality, we need to keep
1597         a pointer to the current intent request in
1598         InjectedBundlePage::didReceiveIntentForFrame() so that
1599         we can reuse it later in
1600         LayoutTestController::sendWebIntentResponse().
1601
1602         The following tests need this functionality:
1603         - webintents/web-intents-failure.html
1604         - webintents/web-intents-reply.html
1605
1606         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
1607         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1608         (WTR::InjectedBundlePage::didReceiveIntentForFrame):
1609         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
1610         (InjectedBundlePage):
1611         (WTR::InjectedBundlePage::currentIntentRequest):
1612         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1613         (WTR::LayoutTestController::sendWebIntentResponse):
1614         (WTR):
1615         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1616         (LayoutTestController):
1617
1618 2012-07-25  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1619
1620         [EFL][WK2][WTR] Alt, Shift and Ctrl keys are not handled by EventSenderProxy.
1621         https://bugs.webkit.org/show_bug.cgi?id=92241
1622
1623         Reviewed by Kenneth Rohde Christiansen.
1624
1625         Alt, Shift and Ctrl keys were not handled by EventSenderProxy.
1626         It caused failure of fast/events/keydown-leftright-keys.html test.
1627
1628         * WebKitTestRunner/efl/EventSenderProxyEfl.cpp: Added handling of Alt, Shift and Ctrl keys.
1629         (WTR::keyName):
1630
1631 2012-07-25  Christophe Dumez  <christophe.dumez@intel.com>
1632
1633         [EFL][WK2] Fix timeout issues when debugging WebProcess.
1634         https://bugs.webkit.org/show_bug.cgi?id=92245
1635
1636         Reviewed by Dirk Pranke.
1637
1638         Fix timeout issues when debugging the WebProcess
1639         under valgrind or gdb.
1640
1641         * Scripts/webkitpy/layout_tests/port/efl.py:
1642         (EflPort.default_timeout_ms):
1643         * WebKitTestRunner/efl/TestControllerEfl.cpp:
1644         (WTR::TestController::platformInitialize):
1645
1646 2012-07-25  Xianzhu Wang  <wangxianzhu@chromium.org>
1647
1648         [Chromium-Android] NRWT driver should not restart between reftests
1649         https://bugs.webkit.org/show_bug.cgi?id=92267
1650
1651         Reviewed by Dirk Pranke.
1652
1653         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1654         (ChromiumAndroidDriver.start): Resets self._pixel_tests when the pixel_tests parameter changes.
1655
1656 2012-07-25  Tommy Widenflycht  <tommyw@google.com>
1657
1658         MediaStream API: Update MediaStreamTrack to match the specification
1659         https://bugs.webkit.org/show_bug.cgi?id=90180
1660
1661         Reviewed by Adam Barth.
1662
1663         Adding a Mock WebMediaStreamCenter to enable better LayoutTests.
1664
1665         * DumpRenderTree/DumpRenderTree.gypi:
1666         * DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp:
1667         (MockWebKitPlatformSupport::createMediaStreamCenter):
1668         * DumpRenderTree/chromium/MockWebMediaStreamCenter.cpp: Added.
1669         (WebKit):
1670         (WebKit::MockWebMediaStreamCenter::MockWebMediaStreamCenter):
1671         (WebKit::MockWebMediaStreamCenter::queryMediaStreamSources):
1672         (WebKit::MockWebMediaStreamCenter::didEnableMediaStreamTrack):
1673         (WebKit::MockWebMediaStreamCenter::didDisableMediaStreamTrack):
1674         (WebKit::MockWebMediaStreamCenter::didStopLocalMediaStream):
1675         (WebKit::MockWebMediaStreamCenter::didCreateMediaStream):
1676         (WebKit::MockWebMediaStreamCenter::constructSDP):
1677         * DumpRenderTree/chromium/MockWebMediaStreamCenter.h: Copied from Tools/DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp.
1678         (WebKit):
1679         (MockWebMediaStreamCenter):
1680         (WebKit::MockWebMediaStreamCenter::MockWebMediaStreamCenter):
1681
1682 2012-07-25  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
1683
1684         [NRWT] Rename baseline_search_path() to default_baseline_search_path()
1685         https://bugs.webkit.org/show_bug.cgi?id=92122
1686
1687         Reviewed by Dirk Pranke.
1688
1689         * Scripts/webkitpy/layout_tests/port/base.py:
1690         (Port.baseline_version_dir):
1691         (Port.baseline_search_path):
1692         (Port):
1693         (Port.default_baseline_search_path):
1694         (Port.expected_baselines):
1695         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1696         (PortTest.test_additional_platform_directory):
1697         * Scripts/webkitpy/layout_tests/port/mac.py:
1698         (MacPort.default_baseline_search_path):
1699         * Scripts/webkitpy/layout_tests/port/qt.py:
1700         (QtPort.default_baseline_search_path):
1701         * Scripts/webkitpy/layout_tests/port/win.py:
1702         (WinPort.default_baseline_search_path):
1703
1704 2012-07-25  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1705
1706         [EFL] Use eina_stringshare_add instead of strdup.
1707         https://bugs.webkit.org/show_bug.cgi?id=92072
1708
1709         Reviewed by Kentaro Hara.
1710
1711         Eina of EFL libraries supports a string functionality that replaces strdup. So, EFL port needs
1712         to replace strdup with eina_stringshare_add function.
1713
1714         DumpRenderTree also needs to be modified according to ewk APIs's parameter change.
1715
1716         * DumpRenderTree/efl/DumpRenderTree.cpp:
1717         (dumpFramesAsText):
1718         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1719         (DumpRenderTreeChrome::onFrameIntentNew):
1720         * DumpRenderTree/efl/DumpRenderTreeView.cpp:
1721         (onJavaScriptPrompt):
1722
1723 2012-07-25  Alexander Shalamov  <alexander.shalamov@intel.com>
1724
1725         [WTR][EFL] Wheel events are not emitted
1726         https://bugs.webkit.org/show_bug.cgi?id=92094
1727
1728         Reviewed by Csaba Osztrogonác.
1729
1730         Wheel events are not emitted, therefore, layout tests that use mouseScrollBy method are failing.
1731
1732         * WebKitTestRunner/efl/EventSenderProxyEfl.cpp:
1733         (WTR::dispatchMouseScrollByEvent):
1734         (WTR::EventSenderProxy::mouseScrollBy):
1735         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp:
1736         (WTR::PlatformWebView::PlatformWebView):
1737         (WTR::PlatformWebView::focus):
1738
1739 2012-07-25  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1740
1741         [EFL][WK2] Add ewk_main.{cpp,h} to EFL WK2
1742         https://bugs.webkit.org/show_bug.cgi?id=92101
1743
1744         Reviewed by Kenneth Rohde Christiansen.
1745
1746         Added a centralized place for general initialization in UI process for EFL WK2.
1747
1748         * MiniBrowser/efl/main.c: Added using of newly added ewk_init() and ewk_shutdown() functions.
1749         (main):
1750         * WebKitTestRunner/efl/main.cpp: Ditto.
1751         (main):
1752
1753 2012-07-24  Benjamin Poulain  <bpoulain@apple.com> && Joseph Pecoraro  <pecoraro@apple.com>
1754
1755         QualifiedName's HashSet should be big enough to hold at least all the static names
1756         https://bugs.webkit.org/show_bug.cgi?id=91891
1757
1758         Reviewed by Darin Adler.
1759
1760         Add a test for WTF::hashTableCapacityForSize.
1761
1762         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1763         * TestWebKitAPI/Tests/WTF/HashSet.cpp: Added.
1764         (InitialCapacityTestHashTraits):
1765         (TestWebKitAPI::testInitialCapacity):
1766         (TestWebKitAPI::generateTestCapacityUpToSize):
1767         (TestWebKitAPI::TEST):
1768
1769 2012-07-24  Adam Barth  <abarth@webkit.org>
1770
1771         The EWS bots get flaky when we hit the failure limit
1772         https://bugs.webkit.org/show_bug.cgi?id=92212
1773
1774         Reviewed by Eric Seidel.
1775
1776         Previously we were missing some failure types, which means we didn't
1777         notice that the bots were hitting the failure limit.
1778
1779         * Scripts/webkitpy/common/net/layouttestresults.py:
1780         (LayoutTestResults.failing_test_results):
1781         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
1782         (LayoutTestResultsTest.test_results_from_string):
1783         (LayoutTestResultsTest.test_tests_matching_failure_types):
1784
1785 2012-07-24  Raphael Kubo da Costa  <rakuco@webkit.org>
1786
1787         [EFL] Unreviewed build fix.
1788
1789         * WebKitTestRunner/efl/EventSenderProxyEfl.cpp: Include unistd.h
1790         for usleep(3).
1791
1792 2012-07-24  Dirk Pranke  <dpranke@chromium.org>
1793
1794         make test-webkitpy at least kinda work in parallel on win32
1795         https://bugs.webkit.org/show_bug.cgi?id=91700
1796
1797         Reviewed by Adam Barth.
1798
1799         This change makes the hoop-jumping we have to do to make running
1800         new-run-webkit-tests in parallel on win32 generic so that we can
1801         re-use it for test-webkitpy as well.
1802
1803         * Scripts/new-run-webkit-tests:
1804         * Scripts/test-webkitpy:
1805         * Scripts/webkitpy/common/multiprocessing_bootstrap.py: Copied from Tools/Scripts/new-run-webkit-tests.
1806         (run):
1807         * Scripts/webkitpy/test/main.py:
1808         (main):
1809         (Tester._log_exception):
1810
1811 2012-07-24  Sheriff Bot  <webkit.review.bot@gmail.com>
1812
1813         Unreviewed, rolling out r123499.
1814         http://trac.webkit.org/changeset/123499
1815         https://bugs.webkit.org/show_bug.cgi?id=92161
1816
1817         Android fails to compile (Requested by zhenyao on #webkit).
1818
1819         * DumpRenderTree/DumpRenderTree.gypi:
1820         * DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp:
1821         (MockWebKitPlatformSupport::createMediaStreamCenter):
1822         * DumpRenderTree/chromium/MockWebKitPlatformSupport.h:
1823         (MockWebKitPlatformSupport):
1824         * DumpRenderTree/chromium/MockWebMediaStreamCenter.cpp: Removed.
1825         * DumpRenderTree/chromium/MockWebMediaStreamCenter.h: Removed.
1826
1827 2012-07-24  Dirk Pranke  <dpranke@chromium.org>
1828
1829         nrwt: never finds binaries in the 'out' dir on chromium win
1830         https://bugs.webkit.org/show_bug.cgi?id=91890
1831
1832         Reviewed by Tony Chang.
1833
1834         Attempt to fix this bug yet again; it turns out that --root and
1835         --build-directory don't have the same meaning (--root should
1836         include the configuration, --build-directory should not).
1837
1838         * Scripts/webkitpy/layout_tests/port/base.py:
1839         (Port._build_path):
1840         * Scripts/webkitpy/layout_tests/port/chromium.py:
1841         (ChromiumPort._static_build_path):
1842         (ChromiumPort.default_results_directory):
1843         (ChromiumPort._build_path):
1844         (ChromiumPort):
1845         (ChromiumPort._build_path_with_configuration):
1846         (ChromiumPort._build_path_with_configuration.does):
1847         (ChromiumPort._path_to_image_diff):
1848         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1849         (ChromiumAndroidPort._build_path):
1850         (ChromiumAndroidPort):
1851         (ChromiumAndroidPort._build_path_with_configuration):
1852         (ChromiumAndroidPort._path_to_driver):
1853         (ChromiumAndroidPort._path_to_forwarder):
1854         (ChromiumAndroidPort._push_executable):
1855         (ChromiumAndroidPort._push_fonts):
1856         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1857         (ChromiumLinuxPort._determine_driver_path_statically):
1858         (ChromiumLinuxPort._modules_to_search_for_symbols):
1859         (ChromiumLinuxPort._path_to_driver):
1860         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
1861         (ChromiumLinuxPortTest.test_build_path):
1862         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1863         (ChromiumMacPort._modules_to_search_for_symbols):
1864         (ChromiumMacPort._path_to_driver):
1865         (ChromiumMacPort._path_to_helper):
1866         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1867         (ChromiumMacPortTest.test_build_path):
1868         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1869         (ChromiumWinPort._path_to_driver):
1870         (ChromiumWinPort._path_to_helper):
1871         (ChromiumWinPort._path_to_image_diff):
1872         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1873         (ChromiumWinTest.test_build_path):
1874         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1875         (parse_args):
1876
1877 2012-07-24  Peter Beverloo  <peter@chromium.org>
1878
1879         Android-chromium layout tests failing with GPU related crash
1880         https://bugs.webkit.org/show_bug.cgi?id=92126
1881
1882         Reviewed by Adam Barth.
1883
1884         Remove support for the "--enable-hardware-gpu" flag from run_webkit_tests,
1885         as it's not being used by any port anymore, and neither is being passed
1886         on to the test runner. Instead, append the flag directly to DRT for the
1887         Chromium Android port, which does rely on this functionality.
1888
1889         The ChromiumAndroidUnitTest tests didn't catch this as the property was
1890         set to True, but it's unable to detect whether they'll actually be used.
1891         No occurrances of "--enable-hardware-gpu" exist in either Tools/Scripts/
1892         or Tools/BuildSlaveSupport/, so nothing upstream seems to rely on the flag.
1893
1894         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1895         (ChromiumAndroidPort.__init__):
1896         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
1897         (ChromiumAndroidPortTest.test_attributes):
1898         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1899         (parse_args):
1900
1901 2012-07-24  Alexander Shalamov  <alexander.shalamov@intel.com>
1902
1903         [WTR][EFL] Multiple mouse clicks are not emitted in EventSenderProxyEfl
1904         https://bugs.webkit.org/show_bug.cgi?id=92107
1905
1906         Reviewed by Kenneth Rohde Christiansen.
1907
1908         Fixes mouse click events in EventSenderProxyEfl.
1909
1910         * WebKitTestRunner/efl/EventSenderProxyEfl.cpp:
1911         (WTR::dispatchMouseDownEvent):
1912         (WTR::EventSenderProxy::dispatchEvent):
1913
1914 2012-07-24  Tommy Widenflycht  <tommyw@google.com>
1915
1916         MediaStream API: Update MediaStreamTrack to match the specification
1917         https://bugs.webkit.org/show_bug.cgi?id=90180
1918
1919         Reviewed by Adam Barth.
1920
1921         Adding a Mock WebMediaStreamCenter to enable better LayoutTests.
1922
1923         * DumpRenderTree/DumpRenderTree.gypi:
1924         * DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp:
1925         (MockWebKitPlatformSupport::createMediaStreamCenter):
1926         * DumpRenderTree/chromium/MockWebMediaStreamCenter.cpp: Added.
1927         (WebKit):
1928         (WebKit::MockWebMediaStreamCenter::MockWebMediaStreamCenter):
1929         (WebKit::MockWebMediaStreamCenter::queryMediaStreamSources):
1930         (WebKit::MockWebMediaStreamCenter::didEnableMediaStreamTrack):
1931         (WebKit::MockWebMediaStreamCenter::didDisableMediaStreamTrack):
1932         (WebKit::MockWebMediaStreamCenter::didStopLocalMediaStream):
1933         (WebKit::MockWebMediaStreamCenter::didCreateMediaStream):
1934         (WebKit::MockWebMediaStreamCenter::constructSDP):
1935         * DumpRenderTree/chromium/MockWebMediaStreamCenter.h: Copied from Tools/DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp.
1936         (WebKit):
1937         (MockWebMediaStreamCenter):
1938         (WebKit::MockWebMediaStreamCenter::MockWebMediaStreamCenter):
1939
1940 2012-07-23  Ryosuke Niwa  <rniwa@webkit.org>
1941
1942         Fix regressions in replay performance tests
1943         https://bugs.webkit.org/show_bug.cgi?id=92052
1944
1945         Reviewed by Dirk Pranke.
1946
1947         Fix regression that accumulated over time to make replay tests work again.
1948
1949         * DumpRenderTree/chromium/WebViewHost.cpp:
1950         (WebViewHost::willSendRequest): Use the main document URL in the request since the main frame's
1951         URL hadn't been updated yet When requesting a new URL on the main frame.
1952         * Scripts/webkitpy/performance_tests/perftest.py:
1953         (ReplayServer.wait_until_ready): Waiting for 6 seconds (3 tries) seems enough.
1954         (ReplayPerfTest.prepare): Pass self._archive_path to run_signle as that is the current contract.
1955         (ReplayPerfTest.run_single): Use the archive path to generate -expected.png path.
1956         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1957         (TestReplayPerfTest.test_run_single):
1958         (TestReplayPerfTest.test_prepare):
1959         (TestReplayPerfTest.test_prepare_calls_run_single.run_single):
1960         (TestReplayPerfTest):
1961
1962 2012-07-24  Simon Hausmann  <simon.hausmann@nokia.com>
1963
1964         [Qt] MSVC: unresolved external symbol __DllMainCRTStartup@12
1965         https://bugs.webkit.org/show_bug.cgi?id=91229
1966
1967         Reviewed by Jocelyn Turcotte.
1968
1969         In order to successfully link a DLL on Windows we need to have at least
1970         one object file (or compilation unit). The forward export header files were
1971         supposed to be that, but unfortunately the rule in win32/default_post.prf for
1972         creating the header files had some bugs, among others that it did an exists()
1973         check on the depending static library. At the time qmake is ran those libraries
1974         do not exist yet and therefore the corresponding extra compiler rules were never
1975         created, resulting in empty OBJECTS/SOURCES.
1976
1977         Even without such an exists() check, qmake extra compilers require the files
1978         referred to in the .input variable to exist at qmake time. In this case the input
1979         files were the static libraries, which do not exist yet.
1980
1981         This patch solves this by using a qmake extra target instead of extra
1982         compiler, which does not have this limitation. The target is referenced
1983         through the extension of GENERATED_SOURCES.
1984
1985         The patch also adds a d/_debug suffix for debug builds, do allow for separate
1986         symbol exports if necessary.
1987
1988         * Scripts/generate-win32-export-forwards: Support multiple input files, i.e.
1989         consider the last argument to be the output file and everything else input.
1990         * qmake/mkspecs/features/win32/default_post.prf:
1991
1992 2012-07-24  Balazs Kelemen  <kbalazs@webkit.org>
1993
1994         [Qt] Add quirks for running the web process in a profiler shell, like valgrind
1995         https://bugs.webkit.org/show_bug.cgi?id=87672
1996
1997         Reviewed by Jocelyn Turcotte.
1998
1999         Add --no-timeout-at-all command line switch to not time out no matter how
2000         long the web process does it's job. --no-timeout is not enough for profiling
2001         purposes because it only affects the timeout when we are waiting for the end
2002         of the test. We use another timeout to control how much we wait for the web
2003         process to initialize before and reset after the test and it could also easily
2004         expire when running under valgrind. We should keep the behavior of --no-timeout
2005         to not disable this timout check because --no-timeout is used by performace tests
2006         and those should not wait for the web process to initialize and reset more than
2007         normal tests.
2008
2009         * WebKitTestRunner/TestController.cpp:
2010         (WTR::TestController::TestController):
2011         (WTR::TestController::initialize):
2012         (WTR::TestController::runUntil):
2013         * WebKitTestRunner/TestController.h:
2014
2015 2012-07-24  Peter Beverloo  <peter@chromium.org>
2016
2017         [Chromium] Fix a class/struct mismatch error in TestDelegate.h (showing up on clang)
2018         https://bugs.webkit.org/show_bug.cgi?id=92097
2019
2020         Unreviewed build fix.
2021
2022         WebContextMenuData is declared as a struct in Source/WebKit/chromium/public/WebContextMenuData.h,
2023         but TestDelegate.h forward declares it as a class.
2024
2025         * DumpRenderTree/chromium/TestRunner/TestDelegate.h:
2026         (WebKit):
2027
2028 2012-07-23  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
2029
2030         [Qt] Fix compilation against namespaced Qt
2031
2032         Reviewed by Simon Hausmann.
2033
2034         * QtTestBrowser/launcherwindow.h:
2035         * WebKitTestRunner/PlatformWebView.h:
2036
2037 2012-07-24  Peter Beverloo  <peter@chromium.org>
2038
2039         Update a few Chromium Android contributors in committers.py
2040         https://bugs.webkit.org/show_bug.cgi?id=92090
2041
2042         Reviewed by Kent Tamura.
2043
2044         Add Sami Kyostila, who has been doing significant work on Chromium's
2045         compositor. Add IRC nicknames for John Mellor and Tien-Ren Chen, and
2046         remove John Grabowski who changed projects, and add my webkit.org alias.
2047
2048         * Scripts/webkitpy/common/config/committers.py:
2049
2050 2012-07-24  Pierre Rossi  <pierre.rossi@gmail.com>
2051
2052         [Qt] Enable touch slider when touch events are enabled
2053         https://bugs.webkit.org/show_bug.cgi?id=91013
2054
2055         Reviewed by Kenneth Rohde Christiansen.
2056
2057         TOUCH_SLIDER guards are used in WebCore, so we should really
2058         be setting the corresponding ENABLE define in features.prf.
2059
2060         * qmake/mkspecs/features/features.prf:
2061
2062 2012-07-24  Ryuan Choi  <ryuan.choi@samsung.com>
2063
2064         [EFL] EWebLauncher should not require config.h in the style check
2065         https://bugs.webkit.org/show_bug.cgi?id=92070
2066
2067         Reviewed by Kentaro Hara.
2068
2069         * Scripts/webkitpy/style/checker.py:
2070         Remove the include rules from the style-check for EWebLauncher because
2071         EWebLauncher is simple EFL application.
2072
2073 2012-07-24  Jochen Eisinger  <jochen@chromium.or>
2074
2075         [Chromium] EventSender shouldn't know about TestShell
2076         https://bugs.webkit.org/show_bug.cgi?id=91449
2077
2078         Reviewed by Adam Barth.
2079
2080         Original patch by Adam Barth.
2081
2082         This patch removes the dependency from EventSender to TestShell.  Most
2083         of EventSender works directly with a WebView, but there are a handful
2084         of references to WebViewHost.  We don't want to re-use WebViewHost in
2085         ContentShell because we want to use the multi-process embedding
2086         environment for WebViews.
2087
2088         To avoid the dependency on WebViewHost, this patch introduces
2089         TestDelegate, which is an abstract base class for WebViewHost.  In
2090         ContentShell, we'll need another object to implement these functions.
2091
2092         * DumpRenderTree/chromium/EventSender.cpp:
2093         (EventSender::EventSender):
2094         (EventSender::keyDown):
2095         (makeMenuItemStringsFor):
2096         (EventSender::contextClick):
2097         * DumpRenderTree/chromium/EventSender.h:
2098         (EventSender):
2099         (EventSender::setDelegate):
2100         (EventSender::setWebView):
2101         (EventSender::webview):
2102         * DumpRenderTree/chromium/TestRunner/TestDelegate.h: Added.
2103         (WebKit):
2104         (TestDelegate):
2105         * DumpRenderTree/chromium/TestShell.cpp:
2106         (TestShell::initialize):
2107         (TestShell::createMainWindow):
2108         (TestShell::~TestShell):
2109         * DumpRenderTree/chromium/WebViewHost.cpp:
2110         (WebViewHost::fillSpellingSuggestionList):
2111         * DumpRenderTree/chromium/WebViewHost.h:
2112         (WebViewHost):
2113
2114 2012-07-24  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2115
2116         Make WebIntents and WebIntentsTag a build option for the build-webkit script
2117         https://bugs.webkit.org/show_bug.cgi?id=91841
2118
2119         Reviewed by Philippe Normand.
2120
2121         And it will also be disabled when building with --minimal.
2122
2123         * Scripts/webkitperl/FeatureList.pm:
2124
2125 2012-07-23  Dirk Pranke  <dpranke@chromium.org>
2126
2127         Unreviewed, rolling out r123360.
2128         http://trac.webkit.org/changeset/123360
2129         https://bugs.webkit.org/show_bug.cgi?id=91890
2130
2131         broke the chromium bots
2132
2133         * Scripts/webkitpy/layout_tests/port/chromium.py:
2134         (ChromiumPort._static_build_path):
2135         (ChromiumPort.default_results_directory):
2136         (ChromiumPort._build_path):
2137         (ChromiumPort._path_to_image_diff):
2138         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2139         (ChromiumAndroidPort._build_path):
2140         (ChromiumAndroidPort._path_to_driver):
2141         (ChromiumAndroidPort._path_to_forwarder):
2142         (ChromiumAndroidPort._push_executable):
2143         (ChromiumAndroidPort._push_fonts):
2144         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2145         (ChromiumLinuxPort._determine_driver_path_statically):
2146         (ChromiumLinuxPort._modules_to_search_for_symbols):
2147         (ChromiumLinuxPort._path_to_driver):
2148         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
2149         (ChromiumLinuxPortTest.test_build_path):
2150         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2151         (ChromiumMacPort._modules_to_search_for_symbols):
2152         (ChromiumMacPort._path_to_driver):
2153         (ChromiumMacPort._path_to_helper):
2154         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
2155         (ChromiumMacPortTest.test_build_path):
2156         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2157         (ChromiumWinPort._path_to_driver):
2158         (ChromiumWinPort._path_to_helper):
2159         (ChromiumWinPort._path_to_image_diff):
2160         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2161         (ChromiumWinTest.test_build_path):
2162
2163 2012-07-23  Adam Barth  <abarth@webkit.org>
2164
2165         [Chromium] TestInterfaces should be responsible for owning and binding AccessibilityController and TextInputController
2166         https://bugs.webkit.org/show_bug.cgi?id=91459
2167
2168         Reviewed by Ryosuke Niwa.
2169
2170         Rather than having TestShell own and bind each of these JavaScript APIs
2171         individually, TestShell should delegate that work to TestInterfaces.
2172         This patch moves AccessibilityController and TextInputController to be
2173         the responsibility of TestInterfaces.
2174
2175         * DumpRenderTree/chromium/TestRunner/TestInterfaces.cpp:
2176         (TestInterfaces::TestInterfaces):
2177         (TestInterfaces::~TestInterfaces):
2178         (TestInterfaces::setWebView):
2179         (TestInterfaces::bindTo):
2180         (TestInterfaces::resetAll):
2181         * DumpRenderTree/chromium/TestRunner/TestInterfaces.h:
2182         (WebKit):
2183         (TestInterfaces):
2184         (TestInterfaces::accessibilityController):
2185         * DumpRenderTree/chromium/TestShell.cpp:
2186         (TestShell::initialize):
2187         (TestShell::createMainWindow):
2188         (TestShell::~TestShell):
2189         (TestShell::resetTestController):
2190         (TestShell::bindJSObjectsToWindow):
2191         * DumpRenderTree/chromium/TestShell.h:
2192         (TestShell::accessibilityController):
2193         (TestShell):
2194
2195 2012-07-23  Balazs Kelemen  <kbalazs@webkit.org>
2196
2197         nrwt swallows early exceptions
2198         https://bugs.webkit.org/show_bug.cgi?id=91884
2199
2200         Reviewed by Dirk Pranke.
2201
2202         Put everything in a try block so we won't miss any error.
2203
2204         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2205         (run):
2206         (main):
2207
2208 2012-07-23  Christophe Dumez  <christophe.dumez@intel.com>
2209
2210         [WK2] WebKitTestRunner needs to print information about Web intents
2211         https://bugs.webkit.org/show_bug.cgi?id=90873
2212
2213         Reviewed by Anders Carlsson.
2214
2215         Print out information about new intents requests and intent
2216         services registrations in WebKitTestRunner. This is expected
2217         by tests under webintents/*.
2218
2219         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2220         (WTR::InjectedBundlePage::InjectedBundlePage):
2221         (WTR::InjectedBundlePage::didReceiveIntentForFrame):
2222         (WTR):
2223         (WTR::InjectedBundlePage::registerIntentServiceForFrame):
2224         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
2225         (InjectedBundlePage):
2226
2227 2012-07-23  Dirk Pranke  <dpranke@chromium.org>
2228
2229         nrwt: never finds binaries in the 'out' dir on chromium win
2230         https://bugs.webkit.org/show_bug.cgi?id=91890
2231
2232         Reviewed by Tony Chang.
2233
2234         We were figuring out which directory look in for binaries by
2235         testing for the base directory (the directory above
2236         Debug/Release). In chromium-win's case, we look in src/build,
2237         which always exists because there are checked-in files in it,
2238         which means we'd always pick that directory over src/out. All of
2239         the other ports' build_path() implementation was including
2240         Debug/Release. If we matched that, we wouldn't have a problem,
2241         so this change fixes that and updates all of the callers of the
2242         chromium ports' implementation to not pass configuration as part
2243         of the path to look up; we still need to pass configuration in
2244         some cases (to test if the build is out of date between debug
2245         and release) so the implementation gets slightly more
2246         complicated.
2247
2248         * Scripts/webkitpy/layout_tests/port/chromium.py:
2249         (ChromiumPort._static_build_path):
2250         (ChromiumPort.default_results_directory):
2251         (ChromiumPort._build_path):
2252         (ChromiumPort._path_to_image_diff):
2253         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2254         (ChromiumAndroidPort._build_path):
2255         (ChromiumAndroidPort._path_to_driver):
2256         (ChromiumAndroidPort._path_to_forwarder):
2257         (ChromiumAndroidPort._push_executable):
2258         (ChromiumAndroidPort._push_fonts):
2259         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2260         (ChromiumLinuxPort._determine_driver_path_statically):
2261         (ChromiumLinuxPort._modules_to_search_for_symbols):
2262         (ChromiumLinuxPort._path_to_driver):
2263         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
2264         (ChromiumLinuxPortTest.test_build_path):
2265         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2266         (ChromiumMacPort._modules_to_search_for_symbols):
2267         (ChromiumMacPort._path_to_driver):
2268         (ChromiumMacPort._path_to_helper):
2269         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
2270         (ChromiumMacPortTest.test_build_path):
2271         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2272         (ChromiumWinPort._path_to_driver):
2273         (ChromiumWinPort._path_to_helper):
2274         (ChromiumWinPort._path_to_image_diff):
2275         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2276         (ChromiumWinTest.test_build_path):
2277
2278 2012-07-23  Sheriff Bot  <webkit.review.bot@gmail.com>
2279
2280         Unreviewed, rolling out r123339.
2281         http://trac.webkit.org/changeset/123339
2282         https://bugs.webkit.org/show_bug.cgi?id=92006
2283
2284         massive media tests failure (Requested by philn on #webkit).
2285
2286         * Scripts/webkitdirs.pm:
2287         (buildAutotoolsProject):
2288         * gtk/jhbuild.modules:
2289
2290 2012-07-23  Tommy Widenflycht  <tommyw@google.com>
2291
2292         [chromium] MediaStream API: Clean up the MockWebKitPlatformSupport object at shutdown
2293         https://bugs.webkit.org/show_bug.cgi?id=91857
2294
2295         Reviewed by Adam Barth.
2296
2297         Made a few changes so that the destructor could be called at shutdown.
2298
2299         * DumpRenderTree/chromium/DumpRenderTree.cpp:
2300         (WebKitSupportTestEnvironment::WebKitSupportTestEnvironment):
2301         (WebKitSupportTestEnvironment):
2302         * DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp:
2303         (MockWebKitPlatformSupport::create):
2304         * DumpRenderTree/chromium/MockWebKitPlatformSupport.h:
2305         (MockWebKitPlatformSupport):
2306         (MockWebKitPlatformSupport::~MockWebKitPlatformSupport):
2307
2308 2012-07-23  Philippe Normand  <pnormand@igalia.com>
2309
2310         [GTK][jhbuild] Switch to GStreamer 0.11 build
2311         https://bugs.webkit.org/show_bug.cgi?id=91727
2312
2313         Reviewed by Gustavo Noronha Silva.
2314
2315         Switch build-webkit --gtk to GStreamer 0.11 support and build the
2316         necessary GStreamer git modules from JHBuild.
2317
2318         * Scripts/webkitdirs.pm:
2319         (buildAutotoolsProject): Build WebKit with GStreamer 0.11 support.
2320         * gtk/jhbuild.modules: GStreamer build support.
2321
2322 2012-07-23  Christophe Dumez  <christophe.dumez@intel.com>
2323
2324         WebKit2 needs layoutTestController.setAlwaysAcceptCookies
2325         https://bugs.webkit.org/show_bug.cgi?id=42778
2326
2327         Reviewed by Kenneth Rohde Christiansen.
2328
2329         Add support for layoutTestController.setAlwaysAcceptCookies()
2330         since it is required by some tests.
2331
2332         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
2333         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
2334         (WTR::LayoutTestController::setAlwaysAcceptCookies):
2335         (WTR):
2336         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2337         (LayoutTestController):
2338
2339 2012-07-23  Kent Tamura  <tkent@chromium.org>
2340
2341         Rename ENABLE_METER_TAG and ENABLE_PROGRESS_TAG to ENABLE_METER_ELEMENT and ENABLE_PROGRESS_ELEMENT respectively
2342         https://bugs.webkit.org/show_bug.cgi?id=91941
2343
2344         Reviewed by Kentaro Hara.
2345
2346         A flag name for an elmement should be ENABLE_*_ELEMENT.
2347
2348         * Scripts/webkitperl/FeatureList.pm:
2349         * qmake/mkspecs/features/features.pri:
2350
2351 2012-07-22  Kent Tamura  <tkent@chromium.org>
2352
2353         Rename ENABLE_DETAILS to ENABLE_DETAILS_ELEMENT
2354         https://bugs.webkit.org/show_bug.cgi?id=91928
2355
2356         Reviewed by Kentaro Hara.
2357
2358         A flag name for an elmement should be ENABLE_*_ELEMENT.
2359
2360         * Scripts/webkitperl/FeatureList.pm:
2361         * qmake/mkspecs/features/features.pri:
2362
2363 2012-07-20  Kent Tamura  <tkent@chromium.org>
2364
2365         Rename ENABLE_DATALIST to ENABLE_DATALIST_ELEMENT
2366         https://bugs.webkit.org/show_bug.cgi?id=91846
2367
2368         Reviewed by Kentaro Hara.
2369
2370         A flag name for an elmement should be ENABLE_*_ELEMENT.
2371
2372         * Scripts/webkitperl/FeatureList.pm:
2373         * qmake/mkspecs/features/features.pri:
2374
2375 2012-07-20  Xianzhu Wang  <wangxianzhu@chromium.org>
2376
2377         [Chromium-Android] Store test pipes in /data/local/tmp
2378         https://bugs.webkit.org/show_bug.cgi?id=91909
2379
2380         In the previous change (bug 91873, r123258) we placed the pipes in /sdcard/native_tests.
2381         However after tests we found some devices don't allow the app to create pipes there.
2382         So /data/local/tmp is a better place for the pipes.
2383
2384         Reviewed by Adam Barth.
2385
2386         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2387
2388 2012-07-20  Xianzhu Wang  <wangxianzhu@chromium.org>
2389
2390         [Chomium-Android] Place DRT input/output pipes where both app and unrooted shell can read/write
2391         https://bugs.webkit.org/show_bug.cgi?id=91873
2392
2393         Reviewed by Adam Barth.
2394
2395         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2396
2397 2012-07-20  Han Shen  <shenhan@google.com>
2398
2399         [Chromium] Compilation fails under gcc 4.7
2400         https://bugs.webkit.org/show_bug.cgi?id=90227
2401
2402         Reviewed by Tony Chang.
2403
2404         Disable warnings about c++0x compatibility in gcc newer than 4.6.
2405
2406         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2407
2408 2012-07-20  Dirk Pranke  <dpranke@chromium.org>
2409
2410         REGRESSION: run-perf-tests --pause-before-testing is broken
2411         https://bugs.webkit.org/show_bug.cgi?id=91789
2412
2413         Reviewed by Ryosuke Niwa.
2414
2415         Turns out start() wasn't actually starting things :(. Fixed, and
2416         added tests.
2417
2418         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
2419         (ChromiumAndroidDriverTest.test_read_prompt):
2420         * Scripts/webkitpy/layout_tests/port/driver.py:
2421         (Driver.__init__):
2422         (Driver._start):
2423         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
2424         (DriverTest.test_read_block):
2425         (DriverTest.test_read_binary_block):
2426         (DriverTest.test_read_base64_block):
2427         (DriverTest.test_stop_cleans_up_properly):
2428         (DriverTest.test_two_starts_cleans_up_properly):
2429         (DriverTest.test_start_actually_starts):
2430         (MockServerProcess.__init__):
2431         (MockServerProcess.start):
2432
2433 2012-07-20  Jochen Eisinger  <jochen@chromium.org>
2434
2435         [chromium] Don't include WebCore headers in TestInterfaces so it's safe to include from outside of WebCore
2436         https://bugs.webkit.org/show_bug.cgi?id=91847
2437
2438         Reviewed by Adam Barth.
2439
2440         This allows for including the TestIntefaces.h from content shell
2441         without introducing a dependency to WebCore/wtf
2442
2443         * DumpRenderTree/chromium/TestRunner/TestInterfaces.cpp:
2444         (TestInterfaces::Internal):
2445         (TestInterfaces::Internal::Internal):
2446         (TestInterfaces::Internal::~Internal):
2447         (TestInterfaces::Internal::bindTo):
2448         (TestInterfaces::Internal::resetAll):
2449         (TestInterfaces::TestInterfaces):
2450         (TestInterfaces::~TestInterfaces):
2451         (TestInterfaces::bindTo):
2452         (TestInterfaces::resetAll):
2453         * DumpRenderTree/chromium/TestRunner/TestInterfaces.h:
2454         (TestInterfaces):
2455
2456 2012-07-20  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
2457
2458         [Qt] Make WTR use the same color palette as DRT
2459         https://bugs.webkit.org/show_bug.cgi?id=91870
2460
2461         Reviewed by Noam Rosenthal.
2462
2463         Make sure that WebKitTestRunner and DumpRenderTree use the same palette: the
2464         standard from QWindowsStyle.
2465
2466         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
2467         (WTR::activateFonts): Moved code to platformInitialize since setting style is not
2468         exactly font related.
2469         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
2470         (WTR::InjectedBundle::platformInitialize): Force the usage of the style's
2471         palette.  Because at this point QApplication was already instantiated, the
2472         default palette was already initialized to be the system one, so we need to
2473         explicitly set the palette to be the standard from style. This is not needed in
2474         DRT because there we set the style before instantiating QApplication.
2475
2476 2012-07-20  Tony Chang  <tony@chromium.org>
2477
2478         Baseline align text in the flakiness dashboard toolbar
2479         https://bugs.webkit.org/show_bug.cgi?id=91784
2480
2481         Reviewed by Ojan Vafai.
2482
2483         Baseline align the flex items.
2484
2485         * TestResultServer/static-dashboards/flakiness_dashboard.css:
2486         (.forms): Add box-align: baseline
2487         (.forms > *): Use block display to force flex items rather than box.
2488         (#tests-form): Add box-align: baseline
2489         (#tests-form > *): Use block display to force flex items rather than box.
2490
2491 2012-07-20  Sheriff Bot  <webkit.review.bot@gmail.com>
2492
2493         Unreviewed, rolling out r123220.
2494         http://trac.webkit.org/changeset/123220
2495         https://bugs.webkit.org/show_bug.cgi?id=91880
2496
2497         broke Debug build (Requested by philn on #webkit).
2498
2499         * Scripts/webkitdirs.pm:
2500         (buildAutotoolsProject):
2501         * gtk/jhbuild.modules:
2502
2503 2012-07-20  Philippe Normand  <pnormand@igalia.com>
2504
2505         [GTK][jhbuild] Switch to GStreamer 0.11 build
2506         https://bugs.webkit.org/show_bug.cgi?id=91727
2507
2508         Reviewed by Gustavo Noronha Silva.
2509
2510         Switch build-webkit --gtk to GStreamer 0.11 support and build the
2511         necessary GStreamer git modules from JHBuild.
2512
2513         * Scripts/webkitdirs.pm:
2514         (buildAutotoolsProject): Build WebKit with GStreamer 0.11 support.
2515         * gtk/jhbuild.modules: GStreamer build support.
2516
2517 2012-07-20  Christophe Dumez  <christophe.dumez@intel.com>
2518
2519         [EFL] Proxy configuration should honor the no_proxy environment variable
2520         https://bugs.webkit.org/show_bug.cgi?id=91747
2521
2522         Reviewed by Kenneth Rohde Christiansen.
2523
2524         Add ProxyResolverSoup to the list of classes that contain GObjects
2525         in style checker script so that it does not complain about naming
2526         convention issues that are mandated by GObject.
2527
2528         * Scripts/webkitpy/style/checker.py:
2529
2530 2012-07-20  Zeno Albisser  <zeno@webkit.org>
2531
2532         [Qt][WK2] Add configure tests for Xrender and GLX.
2533         https://bugs.webkit.org/show_bug.cgi?id=91845
2534
2535         GraphicsSurface on linux requires Xcomposite, Xrender
2536         and GLX. If any of these requirements is not met,
2537         we should disable GraphicsSurface.
2538
2539         Reviewed by Kenneth Rohde Christiansen.
2540
2541         * qmake/config.tests/glx/glx.cpp: Added.
2542         (main):
2543         * qmake/config.tests/glx/glx.pro: Added.
2544         * qmake/config.tests/libXrender/libXrender.cpp: Added.
2545         (main):
2546         * qmake/config.tests/libXrender/libXrender.pro: Added.
2547         * qmake/configure.pri:
2548         * qmake/mkspecs/features/features.prf:
2549
2550 2012-07-20  Kangil Han  <kangil.han@samsung.com>
2551
2552         [EFL][WK2][WTR] Implement EventSenderProxy
2553         https://bugs.webkit.org/show_bug.cgi?id=91731
2554
2555         Reviewed by Kentaro Hara.
2556
2557         This is an initial EFL port implementation of EventSenderProxy.
2558
2559         * WebKitTestRunner/EventSenderProxy.h:
2560         (WTR):
2561         (EventSenderProxy):
2562         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
2563         * WebKitTestRunner/PlatformEfl.cmake:
2564         * WebKitTestRunner/TestController.cpp:
2565         (WTR::TestController::TestController):
2566         (WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
2567         * WebKitTestRunner/efl/EventSenderProxyEfl.cpp: Added.
2568         (WTR):
2569         (WTREvent):
2570         (WTR::WTREvent::WTREvent):
2571         (WTR::KeyEventInfo::KeyEventInfo):
2572         (KeyEventInfo):
2573         (WTR::evasMouseButton):
2574         (WTR::setEvasModifiers):
2575         (WTR::dispatchMouseDownEvent):
2576         (WTR::dispatchMouseUpEvent):
2577         (WTR::dispatchMouseMoveEvent):
2578         (WTR::dispatchMouseScrollByEvent):
2579         (WTR::keyPadName):
2580         (WTR::keyName):
2581         (WTR::EventSenderProxy::EventSenderProxy):
2582         (WTR::EventSenderProxy::updateClickCountForButton):
2583         (WTR::EventSenderProxy::dispatchEvent):
2584         (WTR::EventSenderProxy::replaySavedEvents):
2585         (WTR::EventSenderProxy::sendOrQueueEvent):
2586         (WTR::EventSenderProxy::mouseDown):
2587         (WTR::EventSenderProxy::mouseUp):
2588         (WTR::EventSenderProxy::mouseMoveTo):
2589         (WTR::EventSenderProxy::mouseScrollBy):
2590         (WTR::EventSenderProxy::leapForward):
2591         (WTR::EventSenderProxy::keyDown):
2592         (WTR::EventSenderProxy::addTouchPoint):
2593         (WTR::EventSenderProxy::updateTouchPoint):
2594         (WTR::EventSenderProxy::setTouchModifier):
2595         (WTR::EventSenderProxy::touchStart):
2596         (WTR::EventSenderProxy::touchMove):
2597         (WTR::EventSenderProxy::touchEnd):
2598         (WTR::EventSenderProxy::touchCancel):
2599         (WTR::EventSenderProxy::clearTouchPoints):
2600         (WTR::EventSenderProxy::releaseTouchPoint):
2601         (WTR::EventSenderProxy::cancelTouchPoint):
2602         (WTR::EventSenderProxy::setTouchPointRadius):
2603
2604 2012-07-20  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
2605
2606         Unreviewed, rolling out r123085.
2607         http://trac.webkit.org/changeset/123085
2608         https://bugs.webkit.org/show_bug.cgi?id=91719
2609
2610         r123085 causes crashes on EFL layout test bot.
2611
2612         * WebKitTestRunner/CMakeLists.txt:
2613         * WebKitTestRunner/InjectedBundle/efl/LayoutTestControllerEfl.cpp:
2614         (WTR::LayoutTestController::platformInitialize):
2615
2616 2012-07-20  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
2617
2618         [WTR] Add *explicit* keyword to constructor which has a parameter
2619         https://bugs.webkit.org/show_bug.cgi?id=91745
2620
2621         Reviewed by Adam Barth.
2622
2623         Add *explicit* keyword to contructor which has a parameter in order to avoid implicit type conversion.
2624
2625         * WebKitTestRunner/EventSenderProxy.h:
2626         (EventSenderProxy):
2627         * WebKitTestRunner/TestInvocation.h:
2628         (TestInvocation):
2629
2630 2012-07-19  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
2631
2632         [EFL] [WK2] Add methods to get/set a custom text encoding
2633         https://bugs.webkit.org/show_bug.cgi?id=90604
2634
2635         Reviewed by Kenneth Rohde Christiansen.
2636
2637         Added support for setting custom character encoding in MiniBrowser.
2638         Use 'F3' to set the custom character encoding.
2639
2640         * MiniBrowser/efl/main.c:
2641         (on_key_down):
2642
2643 2012-07-19  Dirk Pranke  <dpranke@chromium.org>
2644
2645         webkitpy: executive_unittest still failing when run in parallel
2646         https://bugs.webkit.org/show_bug.cgi?id=91795
2647
2648         Reviewed by Adam Barth.
2649
2650         It turns out that test_kill_process and test_kill_all are both
2651         launching the same process, and so if they're run at the same
2652         time, test_kill_all interferes w/ test_kill_process.
2653
2654         Merging the two tests solves the issue :).
2655
2656         * Scripts/webkitpy/common/system/executive_unittest.py:
2657         (ExecutiveTest.test_kill_process):
2658         (ExecutiveTest._assert_windows_image_name):
2659         (ExecutiveTest):
2660         (ExecutiveTest.test_windows_image_name):
2661
2662 2012-07-19  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
2663
2664         [EFL] [WK2] Add a callback to handle delete request in MiniBrowser
2665         https://bugs.webkit.org/show_bug.cgi?id=91483
2666
2667         Reviewed by Dirk Pranke.
2668
2669         Add a callback to handle delete request and quit the main loop when
2670         window is asked to be deleted so that all the resources allocated
2671         for the window will be freed properly.
2672
2673         * MiniBrowser/efl/main.c:
2674         (closeWindow):
2675         (browserCreate):
2676
2677 2012-07-19  No'am Rosenthal  <noam.rosenthal@nokia.com>
2678
2679         [Qt] Enable CSS shaders in Qt (software mode)
2680         https://bugs.webkit.org/show_bug.cgi?id=85140
2681
2682         Reviewed by Simon Hausmann.
2683
2684         Enable CSS_SHADERS when WEBGL and CSS_FILTERS are enabled.
2685
2686         * qmake/mkspecs/features/features.prf:
2687
2688 2012-07-19  No'am Rosenthal  <noam.rosenthal@nokia.com>
2689
2690         [Qt] Enable CSS shaders in Qt (software mode)
2691         https://bugs.webkit.org/show_bug.cgi?id=85140
2692
2693         Reviewed by Simon Hausmann.
2694
2695         Enable CSS_SHADERS when WEBGL and CSS_FILTERS are enabled.
2696
2697         * qmake/mkspecs/features/features.prf:
2698
2699 2012-07-19  Dirk Pranke  <dpranke@chromium.org>
2700
2701         test-webkitpy: fix executive_unittest nits, including a missing 'errno' import
2702         https://bugs.webkit.org/show_bug.cgi?id=91770
2703
2704         Reviewed by Adam Barth.
2705
2706         The missing errno import is causing a flaky failure when run in
2707         parallel on the bots.
2708
2709         * Scripts/webkitpy/common/system/executive_unittest.py:
2710         (ExecutiveTest.assert_interpreter_for_content):
2711
2712 2012-07-19  Oliver Hunt  <oliver@apple.com>
2713
2714         Make style bot handle underscores in names more reasonably
2715         https://bugs.webkit.org/show_bug.cgi?id=91777
2716
2717         Reviewed by Gavin Barraclough.
2718
2719         This patch adds a naming/readability/underscores category for style errors.
2720         That allows us to use underscores in those portions of JSC (and potentially
2721         gtk?) without turning off all naming checks for the effected directories.
2722         We also filter out the DFG_OPERATION warning, and a few other minor changes
2723         to make the style bot less noisy when running over JSC at least.
2724
2725         * Scripts/webkitpy/style/checker.py:
2726         * Scripts/webkitpy/style/checkers/cpp.py:
2727         (check_identifier_name_in_declaration):
2728         (CppChecker):
2729         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
2730         (WebKitStyleTest.test_names):
2731
2732 2012-07-19  Dirk Pranke  <dpranke@chromium.org>
2733
2734         config_standalone.py is crashing on apple win's cygwin port
2735         https://bugs.webkit.org/show_bug.cgi?id=91682
2736
2737         Reviewed by Adam Roben.
2738
2739         Work around a weird interaction of ctypes and cygwin that
2740         is causing unit tests to crash for me locally.
2741
2742         * Scripts/webkitpy/common/system/executive.py:
2743         (Executive._win32_check_running_pid):
2744
2745 2012-07-19  Xabier Rodriguez Calvar  <calvaris@igalia.com>
2746
2747         [GTK] WebKit test runner ignores all system environment variables
2748         https://bugs.webkit.org/show_bug.cgi?id=90284
2749
2750         Reviewed by Dirk Pranke.
2751
2752         Added support in unit test framework to pass a custom environment
2753         variable to the tests by using the --additional-env-var parameter.
2754
2755         * Scripts/webkitpy/layout_tests/port/base.py:
2756         (Port.to.setup_environ_for_server): Adding environment variables
2757         that come through the additional_env_var option.
2758         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2759         (PortTest.test_additional_env_var): Added unit test for the
2760         additional_env_var option
2761         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2762         (parse_args): Added the --additional-env-var parameter.
2763
2764 2012-07-19  Joshua Bell  <jsbell@chromium.org>
2765
2766         Roll out r121610 and r122487 which may have been causing flaky crashes
2767         https://bugs.webkit.org/show_bug.cgi?id=91637
2768
2769         Reviewed by Kentaro Hara.
2770
2771         Flaky crashes started in random tests following r121610. It's not clear
2772         that r121610 is to blame, but we'd like to prove or disprove it.
2773         If this doesn't resolve the crashes, this patch should be rolled out.
2774
2775         * DumpRenderTree/DumpRenderTree.gypi:
2776         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
2777         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
2778         (PluginTest):
2779         * DumpRenderTree/TestNetscapePlugIn/Tests/LeakWindowScriptableObject.cpp: Removed.
2780
2781 2012-07-19  Pierre Rossi  <pierre.rossi@gmail.com>
2782
2783         [Qt][WK2] Map the elementRect exposed in QML to the view for popup menu
2784         https://bugs.webkit.org/show_bug.cgi?id=91645
2785
2786         Reviewed by Simon Hausmann.
2787
2788         Also improved the logic slightly for edge cases.
2789
2790         * MiniBrowser/qt/qml/ItemSelector.qml:
2791
2792 2012-07-19  Christophe Dumez  <christophe.dumez@intel.com>
2793
2794         [EFL] Bump libsoup dependency to v2.39.4.1 to fix cookie issues
2795         https://bugs.webkit.org/show_bug.cgi?id=91741
2796
2797         Reviewed by Kenneth Rohde Christiansen.
2798
2799         Bump libsoup dependency in jhbuild to v2.39.4.1 to fix issues
2800         with cookies in non-suffixed/private domains [GNOME #679230].
2801
2802         * efl/jhbuild.modules:
2803
2804 2012-07-19  Simon Hausmann  <simon.hausmann@nokia.com>
2805
2806         [Qt] Remove support for Qt versions before 4.8.0
2807         https://bugs.webkit.org/show_bug.cgi?id=91730
2808
2809         Reviewed by Kenneth Rohde Christiansen.
2810
2811         Remove Qt version check #ifdefs for Qt versions before 4.8.0.
2812
2813         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2814         (LayoutTestController::reset):
2815         (LayoutTestController::setAlwaysBlockCookies):
2816         * DumpRenderTree/qt/LayoutTestControllerQt.h:
2817         (LayoutTestController):
2818
2819 2012-07-19  Mario Sanchez Prada  <msanchez@igalia.com>
2820
2821         [GTK] Enable MHTML support by default at build time
2822         https://bugs.webkit.org/show_bug.cgi?id=89987
2823
2824         Reviewed by Carlos Garcia Campos.
2825
2826         Enable MHTML support by default for GTK in build-webkit.
2827
2828         * Scripts/webkitperl/FeatureList.pm:
2829
2830 2012-07-19  Zoltan Nyul  <zoltan.nyul@intel.com>
2831
2832         [EFL][WK2] Using different cache directory for each WTR process
2833         https://bugs.webkit.org/show_bug.cgi?id=91719
2834
2835         Reviewed by Kenneth Rohde Christiansen.
2836
2837         The testrunner script may be running multiple processes in
2838         parallel, and it makes appcache tests fail if they are using the same
2839         directory. I modified the EFL's LayoutTestController to use the
2840         DUMPRENDERTREE_TEMP for application cache directory because it's
2841         different for each process.
2842
2843         * WebKitTestRunner/CMakeLists.txt:
2844         * WebKitTestRunner/InjectedBundle/efl/LayoutTestControllerEfl.cpp:
2845         (WTR::LayoutTestController::platformInitialize):
2846
2847 2012-07-19  Mario Sanchez Prada  <msanchez@igalia.com>
2848
2849         [WK2][GTK] Complete implementation of AccessibilityController in WKTR for GTK
2850         https://bugs.webkit.org/show_bug.cgi?id=89224
2851
2852         Reviewed by Chris Fleizach.
2853
2854         Implement logAccessibilityEvents() and resetToConsistentState(),
2855         so we can ask WKTR from the layout tests for logging all the
2856         accessibility related events into the result.
2857
2858         * WebKitTestRunner/GNUmakefile.am:
2859         * WebKitTestRunner/InjectedBundle/AccessibilityController.cpp:
2860         (WTR::AccessibilityController::AccessibilityController):
2861         (WTR):
2862         (WTR::AccessibilityController::logAccessibilityEvents):
2863         (WTR::AccessibilityController::resetToConsistentState):
2864         * WebKitTestRunner/InjectedBundle/AccessibilityController.h:
2865         (AccessibilityController):
2866         * WebKitTestRunner/InjectedBundle/gtk/AccessibilityControllerGtk.cpp: Added.
2867         (WTR):
2868         (WTR::printAccessibilityEvent):
2869         (WTR::axObjectEventListener):
2870         (WTR::AccessibilityController::logAccessibilityEvents):
2871         (WTR::AccessibilityController::resetToConsistentState):
2872
2873         Ensure we reset the accessibility controller to a consistent
2874         whenever a test has finished running.
2875
2876         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
2877         (WTR::InjectedBundle::done):
2878
2879 2012-07-19  Dominik Röttsches  <dominik.rottsches@intel.com>
2880
2881         [EFL][Wk2] Add a bot that runs WTR / WebKit2 tests
2882         https://bugs.webkit.org/show_bug.cgi?id=90788
2883
2884         Reviewed by Kentaro Hara.
2885
2886         Adding EFL 64bit WK2 BuildAndTest bot. At this point, we cannot transfer
2887         the build product from one bot to the other due to link path issues with jhbuild,
2888         so we do a full build and test cycle for WebKit2 on this bot.
2889
2890         * BuildSlaveSupport/build.webkit.org-config/config.json: Adding the new slave.
2891         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Adding a new factory for building and testing WebKit2.
2892         (BuildAndTestWebKit2Factory):
2893
2894 2012-07-18  Mihnea Ovidenie  <mihnea@adobe.com>
2895
2896         Unreviewed. Adding Andrei Bucur, Mihai Balan and Raul Hudea to the contributors list.
2897         They are Adobe contributors working on CSSRegions.
2898
2899         * Scripts/webkitpy/common/config/committers.py:
2900
2901 2012-07-18  Christophe Dumez  <christophe.dumez@intel.com>
2902
2903         [CMake][EFL] Should try to find the same libsoup version pulled by jhbuild
2904         https://bugs.webkit.org/show_bug.cgi?id=91626
2905
2906         Reviewed by Kenneth Rohde Christiansen.
2907
2908         Update jhbuild to pull libsoup 2.39.3 tarball from Gnome FTP instead
2909         of fetching a tag from Gnome GIT. We need to do this to map the version
2910         of libsoup that is required by our CMake config.
2911
2912         * efl/jhbuild.modules:
2913
2914 2012-07-18  MORITA Hajime  <morrita@google.com>
2915
2916         Add window resize benchmark
2917         https://bugs.webkit.org/show_bug.cgi?id=91070
2918
2919         Reviewed by Ryosuke Niwa.
2920
2921         Added some more output filters to take care of html5.html output.
2922
2923         * Scripts/webkitpy/performance_tests/perftest.py:
2924         (PerfTest):
2925
2926 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
2927
2928         nrwt: merge WebKitDriver into Driver
2929         https://bugs.webkit.org/show_bug.cgi?id=91696
2930
2931         Reviewed by Adam Barth.
2932
2933         As part of merging WebKitPort into Port, we also should merge
2934         WebKitDriver into Driver, since all of the non-test Drivers now
2935         inherit from WebKitDriver.
2936
2937         There are no functional changes in this patch, and all the changes should
2938         be covered by the existing tests.
2939
2940         * Scripts/webkitpy/layout_tests/port/base.py:
2941         (Port._driver_class):
2942         * Scripts/webkitpy/layout_tests/port/chromium.py:
2943         (ChromiumPort._check_file_exists):
2944         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2945         (ChromiumAndroidDriver):
2946         (ChromiumAndroidDriver.__init__):
2947         (ChromiumAndroidDriver._drt_cmd_line):
2948         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
2949         (ChromiumAndroidDriverTest.test_read_prompt):
2950         * Scripts/webkitpy/layout_tests/port/driver.py:
2951         (Driver):
2952         (Driver.__init__):
2953         (Driver.__del__):
2954         (Driver.run_test):
2955         (Driver.has_crashed):
2956         (Driver.start):
2957         (Driver._start):
2958         (Driver.stop):
2959         (Driver.cmd_line):
2960         (Driver._check_for_driver_crash):
2961         (Driver._command_from_driver_input):
2962         (Driver._read_first_block):
2963         (Driver._read_optional_image_block):
2964         (Driver._read_header):
2965         (Driver._process_stdout_line):
2966         (Driver._strip_eof):
2967         (Driver._read_block):
2968         (ContentBlock):
2969         (ContentBlock.__init__):
2970         (ContentBlock.decode_content):
2971         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
2972         (DriverTest.make_port):
2973         (DriverTest._assert_wrapper):
2974         (DriverTest.test_read_block):
2975         (DriverTest):
2976         (DriverTest.test_read_binary_block):
2977         (DriverTest.test_read_base64_block):
2978         (DriverTest.test_no_timeout):
2979         (DriverTest.test_check_for_driver_crash):
2980         (DriverTest.test_check_for_driver_crash.FakeServerProcess):
2981         (DriverTest.test_check_for_driver_crash.FakeServerProcess.__init__):
2982         (DriverTest.test_check_for_driver_crash.FakeServerProcess.pid):
2983         (DriverTest.test_check_for_driver_crash.FakeServerProcess.name):
2984         (DriverTest.test_check_for_driver_crash.FakeServerProcess.has_crashed):
2985         (DriverTest.test_check_for_driver_crash.FakeServerProcess.stop):
2986         (DriverTest.test_check_for_driver_crash.assert_crash):
2987         (DriverTest.test_creating_a_port_does_not_write_to_the_filesystem):
2988         (DriverTest.test_stop_cleans_up_properly):
2989         (DriverTest.test_two_starts_cleans_up_properly):
2990         (MockServerProcess):
2991         (MockServerProcess.__init__):
2992         (MockServerProcess.has_crashed):
2993         (MockServerProcess.read_stdout_line):
2994         (MockServerProcess.read_stdout):
2995         (MockServerProcess.read_either_stdout_or_stderr_line):
2996         (MockServerProcess.start):
2997         (MockServerProcess.stop):
2998         (MockServerProcess.kill):
2999         * Scripts/webkitpy/layout_tests/port/gtk.py:
3000         * Scripts/webkitpy/layout_tests/port/webkit.py:
3001         (WebKitPort._read_image_diff):
3002         (WebKitPort.skipped_layout_tests):
3003         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3004         (test_path_to_apache_config_file):
3005         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
3006         (XvfbDriver):
3007         (XvfbDriver.stop):
3008
3009 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
3010
3011         webkitpy.common.system.executive_unittest.ExecutiveTest.test_kill_process is flaky
3012         https://bugs.webkit.org/show_bug.cgi?id=91698
3013
3014         Unreviewed, build fix.
3015
3016         This is a workaround for the test failing (apparently mostly?)
3017         when run in parallel with other tests. The underlying bug still
3018         needs to be diagnosed as we shouldn't be getting ECHILD from
3019         process.wait().
3020
3021         * Scripts/webkitpy/common/system/executive_unittest.py:
3022         (ExecutiveTest.test_kill_process):
3023
3024 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
3025
3026         Disable running python unit tests in parallel on win32 for now.
3027         https://bugs.webkit.org/show_bug.cgi?id=91697
3028
3029         Unreviewed, build fix.
3030
3031         * Scripts/webkitpy/test/main.py:
3032         (Tester._parse_args):
3033
3034 2012-07-18  Dave Tu  <dtu@chromium.org>
3035
3036         [chromium] Add gl_tests to flakiness dashboard.
3037         https://bugs.webkit.org/show_bug.cgi?id=91680
3038
3039         Reviewed by Dirk Pranke.
3040
3041         * TestResultServer/static-dashboards/builders.js:
3042         (loadBuildersList):
3043         * TestResultServer/static-dashboards/dashboard_base.js:
3044         (currentBuilderGroupCategory):
3045
3046 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
3047
3048         fix python unit tests failing under cygwin
3049         https://bugs.webkit.org/show_bug.cgi?id=91678
3050
3051         Reviewed by Adam Barth.
3052
3053         Fix a case where executive.kill_process was failing under cygwin
3054         (apple win port) in an unexpected way.
3055         
3056         * Scripts/webkitpy/common/system/executive.py:
3057         (Executive.kill_process):
3058
3059 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
3060
3061         test-webkitpy: run tests in parallel by default
3062         https://bugs.webkit.org/show_bug.cgi?id=91422
3063
3064         Reviewed by Adam Barth.
3065
3066         We use multiprocessing.cpu_count() for the default number of
3067         jobs to run; memory overhead should be very low, so this should
3068         be fine.
3069
3070         * Scripts/webkitpy/test/main.py:
3071         (Tester._parse_args):
3072
3073 2012-07-18  Ryosuke Niwa  <rniwa@webkit.org>
3074
3075         Add Pravin D to the list of contributors.
3076
3077         * Scripts/webkitpy/common/config/committers.py:
3078
3079 2012-07-18  Balazs Kelemen  <kbalazs@webkit.org>
3080
3081         [Qt] feature detection in orwt/nrwt does not work with force_static_libs_as_shared
3082         https://bugs.webkit.org/show_bug.cgi?id=91514
3083
3084         Reviewed by Dirk Pranke.
3085
3086         Analyzing libQtWebKit.so is not enough in the case of force_static_libs_as_shared.
3087         We need to analyze all the shared library or at least the WebCore one.
3088
3089         * Scripts/webkitdirs.pm:
3090         (builtDylibPathForName):
3091         In orwt it is hard coded to search for symbols in the WebCore library.
3092         While in theory it is possible that symbols for a given feature are not
3093         located in that, it doesn't happen in practice, so returning the path of
3094         the WebCore library for a force_static_libs_as_shared build is enough to
3095         fix the bug and it doesn't require a bigger refactoring. For a default
3096         build we still return the path of the QtWebKit lib.
3097         * Scripts/webkitpy/layout_tests/port/qt.py:
3098         (QtPort._modules_to_search_for_symbols):
3099         Enumerate all dynamic libraries, not just the QtWebKit one.
3100
3101 2012-07-18  Tony Chang  <tony@chromium.org>
3102
3103         [chromium] Unreviewed, more compile fixes on Chromium Win.
3104
3105         * DumpRenderTree/chromium/TestShellWin.cpp:
3106         (TestShell::waitTestFinished):
3107         * DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
3108
3109 2012-07-18  Christophe Dumez  <christophe.dumez@intel.com>
3110
3111         [EFL] Add central error management to EFL port
3112         https://bugs.webkit.org/show_bug.cgi?id=91598
3113
3114         Reviewed by Kenneth Rohde Christiansen.
3115
3116         Map WebKitNetworkError to NSURLErrorDomain when
3117         printing in DumpRenderTree so that the output
3118         matches the expected one.
3119
3120         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
3121         (descriptionSuitableForTestResult):
3122
3123 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
3124
3125         nrwt: start merging port/webkit.py into port/base.py
3126         https://bugs.webkit.org/show_bug.cgi?id=91559
3127
3128         Reviewed by Ojan Vafai.
3129
3130         Since all the non-test port implementations now derive from
3131         WebKitPort, there's no real point in keeping WebKitPort distinct
3132         from Port. This patch starts merging the two by moving nearly
3133         all of the webkit implementations of routines with no default
3134         behavior into base.py. The few that didn't move rely on
3135         additional infrastructure that should be refactored differently
3136         (like the image diffing, which should probably be its own
3137         class) and deserve their own patches.
3138
3139         This patch should just be moving code around, and require no
3140         additional tests; in fact, we can delete the tests that were
3141         asserting virtual methods in the base class.
3142
3143         * Scripts/webkitpy/layout_tests/port/base.py:
3144         (Port.baseline_search_path):
3145         (Port.check_build):
3146         (Port):
3147         (Port._check_driver):
3148         (Port._check_port_build):
3149         (Port.check_image_diff):
3150         (Port.driver_name):
3151         (Port.default_results_directory):
3152         (Port.to.setup_environ_for_server):
3153         (Port._path_to_apache):
3154         (Port._is_redhat_based):
3155         (Port._is_debian_based):
3156         (Port._apache_config_file_name_for_platform):
3157         (Port._path_to_apache_config_file):
3158         (Port._build_path):
3159         (Port._path_to_driver):
3160         (Port._path_to_webcore_library):
3161         (Port._path_to_helper):
3162         (Port._path_to_image_diff):
3163         (Port._path_to_wdiff):
3164         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3165         (PortTest.test_httpd_returns_error_code):
3166         * Scripts/webkitpy/layout_tests/port/webkit.py:
3167         (WebKitPort):
3168         (WebKitPort._build_driver_flags):
3169         (WebKitPort._read_image_diff):
3170         (WebKitPort.skipped_layout_tests):
3171
3172 2012-07-18  Tommy Widenflycht  <tommyw@google.com>
3173
3174         [chromium] MediaStream API: Moving the mock create* WebRTC calls into a shadow Platform class
3175         https://bugs.webkit.org/show_bug.cgi?id=86215
3176
3177         Reviewed by Adam Barth.
3178
3179         Adding a shadow Platform object that is used to override some WebKit::Platform funtions to
3180         instead create mock objects. No actual mock objects created yet.
3181
3182         * DumpRenderTree/DumpRenderTree.gypi:
3183         * DumpRenderTree/chromium/DumpRenderTree.cpp:
3184         (WebKitSupportTestEnvironment::WebKitSupportTestEnvironment):
3185         * DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp: Copied from Tools/DumpRenderTree/chromium/config.h.
3186         (MockWebKitPlatformSupport::create):
3187         (MockWebKitPlatformSupport::MockWebKitPlatformSupport):
3188         (MockWebKitPlatformSupport::cryptographicallyRandomValues):
3189         (MockWebKitPlatformSupport::createMediaStreamCenter):
3190         * DumpRenderTree/chromium/MockWebKitPlatformSupport.h: Copied from Tools/DumpRenderTree/chromium/config.h.
3191         (MockWebKitPlatformSupport):
3192         * DumpRenderTree/chromium/config.h:
3193
3194 2012-07-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
3195
3196         [CMake][EFL] Build and run TestWebKitAPI unit tests
3197         https://bugs.webkit.org/show_bug.cgi?id=90671
3198
3199         Reviewed by Daniel Bates.
3200
3201         This patch enables almost for free a significant amount of test for the
3202         WebKit C API, WTF and KURL. It makes easy for CMake based ports to
3203         enable TestWebKitAPI since the missing bits to implement are
3204         relatively simple.
3205
3206         * CMakeLists.txt:
3207         * TestWebKitAPI/CMakeLists.txt: Added.
3208         * TestWebKitAPI/PlatformEfl.cmake: Added.
3209         * TestWebKitAPI/PlatformWebView.h:
3210         * TestWebKitAPI/config.h:
3211         * TestWebKitAPI/efl/InjectedBundleController.cpp: Added.
3212         (TestWebKitAPI):
3213         (TestWebKitAPI::InjectedBundleController::platformInitialize):
3214         * TestWebKitAPI/efl/PlatformUtilities.cpp: Added.
3215         (TestWebKitAPI):
3216         (Util):
3217         (TestWebKitAPI::Util::run):
3218         (TestWebKitAPI::Util::sleep):
3219         (TestWebKitAPI::Util::createURLForResource):
3220         (TestWebKitAPI::Util::createInjectedBundlePath):
3221         (TestWebKitAPI::Util::URLForNonExistentResource):
3222         * TestWebKitAPI/efl/PlatformWebView.cpp: Added.
3223         (TestWebKitAPI):
3224         (TestWebKitAPI::initEcoreEvas):
3225         (TestWebKitAPI::PlatformWebView::PlatformWebView):
3226         (TestWebKitAPI::PlatformWebView::~PlatformWebView):
3227         (TestWebKitAPI::PlatformWebView::page):
3228         * TestWebKitAPI/efl/main.cpp: Added.
3229         (checkForUseX11WindowArgument):
3230         (main):
3231
3232 2012-07-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
3233
3234         [CMake][EFL] Building jsc causes reconfiguration
3235         https://bugs.webkit.org/show_bug.cgi?id=91387
3236
3237         Reviewed by Daniel Bates.
3238
3239         We should remove CMakeCache only when running build-webkit script,
3240         otherwise it will cause a reconfiguration every time someone calls
3241         generateBuildSystemFromCMakeProject(). We were re-building "jsc" and
3242         not running WebKit2 unit tests on the bots because the project was
3243         reconfigured with default values by the scripts that run these tests.
3244
3245         * Scripts/build-webkit:
3246         * Scripts/webkitdirs.pm:
3247         (removeCMakeCache):
3248         (generateBuildSystemFromCMakeProject):
3249
3250 2012-07-18  Seokju Kwon  <seokju.kwon@samsung.com>
3251
3252         [EFL][DRT] Add support for Web Inspector in WebKit-EFL DRT
3253         https://bugs.webkit.org/show_bug.cgi?id=87935
3254
3255         Reviewed by Andreas Kling.
3256
3257         Web Inspector will be shown when path or url contains "inspector/".
3258         Dumprendertree should wait util web inspector resources are loaded totally
3259         and handle the signals for creating or removing a view of web inspector.
3260         ("inspector,view,create" and "inspector,view,close")
3261
3262         * DumpRenderTree/efl/DumpRenderTree.cpp:
3263         (shouldOpenWebInspector):
3264         (createLayoutTestController):
3265         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
3266         (DumpRenderTreeChrome::createView):
3267         (DumpRenderTreeChrome::createWebInspectorView):
3268         (DumpRenderTreeChrome::removeWebInspectorView):
3269         (DumpRenderTreeChrome::waitInspectorLoadFinished):
3270         (DumpRenderTreeChrome::onInspectorViewCreate):
3271         (DumpRenderTreeChrome::onInspectorViewClose):
3272         (DumpRenderTreeChrome::onInspectorFrameLoadFinished):
3273         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
3274         (DumpRenderTreeChrome):
3275         * DumpRenderTree/efl/DumpRenderTreeView.cpp:
3276         (onConsoleMessage):
3277         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
3278         (LayoutTestController::showWebInspector):
3279         (LayoutTestController::closeWebInspector):
3280         (LayoutTestController::evaluateInWebInspector):
3281
3282 2012-07-18  Simon Hausmann  <simon.hausmann@nokia.com>
3283
3284         [ANGLE] On QT, use Bison and Flex during ANGLE build
3285         https://bugs.webkit.org/show_bug.cgi?id=91108
3286
3287         Reviewed by Kenneth Rohde Christiansen.
3288
3289         * qmake/mkspecs/features/default_post.prf: Add support for variable_out to our generators, to allow
3290         generating not only for SOURCES but also ANGLE_SOURCES (in this bug)
3291
3292 2012-07-18  Balazs Kelemen  <kbalazs@webkit.org>
3293
3294         [Qt] plugin tests should not be disabled for WebKit1
3295         https://bugs.webkit.org/show_bug.cgi?id=91604
3296
3297         Reviewed by Simon Hausmann.
3298
3299         Instead of not building TestNetscapePlugIn, we could programatically
3300         disable actually loading it from WTR until https://bugs.webkit.org/show_bug.cgi?id=86620
3301         has been solved, so we can still test plugins on WebKit1.
3302
3303         * Tools.pro:
3304         * WebKitTestRunner/TestController.cpp:
3305         (WTR::TestController::initialize):
3306         * WebKitTestRunner/qt/TestControllerQt.cpp:
3307         (WTR::TestController::initializeTestPluginDirectory):
3308
3309 2012-07-18  Mario Sanchez Prada  <msanchez@igalia.com>
3310
3311         [WK2][GTK] Implement AccessibilityUIElement in WKTR for GTK
3312         https://bugs.webkit.org/show_bug.cgi?id=89223
3313
3314         Reviewed by Chris Fleizach.
3315
3316         Implemented AccessibilityUIElement in WKTR for GTK, based in the
3317         implementation present in DumpRenderTree.
3318
3319         Added new files in gtk/ and updated the cross-platform ones.
3320
3321         * WebKitTestRunner/GNUmakefile.am:
3322         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.cpp:
3323         (WTR):
3324         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
3325         (AccessibilityUIElement):
3326         * WebKitTestRunner/InjectedBundle/gtk/AccessibilityUIElementGtk.cpp: Added.
3327         (WTR):
3328         (WTR::attributeSetToString):
3329         (WTR::checkElementState):
3330         (WTR::indexRangeInTable):
3331         (WTR::alterCurrentValue):
3332         (WTR::AccessibilityUIElement::AccessibilityUIElement):
3333         (WTR::AccessibilityUIElement::~AccessibilityUIElement):
3334         (WTR::AccessibilityUIElement::isEqual):
3335         (WTR::AccessibilityUIElement::getChildren):
3336         (WTR::AccessibilityUIElement::getChildrenWithRange):
3337         (WTR::AccessibilityUIElement::childrenCount):
3338         (WTR::AccessibilityUIElement::elementAtPoint):
3339         (WTR::AccessibilityUIElement::indexOfChild):
3340         (WTR::AccessibilityUIElement::childAtIndex):
3341         (WTR::AccessibilityUIElement::linkedUIElementAtIndex):
3342         (WTR::AccessibilityUIElement::ariaOwnsElementAtIndex):
3343         (WTR::AccessibilityUIElement::ariaFlowToElementAtIndex):
3344         (WTR::AccessibilityUIElement::disclosedRowAtIndex):
3345         (WTR::AccessibilityUIElement::rowAtIndex):
3346         (WTR::AccessibilityUIElement::selectedChildAtIndex):
3347         (WTR::AccessibilityUIElement::selectedChildrenCount):
3348         (WTR::AccessibilityUIElement::selectedRowAtIndex):
3349         (WTR::AccessibilityUIElement::titleUIElement):
3350         (WTR::AccessibilityUIElement::parentElement):
3351         (WTR::AccessibilityUIElement::disclosedByRow):
3352         (WTR::AccessibilityUIElement::attributesOfLinkedUIElements):
3353         (WTR::AccessibilityUIElement::attributesOfDocumentLinks):
3354         (WTR::AccessibilityUIElement::attributesOfChildren):
3355         (WTR::AccessibilityUIElement::allAttributes):
3356         (WTR::AccessibilityUIElement::stringAttributeValue):
3357         (WTR::AccessibilityUIElement::numberAttributeValue):
3358         (WTR::AccessibilityUIElement::uiElementAttributeValue):
3359         (WTR::AccessibilityUIElement::boolAttributeValue):
3360         (WTR::AccessibilityUIElement::isAttributeSettable):
3361         (WTR::AccessibilityUIElement::isAttributeSupported):
3362         (WTR::AccessibilityUIElement::parameterizedAttributeNames):
3363         (WTR::AccessibilityUIElement::role):
3364         (WTR::AccessibilityUIElement::subrole):
3365         (WTR::AccessibilityUIElement::roleDescription):
3366         (WTR::AccessibilityUIElement::title):
3367         (WTR::AccessibilityUIElement::description):
3368         (WTR::AccessibilityUIElement::orientation):
3369         (WTR::AccessibilityUIElement::stringValue):
3370         (WTR::AccessibilityUIElement::language):
3371         (WTR::AccessibilityUIElement::helpText):
3372         (WTR::AccessibilityUIElement::x):
3373         (WTR::AccessibilityUIElement::y):
3374         (WTR::AccessibilityUIElement::width):
3375         (WTR::AccessibilityUIElement::height):
3376         (WTR::AccessibilityUIElement::clickPointX):
3377         (WTR::AccessibilityUIElement::clickPointY):
3378         (WTR::AccessibilityUIElement::intValue):
3379         (WTR::AccessibilityUIElement::minValue):
3380         (WTR::AccessibilityUIElement::maxValue):
3381         (WTR::AccessibilityUIElement::valueDescription):
3382         (WTR::AccessibilityUIElement::insertionPointLineNumber):
3383         (WTR::AccessibilityUIElement::isActionSupported):
3384         (WTR::AccessibilityUIElement::isEnabled):
3385         (WTR::AccessibilityUIElement::isRequired):
3386         (WTR::AccessibilityUIElement::isFocused):
3387         (WTR::AccessibilityUIElement::isSelected):
3388         (WTR::AccessibilityUIElement::isExpanded):
3389         (WTR::AccessibilityUIElement::isChecked):
3390         (WTR::AccessibilityUIElement::hierarchicalLevel):
3391         (WTR::AccessibilityUIElement::speak):
3392         (WTR::AccessibilityUIElement::ariaIsGrabbed):
3393         (WTR::AccessibilityUIElement::ariaDropEffects):
3394         (WTR::AccessibilityUIElement::lineForIndex):
3395         (WTR::AccessibilityUIElement::rangeForLine):
3396         (WTR::AccessibilityUIElement::rangeForPosition):
3397         (WTR::AccessibilityUIElement::boundsForRange):
3398         (WTR::AccessibilityUIElement::stringForRange):
3399         (WTR::AccessibilityUIElement::attributedStringForRange):
3400         (WTR::AccessibilityUIElement::attributedStringRangeIsMisspelled):
3401         (WTR::AccessibilityUIElement::uiElementForSearchPredicate):
3402         (WTR::AccessibilityUIElement::attributesOfColumnHeaders):
3403         (WTR::AccessibilityUIElement::attributesOfRowHeaders):
3404         (WTR::AccessibilityUIElement::attributesOfColumns):
3405         (WTR::AccessibilityUIElement::attributesOfRows):
3406         (WTR::AccessibilityUIElement::attributesOfVisibleCells):
3407         (WTR::AccessibilityUIElement::attributesOfHeader):
3408         (WTR::AccessibilityUIElement::rowCount):
3409         (WTR::AccessibilityUIElement::columnCount):
3410         (WTR::AccessibilityUIElement::indexInTable):
3411         (WTR::AccessibilityUIElement::rowIndexRange):
3412         (WTR::AccessibilityUIElement::columnIndexRange):
3413         (WTR::AccessibilityUIElement::cellForColumnAndRow):
3414         (WTR::AccessibilityUIElement::horizontalScrollbar):
3415         (WTR::AccessibilityUIElement::verticalScrollbar):
3416         (WTR::AccessibilityUIElement::selectedTextRange):
3417         (WTR::AccessibilityUIElement::setSelectedTextRange):
3418         (WTR::AccessibilityUIElement::increment):
3419         (WTR::AccessibilityUIElement::decrement):
3420         (WTR::AccessibilityUIElement::showMenu):
3421         (WTR::AccessibilityUIElement::press):
3422         (WTR::AccessibilityUIElement::setSelectedChild):
3423         (WTR::AccessibilityUIElement::accessibilityValue):
3424         (WTR::AccessibilityUIElement::documentEncoding):
3425         (WTR::AccessibilityUIElement::documentURI):
3426         (WTR::AccessibilityUIElement::url):
3427         (WTR::AccessibilityUIElement::addNotificationListener):
3428         (WTR::AccessibilityUIElement::removeNotificationListener):
3429         (WTR::AccessibilityUIElement::isFocusable):
3430         (WTR::AccessibilityUIElement::isSelectable):
3431         (WTR::AccessibilityUIElement::isMultiSelectable):
3432         (WTR::AccessibilityUIElement::isVisible):
3433         (WTR::AccessibilityUIElement::isOffScreen):
3434         (WTR::AccessibilityUIElement::isCollapsed):
3435         (WTR::AccessibilityUIElement::isIgnored):
3436         (WTR::AccessibilityUIElement::hasPopup):
3437         (WTR::AccessibilityUIElement::takeFocus):
3438         (WTR::AccessibilityUIElement::takeSelection):
3439         (WTR::AccessibilityUIElement::addSelection):
3440         (WTR::AccessibilityUIElement::removeSelection):
3441         (WTR::AccessibilityUIElement::textMarkerRangeForElement):
3442         (WTR::AccessibilityUIElement::textMarkerRangeLength):
3443         (WTR::AccessibilityUIElement::previousTextMarker):
3444         (WTR::AccessibilityUIElement::nextTextMarker):
3445         (WTR::AccessibilityUIElement::stringForTextMarkerRange):
3446         (WTR::AccessibilityUIElement::textMarkerRangeForMarkers):
3447         (WTR::AccessibilityUIElement::startTextMarkerForTextMarkerRange):
3448         (WTR::AccessibilityUIElement::endTextMarkerForTextMarkerRange):
3449         (WTR::AccessibilityUIElement::textMarkerForPoint):
3450         (WTR::AccessibilityUIElement::accessibilityElementForTextMarker):
3451         (WTR::AccessibilityUIElement::attributedStringForTextMarkerRangeContainsAttribute):
3452         (WTR::AccessibilityUIElement::indexForTextMarker):
3453         (WTR::AccessibilityUIElement::isTextMarkerValid):
3454         (WTR::AccessibilityUIElement::textMarkerForIndex):
3455
3456         Add documentEncoding and documentURI to AccessibilityUIElement.idl.
3457
3458         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityUIElement.idl:
3459
3460 2012-07-18  Alexis Menard  <alexis.menard@openbossa.org>
3461
3462         [EFL] Build fix in WebKitTestRunner.
3463         https://bugs.webkit.org/show_bug.cgi?id=91567
3464
3465         Reviewed by Kentaro Hara.
3466
3467         sleep() is defined in unistd.h, we need to include it.
3468
3469         * WebKitTestRunner/efl/TestControllerEfl.cpp:
3470
3471 2012-07-18  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
3472
3473         [NRWT] Unreviewed gardening after r122913
3474         https://bugs.webkit.org/show_bug.cgi?id=91601
3475
3476         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
3477         (TrivialMockPort.process_kill_time):
3478
3479 2012-07-17  Xianzhu Wang  <wangxianzhu@chromium.org>
3480
3481         [Chromium-Android] Run ref tests together to avoid expensive driver restarts
3482         https://bugs.webkit.org/show_bug.cgi?id=91533
3483
3484         Reviewed by Dirk Pranke.
3485
3486         Though DriverProxy maintains two drivers to support pixel tests and non-pixel tests,
3487         chromium-android uses another way because it can't support multiple drivers.
3488         It restarts the driver when pixel-test mode changes (e.g. when running a ref test after
3489         a normal test in --no-pixel-tests mode). However restarting driver is expensive on
3490         Android (several seconds each time). To reduce the cost, a command line option
3491         '--shard-ref-tests' is added to group ref tests in dedicated shards.
3492         The option is by default enabled on Android.
3493
3494         Will remove the option once DRT supports switching pixel test mode during one run.
3495         (https://bugs.webkit.org/show_bug.cgi?id=91538, https://bugs.webkit.org/show_bug.cgi?id=91539)
3496
3497         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3498         (Manager._shard_tests):
3499         (Manager._shard_in_two):
3500         (Manager._shard_by_directory):
3501         (Manager._run_tests):
3502         * Scripts/webkitpy/layout_tests/controllers/worker.py:
3503         (Worker._update_test_input):
3504         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3505         (ChromiumAndroidPort.__init__):
3506         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3507         (parse_args):
3508
3509 2012-07-17  Don Olmstead  <don.olmstead@am.sony.com>
3510
3511         NRWT The time before server_process kills DRT should be variable
3512         https://bugs.webkit.org/show_bug.cgi?id=91542
3513
3514         Reviewed by Dirk Pranke.
3515
3516         Added the ability to configure the amount of time
3517         server_process waits before killing the DRT process
3518         during the call to stop.
3519
3520         * Scripts/webkitpy/layout_tests/port/base.py:
3521         (Port.variable.process_kill_time):
3522         * Scripts/webkitpy/layout_tests/port/server_process.py:
3523         (ServerProcess.stop):
3524
3525 2012-07-17  Alexis Menard  <alexis.menard@openbossa.org>
3526
3527         webkitdirs.pm should fallback to uname -m if arch is not present.
3528         https://bugs.webkit.org/show_bug.cgi?id=91543
3529
3530         Reviewed by Martin Robinson.
3531
3532         It seems that coreutils is moving away from the 'arch' command. Archlinux
3533         for example doesn't ship it anymore (coreutils 8.17). We can then fallback
3534         to 'uname -m' to find out the architecture.
3535
3536         * Scripts/webkitdirs.pm:
3537         (determineArchitecture):
3538
3539 2012-07-17  Dirk Pranke  <dpranke@chromium.org>
3540
3541         nrwt: rename printer.print_update to printer.write_update to match metered_stream
3542         https://bugs.webkit.org/show_bug.cgi?id=91557
3543
3544         Reviewed by Ojan Vafai.
3545
3546         Now that all of the printing logic is in printer.py, I want the
3547         public interface to printer to follow the meteredstream
3548         interface where possible. renaming write_update() gets us close.
3549
3550         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3551         (Manager._run_tests):
3552         (Manager._set_up_run):
3553         (Manager.start_servers_with_lock):
3554         (Manager.stop_servers_with_lock):
3555         (Manager._clobber_old_results):
3556         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3557         (run):
3558         * Scripts/webkitpy/layout_tests/views/printing.py:
3559         (Printer.write_update):
3560         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
3561         (Testprinter.test_write_update):
3562
3563 2012-07-17  Tony Chang  <tony@chromium.org>
3564
3565         Fix a bug where optimize-baselines would incorrectly fail to optimize
3566         https://bugs.webkit.org/show_bug.cgi?id=91551
3567
3568         Reviewed by Adam Barth.
3569
3570         In some cases, optimize-baseline would correctly optimize the results, but because
3571         we weren't filtering the virtual ports out of _results_by_port_name, we thought
3572         that we had failed to optimize.
3573
3574         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
3575         (BaselineOptimizer._find_optimal_result_placement): No virtual filtering here.
3576         (BaselineOptimizer._filtered_results_by_port_name): New function that filters out virtual directories.
3577         (BaselineOptimizer.optimize): Filter out virtual ports.
3578         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
3579         (TestBaselineOptimizer._move_baselines): Add a stub so we don't actual move results.
3580         (BaselineOptimizerTest._assertOptimization): Call optimize and verify that the right files were moved.
3581         (BaselineOptimizerTest._assertOptimizationFailed): Add a method for when optimization should fail.
3582         (BaselineOptimizerTest.test_common_directory_includes_root): Update since this test should fail.
3583         (BaselineOptimizerTest.test_virtual_ports_filtered): New test case that demonstrates the bug.
3584
3585 2012-07-17  Xianzhu Wang  <wangxianzhu@chromium.org>
3586
3587         [Chromium] Add --encode-binary command line option for DRT
3588         https://bugs.webkit.org/show_bug.cgi?id=91532
3589
3590         Reviewed by Adam Barth.
3591
3592         When the option presents, DRT will encode binary output data in base64.
3593         ChromiumAndroidPort will pass the option because 'adb shell' doesn't support
3594         binary data output.
3595
3596         This change is a refactoring to replace the original hard-coded code for Android.
3597
3598         * DumpRenderTree/chromium/DumpRenderTree.cpp:
3599         (main):
3600         * DumpRenderTree/chromium/TestEventPrinter.cpp:
3601         (TestEventPrinter::TestEventPrinter):
3602         (TestEventPrinter::handleAudio): New function replacing the original handleAudioHeader
3603         (TestEventPrinter::handleImage):
3604         (TestEventPrinter::handleBinary): Extracted common function for handleAudio and handleImage
3605         * DumpRenderTree/chromium/TestEventPrinter.h:
3606         (TestEventPrinter):
3607         * DumpRenderTree/chromium/TestShell.cpp:
3608         (TestShell::TestShell):
3609         (TestShell::initialize):
3610         (TestShell::dump):
3611         * DumpRenderTree/chromium/TestShell.h:
3612         (TestShell):
3613         (TestShell::setEncodeBinary):
3614         * Scripts/webkitpy/layout_tests/port/chromium_android.py: Add --encode-binary to additional_drt_flags
3615         (ChromiumAndroidPort.__init__):
3616
3617 2012-07-17  Dirk Pranke  <dpranke@chromium.org>
3618
3619         nrwt: move per-test result output into printing.py
3620         https://bugs.webkit.org/show_bug.cgi?id=91465
3621
3622         Reviewed by Ojan Vafai.
3623
3624         This moves the last substantive logging/printing code from the
3625         manager into printing.py. Subsequent patches in this thread can
3626         largely focus on cleaning up printing.py and will leave the rest
3627         of the code alone.
3628
3629         This patch also removes manager.update() and
3630         manager.update_summary(), two routines that were no longer being
3631         used or needed (they have been dead for a long time but I didn't
3632         notice them until now).
3633
3634         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3635         (Manager.results_directory):
3636         (Manager._look_for_new_crash_logs):