c74042e760a9edf23ca4fd21f278cd6b0beb0a29
[WebKit-https.git] / Tools / ChangeLog
1 2012-07-31  Jochen Eisinger  <jochen@chromium.org>
2
3         Unreviewed. Update watchlist
4
5         * Scripts/webkitpy/common/config/watchlist: Turned out to be less useful than I had hoped
6
7 2012-07-31  Kent Tamura  <tkent@chromium.org>
8
9         Unreviewed, update watchlist
10
11         * Scripts/webkitpy/common/config/watchlist:
12         - Improve "Forms" filename pattern.
13         - tkent stops watching Chromium DRT. It was not so helpful.
14
15 2012-07-31  Martin Robinson  <mrobinson@igalia.com>
16
17         [jhbuild] Also try to load jhbuild Python module from the source checkout
18
19         Reviewed by Philippe Normand.
20
21         JHBuild also seems to have an installation mode where it prefers to load its
22         Python modules from the source checkout. This seems to be used particularly
23         on the GTK+ 64-bit release bot. We should fall back to this mode.
24
25         * jhbuild/jhbuildutils.py:
26         (enter_jhbuild_environment_if_available): Use the source checkout if there
27         is no installation path.
28
29 2012-07-30  Ryosuke Niwa  <rniwa@webkit.org>
30
31         Revert r124153 and land the "right" build fix per ap's comment on the bug 92676.
32
33         * MiniBrowser/win/BrowserView.cpp:
34         (BrowserView::create):
35
36 2012-07-30  Sam Weinig  <sam@webkit.org>
37
38         Rename WebKit2.h to WebKit2_C.h
39         https://bugs.webkit.org/show_bug.cgi?id=92704
40
41         Reviewed by Dan Bernstein.
42
43         * MiniBrowser/mac/MiniBrowser_Prefix.pch:
44         * MiniBrowser/win/BrowserView.h:
45         * TestWebKitAPI/config.h:
46         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
47         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
48         * WebKitTestRunner/WebKitTestRunnerPrefix.h:
49         * WebKitTestRunner/config.h:
50         Update for changed header file name.
51
52 2012-07-30  Dirk Pranke  <dpranke@chromium.org>
53
54         chromium win compile broken after removing webkit.py
55         https://bugs.webkit.org/show_bug.cgi?id=92549
56
57         Reviewed by Ryosuke Niwa.
58
59         Remove the stub file for webkit.py now that it's no longer
60         necessary.
61
62         * Scripts/webkitpy/layout_tests/port/webkit.py: Removed.
63
64 2012-07-30  Ryosuke Niwa  <rniwa@webkit.org>
65
66         Windows build fix attempt after r124092.
67
68         * MiniBrowser/win/BrowserView.cpp:
69         (BrowserView::create):
70
71 2012-07-30  Ryosuke Niwa  <rniwa@webkit.org>
72
73         Remove erroneously committed debugging print call from a test.
74
75         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
76         (test_run_generates_results_page):
77
78 2012-07-30  Arnaud Renevier  <a.renevier@sisa.samsung.com>
79
80         webkit-patch: system keyring is not used to read my password
81         https://bugs.webkit.org/show_bug.cgi?id=92532
82
83         Reviewed by Dirk Pranke.
84
85         In case no username can be fetched from environment, git or keychain,
86         prompt for username, and try to get password from keyring associated
87         with that username.
88
89         * Scripts/webkitpy/common/net/credentials.py:
90         (Credentials.read_credentials):
91         * Scripts/webkitpy/common/net/credentials_unittest.py:
92         (test_keyring_without_git_repo_nor_keychain):
93         (test_keyring_without_git_repo_nor_keychain.MockKeyring):
94         (test_keyring_without_git_repo_nor_keychain.MockKeyring.get_password):
95         (test_keyring_without_git_repo_nor_keychain.FakeCredentials):
96         (test_keyring_without_git_repo_nor_keychain.FakeCredentials._credentials_from_keychain):
97         (test_keyring_without_git_repo_nor_keychain.FakeCredentials._credentials_from_environment):
98         (test_keyring_without_git_repo_nor_keychain.FakeUser):
99         (test_keyring_without_git_repo_nor_keychain.FakeUser.prompt):
100         (test_keyring_without_git_repo_nor_keychain.FakeUser.prompt_password):
101
102 2012-07-30  Dirk Pranke  <dpranke@chromium.org>
103
104         nrwt: move the code that identifies the chunk of tests to run into finder
105         https://bugs.webkit.org/show_bug.cgi?id=92694
106
107         Reviewed by Ryosuke Niwa.
108
109         Another patch to make manager.py smaller and clearer; this patch moves
110         the handling of breaking the list of tests into chunks (--run-part,
111         --run-chunk) from manager.py to finder.py.
112
113         There are no functional changes and this is covered by existing tests.
114
115         * Scripts/webkitpy/layout_tests/controllers/finder.py:
116         (Finder.__init__):
117         (Finder._read_test_files):
118         (Finder):
119         (Finder.split_into_chunks_if_necessary):
120         * Scripts/webkitpy/layout_tests/controllers/manager.py:
121         (Manager.__init__):
122         (Manager._collect_tests):
123         (Manager._parse_expectations):
124         (Manager.prepare_lists_and_print_output):
125
126 2012-07-30  Dirk Pranke  <dpranke@chromium.org>
127
128         nrwt: split test-finding code out from manager.py
129         https://bugs.webkit.org/show_bug.cgi?id=92693
130
131         Reviewed by Ryosuke Niwa.
132
133         In the interest of making manager.py smaller, this patch moves
134         the code that actually takes the command line arguments and
135         --test-file lists of tests to run and expands them into an
136         actual list of tests out into a separate module.
137
138         * Scripts/webkitpy/layout_tests/controllers/finder.py: Added.
139         (LayoutTestFinder):
140         (LayoutTestFinder.__init__):
141         (LayoutTestFinder.find_tests):
142         (LayoutTestFinder._strip_test_dir_prefixes):
143         (LayoutTestFinder._strip_test_dir_prefix):
144         (LayoutTestFinder._read_test_names_from_file):
145         * Scripts/webkitpy/layout_tests/controllers/manager.py:
146         (Manager._collect_tests):
147         (Manager._handle_finished_test):
148         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
149         (strip_comments):
150
151 2012-07-30  Adam Barth  <abarth@webkit.org>
152
153         commit-queue is corrupting ChangeLogs
154         https://bugs.webkit.org/show_bug.cgi?id=92681
155
156         Reviewed by Tony Chang.
157
158         Some commits made with the commit-queue are ending up with duplicate
159         ChangeLog entries. I've hot-patched the servers to not use a special
160         merge driver for ChangeLogs, which should fix this issue. This patch
161         removes the merge driver from our cold-boot.sh script so that we won't
162         use it on new commit-queue instances.
163
164         * EWSTools/cold-boot.sh:
165
166 2012-07-28  Sam Weinig  <sam@webkit.org>
167
168         Add ability to load from a string to the ObjC WK API
169         https://bugs.webkit.org/show_bug.cgi?id=92590
170
171         Reviewed by Dan Bernstein.
172
173         Add tests for [WKBrowsingContextController loadHTMLString:baseURL:]. We will be able
174         to greatly improve these tests (to test more than just not crashing) when methods to
175         access page content are added (soon!).
176
177         Adds:
178             Test: WKBrowsingContextLoadDelegateTest_SimpleLoadOfHTMLString
179             Test: WKBrowsingContextLoadDelegateTest_SimpleLoadOfHTMLString_NilBaseURL
180             Test: WKBrowsingContextLoadDelegateTest_SimpleLoadOfHTMLString_NilHTMLStringAndBaseURL
181
182         * TestWebKitAPI/Tests/WebKit2ObjC/WKBrowsingContextLoadDelegateTest.mm:
183         Adds tests and refactors delegates to not use global state.
184
185 2012-07-30  Thiago Marcos P. Santos  <thiago.santos@intel.com>
186
187         [CMake] TestWebKitAPI bundle should link with WTF
188         https://bugs.webkit.org/show_bug.cgi?id=92616
189
190         Reviewed by Antonio Gomes.
191
192         Fixes build when shared core is disabled.
193
194         * TestWebKitAPI/CMakeLists.txt:
195
196 2012-07-30  Dirk Pranke  <dpranke@chromium.org>
197
198         nrwt: clean up handling of 'expected' stats
199         https://bugs.webkit.org/show_bug.cgi?id=92527
200
201         Reviewed by Tony Chang.
202
203         This patch alters the way we compute and log the "expected"
204         results and how we treat skipped tests; we will now log the
205         number of skipped tests separately from the categories, e.g.:
206
207         Found 31607 tests; running 24464.
208         Expect: 23496 passes   (23496 now,    0 wontfix)
209         Expect:   548 failures (  543 now,    5 wontfix)
210         Expect:   420 flaky    (  245 now,  175 wontfix)
211
212         (so that the "expect" totals add up to the "running" totals);
213         in addition, the totals in the one-line-progress reflect the
214         number of tests we will actually run. If --iterations or
215         --repeat-each are specified, the number of tests we run are
216         multiplied as appropriate, but the "expect" numbers are
217         unchanged, since we don't count multiple invocations of the same
218         test multiple times. In addition, if we are using --run-part or
219         --run-chunk, the tests we don't run are treated as skipped
220         for consistency. We will also log the values for --iterations
221         and --repeat each as part of the found/running line.
222
223         Previously the code had parsed and re-parsed the
224         TestExpectations files several times in an attempt to come up
225         with some sane statistics, but this was expensive and lead to
226         confusing layer; treating files as skipped in the way described
227         above is more consistent and cleaner.
228
229         * Scripts/webkitpy/layout_tests/controllers/manager.py:
230         (Manager._split_into_chunks_if_necessary):
231         (Manager.prepare_lists_and_print_output):
232         (Manager.run):
233         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
234         (ManagerTest.test_interrupt_if_at_failure_limits):
235         (ManagerTest.test_update_summary_with_result):
236         (ManagerTest.test_look_for_new_crash_logs):
237         (ResultSummaryTest.get_result_summary):
238         * Scripts/webkitpy/layout_tests/models/result_summary.py:
239         (ResultSummary.__init__):
240         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
241         (TestExpectationParser.expectation_for_skipped_test):
242         (TestExpectations.__init__):
243         (TestExpectations.add_skipped_tests):
244           Here we make add_skipped_tests() public, so that we can update
245           the expectations for tests that we are skipping due to
246           --run-part or --run-chunk; we use the wontfix flag so that
247           the tests that are intentionally skipped aren't considered
248           "fixable".
249         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
250         (SkippedTests.check):
251         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
252         (parse_args):
253         * Scripts/webkitpy/layout_tests/views/printing.py:
254         (Printer.print_found):
255         (Printer):
256         (Printer.print_expected):
257         (Printer._print_result_summary):
258         (Printer._print_result_summary_entry):
259           Here we split out printing the number of tests found and run
260           from the expected results, to be clearer and so that we don't
261           have to reparse the expectations to update the stats.
262         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
263         (Testprinter.get_result_summary):
264
265 2012-07-30  Sadrul Habib Chowdhury  <sadrul@chromium.org>
266
267         Propagate gesture events to plugins.
268         https://bugs.webkit.org/show_bug.cgi?id=92281
269
270         Reviewed by Adam Barth.
271
272         Update TestWebPlugin to receive events, and output logs for the events.
273
274         * DumpRenderTree/chromium/TestWebPlugin.cpp:
275         (TestWebPlugin::handleInputEvent):
276         * DumpRenderTree/chromium/TestWebPlugin.h:
277         (TestWebPlugin::acceptsInputEvents):
278         (TestWebPlugin):
279
280 2012-07-30  Raphael Kubo da Costa  <rakuco@webkit.org>
281
282         [jhbuild] gnutls 2.12.14 does not build with glibc 2.16.0
283         https://bugs.webkit.org/show_bug.cgi?id=90643
284
285         Reviewed by Gustavo Noronha Silva.
286
287         The 2.x series of gnutls has a version of Gnulib that is
288         incompatible with glibc 2.16.0, which now does not define gets()
289         by default.
290
291         After talking to kov, mrobinson and philn, it looks like it makes
292         sense to simply remove gnutls from jhbuild.modules and rely on the
293         version installed on the system being recent enough.
294
295         This means at least version 2.12.8 for PKCS11 support, or at the
296         very least 2.11.0 (these are the minimum versions enforced by
297         glib-networking's configure.ac).
298
299         * efl/jhbuild.modules: Do not build gnutls anymore, make
300         glib-networking depend on glib instead, and move the libgcrypt
301         dependency to eet, which is what currently needs it.
302         * gtk/jhbuild.modules: Do not build gnutls anymore.
303
304 2012-07-30  Dirk Pranke  <dpranke@chromium.org>
305
306         nrwt: --verbose logging does not work right on windows
307         https://bugs.webkit.org/show_bug.cgi?id=92673
308
309         Unreviewed, build fix.
310
311         The --verbose log level is not being propagated to the worker
312         processes on windows properly; this has been broken ever since I
313         refactored logging into the message pool directly :(.
314
315         Also, fixing this revealed that outputcapture wasn't saving and
316         restoring log levels correctly, so I had to fix that as well.
317
318         * Scripts/webkitpy/common/message_pool.py:
319         (_MessagePool._start_workers):
320         (_MessagePool._worker_log_level):
321         (_Worker.__init__):
322         (_Worker._set_up_logging):
323         (_WorkerLogHandler.__init__):
324         * Scripts/webkitpy/common/system/outputcapture.py:
325         (OutputCapture.capture_output):
326         (OutputCapture.restore_output):
327
328 2012-07-30  Dirk Pranke  <dpranke@chromium.org>
329
330         After r123895, new-run-webkit-tests can fail with TypeError when a test crashes
331         https://bugs.webkit.org/show_bug.cgi?id=92664
332
333         Reviewed by Michael Saboff.
334
335         Updated to handle crash_logs being returned as a tuple of (stderr, crash_log).
336
337         * Scripts/webkitpy/layout_tests/port/mac.py:
338         (MacPort.look_for_new_crash_logs):
339
340 2012-07-30  Stephen White  <senorblanco@chromium.org>
341
342         Added a SkiaGraphics definition to the watchlist file; cc'ed myself
343         to that group.  Unreviewed.
344
345         * Scripts/webkitpy/common/config/watchlist:
346
347 2012-07-30  Sheriff Bot  <webkit.review.bot@gmail.com>
348
349         Unreviewed, rolling out r123966 and r123967.
350         http://trac.webkit.org/changeset/123966
351         http://trac.webkit.org/changeset/123967
352         https://bugs.webkit.org/show_bug.cgi?id=92656
353
354         This patch is causing assertion failures on the debug bot
355         (also rolling out a dependent patch) (Requested by mrobinson
356         on #webkit).
357
358         * gtk/generate-gtkdoc:
359         (get_webkit2_options):
360
361 2012-07-30  Martin Robinson  <mrobinson@igalia.com>
362
363         [GTK] Add a non-subprocess jhbuild environment and use it for run-gtk-tests
364         https://bugs.webkit.org/show_bug.cgi?id=92626
365
366         Reviewed by Gustavo Noronha Silva.
367
368         Instead of always shelling out to enter a jhbuild environment, add a Python
369         hook for entering a jhbuild environment directly. This avoids requiring a
370         script to be wrapped in jhbuild.
371
372         * GNUmakefile.am: Remove gtk/run-api-tests from the distribution list.
373         * Scripts/run-gtk-tests: Move most of the logic from run-api-tests here
374         with additional code for entering a jhbuild environment directly.
375         * gtk/run-api-tests: Removed.
376         * jhbuild/jhbuild-wrapper: Use the new helper methods defined in jhbuildutils.py.
377         * jhbuild/jhbuildutils.py: Add some new helper methods here including
378         a method which can transform the current environment into a jhbuild-ified one.
379
380 2012-07-30  Alexander Pavlov  <apavlov@chromium.org>
381
382         Unreviewed, upgrade apavlov to reviewer.
383         http://lists.webkit.org/mailman/private/webkit-committers/2012-July/000184.html
384
385         * Scripts/webkitpy/common/config/committers.py:
386
387 2012-07-30  Robin Cao  <robin.cao@torchmobile.com.cn>
388
389         [BlackBerry] Adapt to changes in the SharedArray platform API
390         https://bugs.webkit.org/show_bug.cgi?id=92631
391
392         Reviewed by Rob Buis.
393
394         Adapt to changes in the SharedArray platform API. No behavioural change.
395
396         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
397         (LayoutTestController::webHistoryItemCount):
398
399 2012-07-30  Balazs Kelemen  <kbalazs@webkit.org>
400
401         [Qt][NRWT] REGRESSION(123729): Forcing pixel tests with -p doesn't work
402         https://bugs.webkit.org/show_bug.cgi?id=92627
403
404         Reviewed by Noam Rosenthal.
405
406         Remove the default behavior I added in r123729 so that we run all tests as pixel
407         test with -p. It was intentional but it seems like we don't have consensus on
408         that we want it.
409
410         * Scripts/webkitpy/layout_tests/port/qt.py:
411         (QtPort._supports_switching_pixel_tests_per_test):
412
413 2012-07-30  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
414
415         [EFL][WTR] Enforce SHIFT modifier to upper case keys in event sender.
416         https://bugs.webkit.org/show_bug.cgi?id=92366
417
418         Reviewed by Antonio Gomes.
419
420         Add SHIFT modifier for capital letters as VK_
421         codes do not distinguish between lows and caps and also to be consistent
422         with other ports which send caps as lowercase letters with the shift key down.
423
424         * WebKitTestRunner/efl/EventSenderProxyEfl.cpp:
425         (WTR::EventSenderProxy::keyDown):
426
427 2012-07-29  Raphael Kubo da Costa  <rakuco@webkit.org>
428
429         [EFL][jhbuild] Make run-with-jhbuild become the jhbuild process itself.
430         https://bugs.webkit.org/show_bug.cgi?id=92592
431
432         Reviewed by Eric Seidel.
433
434         Adapt to the changes made to GTK+'s run-with-jhbuild in r123979
435         and make EFL's run-with-jhbuild call os.execve() instead of
436         launching a subprocess as well.
437
438         This should make it possible to interrupt execution of jhbuild
439         cleanly.
440
441         * efl/run-with-jhbuild:
442
443 2012-07-28  Ryosuke Niwa  <rniwa@webkit.org>
444
445         run-perf-tests should generate a results page
446         https://bugs.webkit.org/show_bug.cgi?id=92575
447
448         Reviewed by Eric Seidel.
449
450         Added the ability to generate a results page that summarizes performance test results from
451         multiple runs of run-perf-tests when --output-json-path is specified and --test-results-server
452         is not specified. We cannot generate a results page when --test-results-server is specified
453         because perf-o-matic cannot parse the new JSON format.
454
455         The new JSON format is simply an array of the old JSON output. This JSON is then merged into
456         results-template.html along with jQuery and copied as a HTML file of the same filename as
457         the output JSON file with a .html extension.
458
459         We merge all scripts and the JSON output into the results page as opposed to including them
460         with the src content attribute to make it standalone so that we may post it on Bugzilla or
461         e-mail it to someone else without breaking it.
462
463         * Scripts/webkitpy/performance_tests/perftestsrunner.py:
464         (PerfTestsRunner.run):
465         (PerfTestsRunner._generate_json): Merge "contents" with the existing JSON file when generating
466         a results page, and generate the results page from results-template.html merged with jQuery and
467         the generated JSON output.
468         * Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
469         (_test_run_with_json_output): Extracted from test_run_with_upload_json to be shared with
470         several test cases.
471         (_test_run_with_json_output.mock_upload_json): Ditto.
472         (test_run_with_json_output): Refactored to use _test_run_with_json_output.
473         (test_run_generates_results_page): Added. Tests the new feature.
474         (test_run_with_json_source): Refactored to use _test_run_with_json_output.
475         (test_run_with_multiple_repositories): Ditto.
476         (test_run_with_upload_json): Ditto.
477
478 2012-07-29  Rik Cabanier  <cabanier@adobe.com>
479
480         Add ENABLE_CSS_COMPOSITING flag
481         https://bugs.webkit.org/show_bug.cgi?id=92553
482
483         Reviewed by Dirk Schulze.
484
485         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
486
487         * Scripts/webkitperl/FeatureList.pm:
488
489 2012-07-29  Gustavo Noronha Silva  <gns@gnome.org>
490
491         Reviewed by Martin Robinson.
492
493         Make the jhbuild wrapper scripts become the jhbuild process
494         instead of starting it as a child. This should make it possible to
495         interrupt execution of jhbuild cleanly.
496
497         * gtk/run-with-jhbuild:
498         * jhbuild/jhbuild-wrapper:
499
500 2012-07-29  Mike West  <mkwst@chromium.org>
501
502         Extend `application/x-webkit-test-netscape` plugins to better support multiple frames.
503         https://bugs.webkit.org/show_bug.cgi?id=92478
504
505         Reviewed by Adam Barth.
506
507         DumpRenderTree currently understands
508         `<object src="data:application/x-webkit-test-netscape,alertwhenloaded">`.
509         It would be useful if it understood `logifloaded`, which could give
510         more context regarding test expectations, and can be written in such
511         as way as to support plugins loaded into multiple iframes in a single
512         test.
513
514         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
515         (NPP_New):
516             Adding support for `logifloaded`.
517
518 2012-07-28  Peter Kasting  <pkasting@google.com>
519
520         Fix Chromium/Win compile.
521         https://bugs.webkit.org/show_bug.cgi?id=88787
522
523         Unreviewed, build fix.
524
525         Chromium/Win needs _USE_MATH_DEFINES but doesn't define PLATFORM(WIN).
526         The other files in WebKit that define this macro (both in
527         WebCore/platform/graphics/cg) do it by unconditionally #defining it to 1
528         above the #include of "config.h", so copy that pattern here.
529
530         * TestWebKitAPI/Tests/WTF/MediaTime.cpp:
531
532 2012-07-28  Mario Sanchez Prada  <msanchez@igalia.com>
533
534         [WK2][GTK] Implement a new spell checker API for WebKit2GTK+
535         https://bugs.webkit.org/show_bug.cgi?id=90268
536
537         Reviewed by Martin Robinson.
538
539         Ignore WebKitTextChecker.* private files for gtk-doc.
540
541         * gtk/generate-gtkdoc:
542         (get_webkit2_options):
543
544 2012-07-28  Philippe Normand  <pnormand@igalia.com>
545
546         [GTK] Transparent Media controls timeline and panel
547         https://bugs.webkit.org/show_bug.cgi?id=85279
548
549         Reviewed by Martin Robinson.
550
551         * gtk/jhbuild.modules: Provide gnome-themes-standard build
552         support, this is necessary so the media controls panel colors are
553         properly initialized from the GTK+ theme.
554
555 2012-07-27  Kihong Kwon  <kihong.kwon@samsung.com>
556
557         [EFL] Support for HTML media capture
558         https://bugs.webkit.org/show_bug.cgi?id=91842
559
560         Reviewed by Ryosuke Niwa.
561
562         Add HTML media capture feature to the feature list.
563         In addition, enable that on the Efl port.
564
565         * Scripts/webkitperl/FeatureList.pm:
566
567 2012-07-27  Dirk Pranke  <dpranke@chromium.org>
568
569         "webkit-patch rebaseline <test>" fails with exit_code 1
570         https://bugs.webkit.org/show_bug.cgi?id=91560
571
572         Reviewed by Adam Barth.
573
574         webkit-patch rebaseline-json crashes in an svn checkout if it
575         tries to add files and there are no files to add; arguably
576         svn.add() should be robust against this, but rebaseline-json
577         shouldn't be calling it, either. This patch fixes the latter and
578         adds better debugging info to garden-o-matic so we can see what
579         caused the crash.
580
581         * Scripts/webkitpy/tool/commands/rebaseline.py:
582         (AbstractParallelRebaselineCommand._rebaseline):
583         * Scripts/webkitpy/tool/servers/gardeningserver.py:
584         (GardeningHTTPRequestHandler.rebaselineall):
585         (GardeningHTTPRequestHandler.rebaselineall.error_handler):
586         (GardeningHTTPRequestHandler):
587
588 2012-07-27  Brady Eidson  <beidson@apple.com>
589
590         Plugins should not be allowed to override standard properties/attributes in non-standard worlds
591         <rdar://problem/11975252> and https://bugs.webkit.org/show_bug.cgi?id=92519
592
593         Reviewed by Anders Carlsson.
594
595         Add a test plugin which overrides all properties it is asked about:
596         * DumpRenderTree/TestNetscapePlugIn/Tests/PluginScriptableObjectOverridesAllProperties.cpp: Added.
597         (PluginScriptableObjectOverridesAllProperties):
598         (PluginScriptableObjectOverridesAllProperties::PluginScriptableObjectOverridesAllProperties):
599         (PluginObject):
600         (PluginScriptableObjectOverridesAllProperties::PluginObject::PluginObject):
601         (PluginScriptableObjectOverridesAllProperties::PluginObject::~PluginObject):
602         (PluginScriptableObjectOverridesAllProperties::PluginObject::hasProperty):
603         (PluginScriptableObjectOverridesAllProperties::PluginObject::getProperty):
604         (PluginScriptableObjectOverridesAllProperties::NPP_GetValue):
605
606         Expose NPN_MemAlloc to the test plugin:
607         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
608         (PluginTest::NPN_MemAlloc):
609         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
610         (PluginTest):
611
612         * DumpRenderTree/DumpRenderTree.gypi:
613         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
614         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
615         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
616
617 2012-07-27  Rafael Brandao  <rafael.lobo@openbossa.org>
618
619         "webkit-patch upload" should strip colors from patch if necessary
620         https://bugs.webkit.org/show_bug.cgi?id=92306
621
622         Reviewed by Adam Barth.
623
624         * Scripts/webkitpy/common/checkout/scm/git.py: Explicity remove colors from diff.
625
626 2012-07-27  Eric Seidel  <eric@webkit.org>
627
628         clean-pending-commit throws exception when encountering Ossy
629         https://bugs.webkit.org/show_bug.cgi?id=92534
630
631         Reviewed by Adam Barth.
632
633         Mark a couple strings as unicode so that reviewers with unicode names do not cause
634         clean-pending-commit to barf.
635
636         * Scripts/webkitpy/tool/commands/upload.py:
637         (CleanPendingCommit._flags_to_clear_on_patch):
638         (CleanPendingCommit.execute):
639         (AssignToCommitter._assign_bug_to_last_patch_attacher):
640
641 2012-07-27  Dirk Pranke  <dpranke@chromium.org>
642
643         chromium win compile broken after removing webkit.py
644         https://bugs.webkit.org/show_bug.cgi?id=92549
645
646         Unreviewed, build fix.
647
648         It turns out that this file is listed in the chromium downstream
649         'browser_tests.isolate' file, and so deleting it broke gyp.
650         Adding a stub file back in until we can remove the entry from
651         the file.
652
653         * Scripts/webkitpy/layout_tests/port/webkit.py: Added.
654
655 2012-07-27  Dirk Pranke  <dpranke@chromium.org>
656
657         nrwt: remove 'webkit.py' and the 'WebKitPort'
658         https://bugs.webkit.org/show_bug.cgi?id=92450
659
660         Reviewed by Ryosuke Niwa.
661
662         This patch finishes the work of merging WebKitPort into Port.
663         Now Port is way too big :).
664
665         This is all cutting and pasting; no new functionality and
666         no changes in the tests.
667
668         * Scripts/webkitpy/layout_tests/port/apple.py:
669         (ApplePort):
670         (ApplePort.__init__):
671         * Scripts/webkitpy/layout_tests/port/base.py:
672         (Port.skipped_layout_tests):
673         (Port):
674         (Port._port_flag_for_scripts):
675         (Port._arguments_for_configuration):
676         (Port._run_script):
677         (Port._build_driver):
678         (Port._build_driver_flags):
679         (Port._tests_for_other_platforms):
680         (Port._runtime_feature_list):
681         (Port.nm_command):
682         (Port._modules_to_search_for_symbols):
683         (Port._symbols_string):
684         (Port._missing_feature_to_skipped_tests):
685         (Port._missing_symbol_to_skipped_tests):
686         (Port._has_test_in_directories):
687         (Port._skipped_tests_for_unsupported_features):
688         (Port._wk2_port_name):
689         (Port._skipped_file_search_paths):
690         * Scripts/webkitpy/layout_tests/port/chromium.py:
691         (ChromiumPort):
692         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
693         * Scripts/webkitpy/layout_tests/port/chromium_port_testcase.py:
694         (ChromiumPortTestCase.test_check_build):
695         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
696         * Scripts/webkitpy/layout_tests/port/efl.py:
697         (EflPort):
698         (EflPort.__init__):
699         * Scripts/webkitpy/layout_tests/port/gtk.py:
700         (GtkPort):
701         (GtkPort.setup_environ_for_server):
702         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
703         (TestWebKitPort):
704         (TestWebKitPort.__init__):
705         (TestWebKitPort.all_test_configurations):
706         (TestWebKitPort._symbols_string):
707         (TestWebKitPort._tests_for_other_platforms):
708         (TestWebKitPort._tests_for_disabled_features):
709         (PortTestCase):
710         (PortTestCase.test_diff_image):
711         (PortTestCase.test_path_to_test_expectations_file):
712         (PortTestCase.test_skipped_directories_for_symbols):
713         (test_skipped_directories_for_features):
714         (test_skipped_directories_for_features_no_matching_tests_in_test_list):
715         (test_skipped_tests_for_unsupported_features_empty_test_list):
716         (test_skipped_layout_tests):
717         (test_skipped_file_search_paths):
718         (test_root_option):
719         (test_test_expectations):
720         (test_build_driver):
721         (_assert_config_file_for_platform):
722         (test_linux_distro_detection):
723         (test_apache_config_file_name_for_platform):
724         (test_path_to_apache_config_file):
725         * Scripts/webkitpy/layout_tests/port/qt.py:
726         (QtPort):
727         (QtPort.__init__):
728         (QtPort.setup_environ_for_server):
729         * Scripts/webkitpy/layout_tests/port/webkit.py: Removed.
730         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py: Removed.
731         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
732         (get_test_baselines.AllPlatformsPort):
733         (get_test_baselines.AllPlatformsPort.__init__):
734         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
735         (get_test_config.TestMacPort):
736
737 2012-07-27  Christophe Dumez  <christophe.dumez@intel.com>
738
739         [WK2][WTR] LayoutTestController.deliverWebIntent() needs to be implemented
740         https://bugs.webkit.org/show_bug.cgi?id=92228
741
742         Reviewed by Anders Carlsson.
743
744         Implement LayoutTestController.deliverWebIntent() to
745         allow test cases to deliver a Web intent and test this
746         functionality.
747
748         This functionality is needed by:
749         - webintents/web-intents-delivery.html
750
751         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
752         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
753         (WTR::LayoutTestController::deliverWebIntent):
754         (WTR):
755         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
756         (LayoutTestController):
757
758 2012-07-27  Jer Noble  <jer.noble@apple.com>
759
760         Unreviewed build fix.
761
762         Wrap MSVC-specific fixes is a COMPILER() check, not a PLATFORM() check.
763
764         * TestWebKitAPI/Tests/WTF/MediaTime.cpp:
765
766 2012-07-27  Xianzhu Wang  <wangxianzhu@chromium.org>
767
768         [Chromium-Android] NRWT better handling of DRT deadlocks and crashes
769         https://bugs.webkit.org/show_bug.cgi?id=92299
770
771         Reviewed by Dirk Pranke.
772
773         1. Changed the deadlock detector thread target from a method of driver to a function to avoid its reference to the
774            current driver (see the bug for the problem of the reference.
775         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.
776         3. Extracted common loop until timeout logic into ChromiumAndroidDriver._loop_with_timeout().
777         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.
778
779         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
780         (ChromiumAndroidPort.__init__):
781         (ChromiumAndroidPort._run_adb_command): Reduced the verbose output.
782         (ChromiumAndroidDriver.__init__):
783         (ChromiumAndroidDriver._loop_with_timeout): Extracted common loop until timeout logic here.
784         (ChromiumAndroidDriver._all_pipes_created):
785         (ChromiumAndroidDriver._remove_all_pipes):
786         (ChromiumAndroidDriver._start): Changed the retry loop style.
787         (ChromiumAndroidDriver._start_once): Changed the loop style and code related to deadlock detector.
788         (ChromiumAndroidDriver._start_once.deadlock_detector): Moved the original ChromiumAndroidDriver._deadlock_detector() here.
789         (ChromiumAndroidDriver.stop): Changed the loop style.
790
791 2012-07-27  Dirk Pranke  <dpranke@chromium.org>
792
793         nrwt: move collect_tests(), parse_expectations() into manager
794         https://bugs.webkit.org/show_bug.cgi?id=92423
795
796         Reviewed by Ojan Vafai.
797
798         This is the first in a series of patches refactoring
799         run_webkit_tests.py and manager.py to make them more
800         maintainable; all this patch does is move a couple of functions
801         into manager.run_tests() so manager's public interface is
802         simple.
803
804         * Scripts/webkitpy/layout_tests/controllers/manager.py:
805         (Manager._collect_tests):
806         (Manager._parse_expectations):
807         (Manager._split_into_chunks_if_necessary):
808         (Manager.run):
809         (read_test_files):
810         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
811         (ManagerTest.test_http_locking):
812         (ManagerTest.test_perf_locking):
813         (ManagerTest.integration_test_needs_servers.get_manager_with_tests):
814         (ManagerTest.test_look_for_new_crash_logs.get_manager_with_tests):
815         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
816         (run):
817
818 2012-07-27  Dirk Pranke  <dpranke@chromium.org>
819
820         nrwt: print filtered stderr after a crash correctly
821         https://bugs.webkit.org/show_bug.cgi?id=92428
822
823         Reviewed by Adam Barth.
824
825         Some ports will run the stderr output from DRT/WTR through a
826         filter after a crash in order to print a more useful stack
827         trace; while that gets written to the crash log correctly, we
828         don't log it to the debug output properly. This change fixes
829         that and makes it clearer that we're logging stderr, not the
830         crash log.
831
832         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
833         (SingleTestRunner._handle_error):
834         * Scripts/webkitpy/layout_tests/port/base.py:
835         (Port._get_crash_log):
836         * Scripts/webkitpy/layout_tests/port/chromium.py:
837         (ChromiumPort._get_crash_log):
838         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
839         (ChromiumAndroidPortTest.test_get_crash_log):
840         * Scripts/webkitpy/layout_tests/port/driver.py:
841         (Driver.run_test):
842         * Scripts/webkitpy/layout_tests/port/gtk.py:
843         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
844         * Scripts/webkitpy/layout_tests/port/mac.py:
845         (MacPort._get_crash_log):
846         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
847         (PortTestCase.test_get_crash_log):
848
849 2012-07-27  Dirk Pranke  <dpranke@chromium.org>
850
851         nrwt: move image diffing code to a separate module
852         https://bugs.webkit.org/show_bug.cgi?id=92447
853
854         Reviewed by Ryosuke Niwa.
855
856         This patch moves the code to talk to ImageDiff into its own
857         module, and adds more tests for it. In addition, the patch
858         modifies diff_image() so that we don't automatically stop
859         ImageDiff after a single invocation, and thus subsequent
860         diffs may be slightly faster. (Note that the chromium ports
861         don't use any of this code; that is not changed by this patch).
862
863         The main motivation for this change is to move more "generic"
864         code out of the port/* classes, and in particular to move more
865         code out of webkit.py so that we can eventually eliminate it by
866         merging it into base.py.
867
868         This patch also splits MockServerProcess out from driver_unittest.py
869         so that it can be re-used.
870
871         * Scripts/webkitpy/layout_tests/port/base.py:
872         (Port.__init__):
873         (Port.diff_image):
874         (Port.clean_up_test_run):
875         * Scripts/webkitpy/layout_tests/port/driver.py:
876         (Driver.__init__):
877         (Driver._start):
878         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
879         (DriverTest.test_stop_cleans_up_properly):
880         (DriverTest.test_two_starts_cleans_up_properly):
881         (DriverTest.test_start_actually_starts):
882         * Scripts/webkitpy/layout_tests/port/efl.py:
883         (EflPort.clean_up_test_run):
884         * Scripts/webkitpy/layout_tests/port/gtk.py:
885         (GtkPort.clean_up_test_run):
886         * Scripts/webkitpy/layout_tests/port/image_diff.py: Added.
887         (ImageDiffer):
888         (ImageDiffer.__init__):
889         (ImageDiffer.diff_image):
890         (ImageDiffer._start):
891         (ImageDiffer._read):
892         (ImageDiffer.stop):
893         * Scripts/webkitpy/layout_tests/port/image_diff_unittest.py: Added.
894         (for):
895         (FakePort):
896         (FakePort.__init__):
897         (FakePort._path_to_image_diff):
898         (FakePort.setup_environ_for_server):
899         (TestImageDiffer):
900         (TestImageDiffer.test_diff_image):
901         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
902         (MockDRTPortTest.test_diff_image):
903         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
904         (PortTestCase.test_diff_image__missing_both):
905         (PortTestCase.test_diff_image):
906         (PortTestCase.test_diff_image.make_proc):
907         * Scripts/webkitpy/layout_tests/port/server_process.py:
908         (ServerProcess._start):
909         * Scripts/webkitpy/layout_tests/port/server_process_mock.py: Added.
910         (MockServerProcess):
911         (MockServerProcess.__init__):
912         (MockServerProcess.write):
913         (MockServerProcess.has_crashed):
914         (MockServerProcess.read_stdout_line):
915         (MockServerProcess.read_stdout):
916         (MockServerProcess.pop_all_buffered_stderr):
917         (MockServerProcess.read_either_stdout_or_stderr_line):
918         (MockServerProcess.start):
919         (MockServerProcess.stop):
920         (MockServerProcess.kill):
921         * Scripts/webkitpy/layout_tests/port/webkit.py:
922         (WebKitPort._build_driver_flags):
923         (WebKitPort._symbols_string):
924
925 2012-07-27  Tom Hudson  <hudson@google.com>
926
927         Activate committer bit granted April 6th so I can garden.
928         https://bugs.webkit.org/show_bug.cgi?id=92500
929
930         Reviewed by Jochen Eisinger.
931
932         * Scripts/webkitpy/common/config/committers.py:
933
934 2012-07-27  Jer Noble  <jer.noble@apple.com>
935
936         Support a rational time class for use by media elements.
937         https://bugs.webkit.org/show_bug.cgi?id=88787
938
939         Add unit tests for the WTF::MediaTime class.
940
941         Reviewed by Eric Carlson.
942
943         * TestWebKitAPI/GNUmakefile.am:
944         * TestWebKitAPI/TestWebKitAPI.gypi:
945         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
946         * TestWebKitAPI/Tests/WTF/MediaTime.cpp: Added.
947         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
948         (WTF::operator<<):
949         (TestWebKitAPI):
950         (TestWebKitAPI::TEST):
951
952 2012-07-27  Balazs Kelemen  <kbalazs@webkit.org>
953
954         [NRWT] should have a way to restrict pixel tests for individual directories
955         https://bugs.webkit.org/show_bug.cgi?id=91754
956
957         Unreviewed attempt to fix the chaos I introduced.
958
959         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
960         (WebCore::DumpRenderTree::processLine):
961
962 2012-07-27  YoungTaeck Song  <youngtaeck.song@samsung.com>
963
964         [WK2][EFL] Add an ACCELERATED_COMPOSITING implementation for Efl WebKit2
965         https://bugs.webkit.org/show_bug.cgi?id=91581
966
967         Reviewed by Noam Rosenthal.
968
969         Add a MiniBrowser's option for selecting evas engine.
970
971         * MiniBrowser/efl/main.c:
972         (browserCreate):
973         (main):
974
975 2012-07-27  Sheriff Bot  <webkit.review.bot@gmail.com>
976
977         Unreviewed, rolling out r123869.
978         http://trac.webkit.org/changeset/123869
979         https://bugs.webkit.org/show_bug.cgi?id=92501
980
981         "it did not fix anything but made things even worst"
982         (Requested by kbalazs on #webkit).
983
984         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
985         (WebCore::DumpRenderTree::DumpRenderTree):
986         (WebCore::DumpRenderTree::open):
987         (WebCore::DumpRenderTree::processLine):
988         (WebCore::DumpRenderTree::setDumpPixelsForAllTests):
989         (WebCore::DumpRenderTree::dump):
990         * DumpRenderTree/qt/DumpRenderTreeQt.h:
991         (DumpRenderTree):
992         * DumpRenderTree/qt/main.cpp:
993         (main):
994         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
995         (SingleTestRunner._should_fetch_expected_checksum):
996         (SingleTestRunner._overwrite_baselines):
997         (SingleTestRunner._compare_output):
998         * Scripts/webkitpy/layout_tests/controllers/worker.py:
999         (Worker._update_test_input):
1000         * Scripts/webkitpy/layout_tests/port/base.py:
1001         (Port.supports_switching_pixel_tests_per_test):
1002         (Port):
1003         (Port._supports_switching_pixel_tests_per_test):
1004         (Port.should_run_as_pixel_test):
1005         (Port._should_run_as_pixel_test):
1006         * Scripts/webkitpy/layout_tests/port/driver.py:
1007         (Driver.cmd_line):
1008         (Driver._command_from_driver_input):
1009         * Scripts/webkitpy/layout_tests/port/qt.py:
1010         (QtPort.check_sys_deps):
1011         (QtPort):
1012         (QtPort._supports_switching_pixel_tests_per_test):
1013         (QtPort._should_run_as_pixel_test):
1014         (QtPort._default_pixel_test_directories):
1015         * Scripts/webkitpy/layout_tests/port/test.py:
1016         (TestPort.supports_switching_pixel_tests_per_test):
1017         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1018         (_set_up_derived_options):
1019         (parse_args):
1020         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1021         (MainTest.test_run_singly_actually_runs_tests):
1022         (MainTest.test_pixel_test_directories):
1023         * WebKitTestRunner/TestController.cpp:
1024         (WTR::TestController::TestController):
1025         (WTR::TestController::initialize):
1026         (WTR::TestController::runTest):
1027         * WebKitTestRunner/TestController.h:
1028         (TestController):
1029
1030 2012-07-27  Balazs Kelemen  <kbalazs@webkit.org>
1031
1032         [NRWT] should have a way to restrict pixel tests for individual directories
1033         https://bugs.webkit.org/show_bug.cgi?id=91754
1034
1035         Reviewed by Zoltan Herczeg.
1036
1037         Rollout r123729 because it made Qt debug bots crasy.
1038
1039         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1040         (WebCore::DumpRenderTree::DumpRenderTree):
1041         (WebCore::DumpRenderTree::open):
1042         (WebCore::DumpRenderTree::processLine):
1043         (WebCore::DumpRenderTree::setDumpPixels):
1044         (WebCore::DumpRenderTree::dump):
1045         * DumpRenderTree/qt/DumpRenderTreeQt.h:
1046         (DumpRenderTree):
1047         * DumpRenderTree/qt/main.cpp:
1048         (main):
1049         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
1050         (SingleTestRunner._should_fetch_expected_checksum):
1051         (SingleTestRunner._overwrite_baselines):
1052         (SingleTestRunner._compare_output):
1053         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1054         (Worker._update_test_input):
1055         * Scripts/webkitpy/layout_tests/port/base.py:
1056         (Port.lookup_virtual_test_args):
1057         * Scripts/webkitpy/layout_tests/port/driver.py:
1058         (Driver.cmd_line):
1059         (Driver._command_from_driver_input):
1060         * Scripts/webkitpy/layout_tests/port/qt.py:
1061         (QtPort.check_sys_deps):
1062         * Scripts/webkitpy/layout_tests/port/test.py:
1063         (TestPort.virtual_test_suites):
1064         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1065         (_set_up_derived_options):
1066         (parse_args):
1067         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1068         (MainTest.test_run_singly_actually_runs_tests):
1069         (MainTest.test_missing_and_unexpected_results):
1070         * WebKitTestRunner/TestController.cpp:
1071         (WTR::TestController::TestController):
1072         (WTR::TestController::initialize):
1073         (WTR::TestController::runTest):
1074         * WebKitTestRunner/TestController.h:
1075         (TestController):
1076
1077 2012-07-27  Adam Barth  <abarth@webkit.org>
1078
1079         webkit.review.bot should run clean-review-queue and clean-pending-commit periodically
1080         https://bugs.webkit.org/show_bug.cgi?id=92472
1081
1082         Reviewed by Eric Seidel.
1083
1084         We need to run these command periodically to keep cruft from
1085         accumulating in bugs.webkit.org. Rather than running them manually, we
1086         should just have a bot run them. This patch has the style-queue run the
1087         commands, which admittedly is a bit odd but it doesn't seem worthwhile
1088         to create another bot specifically for this purposes.
1089
1090         * Scripts/webkitpy/tool/commands/queues.py:
1091         (AbstractReviewQueue):
1092         (StyleQueue.begin_work_queue):
1093         (StyleQueue):
1094         (StyleQueue.clean_bugzilla):
1095         * Scripts/webkitpy/tool/commands/queues_unittest.py:
1096         (StyleQueueTest.test_style_queue_with_style_exception):
1097         (test_style_queue_with_watch_list_exception):
1098
1099 2012-07-27  Dominik Röttsches  <dominik.rottsches@intel.com>
1100
1101         [Cairo] Add complex font drawing using HarfbuzzNG
1102         https://bugs.webkit.org/show_bug.cgi?id=91864
1103
1104         Reviewed by Simon Hausmann and Martin Robinson.
1105
1106         Bringing Harfbuzz support to EFL with this patch, so we need HarfBuzz in the pulled in jhbuild dependencies.
1107
1108         * WebKitTestRunner/CMakeLists.txt: Adding additional header include directories.
1109         * efl/jhbuild.modules: Adding source for HarfBuzz release version 0.9.0.
1110
1111 2012-07-27  KwangYong Choi  <ky0.choi@samsung.com>
1112
1113         [WK2][EFL][WTR] TestNetscapePlugin is required.
1114         https://bugs.webkit.org/show_bug.cgi?id=88756
1115
1116         Reviewed by Simon Hausmann.
1117
1118         libTestNetscapePlugin.so should be required for DRT/WTR/EFL.
1119
1120         * CMakeLists.txt:
1121         * DumpRenderTree/TestNetscapePlugin/CMakeList.txt:
1122
1123 2012-07-27  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
1124
1125         [EFL] Memory leaks in EFL DRT
1126         https://bugs.webkit.org/show_bug.cgi?id=92347
1127
1128         Reviewed by Simon Hausmann.
1129
1130         Fix a memory leak in EFL's DRT code by adopting an allocation
1131         of JSString created with JSStringCreateWithUTF8CString.
1132
1133         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1134         (DumpRenderTreeChrome::onWindowObjectCleared):
1135
1136 2012-07-27  Adam Barth  <abarth@webkit.org>
1137
1138         Simplify ExpectedFailures
1139         https://bugs.webkit.org/show_bug.cgi?id=92216
1140
1141         Reviewed by Eric Seidel.
1142
1143         This patch simplifies the ExpectedFailures class we use to remember
1144         which tests are currently failing on the bots. When we wrote this code
1145         originally, we weren't entirely sure how it would work. Now that we
1146         understand it more clearly, we can write the code more clearly.
1147
1148         * Scripts/webkitpy/tool/bot/expectedfailures.py:
1149         (_has_failures):
1150         (_is_trustworthy):
1151         (ExpectedFailures.__init__):
1152         (ExpectedFailures.failures_were_expected):
1153         (ExpectedFailures.unexpected_failures_observed):
1154         (ExpectedFailures.update):
1155         * Scripts/webkitpy/tool/bot/expectedfailures_unittest.py:
1156         (ExpectedFailuresTest._assert_can_trust):
1157         (ExpectedFailuresTest.test_failures_were_expected):
1158         (ExpectedFailuresTest.test_unexpected_failures_observed):
1159         (ExpectedFailuresTest.test_unexpected_failures_observed_when_tree_is_hosed):
1160         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
1161         (PatchAnalysisTask._test):
1162         (PatchAnalysisTask._build_and_test_without_patch):
1163         (PatchAnalysisTask._test_patch):
1164
1165 2012-07-27  Csaba Osztrogonác  <ossy@webkit.org>
1166
1167         [Qt][WK2] REGRESSION(r119127): resetting window.internals settings between tests doesn't work properly
1168         https://bugs.webkit.org/show_bug.cgi?id=88064
1169
1170         Reviewed by Zoltan Herczeg.
1171
1172         Use resetInternalsObject() instead of injectInternalsObject().
1173
1174         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1175         (WTR::InjectedBundlePage::resetAfterTest):
1176
1177 2012-07-26  Ryuan Choi  <ryuan.choi@samsung.com>
1178
1179         [EFL] Remove forwarding headers from MiniBrowser/Efl.
1180         https://bugs.webkit.org/show_bug.cgi?id=91627
1181
1182         Reviewed by Kentaro Hara.
1183
1184         * MiniBrowser/efl/CMakeLists.txt: Remove unnecessary forwarding headers.
1185
1186 2012-07-26  Yoshifumi Inoue  <yosin@chromium.org>
1187
1188         [Forms] Introduce runtime feature flags for input type datetime, datetimelocal, month, time, week
1189         https://bugs.webkit.org/show_bug.cgi?id=92339
1190
1191         Reviewed by Kent Tamura.
1192
1193         This patch enables runtime features for input type datetime, datetime-local,
1194         month, time and week for Chromium DRT.
1195
1196         * DumpRenderTree/chromium/TestShell.cpp:
1197         (TestShell::TestShell): Call enableInputType{DateTime,DateTimeLocal,Month,Time,Week}.
1198
1199 2012-07-26  Jeffrey Pfau  <jpfau@apple.com>
1200
1201         Reloading substitute-data/alternate html string for unreachableURL will add an item to the back-forward-history for each reload
1202         https://bugs.webkit.org/show_bug.cgi?id=84041
1203
1204         Reviewed by Brady Eidson.
1205
1206         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1207         * TestWebKitAPI/Tests/mac/BackForwardList.mm: Added.
1208         (-[BackForwardListTest webView:didFinishLoadForFrame:]):
1209         (-[BackForwardListTest webView:didFailProvisionalLoadWithError:forFrame:]):
1210         (TestWebKitAPI):
1211         (TestWebKitAPI::TEST):
1212
1213 2012-07-26  Sheriff Bot  <webkit.review.bot@gmail.com>
1214
1215         Unreviewed, rolling out r123799.
1216         http://trac.webkit.org/changeset/123799
1217         https://bugs.webkit.org/show_bug.cgi?id=92431
1218
1219         broke the Apple Mac build (Requested by bweinstein on
1220         #webkit).
1221
1222         * DumpRenderTree/chromium/TestWebPlugin.cpp:
1223         * DumpRenderTree/chromium/TestWebPlugin.h:
1224         (TestWebPlugin::acceptsInputEvents):
1225         (TestWebPlugin::handleInputEvent):
1226
1227 2012-07-26  Dirk Pranke  <dpranke@chromium.org>
1228
1229         nrwt: pass --verbose twice on the bots to ensure getting the debug output for now
1230         https://bugs.webkit.org/show_bug.cgi?id=92172
1231
1232         Reviewed by Tony Chang.
1233
1234         Right now new-run-webkit-tests --verbose prints out too much
1235         information (it's like --verbose + debugging). I plan to fix
1236         this in bug 88702, and also clean up and simplify all of the
1237         --print options, but in preparation for these changes I need to
1238         pass --verbose twice to get the same level of output
1239         temporarily (for compatibility).
1240
1241         * Scripts/run-webkit-tests:
1242
1243 2012-07-26  Sadrul Habib Chowdhury  <sadrul@chromium.org>
1244
1245         Propagate gesture events to plugins.
1246         https://bugs.webkit.org/show_bug.cgi?id=92281
1247
1248         Reviewed by Adam Barth.
1249
1250         Update TestWebPlugin to receive events, and output logs for the events.
1251
1252         * DumpRenderTree/chromium/TestWebPlugin.cpp:
1253         (TestWebPlugin::handleInputEvent):
1254         * DumpRenderTree/chromium/TestWebPlugin.h:
1255         (TestWebPlugin::acceptsInputEvents):
1256         (TestWebPlugin):
1257
1258 2012-07-26  Balazs Kelemen  <kbalazs@webkit.org>
1259
1260         [nrwt] get rid of --skip-pixel-test-if-no-baseline
1261         https://bugs.webkit.org/show_bug.cgi?id=92377
1262
1263         Reviewed by Adam Barth.
1264
1265         Remove the option and all code paths that handled it.
1266
1267         * Scripts/webkitpy/layout_tests/port/base.py:
1268         (Port.should_run_as_pixel_test):
1269         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1270         (_set_up_derived_options):
1271         (parse_args):
1272         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1273         (MainTest.test_repeat_each):
1274
1275 2012-07-26  Olivier Blin  <olivier.blin@softathome.com>
1276
1277         Add FastMalloc statistics in window.internals
1278         https://bugs.webkit.org/show_bug.cgi?id=91274
1279
1280         Reviewed by Ryosuke Niwa.
1281
1282         * GNUmakefile.am:
1283
1284 2012-07-26  Olivier Blin  <olivier.blin@softathome.com>
1285
1286         Add FastMalloc statistics in window.internals
1287         https://bugs.webkit.org/show_bug.cgi?id=91274
1288
1289         Reviewed by Ryosuke Niwa.
1290
1291         * GNUmakefile.am:
1292
1293 2012-07-20  Holger Hans Peter Freyther  <holger@moiji-mobile.com>
1294
1295         buildbot: Remove the configuration of hfreyther*
1296         https://bugs.webkit.org/show_bug.cgi?id=91849
1297
1298         Reviewed by Adam Roben.
1299
1300         The machine running these configs was decommissioned and
1301         there is no replacement at this point. Remove them from the
1302         configuration.
1303
1304         * BuildSlaveSupport/build.webkit.org-config/config.json: Remove hfreyther* from
1305         the config. The Qt Linux SH4 and Qt Linux MIPS schedulers are orphaned, remove
1306         them too.
1307
1308 2012-07-26  Balazs Kelemen  <kbalazs@webkit.org>
1309
1310         Unreviewed speculative Mac buildfix after r123729 - 2nd attempt.
1311
1312         * WebKitTestRunner/TestController.cpp:
1313         (WTR::TestController::runTest):
1314
1315 2012-07-26  Balazs Kelemen  <kbalazs@webkit.org>
1316
1317         Unreviewed speculative buildfix after r123729.
1318
1319         * WebKitTestRunner/TestController.cpp:
1320         (WTR::TestController::runTest): Fix sign errors.
1321
1322 2012-07-26  Zoltan Nyul  <zoltan.nyul@intel.com>
1323
1324         [EFL] EFL port should use XDG paths
1325         https://bugs.webkit.org/show_bug.cgi?id=91719
1326
1327         Reviewed by Kenneth Rohde Christiansen.
1328
1329         The appcache/localstorage/webdatabase path should be set in the
1330         ewk_main.cpp instead of the DumpRenderTree, and the XDG_DATA_HOME and
1331         XDG_CACHE_HOME should be set to different directory for each processes
1332         for the layout-tests.
1333
1334         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1335         (DumpRenderTreeChrome::initialize):
1336         * Scripts/webkitpy/layout_tests/port/efl.py:
1337         (EflPort.setup_environ_for_server):
1338         * efl/jhbuild.modules:
1339
1340 2012-07-26  Balazs Kelemen  <kbalazs@webkit.org>
1341
1342         [NRWT] should have a way to restrict pixel tests for individual directories
1343         https://bugs.webkit.org/show_bug.cgi?id=91754
1344
1345         Reviewed by Dirk Pranke.
1346
1347         Added --pixel-test-directories option to nrwt. It implies --pixel-tests.
1348         If the platform supports this option only tests from these directories will
1349         be executed as pixel tests. Furthermore, platforms can have a default set of
1350         pixel test directories. To support the feature the test harness have to be
1351         able to inform the driver whether to dump pixels for each tests. For this
1352         purpose I changed WTR and the Qt DRT to handle a --pixel-test argument from
1353         the standard output and made NRWT use this argument with these drivers.
1354
1355         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1356         (WebCore::DumpRenderTree::DumpRenderTree):
1357         (WebCore::DumpRenderTree::open):
1358         (WebCore::DumpRenderTree::processLine):
1359         (WebCore::DumpRenderTree::setDumpPixelsForAllTests):
1360         (WebCore::DumpRenderTree::dump):
1361         * DumpRenderTree/qt/DumpRenderTreeQt.h:
1362         (DumpRenderTree):
1363         * DumpRenderTree/qt/main.cpp:
1364         (main):
1365         Handle the --pixel-test argument.
1366
1367         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
1368         (SingleTestRunner._should_fetch_expected_checksum):
1369         (SingleTestRunner._overwrite_baselines):
1370         (SingleTestRunner._compare_output):
1371         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1372         (Worker._update_test_input):
1373         * Scripts/webkitpy/layout_tests/port/base.py:
1374         (Port.supports_switching_pixel_tests_per_test): Ports can override this to inform
1375         the harness that --pixel-test argument is supported. True for WTR, otherwise calls
1376         private version that can be overridden by the actual platform. Only true with WTR
1377         and Qt DRT currently. This supposed to exist only temporary until there are ports
1378         that doesn't support the feature in their DRT's.
1379         (Port):
1380         (Port._supports_switching_pixel_tests_per_test):
1381         (Port.should_run_as_pixel_test):
1382         (Port._should_run_as_pixel_test): Ports can override this to define their
1383         default set of pixel tests directories.
1384         * Scripts/webkitpy/layout_tests/port/driver.py:
1385         (Driver.cmd_line): Don't pass --pixel-tests if the platform supports per test
1386         --pixel-test.
1387         (Driver._command_from_driver_input): Pass --pixel-test if it is supported and pixel
1388         tests are enabled.
1389         * Scripts/webkitpy/layout_tests/port/qt.py:
1390         (QtPort.check_sys_deps):
1391         (QtPort):
1392         (QtPort._supports_switching_pixel_tests_per_test):
1393         (QtPort._should_run_as_pixel_test):
1394         (QtPort._default_pixel_test_directories): Only compositing will be our default pixel
1395         test directory for the time being.
1396         * Scripts/webkitpy/layout_tests/port/test.py:
1397         (TestPort.supports_switching_pixel_tests_per_test):
1398         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1399         (_set_up_derived_options):
1400         (parse_args):
1401         Handle the --pixel-test-directory option. It expect a directory below LayoutTests
1402         and can be specified multiple times. Added fixme's that we should improve how we
1403         handle these arguments.
1404
1405         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1406         (MainTest.test_run_singly_actually_runs_tests):
1407         (MainTest.test_pixel_test_directories):
1408         Added an integration test for the feature.
1409
1410         * WebKitTestRunner/TestController.cpp:
1411         (WTR::TestController::TestController):
1412         (WTR::TestController::initialize):
1413         (WTR::TestController::runTest):
1414         * WebKitTestRunner/TestController.h:
1415         (TestController):
1416         Handle the --pixel-test argument.
1417
1418 2012-07-26  Christophe Dumez  <christophe.dumez@intel.com>
1419
1420         [EFL] Regression(r123604): webintents/web-intents-obj-constructor.html is crashing
1421         https://bugs.webkit.org/show_bug.cgi?id=92349
1422
1423         Reviewed by Kentaro Hara.
1424
1425         Fix crash occurring after r123604, which replaced strdup()
1426         usage with eina_stringshare_add(). Unfortunately, the DRT
1427         was not updated accordingly to use eina_stringshare_del()
1428         instead of free() on the values returned by Ewk_Intent
1429         getters and it was causing
1430         webintents/web-intents-obj-constructor.html to crash.
1431
1432         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1433         (DumpRenderTreeChrome::onFrameIntentNew):
1434
1435 2012-07-26  Adam Barth  <abarth@webkit.org>
1436
1437         Unreviewed attempt to fix the chromium-win build.
1438
1439         * DumpRenderTree/chromium/TestRunner/TestRunner.cpp:
1440
1441 2012-07-26  Kaustubh Atrawalkar  <kaustubh@motorola.com>
1442
1443         [DRT] LTC:: pageNumberForElementById() could be moved to Internals
1444         https://bugs.webkit.org/show_bug.cgi?id=92091
1445
1446         Reviewed by Adam Barth.
1447
1448         Move the pageNumberForElementById from LayoutTestCotroller to Internals and
1449         remove the old platform specific implementations as it exclusively tests WebCore functionality.
1450
1451         * DumpRenderTree/LayoutTestController.cpp:
1452         (LayoutTestController::staticFunctions):
1453         * DumpRenderTree/LayoutTestController.h:
1454         (LayoutTestController):
1455         * DumpRenderTree/blackberry/LayoutTestControllerBlackBerry.cpp:
1456         * DumpRenderTree/chromium/LayoutTestController.cpp:
1457         (LayoutTestController::LayoutTestController):
1458         (LayoutTestController::setPrinting):
1459         * DumpRenderTree/chromium/LayoutTestController.h:
1460         (LayoutTestController):
1461         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1462         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1463         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
1464         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1465         * DumpRenderTree/qt/LayoutTestControllerQt.h:
1466         (LayoutTestController):
1467         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1468         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
1469         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
1470         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1471         (WTR):
1472         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1473         (LayoutTestController):
1474
1475 2012-07-25  Adam Barth  <abarth@webkit.org>
1476
1477         TestRunner.a should contain a TestRunner object
1478         https://bugs.webkit.org/show_bug.cgi?id=92313
1479
1480         Reviewed by Kent Tamura.
1481
1482         This patch adds a stub implementation of TestRunner, which is the new
1483         home for our implementation of window.testRunner. A future patch will
1484         migrate functions from LayoutTestController to TestRunner.
1485
1486         * DumpRenderTree/DumpRenderTree.gypi:
1487         * DumpRenderTree/chromium/LayoutTestController.h:
1488         * DumpRenderTree/chromium/TestRunner/TestRunner.cpp: Added.
1489         * DumpRenderTree/chromium/TestRunner/TestRunner.h: Added.
1490
1491 2012-07-25  Wei James  <james.wei@intel.com>
1492
1493         [Chromium] Fix infinite loop issue for chromium android layout test
1494         https://bugs.webkit.org/show_bug.cgi?id=92197
1495
1496         Reviewed by Adam Barth.
1497
1498         The change brought in by r123530 will cause infinite loop when running
1499         Chromium Android Layout Test.
1500         Revert the change for this line only.
1501
1502         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1503         (ChromiumAndroidPort._build_path):
1504
1505 2012-07-25  Adam Barth  <abarth@webkit.org>
1506
1507         Jochen is our a newest reviewer!
1508
1509         * Scripts/webkitpy/common/config/committers.py:
1510
1511 2012-07-25  Jochen Eisinger  <jochen@chromium.org>
1512
1513         [chromium] don't use webkit_support methods directly from test runners but through the test delegate
1514         https://bugs.webkit.org/show_bug.cgi?id=92256
1515
1516         Reviewed by Adam Barth.
1517
1518         This will allow for the content_shell to provide its own implementation
1519         for setting the Gamepad data instead.
1520
1521         * DumpRenderTree/chromium/TestRunner/GamepadController.cpp:
1522         (GamepadController::setDelegate):
1523         (GamepadController::reset):
1524         (GamepadController::connect):
1525         (GamepadController::disconnect):
1526         (GamepadController::setId):
1527         (GamepadController::setButtonCount):
1528         (GamepadController::setButtonData):
1529         (GamepadController::setAxisCount):
1530         (GamepadController::setAxisData):
1531         * DumpRenderTree/chromium/TestRunner/GamepadController.h:
1532         (GamepadController):
1533         * DumpRenderTree/chromium/TestRunner/TestDelegate.h:
1534         (WebKit):
1535         (TestDelegate):
1536         * DumpRenderTree/chromium/TestRunner/TestInterfaces.cpp:
1537         (TestInterfaces::Internal):
1538         (TestInterfaces::Internal::~Internal):
1539         (TestInterfaces::Internal::setDelegate):
1540         (TestInterfaces::setDelegate):
1541         * DumpRenderTree/chromium/TestRunner/TestInterfaces.h:
1542         (TestInterfaces):
1543         * DumpRenderTree/chromium/TestShell.cpp:
1544         (TestShell::createMainWindow):
1545         (TestShell::~TestShell):
1546         * DumpRenderTree/chromium/WebViewHost.cpp:
1547         (WebViewHost::setGamepadData):
1548         * DumpRenderTree/chromium/WebViewHost.h:
1549         (WebViewHost):
1550
1551 2012-07-25  Christophe Dumez  <christophe.dumez@intel.com>
1552
1553         [WK2][WTR] LayoutTestController.sendWebIntentResponse() needs to be implemented
1554         https://bugs.webkit.org/show_bug.cgi?id=92227
1555
1556         Reviewed by Kenneth Rohde Christiansen.
1557
1558         Implement LayoutTestController.sendWebIntentResponse() so
1559         that a test case can reply to a Web intent request and
1560         test this functionality.
1561
1562         In order to support this functionality, we need to keep
1563         a pointer to the current intent request in
1564         InjectedBundlePage::didReceiveIntentForFrame() so that
1565         we can reuse it later in
1566         LayoutTestController::sendWebIntentResponse().
1567
1568         The following tests need this functionality:
1569         - webintents/web-intents-failure.html
1570         - webintents/web-intents-reply.html
1571
1572         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
1573         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1574         (WTR::InjectedBundlePage::didReceiveIntentForFrame):
1575         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
1576         (InjectedBundlePage):
1577         (WTR::InjectedBundlePage::currentIntentRequest):
1578         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1579         (WTR::LayoutTestController::sendWebIntentResponse):
1580         (WTR):
1581         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1582         (LayoutTestController):
1583
1584 2012-07-25  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1585
1586         [EFL][WK2][WTR] Alt, Shift and Ctrl keys are not handled by EventSenderProxy.
1587         https://bugs.webkit.org/show_bug.cgi?id=92241
1588
1589         Reviewed by Kenneth Rohde Christiansen.
1590
1591         Alt, Shift and Ctrl keys were not handled by EventSenderProxy.
1592         It caused failure of fast/events/keydown-leftright-keys.html test.
1593
1594         * WebKitTestRunner/efl/EventSenderProxyEfl.cpp: Added handling of Alt, Shift and Ctrl keys.
1595         (WTR::keyName):
1596
1597 2012-07-25  Christophe Dumez  <christophe.dumez@intel.com>
1598
1599         [EFL][WK2] Fix timeout issues when debugging WebProcess.
1600         https://bugs.webkit.org/show_bug.cgi?id=92245
1601
1602         Reviewed by Dirk Pranke.
1603
1604         Fix timeout issues when debugging the WebProcess
1605         under valgrind or gdb.
1606
1607         * Scripts/webkitpy/layout_tests/port/efl.py:
1608         (EflPort.default_timeout_ms):
1609         * WebKitTestRunner/efl/TestControllerEfl.cpp:
1610         (WTR::TestController::platformInitialize):
1611
1612 2012-07-25  Xianzhu Wang  <wangxianzhu@chromium.org>
1613
1614         [Chromium-Android] NRWT driver should not restart between reftests
1615         https://bugs.webkit.org/show_bug.cgi?id=92267
1616
1617         Reviewed by Dirk Pranke.
1618
1619         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1620         (ChromiumAndroidDriver.start): Resets self._pixel_tests when the pixel_tests parameter changes.
1621
1622 2012-07-25  Tommy Widenflycht  <tommyw@google.com>
1623
1624         MediaStream API: Update MediaStreamTrack to match the specification
1625         https://bugs.webkit.org/show_bug.cgi?id=90180
1626
1627         Reviewed by Adam Barth.
1628
1629         Adding a Mock WebMediaStreamCenter to enable better LayoutTests.
1630
1631         * DumpRenderTree/DumpRenderTree.gypi:
1632         * DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp:
1633         (MockWebKitPlatformSupport::createMediaStreamCenter):
1634         * DumpRenderTree/chromium/MockWebMediaStreamCenter.cpp: Added.
1635         (WebKit):
1636         (WebKit::MockWebMediaStreamCenter::MockWebMediaStreamCenter):
1637         (WebKit::MockWebMediaStreamCenter::queryMediaStreamSources):
1638         (WebKit::MockWebMediaStreamCenter::didEnableMediaStreamTrack):
1639         (WebKit::MockWebMediaStreamCenter::didDisableMediaStreamTrack):
1640         (WebKit::MockWebMediaStreamCenter::didStopLocalMediaStream):
1641         (WebKit::MockWebMediaStreamCenter::didCreateMediaStream):
1642         (WebKit::MockWebMediaStreamCenter::constructSDP):
1643         * DumpRenderTree/chromium/MockWebMediaStreamCenter.h: Copied from Tools/DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp.
1644         (WebKit):
1645         (MockWebMediaStreamCenter):
1646         (WebKit::MockWebMediaStreamCenter::MockWebMediaStreamCenter):
1647
1648 2012-07-25  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
1649
1650         [NRWT] Rename baseline_search_path() to default_baseline_search_path()
1651         https://bugs.webkit.org/show_bug.cgi?id=92122
1652
1653         Reviewed by Dirk Pranke.
1654
1655         * Scripts/webkitpy/layout_tests/port/base.py:
1656         (Port.baseline_version_dir):
1657         (Port.baseline_search_path):
1658         (Port):
1659         (Port.default_baseline_search_path):
1660         (Port.expected_baselines):
1661         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1662         (PortTest.test_additional_platform_directory):
1663         * Scripts/webkitpy/layout_tests/port/mac.py:
1664         (MacPort.default_baseline_search_path):
1665         * Scripts/webkitpy/layout_tests/port/qt.py:
1666         (QtPort.default_baseline_search_path):
1667         * Scripts/webkitpy/layout_tests/port/win.py:
1668         (WinPort.default_baseline_search_path):
1669
1670 2012-07-25  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1671
1672         [EFL] Use eina_stringshare_add instead of strdup.
1673         https://bugs.webkit.org/show_bug.cgi?id=92072
1674
1675         Reviewed by Kentaro Hara.
1676
1677         Eina of EFL libraries supports a string functionality that replaces strdup. So, EFL port needs
1678         to replace strdup with eina_stringshare_add function.
1679
1680         DumpRenderTree also needs to be modified according to ewk APIs's parameter change.
1681
1682         * DumpRenderTree/efl/DumpRenderTree.cpp:
1683         (dumpFramesAsText):
1684         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
1685         (DumpRenderTreeChrome::onFrameIntentNew):
1686         * DumpRenderTree/efl/DumpRenderTreeView.cpp:
1687         (onJavaScriptPrompt):
1688
1689 2012-07-25  Alexander Shalamov  <alexander.shalamov@intel.com>
1690
1691         [WTR][EFL] Wheel events are not emitted
1692         https://bugs.webkit.org/show_bug.cgi?id=92094
1693
1694         Reviewed by Csaba Osztrogonác.
1695
1696         Wheel events are not emitted, therefore, layout tests that use mouseScrollBy method are failing.
1697
1698         * WebKitTestRunner/efl/EventSenderProxyEfl.cpp:
1699         (WTR::dispatchMouseScrollByEvent):
1700         (WTR::EventSenderProxy::mouseScrollBy):
1701         * WebKitTestRunner/efl/PlatformWebViewEfl.cpp:
1702         (WTR::PlatformWebView::PlatformWebView):
1703         (WTR::PlatformWebView::focus):
1704
1705 2012-07-25  Mikhail Pozdnyakov  <mikhail.pozdnyakov@intel.com>
1706
1707         [EFL][WK2] Add ewk_main.{cpp,h} to EFL WK2
1708         https://bugs.webkit.org/show_bug.cgi?id=92101
1709
1710         Reviewed by Kenneth Rohde Christiansen.
1711
1712         Added a centralized place for general initialization in UI process for EFL WK2.
1713
1714         * MiniBrowser/efl/main.c: Added using of newly added ewk_init() and ewk_shutdown() functions.
1715         (main):
1716         * WebKitTestRunner/efl/main.cpp: Ditto.
1717         (main):
1718
1719 2012-07-24  Benjamin Poulain  <bpoulain@apple.com> && Joseph Pecoraro  <pecoraro@apple.com>
1720
1721         QualifiedName's HashSet should be big enough to hold at least all the static names
1722         https://bugs.webkit.org/show_bug.cgi?id=91891
1723
1724         Reviewed by Darin Adler.
1725
1726         Add a test for WTF::hashTableCapacityForSize.
1727
1728         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1729         * TestWebKitAPI/Tests/WTF/HashSet.cpp: Added.
1730         (InitialCapacityTestHashTraits):
1731         (TestWebKitAPI::testInitialCapacity):
1732         (TestWebKitAPI::generateTestCapacityUpToSize):
1733         (TestWebKitAPI::TEST):
1734
1735 2012-07-24  Adam Barth  <abarth@webkit.org>
1736
1737         The EWS bots get flaky when we hit the failure limit
1738         https://bugs.webkit.org/show_bug.cgi?id=92212
1739
1740         Reviewed by Eric Seidel.
1741
1742         Previously we were missing some failure types, which means we didn't
1743         notice that the bots were hitting the failure limit.
1744
1745         * Scripts/webkitpy/common/net/layouttestresults.py:
1746         (LayoutTestResults.failing_test_results):
1747         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
1748         (LayoutTestResultsTest.test_results_from_string):
1749         (LayoutTestResultsTest.test_tests_matching_failure_types):
1750
1751 2012-07-24  Raphael Kubo da Costa  <rakuco@webkit.org>
1752
1753         [EFL] Unreviewed build fix.
1754
1755         * WebKitTestRunner/efl/EventSenderProxyEfl.cpp: Include unistd.h
1756         for usleep(3).
1757
1758 2012-07-24  Dirk Pranke  <dpranke@chromium.org>
1759
1760         make test-webkitpy at least kinda work in parallel on win32
1761         https://bugs.webkit.org/show_bug.cgi?id=91700
1762
1763         Reviewed by Adam Barth.
1764
1765         This change makes the hoop-jumping we have to do to make running
1766         new-run-webkit-tests in parallel on win32 generic so that we can
1767         re-use it for test-webkitpy as well.
1768
1769         * Scripts/new-run-webkit-tests:
1770         * Scripts/test-webkitpy:
1771         * Scripts/webkitpy/common/multiprocessing_bootstrap.py: Copied from Tools/Scripts/new-run-webkit-tests.
1772         (run):
1773         * Scripts/webkitpy/test/main.py:
1774         (main):
1775         (Tester._log_exception):
1776
1777 2012-07-24  Sheriff Bot  <webkit.review.bot@gmail.com>
1778
1779         Unreviewed, rolling out r123499.
1780         http://trac.webkit.org/changeset/123499
1781         https://bugs.webkit.org/show_bug.cgi?id=92161
1782
1783         Android fails to compile (Requested by zhenyao on #webkit).
1784
1785         * DumpRenderTree/DumpRenderTree.gypi:
1786         * DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp:
1787         (MockWebKitPlatformSupport::createMediaStreamCenter):
1788         * DumpRenderTree/chromium/MockWebKitPlatformSupport.h:
1789         (MockWebKitPlatformSupport):
1790         * DumpRenderTree/chromium/MockWebMediaStreamCenter.cpp: Removed.
1791         * DumpRenderTree/chromium/MockWebMediaStreamCenter.h: Removed.
1792
1793 2012-07-24  Dirk Pranke  <dpranke@chromium.org>
1794
1795         nrwt: never finds binaries in the 'out' dir on chromium win
1796         https://bugs.webkit.org/show_bug.cgi?id=91890
1797
1798         Reviewed by Tony Chang.
1799
1800         Attempt to fix this bug yet again; it turns out that --root and
1801         --build-directory don't have the same meaning (--root should
1802         include the configuration, --build-directory should not).
1803
1804         * Scripts/webkitpy/layout_tests/port/base.py:
1805         (Port._build_path):
1806         * Scripts/webkitpy/layout_tests/port/chromium.py:
1807         (ChromiumPort._static_build_path):
1808         (ChromiumPort.default_results_directory):
1809         (ChromiumPort._build_path):
1810         (ChromiumPort):
1811         (ChromiumPort._build_path_with_configuration):
1812         (ChromiumPort._build_path_with_configuration.does):
1813         (ChromiumPort._path_to_image_diff):
1814         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1815         (ChromiumAndroidPort._build_path):
1816         (ChromiumAndroidPort):
1817         (ChromiumAndroidPort._build_path_with_configuration):
1818         (ChromiumAndroidPort._path_to_driver):
1819         (ChromiumAndroidPort._path_to_forwarder):
1820         (ChromiumAndroidPort._push_executable):
1821         (ChromiumAndroidPort._push_fonts):
1822         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1823         (ChromiumLinuxPort._determine_driver_path_statically):
1824         (ChromiumLinuxPort._modules_to_search_for_symbols):
1825         (ChromiumLinuxPort._path_to_driver):
1826         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
1827         (ChromiumLinuxPortTest.test_build_path):
1828         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1829         (ChromiumMacPort._modules_to_search_for_symbols):
1830         (ChromiumMacPort._path_to_driver):
1831         (ChromiumMacPort._path_to_helper):
1832         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1833         (ChromiumMacPortTest.test_build_path):
1834         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1835         (ChromiumWinPort._path_to_driver):
1836         (ChromiumWinPort._path_to_helper):
1837         (ChromiumWinPort._path_to_image_diff):
1838         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1839         (ChromiumWinTest.test_build_path):
1840         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1841         (parse_args):
1842
1843 2012-07-24  Peter Beverloo  <peter@chromium.org>
1844
1845         Android-chromium layout tests failing with GPU related crash
1846         https://bugs.webkit.org/show_bug.cgi?id=92126
1847
1848         Reviewed by Adam Barth.
1849
1850         Remove support for the "--enable-hardware-gpu" flag from run_webkit_tests,
1851         as it's not being used by any port anymore, and neither is being passed
1852         on to the test runner. Instead, append the flag directly to DRT for the
1853         Chromium Android port, which does rely on this functionality.
1854
1855         The ChromiumAndroidUnitTest tests didn't catch this as the property was
1856         set to True, but it's unable to detect whether they'll actually be used.
1857         No occurrances of "--enable-hardware-gpu" exist in either Tools/Scripts/
1858         or Tools/BuildSlaveSupport/, so nothing upstream seems to rely on the flag.
1859
1860         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
1861         (ChromiumAndroidPort.__init__):
1862         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
1863         (ChromiumAndroidPortTest.test_attributes):
1864         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1865         (parse_args):
1866
1867 2012-07-24  Alexander Shalamov  <alexander.shalamov@intel.com>
1868
1869         [WTR][EFL] Multiple mouse clicks are not emitted in EventSenderProxyEfl
1870         https://bugs.webkit.org/show_bug.cgi?id=92107
1871
1872         Reviewed by Kenneth Rohde Christiansen.
1873
1874         Fixes mouse click events in EventSenderProxyEfl.
1875
1876         * WebKitTestRunner/efl/EventSenderProxyEfl.cpp:
1877         (WTR::dispatchMouseDownEvent):
1878         (WTR::EventSenderProxy::dispatchEvent):
1879
1880 2012-07-24  Tommy Widenflycht  <tommyw@google.com>
1881
1882         MediaStream API: Update MediaStreamTrack to match the specification
1883         https://bugs.webkit.org/show_bug.cgi?id=90180
1884
1885         Reviewed by Adam Barth.
1886
1887         Adding a Mock WebMediaStreamCenter to enable better LayoutTests.
1888
1889         * DumpRenderTree/DumpRenderTree.gypi:
1890         * DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp:
1891         (MockWebKitPlatformSupport::createMediaStreamCenter):
1892         * DumpRenderTree/chromium/MockWebMediaStreamCenter.cpp: Added.
1893         (WebKit):
1894         (WebKit::MockWebMediaStreamCenter::MockWebMediaStreamCenter):
1895         (WebKit::MockWebMediaStreamCenter::queryMediaStreamSources):
1896         (WebKit::MockWebMediaStreamCenter::didEnableMediaStreamTrack):
1897         (WebKit::MockWebMediaStreamCenter::didDisableMediaStreamTrack):
1898         (WebKit::MockWebMediaStreamCenter::didStopLocalMediaStream):
1899         (WebKit::MockWebMediaStreamCenter::didCreateMediaStream):
1900         (WebKit::MockWebMediaStreamCenter::constructSDP):
1901         * DumpRenderTree/chromium/MockWebMediaStreamCenter.h: Copied from Tools/DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp.
1902         (WebKit):
1903         (MockWebMediaStreamCenter):
1904         (WebKit::MockWebMediaStreamCenter::MockWebMediaStreamCenter):
1905
1906 2012-07-23  Ryosuke Niwa  <rniwa@webkit.org>
1907
1908         Fix regressions in replay performance tests
1909         https://bugs.webkit.org/show_bug.cgi?id=92052
1910
1911         Reviewed by Dirk Pranke.
1912
1913         Fix regression that accumulated over time to make replay tests work again.
1914
1915         * DumpRenderTree/chromium/WebViewHost.cpp:
1916         (WebViewHost::willSendRequest): Use the main document URL in the request since the main frame's
1917         URL hadn't been updated yet When requesting a new URL on the main frame.
1918         * Scripts/webkitpy/performance_tests/perftest.py:
1919         (ReplayServer.wait_until_ready): Waiting for 6 seconds (3 tries) seems enough.
1920         (ReplayPerfTest.prepare): Pass self._archive_path to run_signle as that is the current contract.
1921         (ReplayPerfTest.run_single): Use the archive path to generate -expected.png path.
1922         * Scripts/webkitpy/performance_tests/perftest_unittest.py:
1923         (TestReplayPerfTest.test_run_single):
1924         (TestReplayPerfTest.test_prepare):
1925         (TestReplayPerfTest.test_prepare_calls_run_single.run_single):
1926         (TestReplayPerfTest):
1927
1928 2012-07-24  Simon Hausmann  <simon.hausmann@nokia.com>
1929
1930         [Qt] MSVC: unresolved external symbol __DllMainCRTStartup@12
1931         https://bugs.webkit.org/show_bug.cgi?id=91229
1932
1933         Reviewed by Jocelyn Turcotte.
1934
1935         In order to successfully link a DLL on Windows we need to have at least
1936         one object file (or compilation unit). The forward export header files were
1937         supposed to be that, but unfortunately the rule in win32/default_post.prf for
1938         creating the header files had some bugs, among others that it did an exists()
1939         check on the depending static library. At the time qmake is ran those libraries
1940         do not exist yet and therefore the corresponding extra compiler rules were never
1941         created, resulting in empty OBJECTS/SOURCES.
1942
1943         Even without such an exists() check, qmake extra compilers require the files
1944         referred to in the .input variable to exist at qmake time. In this case the input
1945         files were the static libraries, which do not exist yet.
1946
1947         This patch solves this by using a qmake extra target instead of extra
1948         compiler, which does not have this limitation. The target is referenced
1949         through the extension of GENERATED_SOURCES.
1950
1951         The patch also adds a d/_debug suffix for debug builds, do allow for separate
1952         symbol exports if necessary.
1953
1954         * Scripts/generate-win32-export-forwards: Support multiple input files, i.e.
1955         consider the last argument to be the output file and everything else input.
1956         * qmake/mkspecs/features/win32/default_post.prf:
1957
1958 2012-07-24  Balazs Kelemen  <kbalazs@webkit.org>
1959
1960         [Qt] Add quirks for running the web process in a profiler shell, like valgrind
1961         https://bugs.webkit.org/show_bug.cgi?id=87672
1962
1963         Reviewed by Jocelyn Turcotte.
1964
1965         Add --no-timeout-at-all command line switch to not time out no matter how
1966         long the web process does it's job. --no-timeout is not enough for profiling
1967         purposes because it only affects the timeout when we are waiting for the end
1968         of the test. We use another timeout to control how much we wait for the web
1969         process to initialize before and reset after the test and it could also easily
1970         expire when running under valgrind. We should keep the behavior of --no-timeout
1971         to not disable this timout check because --no-timeout is used by performace tests
1972         and those should not wait for the web process to initialize and reset more than
1973         normal tests.
1974
1975         * WebKitTestRunner/TestController.cpp:
1976         (WTR::TestController::TestController):
1977         (WTR::TestController::initialize):
1978         (WTR::TestController::runUntil):
1979         * WebKitTestRunner/TestController.h:
1980
1981 2012-07-24  Peter Beverloo  <peter@chromium.org>
1982
1983         [Chromium] Fix a class/struct mismatch error in TestDelegate.h (showing up on clang)
1984         https://bugs.webkit.org/show_bug.cgi?id=92097
1985
1986         Unreviewed build fix.
1987
1988         WebContextMenuData is declared as a struct in Source/WebKit/chromium/public/WebContextMenuData.h,
1989         but TestDelegate.h forward declares it as a class.
1990
1991         * DumpRenderTree/chromium/TestRunner/TestDelegate.h:
1992         (WebKit):
1993
1994 2012-07-23  Oswald Buddenhagen  <oswald.buddenhagen@nokia.com>
1995
1996         [Qt] Fix compilation against namespaced Qt
1997
1998         Reviewed by Simon Hausmann.
1999
2000         * QtTestBrowser/launcherwindow.h:
2001         * WebKitTestRunner/PlatformWebView.h:
2002
2003 2012-07-24  Peter Beverloo  <peter@chromium.org>
2004
2005         Update a few Chromium Android contributors in committers.py
2006         https://bugs.webkit.org/show_bug.cgi?id=92090
2007
2008         Reviewed by Kent Tamura.
2009
2010         Add Sami Kyostila, who has been doing significant work on Chromium's
2011         compositor. Add IRC nicknames for John Mellor and Tien-Ren Chen, and
2012         remove John Grabowski who changed projects, and add my webkit.org alias.
2013
2014         * Scripts/webkitpy/common/config/committers.py:
2015
2016 2012-07-24  Pierre Rossi  <pierre.rossi@gmail.com>
2017
2018         [Qt] Enable touch slider when touch events are enabled
2019         https://bugs.webkit.org/show_bug.cgi?id=91013
2020
2021         Reviewed by Kenneth Rohde Christiansen.
2022
2023         TOUCH_SLIDER guards are used in WebCore, so we should really
2024         be setting the corresponding ENABLE define in features.prf.
2025
2026         * qmake/mkspecs/features/features.prf:
2027
2028 2012-07-24  Ryuan Choi  <ryuan.choi@samsung.com>
2029
2030         [EFL] EWebLauncher should not require config.h in the style check
2031         https://bugs.webkit.org/show_bug.cgi?id=92070
2032
2033         Reviewed by Kentaro Hara.
2034
2035         * Scripts/webkitpy/style/checker.py:
2036         Remove the include rules from the style-check for EWebLauncher because
2037         EWebLauncher is simple EFL application.
2038
2039 2012-07-24  Jochen Eisinger  <jochen@chromium.or>
2040
2041         [Chromium] EventSender shouldn't know about TestShell
2042         https://bugs.webkit.org/show_bug.cgi?id=91449
2043
2044         Reviewed by Adam Barth.
2045
2046         Original patch by Adam Barth.
2047
2048         This patch removes the dependency from EventSender to TestShell.  Most
2049         of EventSender works directly with a WebView, but there are a handful
2050         of references to WebViewHost.  We don't want to re-use WebViewHost in
2051         ContentShell because we want to use the multi-process embedding
2052         environment for WebViews.
2053
2054         To avoid the dependency on WebViewHost, this patch introduces
2055         TestDelegate, which is an abstract base class for WebViewHost.  In
2056         ContentShell, we'll need another object to implement these functions.
2057
2058         * DumpRenderTree/chromium/EventSender.cpp:
2059         (EventSender::EventSender):
2060         (EventSender::keyDown):
2061         (makeMenuItemStringsFor):
2062         (EventSender::contextClick):
2063         * DumpRenderTree/chromium/EventSender.h:
2064         (EventSender):
2065         (EventSender::setDelegate):
2066         (EventSender::setWebView):
2067         (EventSender::webview):
2068         * DumpRenderTree/chromium/TestRunner/TestDelegate.h: Added.
2069         (WebKit):
2070         (TestDelegate):
2071         * DumpRenderTree/chromium/TestShell.cpp:
2072         (TestShell::initialize):
2073         (TestShell::createMainWindow):
2074         (TestShell::~TestShell):
2075         * DumpRenderTree/chromium/WebViewHost.cpp:
2076         (WebViewHost::fillSpellingSuggestionList):
2077         * DumpRenderTree/chromium/WebViewHost.h:
2078         (WebViewHost):
2079
2080 2012-07-24  Thiago Marcos P. Santos  <thiago.santos@intel.com>
2081
2082         Make WebIntents and WebIntentsTag a build option for the build-webkit script
2083         https://bugs.webkit.org/show_bug.cgi?id=91841
2084
2085         Reviewed by Philippe Normand.
2086
2087         And it will also be disabled when building with --minimal.
2088
2089         * Scripts/webkitperl/FeatureList.pm:
2090
2091 2012-07-23  Dirk Pranke  <dpranke@chromium.org>
2092
2093         Unreviewed, rolling out r123360.
2094         http://trac.webkit.org/changeset/123360
2095         https://bugs.webkit.org/show_bug.cgi?id=91890
2096
2097         broke the chromium bots
2098
2099         * Scripts/webkitpy/layout_tests/port/chromium.py:
2100         (ChromiumPort._static_build_path):
2101         (ChromiumPort.default_results_directory):
2102         (ChromiumPort._build_path):
2103         (ChromiumPort._path_to_image_diff):
2104         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2105         (ChromiumAndroidPort._build_path):
2106         (ChromiumAndroidPort._path_to_driver):
2107         (ChromiumAndroidPort._path_to_forwarder):
2108         (ChromiumAndroidPort._push_executable):
2109         (ChromiumAndroidPort._push_fonts):
2110         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2111         (ChromiumLinuxPort._determine_driver_path_statically):
2112         (ChromiumLinuxPort._modules_to_search_for_symbols):
2113         (ChromiumLinuxPort._path_to_driver):
2114         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
2115         (ChromiumLinuxPortTest.test_build_path):
2116         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2117         (ChromiumMacPort._modules_to_search_for_symbols):
2118         (ChromiumMacPort._path_to_driver):
2119         (ChromiumMacPort._path_to_helper):
2120         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
2121         (ChromiumMacPortTest.test_build_path):
2122         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2123         (ChromiumWinPort._path_to_driver):
2124         (ChromiumWinPort._path_to_helper):
2125         (ChromiumWinPort._path_to_image_diff):
2126         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2127         (ChromiumWinTest.test_build_path):
2128
2129 2012-07-23  Adam Barth  <abarth@webkit.org>
2130
2131         [Chromium] TestInterfaces should be responsible for owning and binding AccessibilityController and TextInputController
2132         https://bugs.webkit.org/show_bug.cgi?id=91459
2133
2134         Reviewed by Ryosuke Niwa.
2135
2136         Rather than having TestShell own and bind each of these JavaScript APIs
2137         individually, TestShell should delegate that work to TestInterfaces.
2138         This patch moves AccessibilityController and TextInputController to be
2139         the responsibility of TestInterfaces.
2140
2141         * DumpRenderTree/chromium/TestRunner/TestInterfaces.cpp:
2142         (TestInterfaces::TestInterfaces):
2143         (TestInterfaces::~TestInterfaces):
2144         (TestInterfaces::setWebView):
2145         (TestInterfaces::bindTo):
2146         (TestInterfaces::resetAll):
2147         * DumpRenderTree/chromium/TestRunner/TestInterfaces.h:
2148         (WebKit):
2149         (TestInterfaces):
2150         (TestInterfaces::accessibilityController):
2151         * DumpRenderTree/chromium/TestShell.cpp:
2152         (TestShell::initialize):
2153         (TestShell::createMainWindow):
2154         (TestShell::~TestShell):
2155         (TestShell::resetTestController):
2156         (TestShell::bindJSObjectsToWindow):
2157         * DumpRenderTree/chromium/TestShell.h:
2158         (TestShell::accessibilityController):
2159         (TestShell):
2160
2161 2012-07-23  Balazs Kelemen  <kbalazs@webkit.org>
2162
2163         nrwt swallows early exceptions
2164         https://bugs.webkit.org/show_bug.cgi?id=91884
2165
2166         Reviewed by Dirk Pranke.
2167
2168         Put everything in a try block so we won't miss any error.
2169
2170         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2171         (run):
2172         (main):
2173
2174 2012-07-23  Christophe Dumez  <christophe.dumez@intel.com>
2175
2176         [WK2] WebKitTestRunner needs to print information about Web intents
2177         https://bugs.webkit.org/show_bug.cgi?id=90873
2178
2179         Reviewed by Anders Carlsson.
2180
2181         Print out information about new intents requests and intent
2182         services registrations in WebKitTestRunner. This is expected
2183         by tests under webintents/*.
2184
2185         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2186         (WTR::InjectedBundlePage::InjectedBundlePage):
2187         (WTR::InjectedBundlePage::didReceiveIntentForFrame):
2188         (WTR):
2189         (WTR::InjectedBundlePage::registerIntentServiceForFrame):
2190         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
2191         (InjectedBundlePage):
2192
2193 2012-07-23  Dirk Pranke  <dpranke@chromium.org>
2194
2195         nrwt: never finds binaries in the 'out' dir on chromium win
2196         https://bugs.webkit.org/show_bug.cgi?id=91890
2197
2198         Reviewed by Tony Chang.
2199
2200         We were figuring out which directory look in for binaries by
2201         testing for the base directory (the directory above
2202         Debug/Release). In chromium-win's case, we look in src/build,
2203         which always exists because there are checked-in files in it,
2204         which means we'd always pick that directory over src/out. All of
2205         the other ports' build_path() implementation was including
2206         Debug/Release. If we matched that, we wouldn't have a problem,
2207         so this change fixes that and updates all of the callers of the
2208         chromium ports' implementation to not pass configuration as part
2209         of the path to look up; we still need to pass configuration in
2210         some cases (to test if the build is out of date between debug
2211         and release) so the implementation gets slightly more
2212         complicated.
2213
2214         * Scripts/webkitpy/layout_tests/port/chromium.py:
2215         (ChromiumPort._static_build_path):
2216         (ChromiumPort.default_results_directory):
2217         (ChromiumPort._build_path):
2218         (ChromiumPort._path_to_image_diff):
2219         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2220         (ChromiumAndroidPort._build_path):
2221         (ChromiumAndroidPort._path_to_driver):
2222         (ChromiumAndroidPort._path_to_forwarder):
2223         (ChromiumAndroidPort._push_executable):
2224         (ChromiumAndroidPort._push_fonts):
2225         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2226         (ChromiumLinuxPort._determine_driver_path_statically):
2227         (ChromiumLinuxPort._modules_to_search_for_symbols):
2228         (ChromiumLinuxPort._path_to_driver):
2229         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
2230         (ChromiumLinuxPortTest.test_build_path):
2231         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2232         (ChromiumMacPort._modules_to_search_for_symbols):
2233         (ChromiumMacPort._path_to_driver):
2234         (ChromiumMacPort._path_to_helper):
2235         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
2236         (ChromiumMacPortTest.test_build_path):
2237         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2238         (ChromiumWinPort._path_to_driver):
2239         (ChromiumWinPort._path_to_helper):
2240         (ChromiumWinPort._path_to_image_diff):
2241         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2242         (ChromiumWinTest.test_build_path):
2243
2244 2012-07-23  Sheriff Bot  <webkit.review.bot@gmail.com>
2245
2246         Unreviewed, rolling out r123339.
2247         http://trac.webkit.org/changeset/123339
2248         https://bugs.webkit.org/show_bug.cgi?id=92006
2249
2250         massive media tests failure (Requested by philn on #webkit).
2251
2252         * Scripts/webkitdirs.pm:
2253         (buildAutotoolsProject):
2254         * gtk/jhbuild.modules:
2255
2256 2012-07-23  Tommy Widenflycht  <tommyw@google.com>
2257
2258         [chromium] MediaStream API: Clean up the MockWebKitPlatformSupport object at shutdown
2259         https://bugs.webkit.org/show_bug.cgi?id=91857
2260
2261         Reviewed by Adam Barth.
2262
2263         Made a few changes so that the destructor could be called at shutdown.
2264
2265         * DumpRenderTree/chromium/DumpRenderTree.cpp:
2266         (WebKitSupportTestEnvironment::WebKitSupportTestEnvironment):
2267         (WebKitSupportTestEnvironment):
2268         * DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp:
2269         (MockWebKitPlatformSupport::create):
2270         * DumpRenderTree/chromium/MockWebKitPlatformSupport.h:
2271         (MockWebKitPlatformSupport):
2272         (MockWebKitPlatformSupport::~MockWebKitPlatformSupport):
2273
2274 2012-07-23  Philippe Normand  <pnormand@igalia.com>
2275
2276         [GTK][jhbuild] Switch to GStreamer 0.11 build
2277         https://bugs.webkit.org/show_bug.cgi?id=91727
2278
2279         Reviewed by Gustavo Noronha Silva.
2280
2281         Switch build-webkit --gtk to GStreamer 0.11 support and build the
2282         necessary GStreamer git modules from JHBuild.
2283
2284         * Scripts/webkitdirs.pm:
2285         (buildAutotoolsProject): Build WebKit with GStreamer 0.11 support.
2286         * gtk/jhbuild.modules: GStreamer build support.
2287
2288 2012-07-23  Christophe Dumez  <christophe.dumez@intel.com>
2289
2290         WebKit2 needs layoutTestController.setAlwaysAcceptCookies
2291         https://bugs.webkit.org/show_bug.cgi?id=42778
2292
2293         Reviewed by Kenneth Rohde Christiansen.
2294
2295         Add support for layoutTestController.setAlwaysAcceptCookies()
2296         since it is required by some tests.
2297
2298         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
2299         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
2300         (WTR::LayoutTestController::setAlwaysAcceptCookies):
2301         (WTR):
2302         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2303         (LayoutTestController):
2304
2305 2012-07-23  Kent Tamura  <tkent@chromium.org>
2306
2307         Rename ENABLE_METER_TAG and ENABLE_PROGRESS_TAG to ENABLE_METER_ELEMENT and ENABLE_PROGRESS_ELEMENT respectively
2308         https://bugs.webkit.org/show_bug.cgi?id=91941
2309
2310         Reviewed by Kentaro Hara.
2311
2312         A flag name for an elmement should be ENABLE_*_ELEMENT.
2313
2314         * Scripts/webkitperl/FeatureList.pm:
2315         * qmake/mkspecs/features/features.pri:
2316
2317 2012-07-22  Kent Tamura  <tkent@chromium.org>
2318
2319         Rename ENABLE_DETAILS to ENABLE_DETAILS_ELEMENT
2320         https://bugs.webkit.org/show_bug.cgi?id=91928
2321
2322         Reviewed by Kentaro Hara.
2323
2324         A flag name for an elmement should be ENABLE_*_ELEMENT.
2325
2326         * Scripts/webkitperl/FeatureList.pm:
2327         * qmake/mkspecs/features/features.pri:
2328
2329 2012-07-20  Kent Tamura  <tkent@chromium.org>
2330
2331         Rename ENABLE_DATALIST to ENABLE_DATALIST_ELEMENT
2332         https://bugs.webkit.org/show_bug.cgi?id=91846
2333
2334         Reviewed by Kentaro Hara.
2335
2336         A flag name for an elmement should be ENABLE_*_ELEMENT.
2337
2338         * Scripts/webkitperl/FeatureList.pm:
2339         * qmake/mkspecs/features/features.pri:
2340
2341 2012-07-20  Xianzhu Wang  <wangxianzhu@chromium.org>
2342
2343         [Chromium-Android] Store test pipes in /data/local/tmp
2344         https://bugs.webkit.org/show_bug.cgi?id=91909
2345
2346         In the previous change (bug 91873, r123258) we placed the pipes in /sdcard/native_tests.
2347         However after tests we found some devices don't allow the app to create pipes there.
2348         So /data/local/tmp is a better place for the pipes.
2349
2350         Reviewed by Adam Barth.
2351
2352         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2353
2354 2012-07-20  Xianzhu Wang  <wangxianzhu@chromium.org>
2355
2356         [Chomium-Android] Place DRT input/output pipes where both app and unrooted shell can read/write
2357         https://bugs.webkit.org/show_bug.cgi?id=91873
2358
2359         Reviewed by Adam Barth.
2360
2361         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2362
2363 2012-07-20  Han Shen  <shenhan@google.com>
2364
2365         [Chromium] Compilation fails under gcc 4.7
2366         https://bugs.webkit.org/show_bug.cgi?id=90227
2367
2368         Reviewed by Tony Chang.
2369
2370         Disable warnings about c++0x compatibility in gcc newer than 4.6.
2371
2372         * DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
2373
2374 2012-07-20  Dirk Pranke  <dpranke@chromium.org>
2375
2376         REGRESSION: run-perf-tests --pause-before-testing is broken
2377         https://bugs.webkit.org/show_bug.cgi?id=91789
2378
2379         Reviewed by Ryosuke Niwa.
2380
2381         Turns out start() wasn't actually starting things :(. Fixed, and
2382         added tests.
2383
2384         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
2385         (ChromiumAndroidDriverTest.test_read_prompt):
2386         * Scripts/webkitpy/layout_tests/port/driver.py:
2387         (Driver.__init__):
2388         (Driver._start):
2389         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
2390         (DriverTest.test_read_block):
2391         (DriverTest.test_read_binary_block):
2392         (DriverTest.test_read_base64_block):
2393         (DriverTest.test_stop_cleans_up_properly):
2394         (DriverTest.test_two_starts_cleans_up_properly):
2395         (DriverTest.test_start_actually_starts):
2396         (MockServerProcess.__init__):
2397         (MockServerProcess.start):
2398
2399 2012-07-20  Jochen Eisinger  <jochen@chromium.org>
2400
2401         [chromium] Don't include WebCore headers in TestInterfaces so it's safe to include from outside of WebCore
2402         https://bugs.webkit.org/show_bug.cgi?id=91847
2403
2404         Reviewed by Adam Barth.
2405
2406         This allows for including the TestIntefaces.h from content shell
2407         without introducing a dependency to WebCore/wtf
2408
2409         * DumpRenderTree/chromium/TestRunner/TestInterfaces.cpp:
2410         (TestInterfaces::Internal):
2411         (TestInterfaces::Internal::Internal):
2412         (TestInterfaces::Internal::~Internal):
2413         (TestInterfaces::Internal::bindTo):
2414         (TestInterfaces::Internal::resetAll):
2415         (TestInterfaces::TestInterfaces):
2416         (TestInterfaces::~TestInterfaces):
2417         (TestInterfaces::bindTo):
2418         (TestInterfaces::resetAll):
2419         * DumpRenderTree/chromium/TestRunner/TestInterfaces.h:
2420         (TestInterfaces):
2421
2422 2012-07-20  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
2423
2424         [Qt] Make WTR use the same color palette as DRT
2425         https://bugs.webkit.org/show_bug.cgi?id=91870
2426
2427         Reviewed by Noam Rosenthal.
2428
2429         Make sure that WebKitTestRunner and DumpRenderTree use the same palette: the
2430         standard from QWindowsStyle.
2431
2432         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
2433         (WTR::activateFonts): Moved code to platformInitialize since setting style is not
2434         exactly font related.
2435         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
2436         (WTR::InjectedBundle::platformInitialize): Force the usage of the style's
2437         palette.  Because at this point QApplication was already instantiated, the
2438         default palette was already initialized to be the system one, so we need to
2439         explicitly set the palette to be the standard from style. This is not needed in
2440         DRT because there we set the style before instantiating QApplication.
2441
2442 2012-07-20  Tony Chang  <tony@chromium.org>
2443
2444         Baseline align text in the flakiness dashboard toolbar
2445         https://bugs.webkit.org/show_bug.cgi?id=91784
2446
2447         Reviewed by Ojan Vafai.
2448
2449         Baseline align the flex items.
2450
2451         * TestResultServer/static-dashboards/flakiness_dashboard.css:
2452         (.forms): Add box-align: baseline
2453         (.forms > *): Use block display to force flex items rather than box.
2454         (#tests-form): Add box-align: baseline
2455         (#tests-form > *): Use block display to force flex items rather than box.
2456
2457 2012-07-20  Sheriff Bot  <webkit.review.bot@gmail.com>
2458
2459         Unreviewed, rolling out r123220.
2460         http://trac.webkit.org/changeset/123220
2461         https://bugs.webkit.org/show_bug.cgi?id=91880
2462
2463         broke Debug build (Requested by philn on #webkit).
2464
2465         * Scripts/webkitdirs.pm:
2466         (buildAutotoolsProject):
2467         * gtk/jhbuild.modules:
2468
2469 2012-07-20  Philippe Normand  <pnormand@igalia.com>
2470
2471         [GTK][jhbuild] Switch to GStreamer 0.11 build
2472         https://bugs.webkit.org/show_bug.cgi?id=91727
2473
2474         Reviewed by Gustavo Noronha Silva.
2475
2476         Switch build-webkit --gtk to GStreamer 0.11 support and build the
2477         necessary GStreamer git modules from JHBuild.
2478
2479         * Scripts/webkitdirs.pm:
2480         (buildAutotoolsProject): Build WebKit with GStreamer 0.11 support.
2481         * gtk/jhbuild.modules: GStreamer build support.
2482
2483 2012-07-20  Christophe Dumez  <christophe.dumez@intel.com>
2484
2485         [EFL] Proxy configuration should honor the no_proxy environment variable
2486         https://bugs.webkit.org/show_bug.cgi?id=91747
2487
2488         Reviewed by Kenneth Rohde Christiansen.
2489
2490         Add ProxyResolverSoup to the list of classes that contain GObjects
2491         in style checker script so that it does not complain about naming
2492         convention issues that are mandated by GObject.
2493
2494         * Scripts/webkitpy/style/checker.py:
2495
2496 2012-07-20  Zeno Albisser  <zeno@webkit.org>
2497
2498         [Qt][WK2] Add configure tests for Xrender and GLX.
2499         https://bugs.webkit.org/show_bug.cgi?id=91845
2500
2501         GraphicsSurface on linux requires Xcomposite, Xrender
2502         and GLX. If any of these requirements is not met,
2503         we should disable GraphicsSurface.
2504
2505         Reviewed by Kenneth Rohde Christiansen.
2506
2507         * qmake/config.tests/glx/glx.cpp: Added.
2508         (main):
2509         * qmake/config.tests/glx/glx.pro: Added.
2510         * qmake/config.tests/libXrender/libXrender.cpp: Added.
2511         (main):
2512         * qmake/config.tests/libXrender/libXrender.pro: Added.
2513         * qmake/configure.pri:
2514         * qmake/mkspecs/features/features.prf:
2515
2516 2012-07-20  Kangil Han  <kangil.han@samsung.com>
2517
2518         [EFL][WK2][WTR] Implement EventSenderProxy
2519         https://bugs.webkit.org/show_bug.cgi?id=91731
2520
2521         Reviewed by Kentaro Hara.
2522
2523         This is an initial EFL port implementation of EventSenderProxy.
2524
2525         * WebKitTestRunner/EventSenderProxy.h:
2526         (WTR):
2527         (EventSenderProxy):
2528         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
2529         * WebKitTestRunner/PlatformEfl.cmake:
2530         * WebKitTestRunner/TestController.cpp:
2531         (WTR::TestController::TestController):
2532         (WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
2533         * WebKitTestRunner/efl/EventSenderProxyEfl.cpp: Added.
2534         (WTR):
2535         (WTREvent):
2536         (WTR::WTREvent::WTREvent):
2537         (WTR::KeyEventInfo::KeyEventInfo):
2538         (KeyEventInfo):
2539         (WTR::evasMouseButton):
2540         (WTR::setEvasModifiers):
2541         (WTR::dispatchMouseDownEvent):
2542         (WTR::dispatchMouseUpEvent):
2543         (WTR::dispatchMouseMoveEvent):
2544         (WTR::dispatchMouseScrollByEvent):
2545         (WTR::keyPadName):
2546         (WTR::keyName):
2547         (WTR::EventSenderProxy::EventSenderProxy):
2548         (WTR::EventSenderProxy::updateClickCountForButton):
2549         (WTR::EventSenderProxy::dispatchEvent):
2550         (WTR::EventSenderProxy::replaySavedEvents):
2551         (WTR::EventSenderProxy::sendOrQueueEvent):
2552         (WTR::EventSenderProxy::mouseDown):
2553         (WTR::EventSenderProxy::mouseUp):
2554         (WTR::EventSenderProxy::mouseMoveTo):
2555         (WTR::EventSenderProxy::mouseScrollBy):
2556         (WTR::EventSenderProxy::leapForward):
2557         (WTR::EventSenderProxy::keyDown):
2558         (WTR::EventSenderProxy::addTouchPoint):
2559         (WTR::EventSenderProxy::updateTouchPoint):
2560         (WTR::EventSenderProxy::setTouchModifier):
2561         (WTR::EventSenderProxy::touchStart):
2562         (WTR::EventSenderProxy::touchMove):
2563         (WTR::EventSenderProxy::touchEnd):
2564         (WTR::EventSenderProxy::touchCancel):
2565         (WTR::EventSenderProxy::clearTouchPoints):
2566         (WTR::EventSenderProxy::releaseTouchPoint):
2567         (WTR::EventSenderProxy::cancelTouchPoint):
2568         (WTR::EventSenderProxy::setTouchPointRadius):
2569
2570 2012-07-20  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
2571
2572         Unreviewed, rolling out r123085.
2573         http://trac.webkit.org/changeset/123085
2574         https://bugs.webkit.org/show_bug.cgi?id=91719
2575
2576         r123085 causes crashes on EFL layout test bot.
2577
2578         * WebKitTestRunner/CMakeLists.txt:
2579         * WebKitTestRunner/InjectedBundle/efl/LayoutTestControllerEfl.cpp:
2580         (WTR::LayoutTestController::platformInitialize):
2581
2582 2012-07-20  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
2583
2584         [WTR] Add *explicit* keyword to constructor which has a parameter
2585         https://bugs.webkit.org/show_bug.cgi?id=91745
2586
2587         Reviewed by Adam Barth.
2588
2589         Add *explicit* keyword to contructor which has a parameter in order to avoid implicit type conversion.
2590
2591         * WebKitTestRunner/EventSenderProxy.h:
2592         (EventSenderProxy):
2593         * WebKitTestRunner/TestInvocation.h:
2594         (TestInvocation):
2595
2596 2012-07-19  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
2597
2598         [EFL] [WK2] Add methods to get/set a custom text encoding
2599         https://bugs.webkit.org/show_bug.cgi?id=90604
2600
2601         Reviewed by Kenneth Rohde Christiansen.
2602
2603         Added support for setting custom character encoding in MiniBrowser.
2604         Use 'F3' to set the custom character encoding.
2605
2606         * MiniBrowser/efl/main.c:
2607         (on_key_down):
2608
2609 2012-07-19  Dirk Pranke  <dpranke@chromium.org>
2610
2611         webkitpy: executive_unittest still failing when run in parallel
2612         https://bugs.webkit.org/show_bug.cgi?id=91795
2613
2614         Reviewed by Adam Barth.
2615
2616         It turns out that test_kill_process and test_kill_all are both
2617         launching the same process, and so if they're run at the same
2618         time, test_kill_all interferes w/ test_kill_process.
2619
2620         Merging the two tests solves the issue :).
2621
2622         * Scripts/webkitpy/common/system/executive_unittest.py:
2623         (ExecutiveTest.test_kill_process):
2624         (ExecutiveTest._assert_windows_image_name):
2625         (ExecutiveTest):
2626         (ExecutiveTest.test_windows_image_name):
2627
2628 2012-07-19  Sudarsana Nagineni  <sudarsana.nagineni@linux.intel.com>
2629
2630         [EFL] [WK2] Add a callback to handle delete request in MiniBrowser
2631         https://bugs.webkit.org/show_bug.cgi?id=91483
2632
2633         Reviewed by Dirk Pranke.
2634
2635         Add a callback to handle delete request and quit the main loop when
2636         window is asked to be deleted so that all the resources allocated
2637         for the window will be freed properly.
2638
2639         * MiniBrowser/efl/main.c:
2640         (closeWindow):
2641         (browserCreate):
2642
2643 2012-07-19  No'am Rosenthal  <noam.rosenthal@nokia.com>
2644
2645         [Qt] Enable CSS shaders in Qt (software mode)
2646         https://bugs.webkit.org/show_bug.cgi?id=85140
2647
2648         Reviewed by Simon Hausmann.
2649
2650         Enable CSS_SHADERS when WEBGL and CSS_FILTERS are enabled.
2651
2652         * qmake/mkspecs/features/features.prf:
2653
2654 2012-07-19  No'am Rosenthal  <noam.rosenthal@nokia.com>
2655
2656         [Qt] Enable CSS shaders in Qt (software mode)
2657         https://bugs.webkit.org/show_bug.cgi?id=85140
2658
2659         Reviewed by Simon Hausmann.
2660
2661         Enable CSS_SHADERS when WEBGL and CSS_FILTERS are enabled.
2662
2663         * qmake/mkspecs/features/features.prf:
2664
2665 2012-07-19  Dirk Pranke  <dpranke@chromium.org>
2666
2667         test-webkitpy: fix executive_unittest nits, including a missing 'errno' import
2668         https://bugs.webkit.org/show_bug.cgi?id=91770
2669
2670         Reviewed by Adam Barth.
2671
2672         The missing errno import is causing a flaky failure when run in
2673         parallel on the bots.
2674
2675         * Scripts/webkitpy/common/system/executive_unittest.py:
2676         (ExecutiveTest.assert_interpreter_for_content):
2677
2678 2012-07-19  Oliver Hunt  <oliver@apple.com>
2679
2680         Make style bot handle underscores in names more reasonably
2681         https://bugs.webkit.org/show_bug.cgi?id=91777
2682
2683         Reviewed by Gavin Barraclough.
2684
2685         This patch adds a naming/readability/underscores category for style errors.
2686         That allows us to use underscores in those portions of JSC (and potentially
2687         gtk?) without turning off all naming checks for the effected directories.
2688         We also filter out the DFG_OPERATION warning, and a few other minor changes
2689         to make the style bot less noisy when running over JSC at least.
2690
2691         * Scripts/webkitpy/style/checker.py:
2692         * Scripts/webkitpy/style/checkers/cpp.py:
2693         (check_identifier_name_in_declaration):
2694         (CppChecker):
2695         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
2696         (WebKitStyleTest.test_names):
2697
2698 2012-07-19  Dirk Pranke  <dpranke@chromium.org>
2699
2700         config_standalone.py is crashing on apple win's cygwin port
2701         https://bugs.webkit.org/show_bug.cgi?id=91682
2702
2703         Reviewed by Adam Roben.
2704
2705         Work around a weird interaction of ctypes and cygwin that
2706         is causing unit tests to crash for me locally.
2707
2708         * Scripts/webkitpy/common/system/executive.py:
2709         (Executive._win32_check_running_pid):
2710
2711 2012-07-19  Xabier Rodriguez Calvar  <calvaris@igalia.com>
2712
2713         [GTK] WebKit test runner ignores all system environment variables
2714         https://bugs.webkit.org/show_bug.cgi?id=90284
2715
2716         Reviewed by Dirk Pranke.
2717
2718         Added support in unit test framework to pass a custom environment
2719         variable to the tests by using the --additional-env-var parameter.
2720
2721         * Scripts/webkitpy/layout_tests/port/base.py:
2722         (Port.to.setup_environ_for_server): Adding environment variables
2723         that come through the additional_env_var option.
2724         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2725         (PortTest.test_additional_env_var): Added unit test for the
2726         additional_env_var option
2727         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2728         (parse_args): Added the --additional-env-var parameter.
2729
2730 2012-07-19  Joshua Bell  <jsbell@chromium.org>
2731
2732         Roll out r121610 and r122487 which may have been causing flaky crashes
2733         https://bugs.webkit.org/show_bug.cgi?id=91637
2734
2735         Reviewed by Kentaro Hara.
2736
2737         Flaky crashes started in random tests following r121610. It's not clear
2738         that r121610 is to blame, but we'd like to prove or disprove it.
2739         If this doesn't resolve the crashes, this patch should be rolled out.
2740
2741         * DumpRenderTree/DumpRenderTree.gypi:
2742         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
2743         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
2744         (PluginTest):
2745         * DumpRenderTree/TestNetscapePlugIn/Tests/LeakWindowScriptableObject.cpp: Removed.
2746
2747 2012-07-19  Pierre Rossi  <pierre.rossi@gmail.com>
2748
2749         [Qt][WK2] Map the elementRect exposed in QML to the view for popup menu
2750         https://bugs.webkit.org/show_bug.cgi?id=91645
2751
2752         Reviewed by Simon Hausmann.
2753
2754         Also improved the logic slightly for edge cases.
2755
2756         * MiniBrowser/qt/qml/ItemSelector.qml:
2757
2758 2012-07-19  Christophe Dumez  <christophe.dumez@intel.com>
2759
2760         [EFL] Bump libsoup dependency to v2.39.4.1 to fix cookie issues
2761         https://bugs.webkit.org/show_bug.cgi?id=91741
2762
2763         Reviewed by Kenneth Rohde Christiansen.
2764
2765         Bump libsoup dependency in jhbuild to v2.39.4.1 to fix issues
2766         with cookies in non-suffixed/private domains [GNOME #679230].
2767
2768         * efl/jhbuild.modules:
2769
2770 2012-07-19  Simon Hausmann  <simon.hausmann@nokia.com>
2771
2772         [Qt] Remove support for Qt versions before 4.8.0
2773         https://bugs.webkit.org/show_bug.cgi?id=91730
2774
2775         Reviewed by Kenneth Rohde Christiansen.
2776
2777         Remove Qt version check #ifdefs for Qt versions before 4.8.0.
2778
2779         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2780         (LayoutTestController::reset):
2781         (LayoutTestController::setAlwaysBlockCookies):
2782         * DumpRenderTree/qt/LayoutTestControllerQt.h:
2783         (LayoutTestController):
2784
2785 2012-07-19  Mario Sanchez Prada  <msanchez@igalia.com>
2786
2787         [GTK] Enable MHTML support by default at build time
2788         https://bugs.webkit.org/show_bug.cgi?id=89987
2789
2790         Reviewed by Carlos Garcia Campos.
2791
2792         Enable MHTML support by default for GTK in build-webkit.
2793
2794         * Scripts/webkitperl/FeatureList.pm:
2795
2796 2012-07-19  Zoltan Nyul  <zoltan.nyul@intel.com>
2797
2798         [EFL][WK2] Using different cache directory for each WTR process
2799         https://bugs.webkit.org/show_bug.cgi?id=91719
2800
2801         Reviewed by Kenneth Rohde Christiansen.
2802
2803         The testrunner script may be running multiple processes in
2804         parallel, and it makes appcache tests fail if they are using the same
2805         directory. I modified the EFL's LayoutTestController to use the
2806         DUMPRENDERTREE_TEMP for application cache directory because it's
2807         different for each process.
2808
2809         * WebKitTestRunner/CMakeLists.txt:
2810         * WebKitTestRunner/InjectedBundle/efl/LayoutTestControllerEfl.cpp:
2811         (WTR::LayoutTestController::platformInitialize):
2812
2813 2012-07-19  Mario Sanchez Prada  <msanchez@igalia.com>
2814
2815         [WK2][GTK] Complete implementation of AccessibilityController in WKTR for GTK
2816         https://bugs.webkit.org/show_bug.cgi?id=89224
2817
2818         Reviewed by Chris Fleizach.
2819
2820         Implement logAccessibilityEvents() and resetToConsistentState(),
2821         so we can ask WKTR from the layout tests for logging all the
2822         accessibility related events into the result.
2823
2824         * WebKitTestRunner/GNUmakefile.am:
2825         * WebKitTestRunner/InjectedBundle/AccessibilityController.cpp:
2826         (WTR::AccessibilityController::AccessibilityController):
2827         (WTR):
2828         (WTR::AccessibilityController::logAccessibilityEvents):
2829         (WTR::AccessibilityController::resetToConsistentState):
2830         * WebKitTestRunner/InjectedBundle/AccessibilityController.h:
2831         (AccessibilityController):
2832         * WebKitTestRunner/InjectedBundle/gtk/AccessibilityControllerGtk.cpp: Added.
2833         (WTR):
2834         (WTR::printAccessibilityEvent):
2835         (WTR::axObjectEventListener):
2836         (WTR::AccessibilityController::logAccessibilityEvents):
2837         (WTR::AccessibilityController::resetToConsistentState):
2838
2839         Ensure we reset the accessibility controller to a consistent
2840         whenever a test has finished running.
2841
2842         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
2843         (WTR::InjectedBundle::done):
2844
2845 2012-07-19  Dominik Röttsches  <dominik.rottsches@intel.com>
2846
2847         [EFL][Wk2] Add a bot that runs WTR / WebKit2 tests
2848         https://bugs.webkit.org/show_bug.cgi?id=90788
2849
2850         Reviewed by Kentaro Hara.
2851
2852         Adding EFL 64bit WK2 BuildAndTest bot. At this point, we cannot transfer
2853         the build product from one bot to the other due to link path issues with jhbuild,
2854         so we do a full build and test cycle for WebKit2 on this bot.
2855
2856         * BuildSlaveSupport/build.webkit.org-config/config.json: Adding the new slave.
2857         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Adding a new factory for building and testing WebKit2.
2858         (BuildAndTestWebKit2Factory):
2859
2860 2012-07-18  Mihnea Ovidenie  <mihnea@adobe.com>
2861
2862         Unreviewed. Adding Andrei Bucur, Mihai Balan and Raul Hudea to the contributors list.
2863         They are Adobe contributors working on CSSRegions.
2864
2865         * Scripts/webkitpy/common/config/committers.py:
2866
2867 2012-07-18  Christophe Dumez  <christophe.dumez@intel.com>
2868
2869         [CMake][EFL] Should try to find the same libsoup version pulled by jhbuild
2870         https://bugs.webkit.org/show_bug.cgi?id=91626
2871
2872         Reviewed by Kenneth Rohde Christiansen.
2873
2874         Update jhbuild to pull libsoup 2.39.3 tarball from Gnome FTP instead
2875         of fetching a tag from Gnome GIT. We need to do this to map the version
2876         of libsoup that is required by our CMake config.
2877
2878         * efl/jhbuild.modules:
2879
2880 2012-07-18  MORITA Hajime  <morrita@google.com>
2881
2882         Add window resize benchmark
2883         https://bugs.webkit.org/show_bug.cgi?id=91070
2884
2885         Reviewed by Ryosuke Niwa.
2886
2887         Added some more output filters to take care of html5.html output.
2888
2889         * Scripts/webkitpy/performance_tests/perftest.py:
2890         (PerfTest):
2891
2892 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
2893
2894         nrwt: merge WebKitDriver into Driver
2895         https://bugs.webkit.org/show_bug.cgi?id=91696
2896
2897         Reviewed by Adam Barth.
2898
2899         As part of merging WebKitPort into Port, we also should merge
2900         WebKitDriver into Driver, since all of the non-test Drivers now
2901         inherit from WebKitDriver.
2902
2903         There are no functional changes in this patch, and all the changes should
2904         be covered by the existing tests.
2905
2906         * Scripts/webkitpy/layout_tests/port/base.py:
2907         (Port._driver_class):
2908         * Scripts/webkitpy/layout_tests/port/chromium.py:
2909         (ChromiumPort._check_file_exists):
2910         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
2911         (ChromiumAndroidDriver):
2912         (ChromiumAndroidDriver.__init__):
2913         (ChromiumAndroidDriver._drt_cmd_line):
2914         * Scripts/webkitpy/layout_tests/port/chromium_android_unittest.py:
2915         (ChromiumAndroidDriverTest.test_read_prompt):
2916         * Scripts/webkitpy/layout_tests/port/driver.py:
2917         (Driver):
2918         (Driver.__init__):
2919         (Driver.__del__):
2920         (Driver.run_test):
2921         (Driver.has_crashed):
2922         (Driver.start):
2923         (Driver._start):
2924         (Driver.stop):
2925         (Driver.cmd_line):
2926         (Driver._check_for_driver_crash):
2927         (Driver._command_from_driver_input):
2928         (Driver._read_first_block):
2929         (Driver._read_optional_image_block):
2930         (Driver._read_header):
2931         (Driver._process_stdout_line):
2932         (Driver._strip_eof):
2933         (Driver._read_block):
2934         (ContentBlock):
2935         (ContentBlock.__init__):
2936         (ContentBlock.decode_content):
2937         * Scripts/webkitpy/layout_tests/port/driver_unittest.py:
2938         (DriverTest.make_port):
2939         (DriverTest._assert_wrapper):
2940         (DriverTest.test_read_block):
2941         (DriverTest):
2942         (DriverTest.test_read_binary_block):
2943         (DriverTest.test_read_base64_block):
2944         (DriverTest.test_no_timeout):
2945         (DriverTest.test_check_for_driver_crash):
2946         (DriverTest.test_check_for_driver_crash.FakeServerProcess):
2947         (DriverTest.test_check_for_driver_crash.FakeServerProcess.__init__):
2948         (DriverTest.test_check_for_driver_crash.FakeServerProcess.pid):
2949         (DriverTest.test_check_for_driver_crash.FakeServerProcess.name):
2950         (DriverTest.test_check_for_driver_crash.FakeServerProcess.has_crashed):
2951         (DriverTest.test_check_for_driver_crash.FakeServerProcess.stop):
2952         (DriverTest.test_check_for_driver_crash.assert_crash):
2953         (DriverTest.test_creating_a_port_does_not_write_to_the_filesystem):
2954         (DriverTest.test_stop_cleans_up_properly):
2955         (DriverTest.test_two_starts_cleans_up_properly):
2956         (MockServerProcess):
2957         (MockServerProcess.__init__):
2958         (MockServerProcess.has_crashed):
2959         (MockServerProcess.read_stdout_line):
2960         (MockServerProcess.read_stdout):
2961         (MockServerProcess.read_either_stdout_or_stderr_line):
2962         (MockServerProcess.start):
2963         (MockServerProcess.stop):
2964         (MockServerProcess.kill):
2965         * Scripts/webkitpy/layout_tests/port/gtk.py:
2966         * Scripts/webkitpy/layout_tests/port/webkit.py:
2967         (WebKitPort._read_image_diff):
2968         (WebKitPort.skipped_layout_tests):
2969         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2970         (test_path_to_apache_config_file):
2971         * Scripts/webkitpy/layout_tests/port/xvfbdriver.py:
2972         (XvfbDriver):
2973         (XvfbDriver.stop):
2974
2975 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
2976
2977         webkitpy.common.system.executive_unittest.ExecutiveTest.test_kill_process is flaky
2978         https://bugs.webkit.org/show_bug.cgi?id=91698
2979
2980         Unreviewed, build fix.
2981
2982         This is a workaround for the test failing (apparently mostly?)
2983         when run in parallel with other tests. The underlying bug still
2984         needs to be diagnosed as we shouldn't be getting ECHILD from
2985         process.wait().
2986
2987         * Scripts/webkitpy/common/system/executive_unittest.py:
2988         (ExecutiveTest.test_kill_process):
2989
2990 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
2991
2992         Disable running python unit tests in parallel on win32 for now.
2993         https://bugs.webkit.org/show_bug.cgi?id=91697
2994
2995         Unreviewed, build fix.
2996
2997         * Scripts/webkitpy/test/main.py:
2998         (Tester._parse_args):
2999
3000 2012-07-18  Dave Tu  <dtu@chromium.org>
3001
3002         [chromium] Add gl_tests to flakiness dashboard.
3003         https://bugs.webkit.org/show_bug.cgi?id=91680
3004
3005         Reviewed by Dirk Pranke.
3006
3007         * TestResultServer/static-dashboards/builders.js:
3008         (loadBuildersList):
3009         * TestResultServer/static-dashboards/dashboard_base.js:
3010         (currentBuilderGroupCategory):
3011
3012 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
3013
3014         fix python unit tests failing under cygwin
3015         https://bugs.webkit.org/show_bug.cgi?id=91678
3016
3017         Reviewed by Adam Barth.
3018
3019         Fix a case where executive.kill_process was failing under cygwin
3020         (apple win port) in an unexpected way.
3021         
3022         * Scripts/webkitpy/common/system/executive.py:
3023         (Executive.kill_process):
3024
3025 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
3026
3027         test-webkitpy: run tests in parallel by default
3028         https://bugs.webkit.org/show_bug.cgi?id=91422
3029
3030         Reviewed by Adam Barth.
3031
3032         We use multiprocessing.cpu_count() for the default number of
3033         jobs to run; memory overhead should be very low, so this should
3034         be fine.
3035
3036         * Scripts/webkitpy/test/main.py:
3037         (Tester._parse_args):
3038
3039 2012-07-18  Ryosuke Niwa  <rniwa@webkit.org>
3040
3041         Add Pravin D to the list of contributors.
3042
3043         * Scripts/webkitpy/common/config/committers.py:
3044
3045 2012-07-18  Balazs Kelemen  <kbalazs@webkit.org>
3046
3047         [Qt] feature detection in orwt/nrwt does not work with force_static_libs_as_shared
3048         https://bugs.webkit.org/show_bug.cgi?id=91514
3049
3050         Reviewed by Dirk Pranke.
3051
3052         Analyzing libQtWebKit.so is not enough in the case of force_static_libs_as_shared.
3053         We need to analyze all the shared library or at least the WebCore one.
3054
3055         * Scripts/webkitdirs.pm:
3056         (builtDylibPathForName):
3057         In orwt it is hard coded to search for symbols in the WebCore library.
3058         While in theory it is possible that symbols for a given feature are not
3059         located in that, it doesn't happen in practice, so returning the path of
3060         the WebCore library for a force_static_libs_as_shared build is enough to
3061         fix the bug and it doesn't require a bigger refactoring. For a default
3062         build we still return the path of the QtWebKit lib.
3063         * Scripts/webkitpy/layout_tests/port/qt.py:
3064         (QtPort._modules_to_search_for_symbols):
3065         Enumerate all dynamic libraries, not just the QtWebKit one.
3066
3067 2012-07-18  Tony Chang  <tony@chromium.org>
3068
3069         [chromium] Unreviewed, more compile fixes on Chromium Win.
3070
3071         * DumpRenderTree/chromium/TestShellWin.cpp:
3072         (TestShell::waitTestFinished):
3073         * DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
3074
3075 2012-07-18  Christophe Dumez  <christophe.dumez@intel.com>
3076
3077         [EFL] Add central error management to EFL port
3078         https://bugs.webkit.org/show_bug.cgi?id=91598
3079
3080         Reviewed by Kenneth Rohde Christiansen.
3081
3082         Map WebKitNetworkError to NSURLErrorDomain when
3083         printing in DumpRenderTree so that the output
3084         matches the expected one.
3085
3086         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
3087         (descriptionSuitableForTestResult):
3088
3089 2012-07-18  Dirk Pranke  <dpranke@chromium.org>
3090
3091         nrwt: start merging port/webkit.py into port/base.py
3092         https://bugs.webkit.org/show_bug.cgi?id=91559
3093
3094         Reviewed by Ojan Vafai.
3095
3096         Since all the non-test port implementations now derive from
3097         WebKitPort, there's no real point in keeping WebKitPort distinct
3098         from Port. This patch starts merging the two by moving nearly
3099         all of the webkit implementations of routines with no default
3100         behavior into base.py. The few that didn't move rely on
3101         additional infrastructure that should be refactored differently
3102         (like the image diffing, which should probably be its own
3103         class) and deserve their own patches.
3104
3105         This patch should just be moving code around, and require no
3106         additional tests; in fact, we can delete the tests that were
3107         asserting virtual methods in the base class.
3108
3109         * Scripts/webkitpy/layout_tests/port/base.py:
3110         (Port.baseline_search_path):
3111         (Port.check_build):
3112         (Port):
3113         (Port._check_driver):
3114         (Port._check_port_build):
3115         (Port.check_image_diff):
3116         (Port.driver_name):
3117         (Port.default_results_directory):
3118         (Port.to.setup_environ_for_server):
3119         (Port._path_to_apache):
3120         (Port._is_redhat_based):
3121         (Port._is_debian_based):
3122         (Port._apache_config_file_name_for_platform):
3123         (Port._path_to_apache_config_file):
3124         (Port._build_path):
3125         (Port._path_to_driver):
3126         (Port._path_to_webcore_library):
3127         (Port._path_to_helper):
3128         (Port._path_to_image_diff):
3129         (Port._path_to_wdiff):
3130         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3131         (PortTest.test_httpd_returns_error_code):
3132         * Scripts/webkitpy/layout_tests/port/webkit.py:
3133         (WebKitPort):
3134         (WebKitPort._build_driver_flags):
3135         (WebKitPort._read_image_diff):
3136         (WebKitPort.skipped_layout_tests):
3137
3138 2012-07-18  Tommy Widenflycht  <tommyw@google.com>
3139
3140         [chromium] MediaStream API: Moving the mock create* WebRTC calls into a shadow Platform class
3141         https://bugs.webkit.org/show_bug.cgi?id=86215
3142
3143         Reviewed by Adam Barth.
3144
3145         Adding a shadow Platform object that is used to override some WebKit::Platform funtions to
3146         instead create mock objects. No actual mock objects created yet.
3147
3148         * DumpRenderTree/DumpRenderTree.gypi:
3149         * DumpRenderTree/chromium/DumpRenderTree.cpp:
3150         (WebKitSupportTestEnvironment::WebKitSupportTestEnvironment):
3151         * DumpRenderTree/chromium/MockWebKitPlatformSupport.cpp: Copied from Tools/DumpRenderTree/chromium/config.h.
3152         (MockWebKitPlatformSupport::create):
3153         (MockWebKitPlatformSupport::MockWebKitPlatformSupport):
3154         (MockWebKitPlatformSupport::cryptographicallyRandomValues):
3155         (MockWebKitPlatformSupport::createMediaStreamCenter):
3156         * DumpRenderTree/chromium/MockWebKitPlatformSupport.h: Copied from Tools/DumpRenderTree/chromium/config.h.
3157         (MockWebKitPlatformSupport):
3158         * DumpRenderTree/chromium/config.h:
3159
3160 2012-07-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
3161
3162         [CMake][EFL] Build and run TestWebKitAPI unit tests
3163         https://bugs.webkit.org/show_bug.cgi?id=90671
3164
3165         Reviewed by Daniel Bates.
3166
3167         This patch enables almost for free a significant amount of test for the
3168         WebKit C API, WTF and KURL. It makes easy for CMake based ports to
3169         enable TestWebKitAPI since the missing bits to implement are
3170         relatively simple.
3171
3172         * CMakeLists.txt:
3173         * TestWebKitAPI/CMakeLists.txt: Added.
3174         * TestWebKitAPI/PlatformEfl.cmake: Added.
3175         * TestWebKitAPI/PlatformWebView.h:
3176         * TestWebKitAPI/config.h:
3177         * TestWebKitAPI/efl/InjectedBundleController.cpp: Added.
3178         (TestWebKitAPI):
3179         (TestWebKitAPI::InjectedBundleController::platformInitialize):
3180         * TestWebKitAPI/efl/PlatformUtilities.cpp: Added.
3181         (TestWebKitAPI):
3182         (Util):
3183         (TestWebKitAPI::Util::run):
3184         (TestWebKitAPI::Util::sleep):
3185         (TestWebKitAPI::Util::createURLForResource):
3186         (TestWebKitAPI::Util::createInjectedBundlePath):
3187         (TestWebKitAPI::Util::URLForNonExistentResource):
3188         * TestWebKitAPI/efl/PlatformWebView.cpp: Added.
3189         (TestWebKitAPI):
3190         (TestWebKitAPI::initEcoreEvas):
3191         (TestWebKitAPI::PlatformWebView::PlatformWebView):
3192         (TestWebKitAPI::PlatformWebView::~PlatformWebView):
3193         (TestWebKitAPI::PlatformWebView::page):
3194         * TestWebKitAPI/efl/main.cpp: Added.
3195         (checkForUseX11WindowArgument):
3196         (main):
3197
3198 2012-07-18  Thiago Marcos P. Santos  <thiago.santos@intel.com>
3199
3200         [CMake][EFL] Building jsc causes reconfiguration
3201         https://bugs.webkit.org/show_bug.cgi?id=91387
3202
3203         Reviewed by Daniel Bates.
3204
3205         We should remove CMakeCache only when running build-webkit script,
3206         otherwise it will cause a reconfiguration every time someone calls
3207         generateBuildSystemFromCMakeProject(). We were re-building "jsc" and
3208         not running WebKit2 unit tests on the bots because the project was
3209         reconfigured with default values by the scripts that run these tests.
3210
3211         * Scripts/build-webkit:
3212         * Scripts/webkitdirs.pm:
3213         (removeCMakeCache):
3214         (generateBuildSystemFromCMakeProject):
3215
3216 2012-07-18  Seokju Kwon  <seokju.kwon@samsung.com>
3217
3218         [EFL][DRT] Add support for Web Inspector in WebKit-EFL DRT
3219         https://bugs.webkit.org/show_bug.cgi?id=87935
3220
3221         Reviewed by Andreas Kling.
3222
3223         Web Inspector will be shown when path or url contains "inspector/".
3224         Dumprendertree should wait util web inspector resources are loaded totally
3225         and handle the signals for creating or removing a view of web inspector.
3226         ("inspector,view,create" and "inspector,view,close")
3227
3228         * DumpRenderTree/efl/DumpRenderTree.cpp:
3229         (shouldOpenWebInspector):
3230         (createLayoutTestController):
3231         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
3232         (DumpRenderTreeChrome::createView):
3233         (DumpRenderTreeChrome::createWebInspectorView):
3234         (DumpRenderTreeChrome::removeWebInspectorView):
3235         (DumpRenderTreeChrome::waitInspectorLoadFinished):
3236         (DumpRenderTreeChrome::onInspectorViewCreate):
3237         (DumpRenderTreeChrome::onInspectorViewClose):
3238         (DumpRenderTreeChrome::onInspectorFrameLoadFinished):
3239         * DumpRenderTree/efl/DumpRenderTreeChrome.h:
3240         (DumpRenderTreeChrome):
3241         * DumpRenderTree/efl/DumpRenderTreeView.cpp:
3242         (onConsoleMessage):
3243         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
3244         (LayoutTestController::showWebInspector):
3245         (LayoutTestController::closeWebInspector):
3246         (LayoutTestController::evaluateInWebInspector):
3247
3248 2012-07-18  Simon Hausmann  <simon.hausmann@nokia.com>
3249
3250         [ANGLE] On QT, use Bison and Flex during ANGLE build
3251         https://bugs.webkit.org/show_bug.cgi?id=91108
3252
3253         Reviewed by Kenneth Rohde Christiansen.
3254
3255         * qmake/mkspecs/features/default_post.prf: Add support for variable_out to our generators, to allow
3256         generating not only for SOURCES but also ANGLE_SOURCES (in this bug)
3257
3258 2012-07-18  Balazs Kelemen  <kbalazs@webkit.org>
3259
3260         [Qt] plugin tests should not be disabled for WebKit1
3261         https://bugs.webkit.org/show_bug.cgi?id=91604
3262
3263         Reviewed by Simon Hausmann.
3264
3265         Instead of not building TestNetscapePlugIn, we could programatically
3266         disable actually loading it from WTR until https://bugs.webkit.org/show_bug.cgi?id=86620
3267         has been solved, so we can still test plugins on WebKit1.
3268
3269         * Tools.pro:
3270         * WebKitTestRunner/TestController.cpp:
3271         (WTR::TestController::initialize):
3272         * WebKitTestRunner/qt/TestControllerQt.cpp:
3273         (WTR::TestController::initializeTestPluginDirectory):
3274
3275 2012-07-18  Mario Sanchez Prada  <msanchez@igalia.com>
3276
3277         [WK2][GTK] Implement AccessibilityUIElement in WKTR for GTK
3278         https://bugs.webkit.org/show_bug.cgi?id=89223
3279
3280         Reviewed by Chris Fleizach.
3281
3282         Implemented AccessibilityUIElement in WKTR for GTK, based in the
3283         implementation present in DumpRenderTree.
3284
3285         Added new files in gtk/ and updated the cross-platform ones.
3286
3287         * WebKitTestRunner/GNUmakefile.am:
3288         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.cpp:
3289         (WTR):
3290         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
3291         (AccessibilityUIElement):
3292         * WebKitTestRunner/InjectedBundle/gtk/AccessibilityUIElementGtk.cpp: Added.
3293         (WTR):
3294         (WTR::attributeSetToString):
3295         (WTR::checkElementState):
3296         (WTR::indexRangeInTable):
3297         (WTR::alterCurrentValue):
3298         (WTR::AccessibilityUIElement::AccessibilityUIElement):
3299         (WTR::AccessibilityUIElement::~AccessibilityUIElement):
3300         (WTR::AccessibilityUIElement::isEqual):
3301         (WTR::AccessibilityUIElement::getChildren):
3302         (WTR::AccessibilityUIElement::getChildrenWithRange):
3303         (WTR::AccessibilityUIElement::childrenCount):
3304         (WTR::AccessibilityUIElement::elementAtPoint):
3305         (WTR::AccessibilityUIElement::indexOfChild):
3306         (WTR::AccessibilityUIElement::childAtIndex):
3307         (WTR::AccessibilityUIElement::linkedUIElementAtIndex):
3308         (WTR::AccessibilityUIElement::ariaOwnsElementAtIndex):
3309         (WTR::AccessibilityUIElement::ariaFlowToElementAtIndex):
3310         (WTR::AccessibilityUIElement::disclosedRowAtIndex):
3311         (WTR::AccessibilityUIElement::rowAtIndex):
3312         (WTR::AccessibilityUIElement::selectedChildAtIndex):
3313         (WTR::AccessibilityUIElement::selectedChildrenCount):
3314         (WTR::AccessibilityUIElement::selectedRowAtIndex):
3315         (WTR::AccessibilityUIElement::titleUIElement):
3316         (WTR::AccessibilityUIElement::parentElement):
3317         (WTR::AccessibilityUIElement::disclosedByRow):
3318         (WTR::AccessibilityUIElement::attributesOfLinkedUIElements):
3319         (WTR::AccessibilityUIElement::attributesOfDocumentLinks):
3320         (WTR::AccessibilityUIElement::attributesOfChildren):
3321         (WTR::AccessibilityUIElement::allAttributes):
3322         (WTR::AccessibilityUIElement::stringAttributeValue):
3323         (WTR::AccessibilityUIElement::numberAttributeValue):
3324         (WTR::AccessibilityUIElement::uiElementAttributeValue):
3325         (WTR::AccessibilityUIElement::boolAttributeValue):
3326         (WTR::AccessibilityUIElement::isAttributeSettable):
3327         (WTR::AccessibilityUIElement::isAttributeSupported):
3328         (WTR::AccessibilityUIElement::parameterizedAttributeNames):
3329         (WTR::AccessibilityUIElement::role):
3330         (WTR::AccessibilityUIElement::subrole):
3331         (WTR::AccessibilityUIElement::roleDescription):
3332         (WTR::AccessibilityUIElement::title):
3333         (WTR::AccessibilityUIElement::description):
3334         (WTR::AccessibilityUIElement::orientation):
3335         (WTR::AccessibilityUIElement::stringValue):
3336         (WTR::AccessibilityUIElement::language):
3337         (WTR::AccessibilityUIElement::helpText):
3338         (WTR::AccessibilityUIElement::x):
3339         (WTR::AccessibilityUIElement::y):
3340         (WTR::AccessibilityUIElement::width):
3341         (WTR::AccessibilityUIElement::height):
3342         (WTR::AccessibilityUIElement::clickPointX):
3343         (WTR::AccessibilityUIElement::clickPointY):
3344         (WTR::AccessibilityUIElement::intValue):
3345         (WTR::AccessibilityUIElement::minValue):
3346         (WTR::AccessibilityUIElement::maxValue):
3347         (WTR::AccessibilityUIElement::valueDescription):
3348         (WTR::AccessibilityUIElement::insertionPointLineNumber):
3349         (WTR::AccessibilityUIElement::isActionSupported):
3350         (WTR::AccessibilityUIElement::isEnabled):
3351         (WTR::AccessibilityUIElement::isRequired):
3352         (WTR::AccessibilityUIElement::isFocused):
3353         (WTR::AccessibilityUIElement::isSelected):
3354         (WTR::AccessibilityUIElement::isExpanded):
3355         (WTR::AccessibilityUIElement::isChecked):
3356         (WTR::AccessibilityUIElement::hierarchicalLevel):
3357         (WTR::AccessibilityUIElement::speak):
3358         (WTR::AccessibilityUIElement::ariaIsGrabbed):
3359         (WTR::AccessibilityUIElement::ariaDropEffects):
3360         (WTR::AccessibilityUIElement::lineForIndex):
3361         (WTR::AccessibilityUIElement::rangeForLine):
3362         (WTR::AccessibilityUIElement::rangeForPosition):
3363         (WTR::AccessibilityUIElement::boundsForRange):
3364         (WTR::AccessibilityUIElement::stringForRange):
3365         (WTR::AccessibilityUIElement::attributedStringForRange):
3366         (WTR::AccessibilityUIElement::attributedStringRangeIsMisspelled):
3367         (WTR::AccessibilityUIElement::uiElementForSearchPredicate):
3368         (WTR::AccessibilityUIElement::attributesOfColumnHeaders):
3369         (WTR::AccessibilityUIElement::attributesOfRowHeaders):
3370         (WTR::AccessibilityUIElement::attributesOfColumns):
3371         (WTR::AccessibilityUIElement::attributesOfRows):
3372         (WTR::AccessibilityUIElement::attributesOfVisibleCells):
3373         (WTR::AccessibilityUIElement::attributesOfHeader):
3374         (WTR::AccessibilityUIElement::rowCount):
3375         (WTR::AccessibilityUIElement::columnCount):
3376         (WTR::AccessibilityUIElement::indexInTable):
3377         (WTR::AccessibilityUIElement::rowIndexRange):
3378         (WTR::AccessibilityUIElement::columnIndexRange):
3379         (WTR::AccessibilityUIElement::cellForColumnAndRow):
3380         (WTR::AccessibilityUIElement::horizontalScrollbar):
3381         (WTR::AccessibilityUIElement::verticalScrollbar):
3382         (WTR::AccessibilityUIElement::selectedTextRange):
3383         (WTR::AccessibilityUIElement::setSelectedTextRange):
3384         (WTR::AccessibilityUIElement::increment):
3385         (WTR::AccessibilityUIElement::decrement):
3386         (WTR::AccessibilityUIElement::showMenu):
3387         (WTR::AccessibilityUIElement::press):
3388         (WTR::AccessibilityUIElement::setSelectedChild):
3389         (WTR::AccessibilityUIElement::accessibilityValue):
3390         (WTR::AccessibilityUIElement::documentEncoding):
3391         (WTR::AccessibilityUIElement::documentURI):
3392         (WTR::AccessibilityUIElement::url):
3393         (WTR::AccessibilityUIElement::addNotificationListener):
3394         (WTR::AccessibilityUIElement::removeNotificationListener):
3395         (WTR::AccessibilityUIElement::isFocusable):
3396         (WTR::AccessibilityUIElement::isSelectable):
3397         (WTR::AccessibilityUIElement::isMultiSelectable):
3398         (WTR::AccessibilityUIElement::isVisible):
3399         (WTR::AccessibilityUIElement::isOffScreen):
3400         (WTR::AccessibilityUIElement::isCollapsed):
3401         (WTR::AccessibilityUIElement::isIgnored):
3402         (WTR::AccessibilityUIElement::hasPopup):
3403         (WTR::AccessibilityUIElement::takeFocus):
3404         (WTR::AccessibilityUIElement::takeSelection):
3405         (WTR::AccessibilityUIElement::addSelection):
3406         (WTR::AccessibilityUIElement::removeSelection):
3407         (WTR::AccessibilityUIElement::textMarkerRangeForElement):
3408         (WTR::AccessibilityUIElement::textMarkerRangeLength):
3409         (WTR::AccessibilityUIElement::previousTextMarker):
3410         (WTR::AccessibilityUIElement::nextTextMarker):
3411         (WTR::AccessibilityUIElement::stringForTextMarkerRange):
3412         (WTR::AccessibilityUIElement::textMarkerRangeForMarkers):
3413         (WTR::AccessibilityUIElement::startTextMarkerForTextMarkerRange):
3414         (WTR::AccessibilityUIElement::endTextMarkerForTextMarkerRange):
3415         (WTR::AccessibilityUIElement::textMarkerForPoint):
3416         (WTR::AccessibilityUIElement::accessibilityElementForTextMarker):
3417         (WTR::AccessibilityUIElement::attributedStringForTextMarkerRangeContainsAttribute):
3418         (WTR::AccessibilityUIElement::indexForTextMarker):
3419         (WTR::AccessibilityUIElement::isTextMarkerValid):
3420         (WTR::AccessibilityUIElement::textMarkerForIndex):
3421
3422         Add documentEncoding and documentURI to AccessibilityUIElement.idl.
3423
3424         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityUIElement.idl:
3425
3426 2012-07-18  Alexis Menard  <alexis.menard@openbossa.org>
3427
3428         [EFL] Build fix in WebKitTestRunner.
3429         https://bugs.webkit.org/show_bug.cgi?id=91567
3430
3431         Reviewed by Kentaro Hara.
3432
3433         sleep() is defined in unistd.h, we need to include it.
3434
3435         * WebKitTestRunner/efl/TestControllerEfl.cpp:
3436
3437 2012-07-18  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
3438
3439         [NRWT] Unreviewed gardening after r122913
3440         https://bugs.webkit.org/show_bug.cgi?id=91601
3441
3442         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
3443         (TrivialMockPort.process_kill_time):
3444
3445 2012-07-17  Xianzhu Wang  <wangxianzhu@chromium.org>
3446
3447         [Chromium-Android] Run ref tests together to avoid expensive driver restarts
3448         https://bugs.webkit.org/show_bug.cgi?id=91533
3449
3450         Reviewed by Dirk Pranke.
3451
3452         Though DriverProxy maintains two drivers to support pixel tests and non-pixel tests,
3453         chromium-android uses another way because it can't support multiple drivers.
3454         It restarts the driver when pixel-test mode changes (e.g. when running a ref test after
3455         a normal test in --no-pixel-tests mode). However restarting driver is expensive on
3456         Android (several seconds each time). To reduce the cost, a command line option
3457         '--shard-ref-tests' is added to group ref tests in dedicated shards.
3458         The option is by default enabled on Android.
3459
3460         Will remove the option once DRT supports switching pixel test mode during one run.
3461         (https://bugs.webkit.org/show_bug.cgi?id=91538, https://bugs.webkit.org/show_bug.cgi?id=91539)
3462
3463         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3464         (Manager._shard_tests):
3465         (Manager._shard_in_two):
3466         (Manager._shard_by_directory):
3467         (Manager._run_tests):
3468         * Scripts/webkitpy/layout_tests/controllers/worker.py:
3469         (Worker._update_test_input):
3470         * Scripts/webkitpy/layout_tests/port/chromium_android.py:
3471         (ChromiumAndroidPort.__init__):
3472         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3473         (parse_args):
3474
3475 2012-07-17  Don Olmstead  <don.olmstead@am.sony.com>
3476
3477         NRWT The time before server_process kills DRT should be variable
3478         https://bugs.webkit.org/show_bug.cgi?id=91542
3479
3480         Reviewed by Dirk Pranke.
3481
3482         Added the ability to configure the amount of time
3483         server_process waits before killing the DRT process
3484         during the call to stop.
3485
3486         * Scripts/webkitpy/layout_tests/port/base.py:
3487         (Port.variable.process_kill_time):
3488         * Scripts/webkitpy/layout_tests/port/server_process.py:
3489         (ServerProcess.stop):
3490
3491 2012-07-17  Alexis Menard  <alexis.menard@openbossa.org>
3492
3493         webkitdirs.pm should fallback to uname -m if arch is not present.
3494         https://bugs.webkit.org/show_bug.cgi?id=91543
3495
3496         Reviewed by Martin Robinson.
3497
3498         It seems that coreutils is moving away from the 'arch' command. Archlinux
3499         for example doesn't ship it anymore (coreutils 8.17). We can then fallback
3500         to 'uname -m' to find out the architecture.
3501
3502         * Scripts/webkitdirs.pm:
3503         (determineArchitecture):
3504
3505 2012-07-17  Dirk Pranke  <dpranke@chromium.org>
3506
3507         nrwt: rename printer.print_update to printer.write_update to match metered_stream
3508         https://bugs.webkit.org/show_bug.cgi?id=91557
3509
3510         Reviewed by Ojan Vafai.
3511
3512         Now that all of the printing logic is in printer.py, I want the
3513         public interface to printer to follow the meteredstream
3514         interface where possible. renaming write_update() gets us close.
3515
3516         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3517         (Manager._run_tests):
3518         (Manager._set_up_run):
3519         (Manager.start_servers_with_lock):
3520         (Manager.stop_servers_with_lock):
3521         (Manager._clobber_old_results):
3522         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3523         (run):
3524         * Scripts/webkitpy/layout_tests/views/printing.py:
3525         (Printer.write_update):
3526         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
3527         (Testprinter.test_write_update):
3528
3529 2012-07-17  Tony Chang  <tony@chromium.org>
3530
3531         Fix a bug where optimize-baselines would incorrectly fail to optimize
3532         https://bugs.webkit.org/show_bug.cgi?id=91551
3533
3534         Reviewed by Adam Barth.
3535
3536         In some cases, optimize-baseline would correctly optimize the results, but because
3537         we weren't filtering the virtual ports out of _results_by_port_name, we thought
3538         that we had failed to optimize.
3539
3540         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
3541         (BaselineOptimizer._find_optimal_result_placement): No virtual filtering here.
3542         (BaselineOptimizer._filtered_results_by_port_name): New function that filters out virtual directories.
3543         (BaselineOptimizer.optimize): Filter out virtual ports.
3544         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
3545         (TestBaselineOptimizer._move_baselines): Add a stub so we don't actual move results.
3546         (BaselineOptimizerTest._assertOptimization): Call optimize and verify that the right files were moved.
3547         (BaselineOptimizerTest._assertOptimizationFailed): Add a method for when optimization should fail.
3548         (BaselineOptimizerTest.test_common_directory_includes_root): Update since this test should fail.
3549         (BaselineOptimizerTest.test_virtual_ports_filtered): New test case that demonstrates the bug.
3550
3551 2012-07-17  Xianzhu Wang  <wangxianzhu@chromium.org>
3552
3553         [Chromium] Add --encode-binary command line option for DRT
3554         https://bugs.webkit.org/show_bug.cgi?id=91532
3555
3556         Reviewed by Adam Barth.
3557
3558         When the option presents, DRT will encode binary output data in base64.
3559         ChromiumAndroidPort will pass the option because 'adb shell' doesn't support
3560         binary data output.
3561
3562         This change is a refactoring to replace the original hard-coded code for Android.
3563
3564         * DumpRenderTree/chromium/DumpRenderTree.cpp:
3565         (main):
3566         * DumpRenderTree/chromium/TestEventPrinter.cpp:
3567         (TestEventPrinter::TestEventPrinter):
3568         (TestEventPrinter::handleAudio): New function replacing the original handleAudioHeader
3569         (TestEventPrinter::handleImage):
3570         (TestEventPrinter::handleBinary): Extracted common function for handleAudio and handleImage
3571         * DumpRenderTree/chromium/TestEventPrinter.h:
3572         (TestEventPrinter):
3573         * DumpRenderTree/chromium/TestShell.cpp:
3574         (TestShell::TestShell):
3575         (TestShell::initialize):
3576         (TestShell::dump):
3577         * DumpRenderTree/chromium/TestShell.h:
3578         (TestShell):
3579         (TestShell::setEncodeBinary):
3580         * Scripts/webkitpy/layout_tests/port/chromium_android.py: Add --encode-binary to additional_drt_flags
3581         (ChromiumAndroidPort.__init__):
3582
3583 2012-07-17  Dirk Pranke  <dpranke@chromium.org>
3584
3585         nrwt: move per-test result output into printing.py
3586         https://bugs.webkit.org/show_bug.cgi?id=91465
3587
3588         Reviewed by Ojan Vafai.
3589
3590         This moves the last substantive logging/printing code from the
3591         manager into printing.py. Subsequent patches in this thread can
3592         largely focus on cleaning up printing.py and will leave the rest
3593         of the code alone.
3594
3595         This patch also removes manager.update() and
3596         manager.update_summary(), two routines that were no longer being
3597         used or needed (they have been dead for a long time but I didn't
3598         notice them until now).
3599
3600         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3601         (Manager.results_directory):
3602         (Manager._look_for_new_crash_logs):
3603         (Manager._update_summary_with_result):
3604         * Scripts/webkitpy/layout_tests/views/printing.py:
3605         (Printer.print_finished_test):
3606
3607 2012-07-17  Dirk Pranke  <dpranke@chromium.org>
3608
3609         nrwt: move config-specific logging to printing.py
3610         https://bugs.webkit.org/show_bug.cgi?id=91450
3611
3612         Reviewed by Ojan Vafai.
3613
3614         More refactoring ... this moves the 'config' output to
3615         printing.py.
3616
3617         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3618         (Manager._resize_shards):
3619         (Manager._run_tests):
3620         (Manager._upload_json_files):
3621         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3622         (ManagerTest.get_options):
3623         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3624         (run):
3625         * Scripts/webkitpy/layout_tests/views/printing.py:
3626         (Printer.print_config):
3627         (Printer.print_workers_and_shards):
3628         (Printer._print_config):
3629         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
3630         (Testprinter.test_fallback_path_in_config):
3631         (Testprinter.test_print_config):
3632
3633 2012-07-17  Dirk Pranke  <dpranke@chromium.org>
3634
3635         nrwt: clean up logging when we're only running one chunk of the tests
3636         https://bugs.webkit.org/show_bug.cgi?id=91447
3637