2c6ca8839b6d30626b9677efd1a9f793190f3d25
[WebKit-https.git] / Tools / ChangeLog
1 2011-07-15  Adam Barth  <abarth@webkit.org>
2
3         NRWT stores the Chromium revision number in full_results.json
4         https://bugs.webkit.org/show_bug.cgi?id=64586
5
6         I have no earthly idea how to test this change.
7
8         * Scripts/webkitpy/layout_tests/controllers/manager.py:
9         * Scripts/webkitpy/layout_tests/port/base.py:
10
11 2011-07-14  Ojan Vafai  <ojan@chromium.org>
12
13         fix remaining style issues in the static-dashboards directory
14         https://bugs.webkit.org/show_bug.cgi?id=64561
15
16         Reviewed by Adam Barth.
17
18         * TestResultServer/static-dashboards/aggregate_results.html:
19         * TestResultServer/static-dashboards/builders.js:
20         * TestResultServer/static-dashboards/dashboard_base.js:
21         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
22         * TestResultServer/static-dashboards/timeline_explorer.html:
23         * TestResultServer/static-dashboards/treemap.html:
24
25 2011-07-14  Yuta Kitamura  <yutak@chromium.org>
26
27         WebSocket: Introduce pywebsocket-0.6b2
28         https://bugs.webkit.org/show_bug.cgi?id=64534
29
30         Reviewed by Kent Tamura.
31
32         * Scripts/webkitpy/thirdparty/__init__.py:
33
34 2011-07-14  Dimitri Glazkov  <dglazkov@chromium.org>
35
36         Introduce TestExpectationsModel, split out of TestExpectations.
37         https://bugs.webkit.org/show_bug.cgi?id=64531
38
39         This is a simple split-and-make-work refactoring, a first step among many.
40
41         Reviewed by Adam Barth.
42
43         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Moved all model-related members
44             out of TestExpectations and into TestExpectationsModel.
45
46 2011-07-14  Eric Seidel  <eric@webkit.org>
47
48         Move webkitpy off of loose mocks
49         https://bugs.webkit.org/show_bug.cgi?id=64508
50
51         Unreviewed.  Fixing a test which fails under test-webkitpy --all (but not in a normal run).
52
53         * Scripts/webkitpy/common/checkout/scm/git.py:
54
55 2011-07-14  Adam Barth  <abarth@webkit.org>
56
57         garden-o-matic should have a "rebaseline" button
58         https://bugs.webkit.org/show_bug.cgi?id=64446
59
60         Reviewed by Ojan Vafai.
61
62         This patch adds a basic Rebaseline button that copies the baselines
63         displayed in the results pane into the appropriate directory in your
64         working copy.
65
66         There are two main limitations:
67
68         1) There is no UI for actually committing the baselines.
69
70         2) The baselines are not optimized for redundancy (meaning you can have
71            identical baselines in both chromium-mac and chromium-win).
72
73         * Scripts/webkitpy/tool/commands/rebaseline.py:
74             - Turns out we need to create the directory for the baseline if it doesn't exist yet.
75         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js:
76             - Add an programatic API to call the server.
77         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
78             - Add the rebaseline button itself.
79         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
80             - Change the CSS so that the Rebaseline and Close buttons can
81               appear in the normal order in the DOM.
82         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
83             - Bind the event and translate the parameters.
84             - Hide/show the rebaseline button, as appropriate.
85         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
86             - Add some helpful utility functions for manipulating failure types.
87         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
88
89 2011-07-14  Ryosuke Niwa  <rniwa@webkit.org>
90
91         REGRESSION: webkit-patch roll-chromium-deps is broken
92         https://bugs.webkit.org/show_bug.cgi?id=64568
93
94         Reviewed by Dirk Pranke
95
96         Use _filesystem.join.
97
98         * Scripts/webkitpy/common/checkout/checkout.py:
99         * Scripts/webkitpy/common/checkout/checkout_unittests.py:
100
101 2011-07-14  Noam Rosenthal  <noam.rosenthal@nokia.com>
102
103         Adding myself to the reviewers list; No review needed.
104
105         * Scripts/webkitpy/common/config/committers.py:
106
107 2011-07-14  Adam Roben  <aroben@apple.com>
108
109         Don't use Element.prototype.classList in TestFailures
110
111         Safari 5 doesn't support it.
112
113         Fixes <http://webkit.org/b/64550> Can't expand flaky tests on TestFailures page in Safari 5
114
115         Reviewed by Daniel Bates.
116
117         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
118         (Element.prototype.hasStyleClass):
119         (Element.prototype.addStyleClass):
120         (Element.prototype.removeStyleClass):
121         (Element.prototype.toggleStyleClass):
122         Added these helper functions which simulate classList functionality.
123
124         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities_unittests.js:
125         Added. Tests for the above.
126
127         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
128         (ViewController.prototype._domForPossiblyFlakyTests): Changed to use
129         toggleStyleClass/hasStyleClass instead of classList.
130
131         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
132         Added Utilities_unittests.js, and reordered the tested files to be in
133         roughly dependency order (i.e., the lowest-level files are imported
134         and tested first).
135
136 2011-07-14  Eric Seidel  <eric@webkit.org>
137
138         NRWT doesn't store the svn revision in full_results.json on chromium-win
139         https://bugs.webkit.org/show_bug.cgi?id=64492
140
141         Unreviewed.  Just fixing my test-webkitpy regression.
142
143         Fix the unit tests.  Unfortunately scm does not use a filesystem
144         object so we can't control the result of detect_scm_system.
145         When detect_scm_system would fail, we would log, which would
146         cause all passing_run integration tests to fail.
147
148         * Scripts/webkitpy/common/checkout/scm/scm.py:
149         * Scripts/webkitpy/layout_tests/controllers/manager.py:
150         * Scripts/webkitpy/layout_tests/port/base.py:
151         * Scripts/webkitpy/layout_tests/port/test.py:
152         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
153
154 2011-07-14  Eric Seidel  <eric@webkit.org>
155
156         NRWT doesn't store the svn revision in full_results.json on chromium-win
157         https://bugs.webkit.org/show_bug.cgi?id=64492
158
159         Reviewed by Ojan Vafai.
160
161         This should fix the bug.  Unfortunately this code is currently
162         impossible to test since it's impossible to mock detect_scm_system at the
163         moment.  I started re-writing scm.detection.py to be mockable, but decided that
164         was best left for another day.
165
166         * Scripts/webkitpy/layout_tests/controllers/manager.py:
167         * Scripts/webkitpy/layout_tests/port/base.py:
168
169 2011-07-14  Ojan Vafai  <ojan@chromium.org>
170
171         fix coding style of dashboard_base.js
172         https://bugs.webkit.org/show_bug.cgi?id=64545
173
174         Reviewed by Adam Roben.
175
176         The code changes to the files other than dashboard_base.js
177         are just fallout from renaming globals in dashboard_base.js.
178
179         * TestResultServer/static-dashboards/aggregate_results.html:
180         * TestResultServer/static-dashboards/builders.js:
181         * TestResultServer/static-dashboards/dashboard_base.js:
182         * TestResultServer/static-dashboards/flakiness_dashboard.html:
183         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
184         * TestResultServer/static-dashboards/timeline_explorer.html:
185         * TestResultServer/static-dashboards/treemap.html:
186
187 2011-07-14  Carlos Garcia Campos  <cgarcia@igalia.com>
188
189         Unreviewed. Fix WebKit2 GTK build after r90953.
190
191         * WebKitTestRunner/GNUmakefile.am:
192
193 2011-07-14  Adam Barth  <abarth@webkit.org>
194
195         garden-o-matic results pane should be more discoverable
196         https://bugs.webkit.org/show_bug.cgi?id=64513
197
198         Reviewed by Eric Seidel.
199
200         This patch causes us to trigger the results pane on mouse clicks
201         instead of mousein.  Also, handle the case of no results details more
202         elegantly.
203
204         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
205         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
206             - Change the binding to "click".
207         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
208             - Handle the "no results URLs" case explicitly instead of never
209               calling the callback.
210         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
211         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
212             - Show a message when there are no result details to show.
213         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
214
215 2011-07-14  Adam Barth  <abarth@webkit.org>
216
217         garden-o-matic should badge TIMEOUT failures
218         https://bugs.webkit.org/show_bug.cgi?id=64435
219
220         Reviewed by Eric Seidel.
221
222         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
223         (.regression .what a[draggable].TIMEOUT::before):
224             - Add TIMEOUT.
225
226 2011-07-14  Adam Barth  <abarth@webkit.org>
227
228         Fix Python exception blocking the commit-queue.
229
230         * Scripts/webkitpy/common/checkout/scm/git.py:
231
232 2011-07-13  Eric Seidel  <eric@webkit.org>
233
234         Move webkitpy off of loose mocks
235         https://bugs.webkit.org/show_bug.cgi?id=64508
236
237         Reviewed by Adam Barth.
238
239         Using Mock has caused us more pain than help.
240         It's possible that there was a cleaner way to use it
241         (maybe Mock(class) instead of inheriting from it?).
242         But for now, I've removed all uses of Mock from mocktool.py.
243
244         I also moved run_command into the only 3 files which call it
245         instead of leaving the deprecated method in executive.py.
246
247         I changed various direct calls to os.* to use filesystem instead.
248
249         * Scripts/webkitpy/common/checkout/checkout.py:
250         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
251         * Scripts/webkitpy/common/checkout/scm/git.py:
252         * Scripts/webkitpy/common/checkout/scm/scm.py:
253         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
254         * Scripts/webkitpy/common/checkout/scm/svn.py:
255         * Scripts/webkitpy/common/system/executive.py:
256         * Scripts/webkitpy/common/system/executive_unittest.py:
257         * Scripts/webkitpy/tool/mocktool.py:
258         * Scripts/webkitpy/tool/steps/cleanworkingdirectory_unittest.py:
259         * Scripts/webkitpy/tool/steps/ensurelocalcommitifneeded.py:
260
261 2011-07-13  Eric Seidel  <eric@webkit.org>
262
263         NRWT doesn't store the svn revision in full_results.json on chromium-win
264         https://bugs.webkit.org/show_bug.cgi?id=64492
265
266         Reviewed by Adam Barth.
267
268         Add a warning message in the revision='' case to help us diagnose
269         what's going wrong on the bot.
270
271         * Scripts/webkitpy/layout_tests/controllers/manager.py:
272
273 2011-07-13  Dirk Pranke  <dpranke@chromium.org>
274
275         new-run-webkit-tests: print baseline search path as part of config output
276         https://bugs.webkit.org/show_bug.cgi?id=64499
277
278         Reviewed by Eric Seidel.
279
280         * Scripts/webkitpy/layout_tests/controllers/manager.py:
281         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
282
283 2011-07-13  Adam Barth  <abarth@webkit.org>
284
285         gardening server should have an API for parsing changelogs
286         https://bugs.webkit.org/show_bug.cgi?id=64495
287
288         Reviewed by Eric Seidel.
289
290         This patch exposes much of the same information from CommitInfo in a
291         dictionary form, which is easier to send over-the-wire as JSON to the
292         frontend.
293
294         * Scripts/webkitpy/common/checkout/checkout.py:
295         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
296         * Scripts/webkitpy/tool/mocktool.py:
297         * Scripts/webkitpy/tool/servers/gardeningserver.py:
298         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
299         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
300
301 2011-07-13  Eric Seidel  <eric@webkit.org>
302
303         REGRESSION: GitTestWithMock.test_create_patch fails
304         https://bugs.webkit.org/show_bug.cgi?id=62945
305
306         Reviewed by Daniel Bates.
307
308         I was not able to reproduce the exact failure seen in the bug,
309         however this test was failing on my machine for other reasons.
310
311         I went through and did an audit of our run_command usage, it's
312         entirely in scm classes after this change.  (Not surprising given
313         that scm.py was the second file ever created in webkit.py.)
314
315         The real bug I'm fixing here is that we were setting executive.should_log
316         when the value had changed to executive._should_log.  Now we set the right one
317         and the test works again.
318
319         * Scripts/webkitpy/common/checkout/checkout.py:
320         * Scripts/webkitpy/common/checkout/scm/git.py:
321         * Scripts/webkitpy/common/checkout/scm/scm.py:
322         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
323         * Scripts/webkitpy/common/checkout/scm/svn.py:
324
325 2011-07-13  Ilya Sherman  <isherman@chromium.org>
326
327         Fix WTF header guard style check
328         https://bugs.webkit.org/show_bug.cgi?id=64488
329
330         Reviewed by David Levin.
331
332         * Scripts/webkitpy/style/checkers/cpp.py:
333         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Verify that we don't *always* suggest "WTF_" as a prefix
334
335 2011-07-13  Ojan Vafai  <ojan@chromium.org>
336
337         bring flakiness_dashboard.html closer to webkit style
338         https://bugs.webkit.org/show_bug.cgi?id=64477
339
340         Reviewed by Adam Barth.
341
342         * TestResultServer/static-dashboards/dashboard_base.js:
343         Fix bad variable name that would hit only when running the tests.
344         * TestResultServer/static-dashboards/flakiness_dashboard.html:
345         No code/logic changes. All moving/removing brackets, indents and moving things
346         to one line.
347
348 2011-07-13  Xan Lopez  <xlopez@igalia.com>
349
350         [GTK] Fix distcheck
351
352         Reviewed by Martin Robinson.
353
354         * WebKitTestRunner/GNUmakefile.am: mark built sources as nodist.
355
356 2011-07-13  Joseph Pecoraro  <joepeck@webkit.org>
357
358         Unreviewed fix. Add a "\n" to a printf that somehow got lost.
359
360         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
361         (WebCore::DumpRenderTree::dumpApplicationCacheQuota):
362
363 2011-07-13  Joseph Pecoraro  <joepeck@webkit.org>
364
365         ApplicationCache Quota Output is Flakey
366         https://bugs.webkit.org/show_bug.cgi?id=64410
367
368         Reviewed by Alexey Proskuryakov.
369
370         Unify the delegate logging for reaching application cache
371         quotas. Truncate the space needed to the nearest 10000
372         for less flakey test output. This also better supports
373         printing NSUInteger on non-64 bit systems by casting to
374         an unsigned long.
375
376         * DumpRenderTree/mac/UIDelegate.mm:
377         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:totalSpaceNeeded:]):
378         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
379         (WebCore::DumpRenderTree::dumpApplicationCacheQuota):
380
381 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
382
383         Remove unused TestExpectations._overrides.
384         https://bugs.webkit.org/show_bug.cgi?id=64470
385
386         Reviewed by Adam Barth.
387
388         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Removed unused member.
389
390 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
391
392         Introduce TestExpectationSerializer.list_to_string.
393         https://bugs.webkit.org/show_bug.cgi?id=64462
394
395         Reviewed by Adam Barth.
396
397         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Added list_to_string and change the relevant callsite to use it.
398         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added tests for it.
399
400 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
401
402         Consolidate expectations parsing code.
403         https://bugs.webkit.org/show_bug.cgi?id=64460
404
405         Reviewed by Adam Barth.
406
407         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Folded TestExpectationParser._split_expectation_string into TestExpectationParser.parse.
408
409 2011-07-12  Brent Fulgham  <bfulgham@webkit.org>
410
411         Standardize WinCairo conditionalized code under PLATFORM macro.
412         https://bugs.webkit.org/show_bug.cgi?id=64377
413
414         Reviewed by Maciej Stachowiak.
415
416         Update compile-time conditions that had previously used #ifdef
417         of WIN_CAIRO to consistently use PLATFORM(WIN_CAIRO).
418
419         * DumpRenderTree/config.h: Switch to PLATFORM(WIN_CAIRO)
420         * DumpRenderTree/win/DumpRenderTreeCairo.vsprops: Remove redundant
421           define of WIN_CAIRO=1.
422         * DumpRenderTree/win/DumpRenderTreeDebugCairoCFLite.vsprops: Add
423           WinCairo.vsprops to set (to match Release build).
424
425 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
426
427         Eliminate TestExpectationsFile.
428         https://bugs.webkit.org/show_bug.cgi?id=64458
429
430         Turns out, we can just use a Python list.
431
432         Reviewed by Adam Barth.
433
434         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Folded TestExpectationsFile.append into TestExpectationParser.parse_list,
435             removed TestExpectationsFile.
436         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Moved tests to better reflect new names, removed iterator test,
437             since there's no more custom iterator machinery.
438
439 2011-07-13  Philippe Normand  <pnormand@igalia.com>
440
441         Unreviewed, added my other email addresses.
442
443         * Scripts/webkitpy/common/config/committers.py:
444
445 2011-07-13  Adam Barth  <abarth@webkit.org>
446
447         Folks should only be listed once.
448
449         * Scripts/webkitpy/common/config/committers.py:
450
451 2011-07-13  Philippe Normand  <pnormand@igalia.com>
452
453         Unreviewed, adding myself as Reviewer.
454
455         * Scripts/webkitpy/common/config/committers.py:
456
457 2011-07-13  Adam Barth  <abarth@webkit.org>
458
459         Fix flickering bug introduced by my previous patch.  I forgot to change
460         the name of the class everywhere.
461
462         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
463
464 2011-07-12  Philippe Normand  <pnormand@igalia.com>
465
466         [GTK] media/media-blocked-by-willsendrequest.html fails
467         https://bugs.webkit.org/show_bug.cgi?id=63699
468
469         Reviewed by Martin Robinson.
470
471         * DumpRenderTree/gtk/DumpRenderTree.cpp:
472         (willSendRequestCallback): Abort the request if explicitely asked
473         by the LayoutTestController's willSendRequestReturnsNull() function.
474
475 2011-07-13  Adam Barth  <abarth@webkit.org>
476
477         Remove "Dismiss" button from garden-o-matic butterbar
478         https://bugs.webkit.org/show_bug.cgi?id=64443
479
480         Reviewed by Dimitri Glazkov.
481
482         We don't have any persistent butter bar messages yet, so the dismiss
483         button is premature (and fairly heavy, visually).
484
485         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
486         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
487
488 2011-07-13  Adam Barth  <abarth@webkit.org>
489
490         Refine garden-o-matic status pane
491         https://bugs.webkit.org/show_bug.cgi?id=64442
492
493         Reviewed by Dimitri Glazkov.
494
495         This patch makes two improvements to the status pane:
496
497         1) We only query the server for failure types that we're expecting.
498            This dramatically reduces the number of HTTP requests, making
499            loading the status pane faster.
500
501         2) The status pane now displays which test and which builder it is
502            showing results for.
503
504         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js:
505         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
506         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
507         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
508         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
509         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
510         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
511
512 2011-07-13  Adam Barth  <abarth@webkit.org>
513
514         garden-o-matic should show test results
515         https://bugs.webkit.org/show_bug.cgi?id=64440
516
517         Reviewed by Adam Roben.
518
519         This patch introduces the results detail pane, which appears at the
520         bottom of the window and contains test results from the bots.
521         Currently, you can activate the pane by mousing over one of the builder
522         names associated with a failing test.
523
524         This is just a first iteration of the UI.  There's no way to resize or
525         zoom in on elements of the details pane, and images likely aren't sized
526         correctly, but it's a place to start.
527
528         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js:
529             - Add constants for our data attributes so we don't typo them!
530         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
531             - Add DOM structure for the details pane.
532         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
533         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
534             - Bind events for showing and hiding the details pane.
535         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
536             - Reduce the number of result types that we query for to avoid
537               overfilling the details pane.  We'll probably need another
538               solution here in the long-term.  For example, we could use tabs
539               to pack more results into the pane.
540         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
541         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
542             - Now that we're using the config package during testing, we need
543               to include it in the testing HTML.
544         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
545         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
546             - Change the DOM structure of the results detail to make it fit
547               nicely in the details pane (rather than flowing freely in the
548               body, as it did before).
549
550 2011-07-13  Adam Roben  <aroben@apple.com>
551
552         Make TestFailures show every time a possibly-flaky test failed, but hide it by default
553
554         It's useful to be able to see every time a flaky test failed to see whether it failed the
555         same way every time. But doing so takes a lot of space, so the list of failures is now
556         collapsed by default and can be revealed using a disclosure triangle.
557
558         Fixes <http://webkit.org/b/64455> TestFailures page doesn't show as much information for
559         flaky tests as I would like, even though the page is already so long
560
561         Reviewed by Dimitri Glazkov.
562
563         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
564         (FlakyLayoutTestDetector.prototype.allFailures): Replaced flakinessExamples with this
565         function. Now returns all failures for the given test.
566
567         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector_unittests.js:
568         Added. This just contains some simple tests of the FlakyLayoutTestDetector class. We'll add
569         more over time.
570
571         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
572         (LayoutTestHistoryAnalyzer.prototype.start): Updated the documentation comment to reflect
573         that we no longer return passing builds for possibly-flaky tests.
574
575         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
576         (.existing-bugs-list, .suspect-revisions-list, .flakiness-examples-list): Make the list of
577         flakiness examples small, too, since it can get quite long.
578
579         (.disclosure-triangle):
580         (.expanded > .disclosure-triangle):
581         Simple styles for the disclosure triangle.
582
583         (.flakiness-examples-list): Collapse the list by default.
584
585         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
586         (ViewController.prototype._displayBuilder): Pass the total number of builds analyzed to
587         _domForPossiblyFlakyTests.
588         (ViewController.prototype._domForPossiblyFlakyTests): Put a disclosure triangle to the left
589         of each test name, and the number of failures to the right. When the disclosure triangle is
590         clicked for the first time, we build up the list of failures and expand the element. After
591         that we just collapse or expand the element on subsequent clicks.
592
593         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
594         Pulled in new tests.
595
596 2011-07-13  Adam Roben  <aroben@apple.com>
597         
598         Teach TestFailures to understand NRWT's output when it exits early due to too many failures
599
600         Fixes <http://webkit.org/b/64456> TestFailures page reports way too many failures when NRWT
601         exits early
602
603         Reviewed by Dimitri Glazkov.
604
605         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
606         (Builder.prototype.getNumberOfFailingTests): Relaxed the "Exiting early" test not to require
607         it to be at the beginning of the line, since NRWT prints a bunch of junk earlier in the
608         line. Tightened up the regex that's used to parse the number of failing tests to require the
609         leading number to be followed by whitespace so that we won't parse the "2011" in
610         "2011-07-13" as a number of failures.
611
612         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
613         Added a test for the above.
614
615 2011-07-12  Dimitri Glazkov  <dglazkov@chromium.org>
616
617         Extract model-like TestExpectationLine and TestExpectationFile from TestExpectations.
618         https://bugs.webkit.org/show_bug.cgi?id=64386
619
620         This is the first step in converting TestExpectations to a real model.
621         * TestExpectationsLine represents a line in the test_expectations.txt file, and
622         * TestExpectationsFile represents the file, which is a collection of lines.
623
624         Reviewed by Adam Barth.
625
626         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
627         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
628         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
629
630 2011-07-13  Xan Lopez  <xlopez@igalia.com>
631
632         [GTK] Do not grab focus too early in DRT.
633
634         Reviewed by Gustavo Noronha.
635
636         It causes a layout to happen and a progress signal to be emitted
637         since r90900, but at this point we don't have a
638         LayoutTestController object and we'll eventually crash. Since we
639         already grab focus at the beginning of runTest() this is
640         redundant, so get rid of it to fix the crash.
641
642         * DumpRenderTree/gtk/DumpRenderTree.cpp:
643         (main): remove call to grab_focus
644
645 2011-07-13  Sheriff Bot  <webkit.review.bot@gmail.com>
646
647         Unreviewed, rolling out r90893 and r90894.
648         http://trac.webkit.org/changeset/90893
649         http://trac.webkit.org/changeset/90894
650         https://bugs.webkit.org/show_bug.cgi?id=64441
651
652         NRWT still doesn't work on qt-mac platform (Requested by Ossy
653         on #webkit).
654
655         * Scripts/run-webkit-tests:
656         (useNewRunWebKitTests):
657
658 2011-07-13  Csaba Osztrogonác  <ossy@webkit.org>
659
660         [Qt] NRWT should pick up the right httpd config file
661         https://bugs.webkit.org/show_bug.cgi?id=64086
662
663         * Scripts/run-webkit-tests: Enable NRWT on qt-mac platform after r90810.
664         (useNewRunWebKitTests):
665
666 2011-07-12  Mark Rowe  <mrowe@apple.com>
667
668         Fix the 32-bit build.
669
670         * DumpRenderTree/mac/UIDelegate.mm:
671         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:totalSpaceNeeded:]):
672         Cast the NSUInteger value to unsigned long to match the format specifier.
673
674 2011-07-12  Adam Barth  <abarth@webkit.org>
675
676         Improve garden-o-matic UI when the bots fail to report revision numbers
677         https://bugs.webkit.org/show_bug.cgi?id=64427
678
679         Reviewed by Dimitri Glazkov.
680
681         At least Win (dbg)(2) seems to fail to report the SVN revision number
682         in full_results.json.  This patch makes garden-o-matic more robust to
683         missing revision numbers.
684
685         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
686         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
687         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
688         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
689         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
690
691 2011-07-12  Adam Barth  <abarth@webkit.org>
692
693         Adjust garden-o-matic layout to use fewer lines
694         https://bugs.webkit.org/show_bug.cgi?id=64422
695
696         Reviewed by Ojan Vafai.
697
698         This patch effectively merges the test name line with the list of
699         builders on which the test fails.  Each test failure now occupies two
700         lines instead of three.
701
702         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
703         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
704         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
705
706 2011-07-12  Adam Barth  <abarth@webkit.org>
707
708         garden-o-matic should label tests failures of new tests as such
709         https://bugs.webkit.org/show_bug.cgi?id=64421
710
711         Reviewed by Ojan Vafai.
712
713         Also, we shouldn't dim failures of new tests, even if we've only seen
714         them once because they're likely to be real problems that need
715         attention.
716
717         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
718         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js: Added.
719         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout_unittests.js: Added.
720         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
721         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
722         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
723         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
724         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
725         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
726         * Scripts/webkitpy/tool/servers/gardeningserver.py:
727
728 2011-07-12  Adam Barth  <abarth@webkit.org>
729
730         garden-o-matic should display how many times we've seen a failure
731         https://bugs.webkit.org/show_bug.cgi?id=64417
732
733         Reviewed by Ojan Vafai.
734
735         This patch adds some UI to display how many times we've seen a given
736         failure, which can be helpful for determining whether that failure is a
737         real failure or a flaky test.
738
739         When a failure has only been seen once (i.e., only a single run on a
740         single bot), we set the opacity of to 50% to avoid distracting the
741         gardener.
742
743         This patch also refactors the failure walker to have a simpler API
744         internally by moving from an object-oriented paradigm to a functional
745         paradigm.
746
747         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
748         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
749         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
750         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
751         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
752         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
753
754 2011-07-12  Chris Rogers  <crogers@google.com>
755
756         Enable Web Audio for chromium DRT
757         https://bugs.webkit.org/show_bug.cgi?id=64409
758
759         Reviewed by James Robinson.
760
761         * DumpRenderTree/chromium/TestShell.cpp:
762         (TestShell::TestShell):
763
764 2011-07-12  Adam Barth  <abarth@webkit.org>
765
766         Tweak some UI in garden-o-matic now that the tree actually has a
767         failure and I can see what this all looks like.
768
769         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
770         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
771         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
772
773 2011-07-12  Joseph Pecoraro  <joepeck@webkit.org>
774
775         Unreviewed. Skipping a few tests which fail due to differing output
776         in recent ApplicationCache quota tests. Reenabling the tests is
777         tracked by <http://webkit.org/b/64410>.
778
779         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
780         (WebCore::DumpRenderTree::dumpApplicationCacheQuota): This should
781         have been a space, not a newline.
782
783 2011-07-12  Adam Barth  <abarth@webkit.org>
784
785         garden-o-matic should display regression ranges
786         https://bugs.webkit.org/show_bug.cgi?id=64407
787
788         Reviewed by Dimitri Glazkov.
789
790         This patch computes a regression range for a failure by intersecting
791         the regression ranges seen by the various bots.  We make the underlying
792         assumption that a test is only failing due to one revision at any given
793         moment.  If that's not true, this code probably explodes.
794
795         The regression ranges appear asynchronously, which might be a jarring
796         UI.  We'll have to experiment to see.
797
798         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
799         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
800         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
801         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
802         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
803         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
804         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
805
806 2011-07-12  Adam Barth  <abarth@webkit.org>
807
808         cr-linux-ews should run pixel tests
809         https://bugs.webkit.org/show_bug.cgi?id=64394
810
811         Reviewed by Ojan Vafai.
812
813         This patch enabled pixel test on the cr-linux-ews.  Running a quick
814         test on the bots, it looks like we have five pixel failures, related to
815         some missing fonts.  I've also made the bots skip failing tests to
816         avoid uploading huge zip files to bugs.webkit.org (with all the exected
817         IMAGE failures).
818
819         * Scripts/webkitpy/common/config/ports.py:
820         * Scripts/webkitpy/common/config/ports_unittest.py:
821
822 2011-07-12  Joseph Pecoraro  <joepeck@webkit.org>
823
824         ApplicationCache update should not immediately fail when reaching per-origin quota
825         https://bugs.webkit.org/show_bug.cgi?id=64177
826
827         Reviewed by Alexey Proskuryakov.
828
829         Add layoutTestController.disallowIncreaseForApplicationCacheQuota()
830         to disable the default behavior of raising the default per-origin
831         quota to 5MB when reached. This allows us to test what happens
832         when the quota is reached and not increased.
833
834         * DumpRenderTree/LayoutTestController.cpp:
835         (LayoutTestController::LayoutTestController):
836         (disallowIncreaseForApplicationCacheQuotaCallback):
837         (LayoutTestController::staticFunctions):
838         * DumpRenderTree/LayoutTestController.h:
839         (LayoutTestController::disallowIncreaseForApplicationCacheQuota):
840         (LayoutTestController::setDisallowIncreaseForApplicationCacheQuota):
841         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
842         (LayoutTestController::reset):
843         * DumpRenderTree/qt/LayoutTestControllerQt.h:
844         (LayoutTestController::disallowIncreaseForApplicationCacheQuota):
845         (LayoutTestController::dumpApplicationCacheDelegateCallbacks):
846         Expose a function to prevent automatically increasing the per-origin
847         quota when the quota is reached. This allows us to test what happens
848         when a user would "disallow" a quota increase.
849
850         * DumpRenderTree/mac/UIDelegate.mm:
851         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:spaceNeeded:]):
852         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
853         (WebCore::DumpRenderTree::dumpApplicationCacheQuota):
854         Respect the new disallowIncrease flag and dump the spaceNeeded values
855         for ports that implement application cache quotas.
856
857 2011-07-12  Ojan Vafai  <ojan@chromium.org>
858
859         fix flakiness dashboard to work with new crash log filenames
860         https://bugs.webkit.org/show_bug.cgi?id=64393
861
862         Reviewed by Adam Barth.
863
864         * TestResultServer/static-dashboards/flakiness_dashboard.html:
865
866 2011-07-12  Adam Barth  <abarth@webkit.org>
867
868         Add a happy animation to garden-o-matic when there are no failures
869         https://bugs.webkit.org/show_bug.cgi?id=64382
870
871         Reviewed by Ojan Vafai.
872
873         As requested by Ojan.
874
875         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
876         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
877         * Scripts/webkitpy/tool/servers/data/gardeningserver/partytime.gif: Added.
878         * Scripts/webkitpy/tool/servers/gardeningserver.py:
879
880 2011-07-12  Adam Barth  <abarth@webkit.org>
881
882         sheriffbot can't perform rollouts
883         https://bugs.webkit.org/show_bug.cgi?id=64370
884
885         Reviewed by Eric Seidel.
886
887         Now that we're not updating the working copy before processing every
888         IRC message, the rollout command needs to update it manually.
889
890         * Scripts/webkitpy/tool/bot/irc_command.py:
891         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
892
893 2011-07-12  Adam Barth  <abarth@webkit.org>
894
895         Fix the grammar in the FIXME comment.
896
897         * Scripts/webkitpy/common/net/resultsjsonparser.py:
898
899 2011-07-12  Eric Seidel  <eric@webkit.org>
900
901         NRWT should open test results page with Safari trunk, not the system provided one on Mac
902         https://bugs.webkit.org/show_bug.cgi?id=64346
903
904         Reviewed by Adam Barth.
905
906         To fix this I implemented Port.show_results_html_file in Mac, Gtk and Qt ports with
907         implementations (mostly) matching those found in old-run-webkit-tests.
908         There are still some minor differences for Qt which Qt hackers may wish to tweak.
909
910         I had to add a WebKitPort._port_flag_for_scripts method (similar to flag() in common.config.ports.py)
911         for the Qt/Gtk ports which always require a flag passed to scripts.
912
913         While trying to test this, I found that FactoryTest.test_chromium_gpu_linux
914         was using a real filesystem (due to assert_platform) which due to
915         global variables in config.py was causing set-webkit-configuration to have
916         an affect on unit test results!  So I fixed this by making FactoryTest.assert_port
917         pass a mock file system whenever calling factory.get.
918
919         Unfortunately TestPort was depending on always being passed a None filesystem
920         and asserting filesystem._tests (only true for unit_test_filesystem()).
921         So I just removed the FactoryTest.test_test and FactoryTest.test_dryrun tests
922         deciding that they were pretty much useless anyway.  If others feel strongly
923         I'm happy to fix this in a different way.
924
925         * Scripts/webkitpy/common/system/executive.py:
926          - default arguments in python are screwy.  They use a single shared
927            instance, so it's better to use argument=None and then argument = argument or Default()
928            if you have any chance of mutating (or returning) the default argument.
929         * Scripts/webkitpy/layout_tests/port/config.py:
930          - This code is wrong.  We don't need to use a global variable here (as far as I can tell).
931            I'm not fixing it in this patch, but I've marked it with a FIXME and we can convert to
932            storing the results of the read on the Config object (which should only be created once during normal operation).
933            Unit tests shouldn't be hitting the disk anyway.  It's possible Config should move off of Port and onto Tool/Host directly.
934         * Scripts/webkitpy/layout_tests/port/factory.py:
935         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
936         * Scripts/webkitpy/layout_tests/port/gtk.py:
937         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py: Added.
938         * Scripts/webkitpy/layout_tests/port/mac.py:
939         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
940         * Scripts/webkitpy/layout_tests/port/qt.py:
941         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
942         * Scripts/webkitpy/layout_tests/port/webkit.py:
943
944 2011-07-12  Adam Roben  <aroben@apple.com>
945
946         Teach TestFailures to recognize when run-webkit-tests gets killed by buildbot
947
948         Fixes <http://webkit.org/b/64358> TestFailures page thinks all tests passed in
949         http://build.webkit.org/builders/Windows%207%20Release%20(Tests)/builds/14672
950
951         Reviewed by Daniel Bates.
952
953         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
954         (Builder.prototype.getNumberOfFailingTests): If run-webkit-tests exited with a non-zero
955         exit status but we didn't find any failure counts, assume that there was some error that
956         caused run-webkit-tests to die early (like being killed by buildbot due to a timeout).
957
958         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
959         Added a new test that shows that we get a failingTestCount of -1 when run-webkit-tests dies
960         early.
961         (runGetNumberOfFailingTestsTest): Moved most code here from the only pre-existing test in this
962         file.
963
964         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
965         (LayoutTestResultsLoader.prototype.start): Bump the cache number so old cached data that was
966         tainted by the bug fixed in this patch will be evicted.
967
968 2011-07-12  Adam Barth  <abarth@webkit.org>
969
970         cr-linux-ews complains about tests that aren't actually failing
971         https://bugs.webkit.org/show_bug.cgi?id=64367
972
973         The underlying problem here is that full_results.json doesn't have
974         enough information to determine whether a given test result was
975         expected because whether an actual result is expected depends on
976         whether full_results.json was generated duing a test run that included
977         pixel tests.
978
979         The right long-term solution is to make full_results.json a complete
980         description of what happened durning a test run.  In this patch
981         however, to stop the spam, this patch makes the jsonresultsparser
982         ignore pixel failures.
983
984         (I'm landing this patch unreviewed in order to stop the bots from
985         spamming.  I'll happily address any review comments in a follow-up
986         patch.)
987
988         * Scripts/webkitpy/common/net/resultsjsonparser.py:
989         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
990
991 2011-07-12  Adam Barth  <abarth@webkit.org>
992
993         Re-work garden-o-matic UI to begin to look like the mocks
994         https://bugs.webkit.org/show_bug.cgi?id=64334
995
996         Reviewed by Dimitri Glazkov.
997
998         This patch changes how we display failures to be more like the mock the
999         dglazkov made. We now can display the same information much more
1000         compactly.
1001
1002         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
1003         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
1004         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
1005         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css: Added.
1006         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
1007         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
1008         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
1009         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
1010         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
1011         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1012
1013 2011-07-11  Adam Roben  <aroben@apple.com>
1014
1015         Extract TestFailures's bug-filing code into two new classes
1016
1017         Fixes <http://webkit.org/b/64300> TestFailures page's new-bug-filing code is a mess!
1018
1019         Reviewed by Darin Adler and Adam Barth.
1020
1021         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js:
1022         (Buildbot.prototype.resultsDirectoryURL): Changed to return a URI-encoded URL. Otherwise the
1023         URL isn't valid (and it's harder to mock this function).
1024
1025         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot_unittests.js:
1026         Added. Just tests the above change (for now).
1027
1028         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
1029         Added a license header, enclosed everything in a closure to avoid polluting the global
1030         namespace, changed the test name to actually describe the passing condition, and replaced
1031         uses of equals() with equal(). (The latter is the actual name of the function, and matches
1032         deepEqual, etc.)
1033
1034         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm.js: Added.
1035         (NewBugForm): This class knows how to construct a <form> used for filing new bugs in
1036         Bugzilla based on some parameters.
1037         (NewBugForm.prototype.domElement): Creates and returns the <form> element. Code came from
1038         ViewController.prototype._domForNewAndExistingBugs.
1039
1040         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm_unittests.js:
1041         Added. Tests the above code.
1042
1043         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm.js: Added.
1044         (TestFailureBugForm): This class knows how to construct a <form> element used for filing new
1045         bugs specifically about test failures. Code came from
1046         ViewController.prototype._domForNewAndExistingBugs.
1047         (TestFailureBugForm.prototype.domElement): Slightly customizes the <form> element returned
1048         by NewBugForm.
1049         (TestFailureBugForm.prototype._computeOperatingSystem):
1050         (TestFailureBugForm.prototype._computePlatform):
1051         (TestFailureBugForm.prototype._createBugTitle):
1052         (TestFailureBugForm.prototype._failingResultsHTMLURL):
1053         (TestFailureBugForm.prototype._failingRevision):
1054         (TestFailureBugForm.prototype._passingRevision):
1055         (TestFailureBugForm.prototype._regressionRangeString):
1056         Code came from ViewController.prototype._domForNewAndExistingBugs. I broke it out into
1057         separate functions to break up the rat's nest a bit.
1058
1059         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm_unittests.js:
1060         Added. Tests the above code.
1061
1062         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1063         (ViewController.prototype._domForNewAndExistingBugs): Moved code from here to
1064         TestFailureBugForm/NewBugForm, and changed this code to use a TestFailureBugForm.
1065
1066         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
1067         NewBugForm/TestFailureBugForm.js.
1068
1069         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
1070         Added new tests and required files.
1071
1072 2011-07-12  Adam Roben  <aroben@apple.com>
1073
1074         Test that no intermediate WTF::Strings are created when concatenating with string literals
1075
1076         Test for <http://webkit.org/b/63330> Concatenating string literals and WTF::Strings using
1077         operator+ is suboptimal
1078
1079         Reviewed by Darin Adler.
1080
1081         * TestWebKitAPI/Tests/WTF/StringOperators.cpp: Added.
1082         (TestWebKitAPI::TEST): Test that a bunch of different string concatenation expressions don't
1083         create any intermediate WTF::Strings while they're being evaluated.
1084
1085         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1086         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
1087         Added new file.
1088
1089 2011-07-12  Eric Seidel  <eric@webkit.org>
1090
1091         [Qt] NRWT should pick up the right httpd config file
1092         https://bugs.webkit.org/show_bug.cgi?id=64086
1093
1094         Reviewed by Adam Barth.
1095
1096         This is more fall-out from the Port class inappropriately encapsulating
1097         both platform and port knowledge.  We need to split out some of this
1098         platform knowledge into a separate class which can be better shared
1099         between ports.
1100
1101         The fix was to move all the _path_to_apache_config_file logic down
1102         into the WebKitPort baseclass so that all of the ports can find the
1103         right apache config file, regardless of what platform they may be running on.
1104
1105         I did not move this down into base.Port (even though I considered it).
1106         Chromium duplicates some of this logic, but since they have separate
1107         subclasses for each port-OS combination (e.g. ChromiumMac, ChromiumWin)
1108         they wouldn't notice the change.  Eventually we'll move this logic
1109         out of Port entirely, and then it will be shared by all ports.
1110
1111         I also cleaned up the http-config logic for ORWT while I was there,
1112         although since we're killing that code, I'm happy to revert that part
1113         of the change if changing it is viewed as needlessly risky.
1114
1115         During this cleanup, I noticed that no ports use the "default" httpd.conf
1116         which is left over from Mac OS X Tiger (and old linux distros) which used
1117         Apache 1.3.  I've removed httpd.conf (and associated support in ORWT) since
1118         we no longer support any ports which use this httpd.conf.
1119
1120         * Scripts/webkitperl/httpd.pm:
1121         (getHTTPDConfigPathForTestDirectory):
1122         * Scripts/webkitpy/layout_tests/port/base.py:
1123         * Scripts/webkitpy/layout_tests/port/gtk.py:
1124         * Scripts/webkitpy/layout_tests/port/mac.py:
1125         * Scripts/webkitpy/layout_tests/port/qt.py:
1126         * Scripts/webkitpy/layout_tests/port/webkit.py:
1127         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1128         * Scripts/webkitpy/layout_tests/port/win.py:
1129         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
1130
1131 2011-07-11  Hironori Bono  <hbono@chromium.org>
1132
1133         Reviewed by Adam Roben.
1134
1135         Implement layoutTestController.setTextDirection for WebKit2 and Windows.
1136         https://bugs.webkit.org/show_bug.cgi?id=61931
1137
1138         This change implements layoutTextController.setTextDirection for WebKit2
1139         and Windows so we can run a layout test added by r87770 on them. Each
1140         implementation adds a wrapper function for Editor::setBaseWritingDirection()
1141         so LayoutTestController can call it.
1142
1143         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1144         (LayoutTestController::setTextDirection): Implemented the binding function
1145         for layoutTestController.setTextDirection (Windows).
1146         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
1147         Added a binding function for layoutTestController.setTextDirection (WebKit2).
1148         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp: ditto.
1149         (WTR::LayoutTestController::setTextDirection):
1150         * WebKitTestRunner/InjectedBundle/LayoutTestController.h: ditto.
1151
1152 2011-07-11  Raphael Kubo da Costa  <kubo@profusion.mobi>
1153
1154         [EFL] ImageDiff: Make sure gEcoreEvas is destroyed at the right time.
1155         https://bugs.webkit.org/show_bug.cgi?id=64293
1156
1157         Reviewed by Kent Tamura.
1158
1159         gEcoreEvas was being destroyed after shutdownEfl() was called, so the
1160         call to ecore_evas_free() failed.
1161
1162         We now shut it down manually before shutdownEfl() is called.
1163
1164         * DumpRenderTree/efl/ImageDiff.cpp:
1165         (main): Destroy gEcoreEvas before shutting down the EFL.
1166
1167 2011-07-11  Ryosuke Niwa  <rniwa@webkit.org>
1168
1169         Make Chromium Mac and Linux tests bot core builders
1170         https://bugs.webkit.org/show_bug.cgi?id=63196
1171
1172         Reviewed by Adam Barth.
1173
1174         Added Chromium Mac and Linux Release Tests bots core builders.
1175
1176         Also rearranged the order of bots so that test bots show up right next to
1177         respective builders.
1178
1179         * BuildSlaveSupport/build.webkit.org-config/config.json:
1180         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
1181         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
1182
1183 2011-07-11  Raphael Kubo da Costa  <kubo@profusion.mobi>
1184
1185         [EFL] DRT: Add code related to font management.
1186         https://bugs.webkit.org/show_bug.cgi?id=63989
1187
1188         Reviewed by Kent Tamura.
1189
1190         The added files are responsible for managing fontconfig and adding the
1191         fonts required by some layout tests to fontconfig's path.
1192
1193         * DumpRenderTree/efl/FontManagement.cpp: Added.
1194         (getFontDirectories):
1195         (getFontFiles):
1196         (addFontDirectories):
1197         (addFontFiles):
1198         (addFontsToEnvironment):
1199         * DumpRenderTree/efl/FontManagement.h: Added.
1200
1201 2011-07-08  Dirk Pranke  <dpranke@chromium.org>
1202
1203         nrwt: linting fixes
1204         https://bugs.webkit.org/show_bug.cgi?id=64225
1205
1206         Reviewed by Eric Siedel.
1207
1208         Miscellaneous linting fixes. The most notable change is that
1209         we add public attributes for user, executive, filesystem, and
1210         options on the Port object, so we don't have to refer to the
1211         "protected" versions all over the place".
1212
1213         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1214         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
1215         * Scripts/webkitpy/layout_tests/controllers/message_broker.py:
1216         * Scripts/webkitpy/layout_tests/controllers/message_broker_unittest.py:
1217         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1218         * Scripts/webkitpy/layout_tests/port/base.py:
1219         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1220
1221 2011-07-11  Ryosuke Niwa  <rniwa@webkit.org>
1222
1223         webkit-patch roll-chromium-deps no longer works
1224         https://bugs.webkit.org/show_bug.cgi?id=64324
1225
1226         Reviewed by James Robinson.
1227
1228         Lower the minimum commit message length to 10.
1229
1230         * Scripts/webkitpy/tool/steps/commit.py:
1231
1232 2011-07-11  Alice Boxhall  <aboxhall@chromium.org>
1233
1234         Convert json_results_generator.py to output version 4 JSON.
1235         https://bugs.webkit.org/show_bug.cgi?id=60869
1236
1237         Reviewed by Ojan Vafai.
1238
1239         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
1240         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
1241         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
1242
1243 2011-07-11  Martin Robinson  <mrobinson@igalia.com>
1244
1245         [GTK] [NRWT] Pixel tests do not work
1246         https://bugs.webkit.org/show_bug.cgi?id=64091
1247
1248         Reviewed by Eric Seidel.
1249
1250         Fix pixel test runs for GTK+ new-run-webkit-tests by exposing the
1251         path to the ImageDiff binary.
1252
1253         * Scripts/webkitpy/layout_tests/port/gtk.py: Implement _path_to_image_diff.
1254
1255 2011-07-11  Dirk Pranke  <dpranke@chromium.org>
1256
1257         Fix hang on win32 in a multiprocessing test that shouldn't have been running.
1258
1259         Unreviewed, build fix.
1260
1261         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1262
1263 2011-07-11  Adam Roben  <aroben@apple.com>
1264
1265         Don't count leaks as test failures on TestFailures's front page
1266
1267         As a bonus, this patch adds our first unit test.
1268
1269         Fixes <http://webkit.org/b/64303> REGRESSION (r90489): TestFailures page says far more tests
1270         are failing on the Leaks bot than actually are
1271
1272         Reviewed by Adam Barth.
1273
1274         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
1275         (Builder.prototype.getNumberOfFailingTests): Exclude lines containing "leak" when summing
1276         failure counts.
1277
1278         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
1279         Added. Contains a simple test to show that leaks aren't counted as test failures.
1280
1281         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
1282         Added. This is the test harness.
1283
1284 2011-06-14  Mike Stegeman  <mrstegeman@gmail.com>
1285
1286         [GTK] Add API to allow setting local storage database path
1287         https://bugs.webkit.org/show_bug.cgi?id=62091
1288
1289         Reviewed by Martin Robinson and Gustavo Noronha.
1290
1291         Expose the path of the localStorage databases through a setting
1292         to allow HTML5 localStorage to be persistent. New setting is:
1293         html5-local-storage-database-path
1294
1295         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1296         (resetDefaultsToConsistentValues):
1297
1298 2011-07-11  Adam Barth  <abarth@webkit.org>
1299
1300         garden-o-matic should be able to rebaseline tests
1301         https://bugs.webkit.org/show_bug.cgi?id=64186
1302
1303         Reviewed by Eric Seidel.
1304
1305         This patch contains a simple HTTP server binding for the
1306         rebaseline-test command.  The frontend will appear in another patch.
1307
1308         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1309         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
1310         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
1311
1312 2011-07-11  Adam Barth  <abarth@webkit.org>
1313
1314         Add some basic UI for showing regression ranges
1315         https://bugs.webkit.org/show_bug.cgi?id=64243
1316
1317         Reviewed by Eric Seidel.
1318
1319         This UI is all temporary.  It exists mostly just to have some buttons
1320         to click to exercise the code.
1321
1322         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
1323         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
1324         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
1325         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
1326         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
1327
1328 2011-07-11  Adam Barth  <abarth@webkit.org>
1329
1330         garden-o-matic should be able to determine which revisions caused a given failure
1331         https://bugs.webkit.org/show_bug.cgi?id=64189
1332
1333         Reviewed by Adam Roben.
1334
1335         Walking the failure history looking for failures turns out to be
1336         slightly tricky because the network requests are asynchronous.
1337         Currently we do all the fetches serially and our cache is unbounded.
1338         We'll probably optimize both those parameters eventually.
1339
1340         This patch also generalizes some functionality in the unit testing
1341         framework to make testing this sort of code easier.
1342
1343         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
1344         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
1345         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
1346
1347 2011-07-11  Adam Barth  <abarth@webkit.org>
1348
1349         Add a webkit-patch command for rebaselining an individual test
1350         https://bugs.webkit.org/show_bug.cgi?id=64246
1351
1352         Reviewed by Eric Seidel.
1353
1354         This patch introduces a command that's able to rebaseline a single
1355         test.  Currently, the command works only with the build.chromium.org
1356         buildbots, but extending it to work with the build.webkit.org bots
1357         shouldn't be that hard.
1358
1359         A complete rebaseling tool should also include an "optimize baselines"
1360         command (which moves/deletes baselines in order to reduce the number of
1361         expected results files), but that will come in a future patch.
1362
1363         Really BuilderToPort should be merged into builders.py, but I'm going
1364         to save that for a future patch as well.  (We need to stop shaving yaks
1365         at some point.)
1366
1367         * Scripts/webkitpy/tool/commands/rebaseline.py:
1368
1369 2011-07-11  Adam Barth  <abarth@webkit.org>
1370
1371         garden-o-matic should be able to roll out patches
1372         https://bugs.webkit.org/show_bug.cgi?id=64185
1373
1374         Reviewed by Eric Seidel.
1375
1376         This gardening server API is simply a binding to the rollout machinery
1377         already present in webkit-patch.
1378
1379         * Scripts/webkitpy/tool/commands/gardenomatic.py:
1380         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1381         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Added.
1382
1383 2011-07-11  Csaba Osztrogonác  <ossy@webkit.org>
1384
1385         [Qt][Mac] Unreviewed fix after r90746.
1386
1387         * Scripts/run-webkit-tests:
1388         (useNewRunWebKitTests): Really disable NRWT for qt-mac platform.
1389
1390 2011-07-11  Adam Roben  <aroben@apple.com>
1391
1392         Teach TestFailures that ORWT's results.html file might be missing due to all tests passing
1393
1394         There are three reasons why we might fail to fetch ORWT's results.html:
1395           - All tests passed, so no results.html was generated
1396           - Some error during the test run caused results.html not to be generated (e.g., ORWT
1397             timed out)
1398           - Some network error occurred when fetching results.html
1399
1400         We were failing to account for the first possibility in some cases. For test runs before
1401         r89610, we first check build.webkit.org/json to determine how many tests failed and whether
1402         ORWT exited early due to too many failures; if all tests passed then we don't fetch
1403         results.html at all. r89610 changed ORWT to put information in results.html about exiting
1404         early due to too many failures, so we no longer needed to check build.webkit.org/json to get
1405         that information, and in r89619 I changed TestFailures to do just that. But I forgot that we
1406         still needed to check build.webkit.org/json to find out if all tests passed!
1407
1408         Now, for test runs after r89610, we check results.html first, and then check
1409         build.webkit.org/json if we fail to fetch results.html. This lets us distinguish between all
1410         tests passing and the error cases.
1411
1412         Fixes <http://webkit.org/b/64280> TestFailures page can't pinpoint that r90699 caused 13
1413         tests to fail on Windows 7 Release (WebKit2 Tests)
1414
1415         Reviewed by David Kilzer.
1416
1417         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
1418         (LayoutTestResultsLoader.prototype.start): Bumped the cache version so that old, buggy
1419         cached data will get evicted. We were marking builds where all tests passed as errors!
1420         (LayoutTestResultsLoader.prototype._fetchAndParseORWTResults): Added success/error callback
1421         parameters to the fetchAndParseResultsHTML helper function, and added a similar
1422         fetchNumberOfFailingTests function that fetches data from build.webkit.org/json (code came
1423         from later in the function). For test runs before r89610, we first check
1424         build.webkit.org/json then check results.html. For builds after r89610, we first check
1425         results.html and then check build.webkit.org/json if we couldn't fetch results.html.
1426
1427 2011-07-11  Csaba Osztrogonác  <ossy@webkit.org>
1428
1429         new-run-webkit-tests does not support qt-arm or qt-4.8 results
1430         https://bugs.webkit.org/show_bug.cgi?id=64071
1431
1432         [Qt] NRWT should pick up the right httpd config file
1433         https://bugs.webkit.org/show_bug.cgi?id=64086
1434
1435         Reviewed by Zoltán Herczeg.
1436
1437         * Scripts/run-webkit-tests: Disable NRWT for qt-mac, qt-arm and qt-4.8 platforms until fix.
1438         (useNewRunWebKitTests):
1439
1440 2011-07-11  Kenichi Ishibashi  <bashi@chromium.org>
1441
1442         Add TestNetscapePlugIn/Tests/FormValue.cpp to DRT build files
1443         https://bugs.webkit.org/show_bug.cgi?id=64248
1444
1445         Reviewed by Kent Tamura.
1446
1447         * DumpRenderTree/DumpRenderTree.gypi: Added FormValue.cpp.
1448         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Ditto.
1449         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro: Ditto.
1450         * GNUmakefile.am: Ditto.
1451
1452 2011-07-10  Kenichi Ishibashi  <bashi@chromium.org>
1453
1454         Let plugins participate in form submission
1455         https://bugs.webkit.org/show_bug.cgi?id=13061
1456
1457         Adds NPPVformValue support to the testing plugin.
1458
1459         Reviewed by Anders Carlsson.
1460
1461         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added FormValue.cpp
1462         * DumpRenderTree/TestNetScapePlugIn/Tests/FormValue.cpp: Added.
1463         (FormValue::FormValue): Ditto.
1464         (FormValue::NPP_GetValue): Ditto.
1465
1466 2011-07-11  Roland Steiner  <rolandsteiner@chromium.org>
1467
1468         Reviewed by Tony Chang.
1469
1470         Complete functions in filesystem.py
1471         https://bugs.webkit.org/show_bug.cgi?id=63528
1472
1473         - missing text file functions added
1474         - functions sorted
1475         - removed 'append' optional parameters (were unused)
1476         - adapted filesystem_mock in the same way
1477
1478         * Scripts/webkitpy/common/system/filesystem.py:
1479         * Scripts/webkitpy/common/system/filesystem_mock.py:
1480         * Scripts/webkitpy/common/system/filesystem_unittest.py:
1481
1482 2011-07-10  Adam Barth  <abarth@webkit.org>
1483
1484         Clean up style in fallback path calculation
1485         https://bugs.webkit.org/show_bug.cgi?id=64239
1486
1487         Reviewed by Daniel Bates.
1488
1489         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1490         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1491         * Scripts/webkitpy/layout_tests/port/mac.py:
1492         * Scripts/webkitpy/layout_tests/port/win.py:
1493
1494 2011-07-10  Balazs Kelemen  <kbalazs@webkit.org>
1495
1496         [Qt][WK2] Implement the rest of WTR::PlatformWebView
1497         https://bugs.webkit.org/show_bug.cgi?id=63630
1498
1499         Reviewed by Kenneth Rohde Christiansen.
1500
1501         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
1502         (WTR::PlatformWebView::windowFrame):
1503         (WTR::PlatformWebView::setWindowFrame):
1504
1505 2011-07-10  Mark Rowe  <mrowe@apple.com>
1506
1507         Teach build-webkit how to find the latest WebKitSystemInterface binary.
1508
1509         * Scripts/build-webkit:
1510
1511 2011-07-09  Adam Roben  <aroben@apple.com>
1512
1513         Teach TestFailures to abbreviate the examples of test flakiness
1514
1515         These lists can get quite long, and it's not really helpful in most cases to have soooooo
1516         many examples of flakiness.
1517
1518         Fixes <http://webkit.org/b/64203> Lists of flaky revisions on TestFailures page can get so
1519         long they're hard to navigate
1520
1521         Reviewed by Dan Bates.
1522
1523         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
1524         (FlakyLayoutTestDetector.prototype.flakinessExamples): If we have more than a certain number
1525         of examples, replace the middle items with a separator.
1526
1527         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
1528         (.flakiness-example-separator): Added styles for the separator.
1529
1530         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1531         (ViewController.prototype._domForPossiblyFlakyTests): Use a vertical ellipsis to represent
1532         the separator.
1533
1534 2011-07-09  Dirk Pranke  <dpranke@chromium.org>
1535
1536         nrwt: stack traces from worker-side exceptions aren't very useful inside test-webkitpy
1537         https://bugs.webkit.org/show_bug.cgi?id=64218
1538
1539         Reviewed by Eric Seidel.
1540
1541         Exceptions aren't picklable and can't be sent across the
1542         manager/worker message queue without losing information. NRWT
1543         handles this by turning the stack trace into a set of strings,
1544         and logging the strings when we receive an exception from the
1545         worker. However, when you are running tests and something
1546         crashes on the worker side, test-webkitpy prints the
1547         manager-side stack trace, which is just confusing and useless.
1548
1549         This patch changes the logic so that exceptions are passed
1550         through as-is when the worker and manager are in the same
1551         process (the --worker-model=inline option). This increases the
1552         code paths slightly but makes crashes much more useful.
1553
1554         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1555         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
1556         * Scripts/webkitpy/layout_tests/controllers/message_broker.py:
1557         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1558         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1559
1560 2011-07-09  Darin Fisher  <darin@chromium.org>
1561
1562         Eliminate bad dependency on gfx::Point.
1563         https://bugs.webkit.org/show_bug.cgi?id=64228
1564
1565         Reviewed by Kent Tamura.
1566
1567         * DumpRenderTree/chromium/EventSender.cpp:
1568         (initMouseEvent):
1569
1570 2011-07-08  Ojan Vafai  <ojan@chromium.org>
1571
1572         pull static dashboard files into the appengine server from the chromium repository
1573         https://bugs.webkit.org/show_bug.cgi?id=64208
1574
1575         Reviewed by Dirk Pranke.
1576
1577         These files belong in the WebKit repo since they are tied primarily to the webkit tests.
1578         They have some extra bits to support chromium's gtests, but that seems fine.
1579         Mainly, this will allow other WebKit hackers to hack on the dashboards.
1580
1581         As a nice side-effect, we can now push the dashboard files when we do appengine pushes
1582         instead of the weird thing we used to do of pulling them from the Chromium repository
1583         and storing them in the appengine datastore. This allows for cleaning up a lot of code
1584         and will likely make the dashboards load a bit faster.
1585
1586         The new JS files don't fully match WebKit style, but I'd like to clean that up in a
1587         followup patch if possible to maintain my sanity with this patch.
1588
1589         * TestResultServer/app.yaml:
1590         * TestResultServer/handlers/dashboardhandler.py: Removed.
1591         * TestResultServer/handlers/menu.py:
1592         * TestResultServer/main.py:
1593         * TestResultServer/model/dashboardfile.py: Removed.
1594         * TestResultServer/static-dashboards/LICENSE.dygraph.txt: Added.
1595         * TestResultServer/static-dashboards/README.dygraph.txt: Added.
1596         * TestResultServer/static-dashboards/README.webtreemap.txt: Added.
1597         * TestResultServer/static-dashboards/aggregate_results.html: Added.
1598         * TestResultServer/static-dashboards/builders.js: Added.
1599         * TestResultServer/static-dashboards/dashboard_base.js: Added.
1600         * TestResultServer/static-dashboards/dygraph-combined.js: Added.
1601         * TestResultServer/static-dashboards/flakiness_dashboard.html: Added.
1602         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js: Added.
1603         * TestResultServer/static-dashboards/timeline_explorer.html: Added.
1604         * TestResultServer/static-dashboards/treemap.html: Added.
1605         * TestResultServer/static-dashboards/webtreemap.css: Added.
1606         * TestResultServer/static-dashboards/webtreemap.js: Added.
1607         * TestResultServer/stylesheets/dashboardfile.css: Removed.
1608         * TestResultServer/stylesheets/menu.css:
1609         * TestResultServer/templates/dashboardfilelist.html: Removed.
1610         * TestResultServer/templates/menu.html:
1611
1612 2011-07-08  Adam Roben  <aroben@apple.com>
1613
1614         Make TestFailures's list of flaky tests look more like the list of non-flaky tests
1615
1616         Fixes <http://webkit.org/b/64204> TestFailures page's flaky tests list is ugly!
1617
1618         Reviewed by Daniel Bates.
1619
1620         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
1621         (#failure-history, #possibly-flaky-tests): Expanded this rule to apply to the list of flaky
1622         tests.
1623         (#failure-history > li, #possibly-flaky-tests > li): Ditto, but moved the 50px left padding
1624         from here...
1625         (#failure-history > li): ...to here.
1626
1627         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1628         (ViewController.prototype._domForPossiblyFlakyTests): Give the list an id attribute for
1629         styling purposes.
1630
1631 2011-07-08  Jeffrey Pfau  <jpfau@apple.com>
1632
1633         Unreviewed, add myself as committer.
1634
1635         * Scripts/webkitpy/common/config/committers.py:
1636
1637 2011-07-08  Adam Barth  <abarth@webkit.org>
1638
1639         TestResultsServer should keep old test results
1640         https://bugs.webkit.org/show_bug.cgi?id=64199
1641
1642         Reviewed by Ojan Vafai.
1643
1644         Having historical data will help us do failure archeology.
1645
1646         * TestResultServer/handlers/testfilehandler.py:
1647         * TestResultServer/model/testfile.py:
1648         * TestResultServer/templates/showfilelist.jsonp: Added.
1649
1650 2011-07-08  Adam Roben  <aroben@apple.com>
1651
1652         Remove commit-log-editor's dependency on Module::Load::Conditional
1653
1654         This module isn't available in Perl 5.8.8 (the version used on Leopard).
1655
1656         Fixes <http://webkit.org/b/64198> REGRESSION (r90583):
1657         webkitpy.common.checkout.checkout_unittest failing on Leopard
1658
1659         Reviewed by Daniel Bates.
1660
1661         * Scripts/commit-log-editor: Use the new loadTermReadKey() function instead of
1662         Module::Load::Conditional::can_load.
1663         (loadTermReadKey): Added. Tries to load Term::ReadKey and returns true if it's successful.
1664
1665 2011-07-08  Adam Roben  <aroben@apple.com>
1666
1667         Make checkout_unittest more robust against files moving around
1668
1669         Fixes <http://webkit.org/b/64197> checkout_unittest contains ugly, fragile code to find the
1670         Scripts directory
1671
1672         Reviewed by Adam Barth.
1673
1674         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1675         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Instantiate a real SCM
1676         object and use it to get the path to the Scripts directory, rather than hard-coding the
1677         relative path from this file to Scripts.
1678
1679 2011-07-08  Adam Roben  <aroben@apple.com>
1680
1681         Make TestFailures remember that run-webkit-tests timed out, even across reloads
1682
1683         Fixes <http://webkit.org/b/64193> TestFailures page incorrectly thinks all tests passed in
1684         http://build.webkit.org/builders/Windows%207%20Release%20(Tests)/builds/14589 after a reload
1685
1686         Reviewed by Daniel Bates.
1687
1688         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
1689         (LayoutTestResultsLoader.prototype.start): Store an "error" attribute in the cached data.
1690         When true, it indicates that there was an error retrieving the results for this build and
1691         that the errorCallback should be called.
1692
1693 2011-07-08  Adam Barth  <abarth@webkit.org>
1694
1695         Teach garden-o-matic how to display test results
1696         https://bugs.webkit.org/show_bug.cgi?id=64141
1697
1698         Reviewed by Ojan Vafai.
1699
1700         This patch includes basic infrastructure for probing build.chromium.org
1701         for test results.  We only handle text and image tests, not anything
1702         complicated like reftests.  Also, we're using the revision/build
1703         independent results store on the server, so we're avoiding that
1704         complication for now.
1705
1706         It's slightly hacky that we need to probe the server to see what kinds
1707         of results exist.  A better solution would be to add CORS support to
1708         the server or to use the local server to help.
1709
1710         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
1711         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
1712         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
1713         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
1714         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
1715         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
1716
1717 2011-07-08  Dirk Pranke  <dpranke@chromium.org>
1718
1719         REGRESSION(90419) NRWT's httpd locking is broken for --child-processes=1
1720         https://bugs.webkit.org/show_bug.cgi?id=64092
1721
1722         Reviewed by Tony Chang.
1723
1724         The code was incorrectly creating empty shards in the
1725         shard_in_two case.
1726
1727         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1728         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1729
1730 2011-07-08  Adam Roben  <aroben@apple.com>
1731
1732         Teach buildbot to figure out how many webkitpy/webkitperl tests failed
1733
1734         Fixes <http://webkit.org/b/64192> It's hard to tell how many test-webkitpy/test-webkitperl
1735         tests failed when looking at build.webkit.org
1736
1737         Reviewed by Eric Seidel.
1738
1739         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1740         (TestWithFailureCount): New class that represents a test build step which has an associated
1741         failure count. Eventually we should move more of our test classes to deriving from this.
1742         (TestWithFailureCount.countFailures): Method for subclasses to override to say how many
1743         failures occurred.
1744
1745         (TestWithFailureCount.commandComplete):
1746         (TestWithFailureCount.evaluateCommand):
1747         (TestWithFailureCount.getText):
1748         (TestWithFailureCount.getText2):
1749         These were all based on RunGtkAPITests.
1750
1751         (RunPythonTests): Changed to inherit from TestWithFailureCount.
1752         (RunPythonTests.countFailures): Parses the test-webkitpy output looking for the count of
1753         failures.
1754         (RunPerlTests): Changed to inherit from TestWithFailureCount.
1755         (RunPerlTests.countFailures): Parses the test-webkitperl output looking for the count of
1756         failures.
1757
1758 2011-07-08  Adam Roben  <aroben@apple.com>
1759
1760         Ensure $CHANGE_LOG_EMAIL_ADDRESS is set when testing webkitpy's commit-log-editor integration
1761
1762         Fixes <http://webkit.org/b/64180> REGRESSION (r90564): test-webkitpy failing on multiple
1763         bots due to commit-log-editor errors
1764
1765         Reviewed by Adam Barth.
1766
1767         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1768         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Set
1769         $CHANGE_LOG_EMAIL_ADDRESS to the patch author's email address. This ensures that
1770         commit-log-editor can find a value for the committer's email, and that the committer and
1771         author email addresses match, which will prevent commit-log-editor from inserting a "Patch
1772         by" line in the commit message.
1773
1774 2011-07-08  Andreas Kling  <kling@webkit.org>
1775
1776         [Qt][WK2] Views should expose QActions for basic navigation.
1777         https://bugs.webkit.org/show_bug.cgi?id=64174
1778
1779         Reviewed by Benjamin Poulain.
1780
1781         Bring the toolbar in Qt's MiniBrowser back to life.
1782
1783         * MiniBrowser/qt/BrowserView.cpp:
1784         (BrowserView::navigationAction):
1785         * MiniBrowser/qt/BrowserView.h:
1786         * MiniBrowser/qt/BrowserWindow.cpp:
1787         (BrowserWindow::BrowserWindow):
1788
1789 2011-07-08  Adam Roben  <aroben@apple.com>
1790
1791         Teach TestFailures how to find test names in commit-log-editor-style commit messages
1792
1793         TestFailures was relying on Trac turning the list of modified files in our commit messages
1794         into an HTML list. But Trac only does this when the list of modified files is indented.
1795         commit-log-editor doesn't indent the file list, so the list wasn't being turned into an HTML
1796         list, which was confusing TestFailures.
1797
1798         TestFailures now does much simpler parsing of the commit message (i.e., just a substring
1799         search) without relying at all on its structure.
1800
1801         Fixes <http://webkit.org/b/64173> TestFailures page fails to blame r90608 for breaking
1802         fast/dom/HTMLProgressElement/progress-element-markup.html on Windows
1803
1804         Reviewed by David Kilzer.
1805
1806         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
1807         (Trac.prototype.getCommitDataForRevisionRange): Instead of trying to parse the commit
1808         message, just return its text.
1809
1810         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1811         (ViewController.prototype._domForRegressionRange): Instead of searching for test names in
1812         each commit's list of modified files, just search for test names anywhere in the commit's
1813         message.
1814
1815 2011-07-08  Adam Barth  <abarth@webkit.org>
1816
1817         sheriffbot is too spammy in IRC
1818         https://bugs.webkit.org/show_bug.cgi?id=64153
1819
1820         Reviewed by Eric Seidel.
1821
1822         Reporting failures in IRC worked well when the tree was greener than it
1823         is today.  Nowadays, this feature mostly just results in spam about
1824         false positives.  If we reach a greener state, we can bring this code
1825         back.
1826
1827         (Another possibility is to restrict this feature to builder bots,
1828         rather than including testers, as we did before this patch.)
1829
1830         * Scripts/webkitpy/tool/commands/sheriffbot.py:
1831         * Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
1832
1833 2011-07-08  Adam Roben  <aroben@apple.com>
1834
1835         Ensure commit-log-editor adds a "Patch by" line when the author and committer are different
1836
1837         Previously we were only adding a "Patch by" line when the ChangeLog contained a "Reviewed
1838         by" line. But some patches (like rollout patches) don't contain that line. Now we always add
1839         "Patch by" to the commit log regardless of the ChangeLog's contents.
1840
1841         Fixes <http://webkit.org/b/64127> Committer for r90588 is commit-queue@webkit.org, but
1842         should have been abarth@webkit.org
1843
1844         Reviewed by Anders Carlsson.
1845
1846         * Scripts/commit-log-editor:
1847         (createCommitMessage): Try to put the "Patch by" line just above the "Reviewed by" line, as
1848         before. If there is no "Reviewed by" line, try to put it just above the first modified file.
1849         If all else fails, put it at the end of the commit message.
1850         (patchAuthorshipString): Added. Code came from createCommitMessage.
1851
1852 2011-07-08  Herczeg Zoltan  <zherczeg@webkit.org>
1853
1854         Reviewed by Andreas Kling.
1855
1856         Adding myself as a reviewer.
1857
1858         * Scripts/webkitpy/common/config/committers.py:
1859
1860 2011-07-07  Kevin Ollivier  <kevino@theolliviers.com>
1861
1862         [wx] Unreviewed build fix, remove old files from the tree and update paths to new ones.
1863
1864         * wx/build: Removed.
1865         * wx/build/build_utils.py: Removed.
1866         * wx/build/settings.py: Removed.
1867         * wx/build/waf_extensions.py: Removed.
1868         * wx/build/wxpresets.py: Removed.
1869         * wx/packaging/build-mac-installer.py:
1870
1871 2011-07-07  Kevin Ollivier  <kevino@theolliviers.com>
1872
1873         [wx] Unreviewed build fix, more fixes for install name issues.
1874
1875         * wx/packaging/build-mac-installer.py:
1876
1877 2011-07-07  Dirk Pranke  <dpranke@chromium.org>
1878
1879         test-webkitpy fails on chromium win
1880         https://bugs.webkit.org/show_bug.cgi?id=64137
1881
1882         Reviewed by Eric Seidel.
1883
1884         test-webkitpy doesn't play nicely with the multiprocessing
1885         module on win32. An earlier change actually reenabled the tests
1886         on win32 by mistake.
1887
1888         This patch also fixes a few cases where path names will trip
1889         things up on win32 (e.g., by testing unix paths on windows).
1890         We do not lose any real coverage here with those fixes.
1891
1892         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
1893         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
1894         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
1895         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1896
1897 2011-07-07  Leandro Pereira  <leandro@profusion.mobi>
1898
1899         [ImageDiff] Calculate/print difference right after reading baseline image.
1900         https://bugs.webkit.org/show_bug.cgi?id=64117
1901
1902         Reviewed by Kent Tamura.
1903
1904         Otherwise, ImageDiff will block on fgets() until the universe ends or you ^C
1905         it.  Whichever happens first.
1906
1907         * DumpRenderTree/efl/ImageDiff.cpp:
1908         (main):
1909
1910 2011-07-07  Raphael Kubo da Costa  <kubo@profusion.mobi>
1911
1912         [EFL] DRT: Leak cairo_t to make the pixel tests stop crashing
1913         https://bugs.webkit.org/show_bug.cgi?id=64107
1914
1915         Reviewed by Kent Tamura.
1916
1917         The pointer is later managed by BitmapContext, but it was being
1918         de-refed and deleted earlier by our RefPtr, causing crashes in all
1919         pixel tests.
1920
1921         We now call leakRef() to make sure it is not removed when our
1922         createBitmapContextFromWebView goes out of context.
1923
1924         * DumpRenderTree/efl/PixelDumpSupportEfl.cpp:
1925         (createBitmapContextFromWebView):
1926
1927 2011-07-07  Adam Roben  <aroben@apple.com>
1928
1929         Update TestFailures's title and header immediately upon navigation
1930
1931         Fixes <http://webkit.org/b/64125> TestFailures page seems slow to react on link clicks
1932
1933         Reviewed by Daniel Bates.
1934
1935         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1936         (ViewController.prototype.loaded): Create and store an h1 element for displaying the page's
1937         title.
1938         (ViewController.prototype._displayBuilder): Set the page's title and clear out any old
1939         content immediately instead of waiting for the first set of results for the history
1940         analyzer. As we receive new results from the analyzer we'll just clear out the main content
1941         area and replace it with the new info.
1942         (ViewController.prototype._displayTesters): Set the page's title and clear out any old
1943         content immediately instead of waiting for the list of testers from the buildbot. Sprinkled
1944         in a little use of our removeAllChildren helper, too.
1945         (ViewController.prototype._setTitle): Added. Set's the page's title and the header text.
1946
1947 2011-07-07  Adam Roben  <aroben@apple.com>
1948
1949         Make Checkout use SCM's Executive instead of conjuring up its own
1950
1951         This will improve integration with the rest of webkitpy, particularly when invoked via
1952         webkit-patch.
1953
1954         Fixes <http://webkit.org/b/64115> REGRESSION (r90564): webkitpy's Checkout class uses
1955         Executive inappropriately
1956
1957         Reviewed by Adam Barth.
1958
1959         * Scripts/webkitpy/common/checkout/checkout.py:
1960         (Checkout.commit_message_for_this_commit): Use SCM.run instead of creating an Executive for
1961         our own use. SCM might have some super-special Executive that it uses under the covers, and
1962         we want to use it, too!
1963
1964         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1965         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Mock the SCM.run
1966         method to call through to Executive.run_command.
1967
1968 2011-07-07  Adam Roben  <aroben@apple.com>
1969
1970         Make Term/ReadKey.pm an optional dependency of commit-log-editor
1971
1972         When Term/ReadKey.pm isn't present, the --regenerate-log switch will be non-functional.
1973
1974         Fixes <http://webkit.org/b/64113> REGRESSION (r90564):
1975         webkitpy.common.checkout.checkout_unittest failing on GTK bots due to missing
1976         Term/ReadKey.pm module
1977
1978         Reviewed by Xan Lopez.
1979
1980         * Scripts/commit-log-editor: Use Module::Load::Conditional::can_load to only load
1981         Term/ReadKey.pm if it's present. If it isn't present, ignore the --regenerate-log switch so
1982         that existing commit messages will be preserved. (Users can manually remove existing commit
1983         messages using git-reset.)
1984
1985 2011-07-07  Adam Roben  <aroben@apple.com>
1986
1987         Keep testing that commit messages containing Unicode are handled properly by webkitpy
1988
1989         Fixes <http://webkit.org/b/64109> REGRESSION (r90571): test-webkitpy no longer tests that we
1990         correctly handle Unicode in commit messages
1991
1992         Reviewed by Eric Seidel.
1993
1994         * Scripts/webkitpy/common/checkout/checkout_unittest.py: Put some Unicode characters in
1995         places in the ChangeLog that will end up being included in the commit message.
1996
1997 2011-07-07  Andreas Kling  <kling@webkit.org>
1998
1999         [Qt][WK2] Remove Symbian code.
2000         https://bugs.webkit.org/show_bug.cgi?id=64101
2001
2002         Reviewed by Benjamin Poulain.
2003
2004         * MiniBrowser/qt/BrowserWindow.cpp:
2005         (BrowserWindow::BrowserWindow):
2006         (BrowserWindow::screenshot):
2007         * MiniBrowser/qt/MiniBrowser.pro:
2008         * MiniBrowser/qt/MiniBrowserApplication.h:
2009         (WindowOptions::WindowOptions):
2010
2011 2011-07-07  Andreas Kling  <kling@webkit.org>
2012
2013         [Qt][WK2] Don't support ridiculous matrix of QT_NO_FEATURE combinations.
2014         https://bugs.webkit.org/show_bug.cgi?id=64099
2015
2016         Reviewed by Benjamin Poulain.
2017
2018         * MiniBrowser/qt/BrowserWindow.cpp:
2019         (BrowserWindow::openFile):
2020         (BrowserWindow::screenshot):
2021         (BrowserWindow::loadURLListFromFile):
2022
2023 2011-07-07  Adam Roben  <aroben@apple.com>
2024
2025         Completely ignore too-many-failures builds in TestFailures in most circumstances
2026
2027         Because a semi-arbitrary subset of tests are run in a too-many-failures build, we can't
2028         really use them to perform regression analysis. The only time we want to pay attention to
2029         too-many-failures builds is when we're trying to explain when the current bout of
2030         too-many-failures started.
2031
2032         Fixes <http://webkit.org/b/64106> TestFailures page sometimes claims a test started failing
2033         in a build that didn't even run it (because it exited early due to too many failues)
2034
2035         Reviewed by David Kilzer.
2036
2037         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
2038         (LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): Removed old, broken
2039         too-many-failures handling that would cause us to blame builds that didn't even run a given
2040         test for breaking it. Instead, skip over all too-many-failures builds unless the most recent
2041         build was itself a too-many-failures build.
2042
2043 2011-07-07  Adam Roben  <aroben@apple.com>
2044
2045         Teach webkitpy's Checkout class to use commit-log-editor to create commit messages
2046
2047         Fixes <http://webkit.org/b/26755> webkit-patch's commit messages are less readable than
2048         commit-log-editor's
2049
2050         Reviewed by David Kilzer.
2051
2052         * Scripts/webkitpy/common/checkout/checkout.py:
2053         (Checkout.commit_message_for_this_commit): Run commit-log-editor, passing it the paths of
2054         the modified ChangeLogs, to generate the commit message, rather than trying to generate one
2055         ourselves.
2056
2057         * Scripts/webkitpy/common/checkout/checkout_unittest.py: Updated the expected commit message
2058         to match commit-log-editor's format.
2059         (CommitMessageForThisCommitTest.setUp): Write the ChangeLogs into Tools and LayoutTests
2060         directories so we can see how the various entries get labeled in the commit message.
2061         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Create a mock SCM
2062         instance that knows how to find commit-log-editor and pass it to our Checkout instance.
2063         Don't bother capturing output, since there shouldn't be any.
2064
2065 2011-07-07  Adam Roben  <aroben@apple.com>
2066
2067         Add a --print-log option to commit-log-editor
2068
2069         When specified, commit-log-editor takes a set of ChangeLog paths on the command line,
2070         generates a commit message from those ChangeLogs, prints it to stdout, and exits.
2071
2072         Prep work for fixing <http://webkit.org/b/26755> webkit-patch's commit messages are less
2073         readable than commit-log-editor's
2074
2075         Reviewed by David Kilzer.
2076
2077         * Scripts/commit-log-editor: Changed to use Getopt::Long to parse options. Added --print-log
2078         option, which calls through to createCommitMessage, prints the result, and exits.
2079         (printUsageAndExit): Renamed from "usage" for clarity. Beefed up the usage statement to
2080         explain commit-log-editor's 2.5 (normal, --print-log, and --help) modes of operation.
2081
2082 2011-07-07  Adam Roben  <aroben@apple.com>
2083
2084         Extract commit-log-editor's commit-message-generation code into a separate function
2085
2086         As a bonus, we now skip a bunch of work in the case where we are generating a commit message
2087         for staged git changes.
2088
2089         Prep work for fixing <http://webkit.org/b/26755> webkit-patch's commit messages are less
2090         readable than commit-log-editor's
2091
2092         Reviewed by David Kilzer.
2093
2094         * Scripts/commit-log-editor: Moved the default definition of $endl up toward the top of the
2095         file so that it will be defined even if we aren't operating on an existing log message.
2096         Moved a bunch of top-level code to generate the commit message from here...
2097         (createCommitMessage): ...to here.
2098
2099 2011-07-07  Carlos Garcia Campos  <cgarcia@igalia.com>
2100
2101         [GTK] Open links in a new window when clicking with the middle button in MiniBrowser
2102         https://bugs.webkit.org/show_bug.cgi?id=63988
2103
2104         Reviewed by Martin Robinson.
2105
2106         * MiniBrowser/gtk/BrowserWindow.c:
2107         (browserWindowConstructed):
2108         (decidePolicyForNavigationAction): Ignore the action if a link is
2109         clicked with the middle buttonm and open the link in a new window.
2110         (browserWindowPolicyClientInit):
2111
2112 2011-07-07  Adam Barth  <abarth@webkit.org>
2113
2114         Remove style_references.py
2115         https://bugs.webkit.org/show_bug.cgi?id=64070
2116
2117         Reviewed by Eric Seidel.
2118
2119         We never really adopted the _references design, and this file is a bit
2120         of an orphan at this point.
2121
2122         * Scripts/check-webkit-style:
2123         * Scripts/webkitpy/style/checker.py:
2124         * Scripts/webkitpy/style/checker_unittest.py:
2125         * Scripts/webkitpy/style/checkers/python.py:
2126         * Scripts/webkitpy/style/checkers/test_expectations.py:
2127         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2128         * Scripts/webkitpy/style/main_unittest.py:
2129         * Scripts/webkitpy/style_references.py: Removed.
2130
2131 2011-07-07  Eric Seidel  <eric@webkit.org>
2132
2133         Fix WebKit2 expected results search paths for Mac and Qt under new-run-webkit-tests
2134         https://bugs.webkit.org/show_bug.cgi?id=64056
2135
2136         Unreviewed.
2137
2138         I had written this unittest prior to landing my change, but
2139         had named it qt_unitest.py, so it wasn't running (nor was
2140         it added to my git repo).
2141
2142         Once I fixe the name of the unittest file, it was very easy
2143         to fix the error in webkit.py (which already had a FIXME).
2144
2145         * Scripts/webkitpy/layout_tests/port/qt_unittest.py: Added.
2146         * Scripts/webkitpy/layout_tests/port/webkit.py:
2147
2148 2011-07-07  Eric Seidel  <eric@webkit.org>
2149
2150         Unreviewed.  Add more debug logging about which test expectations we're using.
2151
2152         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2153         * Scripts/webkitpy/layout_tests/port/webkit.py:
2154
2155 2011-07-06  Adam Barth  <abarth@webkit.org>
2156
2157         Remove unexpected_results.json
2158         https://bugs.webkit.org/show_bug.cgi?id=64065
2159
2160         Reviewed by Eric Seidel.
2161
2162         No one should be using this file any more.  This patch removes it from
2163         the face of the earth.  This patch also removes two NRWT-specific
2164         command line options that were implemented using
2165         unexpected_results.json.
2166
2167         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
2168         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2169         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2170         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2171         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2172         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
2173         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
2174
2175 2011-07-07  Adam Barth  <abarth@webkit.org>
2176
2177         REGRESSION(90520): chromium GPU tests produce output in the LayoutTests directory
2178         https://bugs.webkit.org/show_bug.cgi?id=64067
2179
2180         Reviewed by Eric Seidel.
2181
2182         The author of r90520 forgot to change the chromium_gpu.py implemenation
2183         of this method as well.
2184
2185         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2186
2187 2011-07-06  Eric Seidel  <eric@webkit.org>
2188
2189         Fix WebKit2 expected results search paths for Mac and Qt under new-run-webkit-tests
2190         https://bugs.webkit.org/show_bug.cgi?id=64056
2191
2192         Reviewed by Adam Barth.
2193
2194         This fixes the MacPort's version of baseline_search_path to include
2195         wk2 fallback like WebKitPort's does.  This also re-works the QtPort
2196         to set self._name (as other ports do) so that WebKitPort's default
2197         implementations will understand Qt operating system flavors correctly
2198         (including when running wk2 tests).
2199
2200         In trying to test this, I realized that nearly all of our port-tests
2201         use the default FileSystem(), User() and Executive() objects, which
2202         is really bad, because they expose details of your system in the unittest results!
2203
2204         Once I fixed many of the Port() calls to pass MockFileSystem(), then
2205         MacPort.test_tests_for_other_platforms started failing, and I had to
2206         teach MockFileSystem.glob how to handle directories to make it work again.
2207
2208         I removed the useless PortTestCase.test_baseline_search_path and replaced
2209         it with more useful test_baseline_search_path tests in MacPortTest and QtPortTest.
2210
2211         * Scripts/webkitpy/common/config/ports.py:
2212         * Scripts/webkitpy/common/system/filesystem_mock.py:
2213         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2214         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2215         * Scripts/webkitpy/layout_tests/port/mac.py:
2216         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2217         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2218         * Scripts/webkitpy/layout_tests/port/qt.py:
2219         * Scripts/webkitpy/layout_tests/port/test_files.py:
2220         * Scripts/webkitpy/layout_tests/port/win.py:
2221         * Scripts/webkitpy/to_be_moved/deduplicate_tests.py:
2222         * Scripts/webkitpy/tool/mocktool.py:
2223
2224 2011-07-06  Adam Barth  <abarth@webkit.org>
2225
2226         Wean rebaseline-server off unexpected_results.json
2227         https://bugs.webkit.org/show_bug.cgi?id=64063
2228
2229         Reviewed by Eric Seidel.
2230
2231         In the process of changing this code to use full_results.json, I
2232         noticed that the code was broken (because it wasn't tested).  This
2233         patch also adds test coverage for the broken code.
2234
2235         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
2236         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
2237         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
2238
2239 2011-07-06  Adam Barth  <abarth@webkit.org>
2240
2241         Wean resultsjsonparser off unexpected_results.json
2242         https://bugs.webkit.org/show_bug.cgi?id=64061
2243
2244         Reviewed by Eric Seidel.
2245
2246         We're about to delete unexpected_results.json, so we need to remove all
2247         the users.
2248
2249         * Scripts/webkitpy/common/config/ports.py:
2250         * Scripts/webkitpy/common/net/resultsjsonparser.py:
2251         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
2252         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2253
2254 2011-07-06  Adam Barth  <abarth@webkit.org>
2255
2256         Cleanup result_summary.py
2257         https://bugs.webkit.org/show_bug.cgi?id=64057
2258
2259         Reviewed by Eric Seidel.
2260
2261         This class had a bunch of out-of-date docstrings that no longer make
2262         any sense.
2263
2264         * Scripts/webkitpy/layout_tests/models/result_summary.py:
2265
2266 2011-07-06  Adam Barth  <abarth@webkit.org>
2267
2268         Tweak the ChangeLog for DEPS rolls to say "Unreviewed" so that the
2269         patches can be landed by the commit-queue.
2270
2271         * Scripts/webkitpy/tool/steps/preparechangelogfordepsroll.py:
2272
2273 2011-07-06  Adam Barth  <abarth@webkit.org>
2274
2275         Normalize callers of logging.getLogger to use __name__
2276         https://bugs.webkit.org/show_bug.cgi?id=64054
2277
2278         Reviewed by Eric Seidel.
2279
2280         Most callers were already using __name__.  This patch makes everyone
2281         consistent.
2282
2283         * Scripts/webkitpy/common/checkout/diff_parser.py:
2284         * Scripts/webkitpy/common/checkout/scm/svn.py:
2285         * Scripts/webkitpy/common/net/statusserver.py:
2286         * Scripts/webkitpy/common/system/executive.py:
2287         * Scripts/webkitpy/common/system/file_lock.py:
2288         * Scripts/webkitpy/common/system/user.py:
2289         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2290         * Scripts/webkitpy/layout_tests/models/result_summary.py:
2291         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2292         * Scripts/webkitpy/layout_tests/port/chromium.py:
2293         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2294         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2295         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2296         * Scripts/webkitpy/layout_tests/port/gtk.py:
2297         * Scripts/webkitpy/layout_tests/port/http_lock.py:
2298         * Scripts/webkitpy/layout_tests/port/mac.py:
2299         * Scripts/webkitpy/layout_tests/port/qt.py:
2300         * Scripts/webkitpy/layout_tests/port/server_process.py:
2301         * Scripts/webkitpy/layout_tests/port/webkit.py:
2302         * Scripts/webkitpy/layout_tests/port/win.py:
2303         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
2304         * Scripts/webkitpy/layout_tests/servers/http_server.py:
2305         * Scripts/webkitpy/layout_tests/servers/http_server_base.py:
2306         * Scripts/webkitpy/layout_tests/servers/websocket_server.py:
2307         * Scripts/webkitpy/layout_tests/views/printing.py:
2308         * Scripts/webkitpy/python24/versioning.py:
2309         * Scripts/webkitpy/style/checker.py:
2310         * Scripts/webkitpy/style/checkers/test_expectations.py:
2311         * Scripts/webkitpy/style/patchreader.py:
2312         * Scripts/webkitpy/to_be_moved/update_webgl_conformance_tests.py:
2313
2314 2011-07-06  Adam Barth  <abarth@webkit.org>
2315
2316         Move model classes out of layout_package and into models
2317         https://bugs.webkit.org/show_bug.cgi?id=64053
2318
2319         Reviewed by Eric Seidel.
2320
2321         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2322         * Scripts/webkitpy/common/net/layouttestresults.py:
2323         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
2324         * Scripts/webkitpy/common/net/resultsjsonparser.py:
2325         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
2326         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2327         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2328         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
2329         * Scripts/webkitpy/layout_tests/controllers/worker.py:
2330         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
2331         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2332         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2333         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py: Removed.
2334         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py: Removed.
2335         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py: Removed.
2336         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py: Removed.
2337         * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py: Removed.
2338         * Scripts/webkitpy/layout_tests/layout_package/test_input.py: Removed.
2339         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
2340         * Scripts/webkitpy/layout_tests/layout_package/test_results.py: Removed.
2341         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py: Removed.
2342         * Scripts/webkitpy/layout_tests/models: Added.
2343         * Scripts/webkitpy/layout_tests/models/__init__.py: Added.
2344         * Scripts/webkitpy/layout_tests/models/result_summary.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/result_summary.py.
2345         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_expectations.py.
2346         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py.
2347         * Scripts/webkitpy/layout_tests/models/test_failures.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_failures.py.
2348         * Scripts/webkitpy/layout_tests/models/test_failures_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py.
2349         * Scripts/webkitpy/layout_tests/models/test_input.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_input.py.
2350         * Scripts/webkitpy/layout_tests/models/test_results.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_results.py.
2351         * Scripts/webkitpy/layout_tests/models/test_results_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py.
2352         * Scripts/webkitpy/layout_tests/port/chromium.py:
2353         * Scripts/webkitpy/layout_tests/views/printing.py:
2354         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
2355         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2356         * Scripts/webkitpy/style_references.py:
2357         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
2358         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
2359         * Scripts/webkitpy/tool/bot/flakytestreporter_unittest.py:
2360         * Scripts/webkitpy/tool/commands/queues_unittest.py:
2361         * Scripts/webkitpy/tool/commands/rebaseline.py:
2362
2363 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
2364
2365         nrwt: fix typo preventing http startup on windows
2366         https://bugs.webkit.org/show_bug.cgi?id=64050
2367
2368         Reviewed by Eric Siedel.
2369
2370         NRWT checks whether it needs to start the servers by looking
2371         for "/http/" in the test names. After r90520, the leading slash
2372         is not present, so the check isn't quite right.
2373
2374         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2375         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2376
2377 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
2378
2379         nrwt: remove --use-apache from the command line
2380         https://bugs.webkit.org/show_bug.cgi?id=63358
2381
2382         Reviewed by Adam Barth.
2383
2384         Re-land the fix in bug 63358 - the prior fix was flawed in that
2385         we were trying to start the http server during check_sys_deps()
2386         but not actually setting up the environment properly. This
2387         broken prior to this change, but exposed by it, since the patch
2388         changes the chromium port to call the base class's checks.
2389
2390         * Scripts/webkitpy/common/system/executive.py:
2391         * Scripts/webkitpy/layout_tests/port/base.py:
2392         * Scripts/webkitpy/layout_tests/port/chromium.py:
2393         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2394         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2395         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2396         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
2397         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2398         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2399
2400 2011-07-06  Adam Barth  <abarth@webkit.org>
2401
2402         Move view-related code out of layout_package and into views
2403         https://bugs.webkit.org/show_bug.cgi?id=64051
2404
2405         Reviewed by Eric Seidel.
2406
2407         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2408         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2409         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
2410         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
2411         * Scripts/webkitpy/layout_tests/layout_package/metered_stream.py: Removed.
2412         * Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py: Removed.
2413         * Scripts/webkitpy/layout_tests/layout_package/printing.py: Removed.
2414         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py: Removed.
2415         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2416         * Scripts/webkitpy/layout_tests/views: Added.
2417         * Scripts/webkitpy/layout_tests/views/__init__.py: Added.
2418         * Scripts/webkitpy/layout_tests/views/metered_stream.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/metered_stream.py.
2419         * Scripts/webkitpy/layout_tests/views/metered_stream_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py.
2420         * Scripts/webkitpy/layout_tests/views/printing.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/printing.py.
2421         * Scripts/webkitpy/layout_tests/views/printing_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py.
2422
2423 2011-07-06  Adam Barth  <abarth@webkit.org>
2424
2425         Add goofy comments to explain why we have these files.
2426
2427         * Scripts/webkitpy/layout_tests/controllers/__init__.py:
2428         * Scripts/webkitpy/layout_tests/layout_package/__init__.py:
2429
2430 2011-07-06  Adam Barth  <abarth@webkit.org>
2431
2432         Move a bunch of classes out of layout_tests.layout_package and into layout_tests.controllers
2433         https://bugs.webkit.org/show_bug.cgi?id=64048
2434
2435         Reviewed by Eric Seidel.
2436
2437         * Scripts/webkitpy/layout_tests/controllers: Added.
2438         * Scripts/webkitpy/layout_tests/controllers/__init__.py: Added.
2439         * Scripts/webkitpy/layout_tests/controllers/manager.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager.py.
2440         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py.
2441         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py.
2442         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py.
2443         * Scripts/webkitpy/layout_tests/controllers/message_broker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker.py.
2444         * Scripts/webkitpy/layout_tests/controllers/message_broker_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py.
2445         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py.
2446         * Scripts/webkitpy/layout_tests/controllers/worker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/worker.py.
2447         * Scripts/webkitpy/layout_tests/layout_package/manager.py: Removed.
2448         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py: Removed.
2449         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py: Removed.
2450         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py: Removed.
2451         * Scripts/webkitpy/layout_tests/layout_package/message_broker.py: Removed.
2452         * Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py: Removed.
2453         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
2454         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: Removed.
2455         * Scripts/webkitpy/layout_tests/layout_package/worker.py: Removed.
2456         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2457
2458 2011-07-06  Adam Barth  <abarth@webkit.org>
2459
2460         Attempt to fix unittest on Leopard.  Let's enumerate the keys in sorted
2461         order so the test results are predictable.
2462
2463         * Scripts/webkitpy/tool/bot/irc_command.py:
2464         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
2465
2466 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
2467
2468         Reviewed by Eric Siedel.
2469
2470         nrwt: make the code be consistent about using test names instead of filenames or paths
2471
2472         https://bugs.webkit.org/show_bug.cgi?id=63597
2473
2474         It can be hard to figure out where in the code we're assuming
2475         test names are listed as unix-style relative filenames and where
2476         they are either absolute paths or relative paths following the
2477         host filesystem convention.
2478
2479         This patch changes things so that everything outside of the
2480         Port object uses (and must assume) unix-style relative
2481         filenames (with one exception, which is specifying host-local
2482         filenames as a list of test arguments on the command line).
2483
2484         This should make things clearer and more consistent, and also
2485         removes a lot of path conversion calls.
2486
2487         The changes in this patch outside of port/base.py are largely
2488         cosmetic changes of "filename" to "test". We add a few routines
2489         to the base Port implementation to manipulate test names to
2490         split them, figure out if we're referring to a directory of
2491         tests, etc.
2492
2493         * Scripts/webkitpy/common/net/layouttestresults.py:
2494         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
2495         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2496         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2497         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
2498         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
2499         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
2500         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
2501         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
2502         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
2503         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
2504         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
2505         * Scripts/webkitpy/layout_tests/layout_package/test_input.py:
2506         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
2507         * Scripts/webkitpy/layout_tests/layout_package/test_results.py:
2508         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py:
2509         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
2510         * Scripts/webkitpy/layout_tests/port/base.py:
2511         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2512         * Scripts/webkitpy/layout_tests/port/chromium.py:
2513         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2514         * Scripts/webkitpy/layout_tests/port/dryrun.py:
2515         * Scripts/webkitpy/layout_tests/port/test.py:
2516         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2517         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2518         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
2519         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
2520         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
2521         * Scripts/webkitpy/tool/commands/rebaseline.py:
2522         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
2523
2524 2011-07-06  Chris Guillory  <ctguil@chromium.org>
2525
2526         No review necessary.
2527
2528         Adding myself to committers list in committers.py.
2529         https://bugs.webkit.org/show_bug.cgi?id=63478
2530
2531         * Scripts/webkitpy/common/config/committers.py:
2532
2533 2011-07-06  Johnny Ding  <jnd@chromium.org>
2534
2535         https://bugs.webkit.org/show_bug.cgi?id=63857
2536         Implement createPopupMenu for Chromium DRT.
2537
2538         Reviewed by Tony Chang.
2539
2540         * DumpRenderTree/chromium/WebViewHost.cpp:
2541         (WebViewHost::createPopupMenu):
2542         (WebViewHost::~WebViewHost):
2543         * DumpRenderTree/chromium/WebViewHost.h:
2544
2545 2011-07-06  Mihai Parparita  <mihaip@chromium.org>
2546
2547         Unreviewed, rolling out r90503.
2548         http://trac.webkit.org/changeset/90503
2549         https://bugs.webkit.org/show_bug.cgi?id=63358
2550
2551         Makes Chromium Windows canary bots hang.
2552
2553         * Scripts/webkitpy/layout_tests/port/base.py:
2554         * Scripts/webkitpy/layout_tests/port/chromium.py:
2555         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2556         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2557         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2558         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
2559         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2560         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2561
2562 2011-07-06  Adam Barth  <abarth@webkit.org>
2563
2564         Move buildbot URL in webkitpy to config.urls
2565         https://bugs.webkit.org/show_bug.cgi?id=64041
2566
2567         Reviewed by Eric Seidel.
2568
2569         This patch prepares for using the buildbot class to talk to the
2570         Chromium buildbots.
2571
2572         * Scripts/webkitpy/common/config/urls.py:
2573         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
2574         * Scripts/webkitpy/tool/commands/download_unittest.py:
2575         * Scripts/webkitpy/tool/commands/queries.py:
2576         * Scripts/webkitpy/tool/mocktool.py:
2577         * Scripts/webkitpy/tool/steps/ensurebuildersaregreen.py:
2578
2579 2011-07-06  Kevin Ollivier  <kevino@theolliviers.com>
2580
2581         [wx] Unreviewed build fix, add the correct path to the waf build settings.
2582         
2583         * wx/packaging/build-win-installer.py:
2584
2585 2011-07-06  Adam Barth  <abarth@webkit.org>
2586
2587         Add missing include.
2588
2589         * Scripts/webkitpy/tool/bot/irc_command.py:
2590
2591 2011-07-06  Adam Barth  <abarth@webkit.org>
2592
2593         Add roll-chromium-deps command to sheriff-bot
2594         https://bugs.webkit.org/show_bug.cgi?id=64037
2595
2596         Reviewed by Eric Seidel.
2597
2598         Lo, and the DEPS were rolled on command.
2599
2600         * Scripts/webkitpy/tool/bot/irc_command.py:
2601         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
2602         * Scripts/webkitpy/tool/bot/sheriff.py:
2603         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
2604         * Scripts/webkitpy/tool/commands/roll.py:
2605         * Scripts/webkitpy/tool/commands/roll_unittest.py:
2606         * Scripts/webkitpy/tool/steps/updatechromiumdeps.py:
2607
2608 2011-07-06  Eric Seidel  <eric@webkit.org>
2609
2610         Split Driver, DriverOutput and DriverInput out into a new driver.py file
2611         https://bugs.webkit.org/show_bug.cgi?id=64031
2612
2613         Reviewed by Adam Barth.
2614
2615         This ended up kinda large and I'm happy to break it up if desired.
2616
2617         As part of cleaning up the Port module, I split Driver, DriverInput and DriverOutput
2618         out into its own driver.py file.
2619
2620         I also cleaned up how we import Port objects, by making port/__init__.py
2621         expose Port, Driver, DriverInput, and DriverOutput.  I removed exposure of
2622         some various test objects as they were causing circular dependencies.
2623         base.py is now an implementation detail of port.  No longer is "base" referenced
2624         anywhere in the code.
2625
2626         I also made Driver.__init__ non-virtual so that sub-classers could all share
2627         the data storage on Driver.  This more closely matches how other subclasses
2628         work throughout our python code.
2629
2630         Standard line-unwrapping and doc-string removal as I went through the code.
2631
2632         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2633         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
2634         * Scripts/webkitpy/layout_tests/port/__init__.py:
2635         * Scripts/webkitpy/layout_tests/port/base.py:
2636         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2637         * Scripts/webkitpy/layout_tests/port/chromium.py:
2638         * Scripts/webkitpy/layout_tests/port/driver.py: Added.
2639         * Scripts/webkitpy/layout_tests/port/dryrun.py:
2640         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
2641         * Scripts/webkitpy/layout_tests/port/test.py:
2642         * Scripts/webkitpy/layout_tests/port/webkit.py:
2643         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2644         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
2645         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
2646
2647 2011-07-06  Adam Roben  <aroben@apple.com>
2648
2649         Clear TestFailures's getNumberOfFailingTests cache
2650
2651         Now that we support NRWT, we need to clear out old zero values for the NRWT bots that might
2652         have been cached by old versions of TestFailures.
2653
2654         Fixes <http://webkit.org/b/64029> TestFailures page still doesn't show NRWT bots if you've
2655         used it recently
2656
2657         Reviewed by Adam Barth.
2658
2659         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
2660         (Builder.prototype.getNumberOfFailingTests): Added a .version property to the cache data
2661         that must match the current version. Otherwise we'll ignore the cached data.
2662
2663 2011-06-24  Dirk Pranke  <dpranke@chromium.org>
2664
2665         Reviewed by Adam Barth.
2666
2667         nrwt: remove --use-apache from the command line
2668         https://bugs.webkit.org/show_bug.cgi?id=63358
2669
2670         This change removes the --use-apache command line argument. It
2671         was initially put in when we were trying to get the cygwin
2672         apache instance to work with Chromium win, but that code has
2673         bitrotted and doesn't work at all now.
2674
2675         Arguably we should remove all of the code to allow for a choice
2676         of web servers, but since we may still want to switch off of
2677         LigHTTPd at some point on Windows, I'll leave the rest in for
2678         now (or at least yank it in a different change.)
2679
2680         * Scripts/webkitpy/layout_tests/port/base.py:
2681         * Scripts/webkitpy/layout_tests/port/chromium.py:
2682         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2683         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2684         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
2685         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2686
2687 2011-07-06  Dimitri Glazkov  <dglazkov@chromium.org>
2688
2689         garden-o-matic should use a favicon to indicate current state.
2690         https://bugs.webkit.org/show_bug.cgi?id=64027
2691
2692         Reviewed by Adam Barth.
2693
2694         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-green.png: Added.
2695         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-red.png: Added.
2696         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html: Added blurb about icons.
2697         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js: Added setting icon state.
2698         * Scripts/webkitpy/tool/servers/gardeningserver.py: Allowed the use of favicons.
2699
2700 2011-07-06  Daniel Bates  <dbates@rim.com>
2701
2702         Fix webkitpy error after changeset 90483 <http://trac.webkit.org/changeset/90483>; Add missing
2703         parameter to _fetch_file_from_results lamdba function.
2704
2705         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2706
2707 2011-07-06  Kevin Ollivier  <kevino@theolliviers.com>
2708
2709         [wx] Unreviewed build fix. Generate the proper install_name for dylibs on Mac.
2710
2711         * wx/packaging/build-mac-installer.py:
2712
2713 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2714
2715         [GTK] Only set env. variables in start when strictly needed
2716         https://bugs.webkit.org/show_bug.cgi?id=64026
2717
2718         Reviewed by Eric Seidel.
2719
2720         * Scripts/webkitpy/layout_tests/port/gtk.py: add a
2721         setup_environ_for_server method in GtkPort and set as many
2722         env. variables as possible there.
2723
2724 2011-07-06  Adam Roben  <aroben@apple.com>
2725
2726         Teach TestFailures how to load, parse, and interpret NRWT test results
2727
2728         Fixes <http://webkit.org/b/61877> TestFailures page doesn't show testers that use
2729         new-run-webkit-tests
2730
2731         Reviewed by Adam Barth.
2732
2733         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
2734         (Builder.prototype.failureDiagnosisTextAndURL): Added support for the new 'flaky' failure
2735         type. For now we don't account for image-only flakes (but TestFailures doesn't deal with
2736         pixel tests at all currently).
2737         (Builder.prototype.getNumberOfFailingTests): Relaxed the regex used for parsing the number
2738         of failing tests from the buildbot output. Make sure not to count "new passes" (a new
2739         category introduced by NRWT) as failures.
2740
2741         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
2742         (LayoutTestResultsLoader.prototype.start): Moved a bunch of code from here to
2743         _fetchAndParseORWTResults. This function now attempts to load NRWT results, then falls back
2744         to loading ORWT results.
2745         (LayoutTestResultsLoader.prototype._fetchAndParseNRWTResults): Added. Tries to load and
2746         parse the full_results.json file for the given build.
2747         (LayoutTestResultsLoader.prototype._fetchAndParseORWTResults): Added. Code came from start.
2748         Fixed a bug along the way where we were sometimes calling the error callback instead of the
2749         success callback when all tests passed.
2750
2751         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NRWTResultsParser.js: Added.
2752         (NRWTResultsParser): Do-nothing constructor.
2753         (NRWTResultsParser.prototype.parse): Uses eval() (eww!) to get the test results out of the
2754         JS string, then iterates over all the tests in the results data and builds up a data
2755         structure matching what ORWTResultsParser returns.
2756
2757         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
2758         (Array.prototype.contains):
2759         (String.prototype.contains):
2760         Added these simple helper functions.
2761
2762         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
2763         NRWTResultsParser.js.
2764
2765 2011-07-06  Adam Roben  <aroben@apple.com>
2766
2767         Extract code to parse ORWT's results.html file into its own class
2768
2769         Prep work for fixing <http://webkit.org/b/61877> TestFailures page doesn't show testers that
2770         use new-run-webkit-tests
2771
2772         Reviewed by Adam Barth.
2773
2774         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
2775         (LayoutTestResultsLoader.prototype.start): Moved a bunch of code from here...
2776
2777         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ORWTResultsParser.js: Added.
2778         (ORWTResultsParser):
2779         (ORWTResultsParser.prototype.parse):
2780         ...to here.
2781
2782         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
2783         ORWTResultsParser.js.
2784
2785 2011-07-06  Eric Seidel  <eric@webkit.org>
2786
2787         Unreviewed.  Fix a unittest from hitting the network. :)
2788
2789         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2790
2791 2011-07-06  Eric Seidel  <eric@webkit.org>
2792
2793         Unreviewed.  Fixing an exception in test-webkitpy, caused by bug 64006.
2794
2795         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2796
2797 2011-07-06  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
2798
2799         Run an Xvfb instance to run the API tests on, like we do for the
2800         layout tests.
2801
2802         Reviewed by Xan Lopez.
2803
2804         * Scripts/run-gtk-tests:
2805
2806 2011-07-06  Daniel Bates  <dbates@webkit.org>
2807
2808         Make SCM unit tests faster
2809         https://bugs.webkit.org/show_bug.cgi?id=63883
2810
2811         Speeds up the SCM unit tests by a factor of 4. Currently, we create a
2812         mock SVN repo for each test_ method in SVNTest and GitTest and creating
2813         this repo is expensive.
2814
2815         Instead, it is significantly faster to create the mock SVN repo once
2816         and then perform a filesystem copy of it for each test_ method.
2817
2818         Note, Python 2.7's unittest module implements support for per class and
2819         per module setup and tear down methods which could be used to implement
2820         similar functionality. At the time of writing, test-webkitpy is designed
2821         to support Python 2.5. So, we can't take advantage of these Python 2.7
2822         features :(
2823
2824         Reviewed by Eric Seidel.
2825
2826         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
2827
2828 2011-07-06  Adam Barth  <abarth@webkit.org>
2829
2830         Remove unused parameter from _add_test in test_expectations.py
2831         https://bugs.webkit.org/show_bug.cgi?id=64008
2832
2833         Reviewed by Eric Seidel.
2834
2835         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
2836
2837 2011-07-06  Adam Barth  <abarth@webkit.org>
2838
2839         Use old-run-webkit-tests when --leaks is present
2840         https://bugs.webkit.org/show_bug.cgi?id=64012
2841
2842         Reviewed by Adam Roben.
2843
2844         We'll remove this branch once support of --leaks is added to
2845         new-run-webkit-tests.
2846
2847         * Scripts/run-webkit-tests:
2848
2849 2011-07-06  Benjamin Poulain  <benjamin@webkit.org>
2850
2851         [Qt] [WK2] Add touch mocking to Qt's MiniBrowser
2852         https://bugs.webkit.org/show_bug.cgi?id=63995
2853
2854         Reviewed by Andreas Kling.
2855
2856         The mouse events are intercepted through QCoreApplication::notify(),
2857         and fake touch events are generated for mouse events when necessary.
2858
2859         If touch events are received from the platform, we assume the current
2860         hardware has a touch screen and we stop generating fake events.
2861
2862         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2863         (isTouchEvent):
2864         (isMouseEvent):
2865         (MiniBrowserApplication::MiniBrowserApplication):
2866         (MiniBrowserApplication::notify):
2867         * MiniBrowser/qt/MiniBrowserApplication.h:
2868
2869 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2870
2871         Reviewed by Gustavo Noronha.
2872
2873         * Scripts/webkitpy/layout_tests/port/gtk.py: use Popen correctly.
2874
2875 2011-07-06  Eric Seidel  <eric@webkit.org>
2876
2877         webkit-patch failure-reason does not understand NRWT results
2878         https://bugs.webkit.org/show_bug.cgi?id=64006
2879
2880         Reviewed by Adam Barth.
2881
2882         Adam Barth tells me failure-reason should be deleted soon,
2883         but aroben's fancy new TestFailures/ page doesn't work for
2884         NRWT yet, and I needed to know when
2885         third-party-cookie-relaxing started failing.
2886
2887         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
2888         * Scripts/webkitpy/tool/commands/queries.py:
2889
2890 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2891
2892         [GTK] Add missing environment variables in NWRT
2893         https://bugs.webkit.org/show_bug.cgi?id=64004
2894
2895         Reviewed by Gustavo Noronha Silva.
2896
2897         Add missing environment variables needed by the GTK+ port.
2898
2899         * Scripts/webkitpy/layout_tests/port/gtk.py: ditto.
2900
2901 2011-07-06  Adam Roben  <aroben@apple.com>
2902
2903         Make run-api-tests output intelligible on the bots
2904
2905         run-api-tests's output on the bots was very hard to read for two reasons:
2906           1) It was mixing its own output with gtest's
2907           2) It was using ANSI escape sequences to print colored text, but the bots don't support
2908              that
2909
2910         Now, the --verbose flag turns off almost all of run-api-tests's own output so that gtest's
2911         output will not be obscured. We still print "Timeout" messages even in verbose mode, since
2912         gtest doesn't have any native support for timeouts. Also, when our output is being
2913         redirected to a file, we don't print ANSI escape sequences.
2914
2915         Fixes <http://webkit.org/b/63996> It's very hard to read run-api-tests output on the bots
2916
2917         Reviewed by Sam Weinig.
2918
2919         * Scripts/run-api-tests: Don't use the :constants interface of Term::ANSIColor. We want to
2920         use the colored function instead.
2921         (runAllTestsInSuite): Don't print out suite names in verbose mode; gtest will give us enough
2922         context that they aren't needed.
2923         (runTest): Don't print out test names or pass/fail messages in verbose mode; gtest will do
2924         that for us. Also, use the new possiblyColored function instead of always coloring output.
2925         (possiblyColored): Added. When printing to a tty, returns the string with the appropriate
2926         ANSI color escape sequences added. Otherwise just returns the string unmodified.
2927
2928 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2929
2930         Reviewed by Gustavo Noronha.
2931
2932         Launch Xvfb (mostly) the same way we were launching it in the
2933         bots, for maximum compatibility.
2934
2935         * Scripts/webkitpy/layout_tests/port/gtk.py:
2936
2937 2011-07-06  Andreas Kling  <kling@webkit.org>
2938
2939         [Qt][WK2] Split Qt API into two different web views (touch and desktop)
2940         https://bugs.webkit.org/show_bug.cgi?id=63950
2941
2942         Co-authored with Benjamin Poulain <benjamin@webkit.org>.
2943
2944         Reviewed by Kenneth Rohde Christiansen.
2945
2946         Port WTR/Qt to use QDesktopWebView and force a 800x600 viewport.
2947
2948         The Qt MiniBrowser now defaults to QDesktopWebView but can be started
2949         in QTouchWebView mode by passing -touch on the command line.
2950
2951         A lot of functionality has been removed and/or disabled from
2952         MiniBrowser until we reintroduce the necessary API's.
2953
2954         * MiniBrowser/qt/BrowserView.cpp:
2955         (BrowserView::BrowserView):
2956         (BrowserView::~BrowserView):
2957         (BrowserView::load):
2958         (BrowserView::view):
2959         (BrowserView::touchWebView):
2960         (BrowserView::desktopWebView):
2961         * MiniBrowser/qt/BrowserView.h:
2962         * MiniBrowser/qt/BrowserWindow.cpp:
2963         (newPageFunction):
2964         (BrowserWindow::BrowserWindow):
2965         (BrowserWindow::webView):
2966         (BrowserWindow::newWindow):
2967         (BrowserWindow::showUserAgentDialog):
2968         (BrowserWindow::updateUserAgentList):
2969         * MiniBrowser/qt/BrowserWindow.h:
2970         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2971         (MiniBrowserApplication::handleUserOptions):
2972         * MiniBrowser/qt/MiniBrowserApplication.h:
2973         (WindowOptions::WindowOptions):
2974         * MiniBrowser/qt/UrlLoader.cpp:
2975         (UrlLoader::UrlLoader):
2976         * MiniBrowser/qt/main.cpp:
2977         (main):
2978         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
2979         (WTR::WebView::wkView):
2980         (WTR::WebView::pageRef):
2981         (WTR::WebView::WebView):
2982         (WTR::PlatformWebView::page):
2983         (WTR::PlatformWebView::windowFrame):
2984         (WTR::PlatformWebView::setWindowFrame):
2985
2986 2011-07-06  Eric Seidel  <eric@webkit.org>
2987
2988         Unreviewed..
2989
2990         * Scripts/run-webkit-tests:
2991         (useNewRunWebKitTests):
2992          - Fix the pass-thru of --gtk and --qt for old-run-webkit-tests,
2993            it was previously broken by the existence of isGtk/isQt checks which
2994            removed the --gtk/--qt args from @ARGV.
2995
2996 2011-07-06  Eric Seidel  <eric@webkit.org>
2997
2998         new-run-webkit-tests uses a 35s timeout for Mac and 6s for all other ports
2999         https://bugs.webkit.org/show_bug.cgi?id=63983
3000
3001         Reviewed by Xan Lopez.
3002
3003         All WebKit ports want to use a 35s timeout.  Only chromium currently
3004         uses a 6s timeout.  Eventually we'll move all WebKit ports to 6s too
3005         to further speed up new-run-webkit-tests runs, but that's going to be a while.
3006
3007         * Scripts/webkitpy/layout_tests/port/mac.py:
3008         * Scripts/webkitpy/layout_tests/port/webkit.py:
3009         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3010
3011 2011-07-06  Xan Lopez  <xlopez@igalia.com>
3012
3013         Unreviewed.
3014
3015         NWRT actually seems to work locally, so it should work in the bots
3016         as long as we take care of the TIMEOUT issue (which is being fixed
3017         as I write this). Let's give this another shot and fix whatever
3018         breaks.
3019
3020         * Scripts/run-webkit-tests:
3021         (useNewRunWebKitTests):
3022
3023 2011-07-06  Xan Lopez  <xlopez@igalia.com>
3024
3025         Unreviewed.
3026
3027         Disable NWRT on GTK. At the very least it needs an updated exected
3028         results file since NWRT is more sensitive to timeouts, not sure
3029         what the exact problem is.
3030
3031         * Scripts/run-webkit-tests:
3032         (useNewRunWebKitTests):
3033
3034 2011-07-06  Adam Barth  <abarth@webkit.org>
3035
3036         Fix global variable leak detected by noglobals.
3037
3038         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
3039
3040 2011-07-06  Adam Barth  <abarth@webkit.org>
3041
3042         garden-o-matic shouldn't show flaky tests by default
3043         https://bugs.webkit.org/show_bug.cgi?id=63979
3044
3045         Reviewed by Ojan Vafai.
3046
3047         If any of the actual test results are PASS, the test is flaky and we
3048         can ignore it (at least for now).  If we get awesome at dealing with
3049         failing tests, we might want to tackle flaky tests by refering the user
3050         to the flakiness dashboard, which is way more awesome at dealing with
3051         flaky tests that this tool ever will be.
3052
3053         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
3054         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
3055
3056 2011-07-05  Adam Barth  <abarth@webkit.org>
3057
3058         garden-o-matic should list the failing tests
3059         https://bugs.webkit.org/show_bug.cgi?id=63978
3060
3061         Reviewed by Eric Seidel.
3062
3063         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js: Added.
3064         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
3065         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
3066         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
3067         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
3068         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
3069         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js: Added.
3070         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js: Added.
3071         * Scripts/webkitpy/tool/servers/gardeningserver.py:
3072
3073
3074 2011-07-05  Eric Seidel  <eric@webkit.org>
3075
3076         Re-enable NRWT on Leopard now that bug 63973 is fixed.
3077
3078         Reviewed by Adam Barth.
3079
3080         I also made the should-enable expression a black-list instead
3081         of a white-list since every bot on build.webkit.org is
3082         now using NRWT except windows and WebKit2 bots.
3083
3084         * Scripts/run-webkit-tests:
3085         (useNewRunWebKitTests):
3086
3087 2011-07-05  Eric Seidel  <eric@webkit.org>
3088
3089         Perl string concat failure on Leopard (causes NRWT to fail)
3090         https://bugs.webkit.org/show_bug.cgi?id=63973
3091
3092         Reviewed by Daniel Bates.
3093
3094         Leopard's "xcodebuild -version" has a different format from Snow Leopard/Lion
3095         thus the version parsing logic in webkitdirs.pm seems to fail.
3096
3097         Default to "3.0" instead of undef when we can't parse the version number.
3098         This should send us down the correct path for Leopard.
3099
3100         * Scripts/webkitdirs.pm:
3101         (determineBaseProductDir):
3102
3103 2011-07-05  Kevin Ollivier  <kevino@theolliviers.com>
3104
3105         [wx] Unreviewed build fix after addition of Source/WebCore/xml/parser dir.
3106
3107         * waf/build/settings.py:
3108
3109 2011-07-05  Eric Seidel  <eric@webkit.org>
3110
3111         Disable NRWT for Leopard until https://bugs.webkit.org/show_bug.cgi?id=63973 can be fixed.
3112
3113         Reviewed by Adam Barth.
3114
3115         * Scripts/run-webkit-tests:
3116         (useNewRunWebKitTests):
3117
3118 2011-07-05  Eric Seidel  <eric@webkit.org>
3119
3120         Make run-webkit-tests pass --verbose to new-run-webkit-tests on qt/gtk bots.
3121
3122         Reviewed by Adam Barth.
3123
3124         * Scripts/run-webkit-tests:
3125         (runningOnBuildBot):
3126
3127 2011-07-05  Adam Barth  <abarth@webkit.org>
3128
3129         Add results_unittests.js to garden-o-matic
3130         https://bugs.webkit.org/show_bug.cgi?id=63971
3131
3132         Reviewed by Eric Seidel.
3133
3134         Adding these unit tests forced me to commit to an API for the results
3135         module.  I'm not sure if this is the API we'll end up with, but it's
3136         something not entirely unreasonable.
3137
3138         Also, I've changed results.js over to WebKit style, as discussed previously.
3139
3140         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
3141         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
3142         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js: Added.
3143         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
3144
3145 2011-07-05  Eric Seidel  <eric@webkit.org>
3146
3147         Teach run-webkit-tests how to translate --qt and --gtk for new-run-webkit-tests.
3148
3149         Reviewed by Adam Barth.
3150
3151         * Scripts/run-webkit-tests:
3152
3153 2011-07-05  Adam Barth  <abarth@webkit.org>
3154
3155         Add first unit tests for garden-o-matic
3156         https://bugs.webkit.org/show_bug.cgi?id=63969
3157
3158         Reviewed by Eric Seidel.
3159
3160         I've also reformated base.js to match WebKit style, as discussed in the
3161         previous bug.
3162
3163         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
3164         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js: Added.
3165         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html: Added.
3166
3167 2011-07-05  Eric Seidel  <eric@webkit.org>
3168
3169         Move Qt, Gtk and Leopard to NRWT.
3170
3171         We'll roll out any of these if we see problems.
3172
3173         Reviewed by Adam Barth.
3174
3175         * Scripts/run-webkit-tests:
3176         (useNewRunWebKitTests):
3177
3178 2011-07-05  Adam Barth  <abarth@webkit.org>
3179
3180         garden-o-matic should understand which tests have unexpected failures
3181         https://bugs.webkit.org/show_bug.cgi?id=63965
3182
3183         Reviewed by Eric Seidel.
3184
3185         Currently, this code just logs the list of unexpected failures to the
3186         console.  A future patch will do something useful with this
3187         information.  A testing harness will also come in the next patch.
3188
3189         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js: Added.
3190         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
3191         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
3192         * Scripts/webkitpy/tool/servers/gardeningserver.py:
3193
3194 2011-07-05  Adam Barth  <abarth@webkit.org>
3195
3196         garden-o-matic should know how to fetch test results from the (NRWT) bots
3197         https://bugs.webkit.org/show_bug.cgi?id=63959
3198
3199         Reviewed by Eric Seidel.
3200
3201         Currently, this code just fetches the results and expectations JSON
3202         blob for Linux from the server and logs it to the console.
3203
3204         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
3205         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js: Added.
3206         * Scripts/webkitpy/tool/servers/gardeningserver.py:
3207
3208 2011-07-05  Dirk Pranke  <dpranke@chromium.org>
3209
3210         Reviewed by Eric Seidel.
3211
3212         Python tests are failing on leopard
3213         https://bugs.webkit.org/show_bug.cgi?id=63842
3214
3215         Fix the missing flush() call that appears to be needed by the
3216         logging package in python 2.5. Also fix the regressions
3217         introduced in the run_webkit_test tests when we switched from
3218         thread to processes (since processes aren't available on 2.5).
3219
3220         Also fix a missing "from __future__ import with_statement' in
3221         testfilehandler (unreviewed).
3222
3223         * Scripts/webkitpy/layout_tests/layout_package/metered_stream.py:
3224         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3225         * TestResultServer/handlers/testfilehandler.py:
3226
3227 2011-07-05  Eric Seidel  <eric@webkit.org>
3228
3229         Reviewed by Adam Barth.
3230
3231         re-enabling new-run-webkit-tests on the Snow Leopard WebKit1 bots
3232         now that we believe the stray http-server issue to be solved.
3233
3234         * Scripts/run-webkit-tests:
3235         (useNewRunWebKitTests):
3236
3237 2011-07-05  Adam Barth  <abarth@webkit.org>
3238
3239         Turns out we need to support numerals too because that's what jQuery
3240         uses by default.
3241
3242         * TestResultServer/handlers/testfilehandler.py:
3243
3244 2011-07-05  Dirk Pranke  <dpranke@chromium.org>
3245
3246         nrwt: allow for multiple http shards
3247         https://bugs.webkit.org/show_bug.cgi?id=63116
3248
3249         Reviewed by Tony Chang.
3250
3251         This modifies the sharding logic to support multiple http
3252         shards, but for now we clamp to one shard until we can test
3253         perf impact and flakiness impact.
3254
3255         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3256         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
3257
3258 2011-07-05  Dirk Pranke  <dpranke@chromium.org>
3259
3260         Re-land nrwt: make sharding tests needing locks less hard-coded
3261         https://bugs.webkit.org/show_bug.cgi?id=63112
3262
3263         Reviewed by Ojan Vafai.
3264
3265         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3266         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
3267
3268 2011-07-05  Adam Barth  <abarth@webkit.org>
3269
3270         TestResultServer should support callback parameter for JSON
3271         https://bugs.webkit.org/show_bug.cgi?id=63961
3272
3273         Reviewed by Ojan Vafai.
3274
3275         This makes using jQuery.ajax so much more pleasant.
3276
3277         * TestResultServer/handlers/testfilehandler.py:
3278
3279 2011-07-05  Eric Seidel  <eric@webkit.org>
3280
3281         buildbot needs to understand whether NRWT exited early after having too many failures
3282         https://bugs.webkit.org/show_bug.cgi?id=63839
3283
3284         Reviewed by Adam Barth.
3285
3286         Fix-up the exited early messages printed by NRWT so that
3287         the buildbot can parse them as expected.
3288         It looks for lines using "if line.find('Exiting early') >= 0:"
3289
3290         I also plumbed the "early exit" status through to results.json
3291         in the form of an "interrupted" bool.  It was unclear to me
3292         if results.json already had enough information to compute this bool
3293         itself.  It's possible Ojan could come up with a better fix.
3294
3295         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3296         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
3297         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
3298         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
3299
3300 2011-07-05  Eric Seidel  <eric@webkit.org>
3301
3302         new-run-webkit-tests fails to start http server if one is already running
3303         https://bugs.webkit.org/show_bug.cgi?id=63956
3304
3305         Reviewed by Adam Barth.
3306
3307         * BuildSlaveSupport/kill-old-processes:
3308          - Removed ^M line endings and sorted the process names.
3309          - There are no functional changes to this file.
3310            (except that now that its using unix line endings it can be executed directly!)
3311         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
3312          - Change the name to 'httpd' to match old-run-webkit-test paths.
3313
3314 2011-07-05  Adam Barth  <abarth@webkit.org>
3315
3316         Add basic ajax support to garden-o-matic
3317         https://bugs.webkit.org/show_bug.cgi?id=63874
3318
3319         Reviewed by Eric Seidel.
3320
3321         This patch adds some basic infrastructure to garden-o-matic.
3322         Currently, the infrastructure is just used to make the "quit" command
3323         use Ajax, but in the future, this infrastructure will be used to do
3324         more sophistocated remote proceedure calls.
3325
3326         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
3327         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js: Added.
3328         * Scripts/webkitpy/tool/servers/gardeningserver.py:
3329         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
3330
3331 2011-07-05  Adam Barth  <abarth@webkit.org>
3332
3333         Add trivial garden-o-matic command (with server)
3334         https://bugs.webkit.org/show_bug.cgi?id=63872
3335
3336         Reviewed by Eric Seidel.
3337
3338         This patch adds a trivial "Hello, world" garden-o-matic command,
3339         complete with an HTTP server.  This command re-uses a bunch of code
3340         from the existing rebaseline-server command.  Over time, this command
3341         will grow to be a tool useful for gardening the WebKit tree.
3342
3343         * Scripts/webkitpy/tool/commands/__init__.py:
3344         * Scripts/webkitpy/tool/commands/abstractservercommand.py: Added.
3345         * Scripts/webkitpy/tool/commands/gardenomatic.py: Added.
3346         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
3347         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html: Added.
3348         * Scripts/webkitpy/tool/servers/gardeningserver.py: Added.
3349         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
3350
3351 2011-07-05  Adam Roben  <aroben@apple.com>
3352
3353         Make prepare-ChangeLog include modified Perl functions in its ChangeLog template
3354
3355         This is a very simple first cut. Functions must start with a line that starts with "sub "
3356         and end with a line that starts with a closing brace. No leading whitespace is allowed.
3357         Package names aren't parsed at all.
3358
3359         Fixes <http://webkit.org/b/21591> prepare-ChangeLog should know how to find functions in
3360         Perl files
3361
3362         Reviewed by David Kilzer.
3363
3364         * Scripts/prepare-ChangeLog: Removed redundant code that ignored certain files when
3365         searching for function line ranges. This is already done inside the get_function_line_ranges
3366         function.
3367         (get_function_line_ranges): Cleaned up coding style a little bit. Call
3368         get_function_line_ranges_for_perl for files with .pl and .pm extensions. For files with an
3369         unknown extension or no extension, read the shebang line to try to determine the script
3370         interpreter. Call get_function_line_ranges_for_perl if the interpreter seems to be Perl.
3371         (get_function_line_ranges_for_perl): Added. Does extremely basic parsing of the file to find
3372         lines starting with "sub " or "}".
3373
3374 2011-07-05  Adam Roben  <aroben@apple.com>
3375
3376         Clean up run-api-tests output on Windows
3377
3378         We were mixing run-api-tests output with gtest output, and the result was a mess.
3379
3380         Fixes <http://webkit.org/b/63954> run-api-tests output is very confusing on Windows
3381
3382         Reviewed by Dan Bates.
3383
3384         * Scripts/run-api-tests:
3385         (runTest):
3386         (populateTests):
3387         Made Windows use the formerly-Mac-only codepaths which correctly suppress gtest output
3388         except in verbose mode. Most changes are due to indentation. Use of the arch utility is now
3389         the only Mac-specific thing in these functions. Some other platform-specific code was moved
3390         from here...
3391
3392         (prepareEnvironmentForRunningTestTool):
3393         (testToolPath):
3394         ...to here.
3395
3396 2011-07-04  Gavin Barraclough  <barraclough@apple.com>
3397
3398         Reviewed by Sam Weinig.
3399
3400         https://bugs.webkit.org/show_bug.cgi?id=16652
3401         Firefox and JavaScriptCore differ in Number.toString(integer)
3402
3403         Added forwarding header.
3404
3405         * DumpRenderTree/ForwardingHeaders/wtf/StdLibExtras.h: Added.
3406
3407 2011-07-04  Anders Carlsson  <andersca@apple.com>
3408
3409         Try to fix Qt build.
3410
3411         * DumpRenderTree/TestNetscapePlugIn/Tests/NPRuntimeRemoveProperty.cpp:
3412
3413 2011-07-04  Anders Carlsson  <andersca@apple.com>
3414
3415         NP_RemoveProperty is not called back by Safari when delete npObject.prop is encountered in JavaScript
3416         https://bugs.webkit.org/show_bug.cgi?id=63915
3417         <rdar://problem/7124300>
3418
3419         Reviewed by Sam Weinig.
3420
3421         Extend the NPRuntimeRemoveProperty to handle delete object.property from JavaScript.
3422
3423         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
3424         (PluginTest::NPN_IdentifierIsString):
3425         (PluginTest::NPN_UTF8FromIdentifier):
3426         (PluginTest::NPN_IntFromIdentifier):
3427         (PluginTest::NPN_RetainObject):
3428         (PluginTest::NPN_ReleaseObject):
3429         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
3430         (PluginTest::Object::removeProperty):
3431         (PluginTest::Object::identifierIs):
3432         (PluginTest::Object::NP_RemoveProperty):
3433         (PluginTest::Object::npClass):
3434         * DumpRenderTree/TestNetscapePlugIn/Tests/NPRuntimeRemoveProperty.cpp:
3435         (NPRuntimeRemoveProperty::TestObject::TestObject):
3436         (NPRuntimeRemoveProperty::TestObject::hasProperty):
3437         (NPRuntimeRemoveProperty::TestObject::getProperty):
3438         (NPRuntimeRemoveProperty::TestObject::removeProperty):
3439         (NPRuntimeRemoveProperty::PluginObject::PluginObject):
3440         (NPRuntimeRemoveProperty::PluginObject::~PluginObject):
3441         (NPRuntimeRemoveProperty::PluginObject::hasMethod):
3442         (NPRuntimeRemoveProperty::PluginObject::invoke):
3443         (NPRuntimeRemoveProperty::PluginObject::hasProperty):
3444         (NPRuntimeRemoveProperty::PluginObject::getProperty):
3445         (NPRuntimeRemoveProperty::NPP_GetValue):
3446
3447 2011-07-04  Leandro Pereira  <leandro@profusion.mobi>
3448
3449         Reviewed by Kent Tamura.
3450
3451         [EFL] Add LayoutTestControllerEfl.
3452         https://bugs.webkit.org/show_bug.cgi?id=61974
3453
3454         Only the core features are implemented. Most functions are actually
3455         stubs and will be implemented as soon as its features are needed.
3456         Build system changes will be made as soon as DRT is completely
3457         upstreamed.
3458
3459         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp: Added.
3460         (LayoutTestController::~LayoutTestController):
3461         (LayoutTestController::addDisallowedURL):
3462         (LayoutTestController::clearBackForwardList):
3463         (LayoutTestController::copyDecodedHostName):
3464         (LayoutTestController::copyEncodedHostName):
3465         (LayoutTestController::dispatchPendingLoadRequests):
3466         (LayoutTestController::display):
3467         (LayoutTestController::counterValueForElementById):
3468         (LayoutTestController::keepWebHistory):
3469         (LayoutTestController::computedStyleIncludingVisitedInfo):
3470         (LayoutTestController::nodesFromRect):
3471         (LayoutTestController::layerTreeAsText):
3472         (LayoutTestController::pageNumberForElementById):
3473         (LayoutTestController::numberOfPages):
3474         (LayoutTestController::pageProperty):
3475         (LayoutTestController::isPageBoxVisible):
3476         (LayoutTestController::pageSizeAndMarginsInPixels):
3477         (LayoutTestController::webHistoryItemCount):
3478         (LayoutTestController::workerThreadCount):
3479         (LayoutTestController::notifyDone):
3480         (LayoutTestController::pathToLocalResource):
3481         (LayoutTestController::queueLoad):
3482         (LayoutTestController::setAcceptsEditing):
3483         (LayoutTestController::setAlwaysAcceptCookies):
3484         (LayoutTestController::setCustomPolicyDelegate):
3485         (LayoutTestController::waitForPolicyDelegate):
3486         (LayoutTestController::setScrollbarPolicy):
3487         (LayoutTestController::addOriginAccessWhitelistEntry):
3488         (LayoutTestController::removeOriginAccessWhitelistEntry):
3489         (LayoutTestController::setMainFrameIsFirstResponder):
3490         (LayoutTestController::setTabKeyCyclesThroughElements):
3491         (LayoutTestController::setUseDashboardCompatibilityMode):
3492         (LayoutTestController::setUserStyleSheetEnabled):
3493         (LayoutTestController::setUserStyleSheetLocation):
3494         (LayoutTestController::setValueForUser):
3495         (LayoutTestController::setViewModeMediaFeature):
3496         (LayoutTestController::setWindowIsKey):
3497         (LayoutTestController::setSmartInsertDeleteEnabled):
3498         (waitToDumpWatchdogFired):
3499         (LayoutTestController::setWaitToDump):
3500         (LayoutTestController::windowCount):
3501         (LayoutTestController::setPrivateBrowsingEnabled):
3502         (LayoutTestController::setJavaScriptCanAccessClipboard):
3503         (LayoutTestController::setXSSAuditorEnabled):
3504         (LayoutTestController::setFrameFlatteningEnabled):
3505         (LayoutTestController::setSpatialNavigationEnabled):
3506         (LayoutTestController::setAllowUniversalAccessFromFileURLs):
3507         (LayoutTestController::setAllowFileAccessFromFileURLs):
3508         (LayoutTestController::setAuthorAndUserStylesEnabled):
3509         (LayoutTestController::setAutofilled):
3510         (LayoutTestController::disableImageLoading):
3511         (LayoutTestController::setMockDeviceOrientation):
3512         (LayoutTestController::setMockGeolocationPosition):
3513         (LayoutTestController::setMockGeolocationError):
3514         (LayoutTestController::setGeolocationPermission):
3515         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
3516         (LayoutTestController::addMockSpeechInputResult):
3517         (LayoutTestController::setIconDatabaseEnabled):
3518         (LayoutTestController::setJavaScriptProfilingEnabled):
3519         (LayoutTestController::setSelectTrailingWhitespaceEnabled):
3520         (LayoutTestController::setPopupBlockingEnabled):
3521         (LayoutTestController::setPluginsEnabled):
3522         (LayoutTestController::elementDoesAutoCompleteForElementWithId):
3523         (LayoutTestController::execCommand):
3524         (LayoutTestController::findString):
3525         (LayoutTestController::isCommandEnabled):
3526         (LayoutTestController::setCacheModel):
3527         (LayoutTestController::setPersistentUserStyleSheetLocation):
3528         (LayoutTestController::clearPersistentUserStyleSheet):
3529         (LayoutTestController::clearAllApplicationCaches):
3530         (LayoutTestController::setApplicationCacheOriginQuota):
3531         (LayoutTestController::clearApplicationCacheForOrigin):
3532         (LayoutTestController::localStorageDiskUsageForOrigin):
3533         (LayoutTestController::originsWithApplicationCache):
3534         (LayoutTestController::applicationCacheDiskUsageForOrigin):
3535         (LayoutTestController::clearAllDatabases):
3536         (LayoutTestController::setDatabaseQuota):
3537         (LayoutTestController::originsWithLocalStorage):
3538         (LayoutTestController::deleteAllLocalStorage):
3539         (LayoutTestController::deleteLocalStorageForOrigin):
3540         (LayoutTestController::observeStorageTrackerNotifications):
3541         (LayoutTestController::syncLocalStorage):
3542         (LayoutTestController::setDomainRelaxationForbiddenForURLScheme):
3543         (LayoutTestController::goBack):
3544         (LayoutTestController::setDefersLoading):
3545         (LayoutTestController::setAppCacheMaximumSize):
3546         (LayoutTestController::pauseAnimationAtTimeOnElementWithId):
3547         (LayoutTestController::pauseTransitionAtTimeOnElementWithId):
3548         (LayoutTestController::sampleSVGAnimationForElementAtTime):
3549         (LayoutTestController::numberOfActiveAnimations):
3550         (LayoutTestController::suspendAnimations):
3551         (LayoutTestController::resumeAnimations):
3552         (LayoutTestController::overridePreference):
3553         (LayoutTestController::addUserScript):
3554         (LayoutTestController::addUserStyleSheet):
3555         (LayoutTestController::setDeveloperExtrasEnabled):
3556         (LayoutTestController::setAsynchronousSpellCheckingEnabled):
3557         (LayoutTestController::showWebInspector):
3558         (LayoutTestController::closeWebInspector):
3559         (LayoutTestController::evaluateInWebInspector):
3560         (LayoutTestController::evaluateScriptInIsolatedWorld):
3561         (LayoutTestController::removeAllVisitedLinks):
3562         (LayoutTestController::callShouldCloseOnWebView):
3563         (LayoutTestController::apiTestNewWindowDataLoadBaseURL):
3564         (LayoutTestController::apiTestGoToCurrentBackForwardItem):
3565         (LayoutTestController::setWebViewEditable):
3566         (LayoutTestController::markerTextForListItem):
3567         (LayoutTestController::authenticateSession):
3568         (LayoutTestController::setEditingBehavior):
3569         (LayoutTestController::abortModal):