garden-o-matic should have a "rebaseline" button
[WebKit-https.git] / Tools / ChangeLog
1 2011-07-14  Adam Barth  <abarth@webkit.org>
2
3         garden-o-matic should have a "rebaseline" button
4         https://bugs.webkit.org/show_bug.cgi?id=64446
5
6         Reviewed by Ojan Vafai.
7
8         This patch adds a basic Rebaseline button that copies the baselines
9         displayed in the results pane into the appropriate directory in your
10         working copy.
11
12         There are two main limitations:
13
14         1) There is no UI for actually committing the baselines.
15
16         2) The baselines are not optimized for redundancy (meaning you can have
17            identical baselines in both chromium-mac and chromium-win).
18
19         * Scripts/webkitpy/tool/commands/rebaseline.py:
20             - Turns out we need to create the directory for the baseline if it doesn't exist yet.
21         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js:
22             - Add an programatic API to call the server.
23         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
24             - Add the rebaseline button itself.
25         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
26             - Change the CSS so that the Rebaseline and Close buttons can
27               appear in the normal order in the DOM.
28         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
29             - Bind the event and translate the parameters.
30             - Hide/show the rebaseline button, as appropriate.
31         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
32             - Add some helpful utility functions for manipulating failure types.
33         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
34
35 2011-07-14  Ryosuke Niwa  <rniwa@webkit.org>
36
37         REGRESSION: webkit-patch roll-chromium-deps is broken
38         https://bugs.webkit.org/show_bug.cgi?id=64568
39
40         Reviewed by Dirk Pranke
41
42         Use _filesystem.join.
43
44         * Scripts/webkitpy/common/checkout/checkout.py:
45         * Scripts/webkitpy/common/checkout/checkout_unittests.py:
46
47 2011-07-14  Noam Rosenthal  <noam.rosenthal@nokia.com>
48
49         Adding myself to the reviewers list; No review needed.
50
51         * Scripts/webkitpy/common/config/committers.py:
52
53 2011-07-14  Adam Roben  <aroben@apple.com>
54
55         Don't use Element.prototype.classList in TestFailures
56
57         Safari 5 doesn't support it.
58
59         Fixes <http://webkit.org/b/64550> Can't expand flaky tests on TestFailures page in Safari 5
60
61         Reviewed by Daniel Bates.
62
63         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
64         (Element.prototype.hasStyleClass):
65         (Element.prototype.addStyleClass):
66         (Element.prototype.removeStyleClass):
67         (Element.prototype.toggleStyleClass):
68         Added these helper functions which simulate classList functionality.
69
70         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities_unittests.js:
71         Added. Tests for the above.
72
73         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
74         (ViewController.prototype._domForPossiblyFlakyTests): Changed to use
75         toggleStyleClass/hasStyleClass instead of classList.
76
77         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
78         Added Utilities_unittests.js, and reordered the tested files to be in
79         roughly dependency order (i.e., the lowest-level files are imported
80         and tested first).
81
82 2011-07-14  Eric Seidel  <eric@webkit.org>
83
84         NRWT doesn't store the svn revision in full_results.json on chromium-win
85         https://bugs.webkit.org/show_bug.cgi?id=64492
86
87         Unreviewed.  Just fixing my test-webkitpy regression.
88
89         Fix the unit tests.  Unfortunately scm does not use a filesystem
90         object so we can't control the result of detect_scm_system.
91         When detect_scm_system would fail, we would log, which would
92         cause all passing_run integration tests to fail.
93
94         * Scripts/webkitpy/common/checkout/scm/scm.py:
95         * Scripts/webkitpy/layout_tests/controllers/manager.py:
96         * Scripts/webkitpy/layout_tests/port/base.py:
97         * Scripts/webkitpy/layout_tests/port/test.py:
98         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
99
100 2011-07-14  Eric Seidel  <eric@webkit.org>
101
102         NRWT doesn't store the svn revision in full_results.json on chromium-win
103         https://bugs.webkit.org/show_bug.cgi?id=64492
104
105         Reviewed by Ojan Vafai.
106
107         This should fix the bug.  Unfortunately this code is currently
108         impossible to test since it's impossible to mock detect_scm_system at the
109         moment.  I started re-writing scm.detection.py to be mockable, but decided that
110         was best left for another day.
111
112         * Scripts/webkitpy/layout_tests/controllers/manager.py:
113         * Scripts/webkitpy/layout_tests/port/base.py:
114
115 2011-07-14  Ojan Vafai  <ojan@chromium.org>
116
117         fix coding style of dashboard_base.js
118         https://bugs.webkit.org/show_bug.cgi?id=64545
119
120         Reviewed by Adam Roben.
121
122         The code changes to the files other than dashboard_base.js
123         are just fallout from renaming globals in dashboard_base.js.
124
125         * TestResultServer/static-dashboards/aggregate_results.html:
126         * TestResultServer/static-dashboards/builders.js:
127         * TestResultServer/static-dashboards/dashboard_base.js:
128         * TestResultServer/static-dashboards/flakiness_dashboard.html:
129         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
130         * TestResultServer/static-dashboards/timeline_explorer.html:
131         * TestResultServer/static-dashboards/treemap.html:
132
133 2011-07-14  Carlos Garcia Campos  <cgarcia@igalia.com>
134
135         Unreviewed. Fix WebKit2 GTK build after r90953.
136
137         * WebKitTestRunner/GNUmakefile.am:
138
139 2011-07-14  Adam Barth  <abarth@webkit.org>
140
141         garden-o-matic results pane should be more discoverable
142         https://bugs.webkit.org/show_bug.cgi?id=64513
143
144         Reviewed by Eric Seidel.
145
146         This patch causes us to trigger the results pane on mouse clicks
147         instead of mousein.  Also, handle the case of no results details more
148         elegantly.
149
150         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
151         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
152             - Change the binding to "click".
153         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
154             - Handle the "no results URLs" case explicitly instead of never
155               calling the callback.
156         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
157         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
158             - Show a message when there are no result details to show.
159         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
160
161 2011-07-14  Adam Barth  <abarth@webkit.org>
162
163         garden-o-matic should badge TIMEOUT failures
164         https://bugs.webkit.org/show_bug.cgi?id=64435
165
166         Reviewed by Eric Seidel.
167
168         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
169         (.regression .what a[draggable].TIMEOUT::before):
170             - Add TIMEOUT.
171
172 2011-07-14  Adam Barth  <abarth@webkit.org>
173
174         Fix Python exception blocking the commit-queue.
175
176         * Scripts/webkitpy/common/checkout/scm/git.py:
177
178 2011-07-13  Eric Seidel  <eric@webkit.org>
179
180         Move webkitpy off of loose mocks
181         https://bugs.webkit.org/show_bug.cgi?id=64508
182
183         Reviewed by Adam Barth.
184
185         Using Mock has caused us more pain than help.
186         It's possible that there was a cleaner way to use it
187         (maybe Mock(class) instead of inheriting from it?).
188         But for now, I've removed all uses of Mock from mocktool.py.
189
190         I also moved run_command into the only 3 files which call it
191         instead of leaving the deprecated method in executive.py.
192
193         I changed various direct calls to os.* to use filesystem instead.
194
195         * Scripts/webkitpy/common/checkout/checkout.py:
196         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
197         * Scripts/webkitpy/common/checkout/scm/git.py:
198         * Scripts/webkitpy/common/checkout/scm/scm.py:
199         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
200         * Scripts/webkitpy/common/checkout/scm/svn.py:
201         * Scripts/webkitpy/common/system/executive.py:
202         * Scripts/webkitpy/common/system/executive_unittest.py:
203         * Scripts/webkitpy/tool/mocktool.py:
204         * Scripts/webkitpy/tool/steps/cleanworkingdirectory_unittest.py:
205         * Scripts/webkitpy/tool/steps/ensurelocalcommitifneeded.py:
206
207 2011-07-13  Eric Seidel  <eric@webkit.org>
208
209         NRWT doesn't store the svn revision in full_results.json on chromium-win
210         https://bugs.webkit.org/show_bug.cgi?id=64492
211
212         Reviewed by Adam Barth.
213
214         Add a warning message in the revision='' case to help us diagnose
215         what's going wrong on the bot.
216
217         * Scripts/webkitpy/layout_tests/controllers/manager.py:
218
219 2011-07-13  Dirk Pranke  <dpranke@chromium.org>
220
221         new-run-webkit-tests: print baseline search path as part of config output
222         https://bugs.webkit.org/show_bug.cgi?id=64499
223
224         Reviewed by Eric Seidel.
225
226         * Scripts/webkitpy/layout_tests/controllers/manager.py:
227         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
228
229 2011-07-13  Adam Barth  <abarth@webkit.org>
230
231         gardening server should have an API for parsing changelogs
232         https://bugs.webkit.org/show_bug.cgi?id=64495
233
234         Reviewed by Eric Seidel.
235
236         This patch exposes much of the same information from CommitInfo in a
237         dictionary form, which is easier to send over-the-wire as JSON to the
238         frontend.
239
240         * Scripts/webkitpy/common/checkout/checkout.py:
241         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
242         * Scripts/webkitpy/tool/mocktool.py:
243         * Scripts/webkitpy/tool/servers/gardeningserver.py:
244         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
245         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
246
247 2011-07-13  Eric Seidel  <eric@webkit.org>
248
249         REGRESSION: GitTestWithMock.test_create_patch fails
250         https://bugs.webkit.org/show_bug.cgi?id=62945
251
252         Reviewed by Daniel Bates.
253
254         I was not able to reproduce the exact failure seen in the bug,
255         however this test was failing on my machine for other reasons.
256
257         I went through and did an audit of our run_command usage, it's
258         entirely in scm classes after this change.  (Not surprising given
259         that scm.py was the second file ever created in webkit.py.)
260
261         The real bug I'm fixing here is that we were setting executive.should_log
262         when the value had changed to executive._should_log.  Now we set the right one
263         and the test works again.
264
265         * Scripts/webkitpy/common/checkout/checkout.py:
266         * Scripts/webkitpy/common/checkout/scm/git.py:
267         * Scripts/webkitpy/common/checkout/scm/scm.py:
268         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
269         * Scripts/webkitpy/common/checkout/scm/svn.py:
270
271 2011-07-13  Ilya Sherman  <isherman@chromium.org>
272
273         Fix WTF header guard style check
274         https://bugs.webkit.org/show_bug.cgi?id=64488
275
276         Reviewed by David Levin.
277
278         * Scripts/webkitpy/style/checkers/cpp.py:
279         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Verify that we don't *always* suggest "WTF_" as a prefix
280
281 2011-07-13  Ojan Vafai  <ojan@chromium.org>
282
283         bring flakiness_dashboard.html closer to webkit style
284         https://bugs.webkit.org/show_bug.cgi?id=64477
285
286         Reviewed by Adam Barth.
287
288         * TestResultServer/static-dashboards/dashboard_base.js:
289         Fix bad variable name that would hit only when running the tests.
290         * TestResultServer/static-dashboards/flakiness_dashboard.html:
291         No code/logic changes. All moving/removing brackets, indents and moving things
292         to one line.
293
294 2011-07-13  Xan Lopez  <xlopez@igalia.com>
295
296         [GTK] Fix distcheck
297
298         Reviewed by Martin Robinson.
299
300         * WebKitTestRunner/GNUmakefile.am: mark built sources as nodist.
301
302 2011-07-13  Joseph Pecoraro  <joepeck@webkit.org>
303
304         Unreviewed fix. Add a "\n" to a printf that somehow got lost.
305
306         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
307         (WebCore::DumpRenderTree::dumpApplicationCacheQuota):
308
309 2011-07-13  Joseph Pecoraro  <joepeck@webkit.org>
310
311         ApplicationCache Quota Output is Flakey
312         https://bugs.webkit.org/show_bug.cgi?id=64410
313
314         Reviewed by Alexey Proskuryakov.
315
316         Unify the delegate logging for reaching application cache
317         quotas. Truncate the space needed to the nearest 10000
318         for less flakey test output. This also better supports
319         printing NSUInteger on non-64 bit systems by casting to
320         an unsigned long.
321
322         * DumpRenderTree/mac/UIDelegate.mm:
323         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:totalSpaceNeeded:]):
324         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
325         (WebCore::DumpRenderTree::dumpApplicationCacheQuota):
326
327 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
328
329         Remove unused TestExpectations._overrides.
330         https://bugs.webkit.org/show_bug.cgi?id=64470
331
332         Reviewed by Adam Barth.
333
334         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Removed unused member.
335
336 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
337
338         Introduce TestExpectationSerializer.list_to_string.
339         https://bugs.webkit.org/show_bug.cgi?id=64462
340
341         Reviewed by Adam Barth.
342
343         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Added list_to_string and change the relevant callsite to use it.
344         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added tests for it.
345
346 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
347
348         Consolidate expectations parsing code.
349         https://bugs.webkit.org/show_bug.cgi?id=64460
350
351         Reviewed by Adam Barth.
352
353         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Folded TestExpectationParser._split_expectation_string into TestExpectationParser.parse.
354
355 2011-07-12  Brent Fulgham  <bfulgham@webkit.org>
356
357         Standardize WinCairo conditionalized code under PLATFORM macro.
358         https://bugs.webkit.org/show_bug.cgi?id=64377
359
360         Reviewed by Maciej Stachowiak.
361
362         Update compile-time conditions that had previously used #ifdef
363         of WIN_CAIRO to consistently use PLATFORM(WIN_CAIRO).
364
365         * DumpRenderTree/config.h: Switch to PLATFORM(WIN_CAIRO)
366         * DumpRenderTree/win/DumpRenderTreeCairo.vsprops: Remove redundant
367           define of WIN_CAIRO=1.
368         * DumpRenderTree/win/DumpRenderTreeDebugCairoCFLite.vsprops: Add
369           WinCairo.vsprops to set (to match Release build).
370
371 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
372
373         Eliminate TestExpectationsFile.
374         https://bugs.webkit.org/show_bug.cgi?id=64458
375
376         Turns out, we can just use a Python list.
377
378         Reviewed by Adam Barth.
379
380         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Folded TestExpectationsFile.append into TestExpectationParser.parse_list,
381             removed TestExpectationsFile.
382         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Moved tests to better reflect new names, removed iterator test,
383             since there's no more custom iterator machinery.
384
385 2011-07-13  Philippe Normand  <pnormand@igalia.com>
386
387         Unreviewed, added my other email addresses.
388
389         * Scripts/webkitpy/common/config/committers.py:
390
391 2011-07-13  Adam Barth  <abarth@webkit.org>
392
393         Folks should only be listed once.
394
395         * Scripts/webkitpy/common/config/committers.py:
396
397 2011-07-13  Philippe Normand  <pnormand@igalia.com>
398
399         Unreviewed, adding myself as Reviewer.
400
401         * Scripts/webkitpy/common/config/committers.py:
402
403 2011-07-13  Adam Barth  <abarth@webkit.org>
404
405         Fix flickering bug introduced by my previous patch.  I forgot to change
406         the name of the class everywhere.
407
408         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
409
410 2011-07-12  Philippe Normand  <pnormand@igalia.com>
411
412         [GTK] media/media-blocked-by-willsendrequest.html fails
413         https://bugs.webkit.org/show_bug.cgi?id=63699
414
415         Reviewed by Martin Robinson.
416
417         * DumpRenderTree/gtk/DumpRenderTree.cpp:
418         (willSendRequestCallback): Abort the request if explicitely asked
419         by the LayoutTestController's willSendRequestReturnsNull() function.
420
421 2011-07-13  Adam Barth  <abarth@webkit.org>
422
423         Remove "Dismiss" button from garden-o-matic butterbar
424         https://bugs.webkit.org/show_bug.cgi?id=64443
425
426         Reviewed by Dimitri Glazkov.
427
428         We don't have any persistent butter bar messages yet, so the dismiss
429         button is premature (and fairly heavy, visually).
430
431         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
432         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
433
434 2011-07-13  Adam Barth  <abarth@webkit.org>
435
436         Refine garden-o-matic status pane
437         https://bugs.webkit.org/show_bug.cgi?id=64442
438
439         Reviewed by Dimitri Glazkov.
440
441         This patch makes two improvements to the status pane:
442
443         1) We only query the server for failure types that we're expecting.
444            This dramatically reduces the number of HTTP requests, making
445            loading the status pane faster.
446
447         2) The status pane now displays which test and which builder it is
448            showing results for.
449
450         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js:
451         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
452         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
453         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
454         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
455         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
456         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
457
458 2011-07-13  Adam Barth  <abarth@webkit.org>
459
460         garden-o-matic should show test results
461         https://bugs.webkit.org/show_bug.cgi?id=64440
462
463         Reviewed by Adam Roben.
464
465         This patch introduces the results detail pane, which appears at the
466         bottom of the window and contains test results from the bots.
467         Currently, you can activate the pane by mousing over one of the builder
468         names associated with a failing test.
469
470         This is just a first iteration of the UI.  There's no way to resize or
471         zoom in on elements of the details pane, and images likely aren't sized
472         correctly, but it's a place to start.
473
474         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js:
475             - Add constants for our data attributes so we don't typo them!
476         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
477             - Add DOM structure for the details pane.
478         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
479         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
480             - Bind events for showing and hiding the details pane.
481         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
482             - Reduce the number of result types that we query for to avoid
483               overfilling the details pane.  We'll probably need another
484               solution here in the long-term.  For example, we could use tabs
485               to pack more results into the pane.
486         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
487         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
488             - Now that we're using the config package during testing, we need
489               to include it in the testing HTML.
490         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
491         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
492             - Change the DOM structure of the results detail to make it fit
493               nicely in the details pane (rather than flowing freely in the
494               body, as it did before).
495
496 2011-07-13  Adam Roben  <aroben@apple.com>
497
498         Make TestFailures show every time a possibly-flaky test failed, but hide it by default
499
500         It's useful to be able to see every time a flaky test failed to see whether it failed the
501         same way every time. But doing so takes a lot of space, so the list of failures is now
502         collapsed by default and can be revealed using a disclosure triangle.
503
504         Fixes <http://webkit.org/b/64455> TestFailures page doesn't show as much information for
505         flaky tests as I would like, even though the page is already so long
506
507         Reviewed by Dimitri Glazkov.
508
509         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
510         (FlakyLayoutTestDetector.prototype.allFailures): Replaced flakinessExamples with this
511         function. Now returns all failures for the given test.
512
513         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector_unittests.js:
514         Added. This just contains some simple tests of the FlakyLayoutTestDetector class. We'll add
515         more over time.
516
517         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
518         (LayoutTestHistoryAnalyzer.prototype.start): Updated the documentation comment to reflect
519         that we no longer return passing builds for possibly-flaky tests.
520
521         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
522         (.existing-bugs-list, .suspect-revisions-list, .flakiness-examples-list): Make the list of
523         flakiness examples small, too, since it can get quite long.
524
525         (.disclosure-triangle):
526         (.expanded > .disclosure-triangle):
527         Simple styles for the disclosure triangle.
528
529         (.flakiness-examples-list): Collapse the list by default.
530
531         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
532         (ViewController.prototype._displayBuilder): Pass the total number of builds analyzed to
533         _domForPossiblyFlakyTests.
534         (ViewController.prototype._domForPossiblyFlakyTests): Put a disclosure triangle to the left
535         of each test name, and the number of failures to the right. When the disclosure triangle is
536         clicked for the first time, we build up the list of failures and expand the element. After
537         that we just collapse or expand the element on subsequent clicks.
538
539         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
540         Pulled in new tests.
541
542 2011-07-13  Adam Roben  <aroben@apple.com>
543         
544         Teach TestFailures to understand NRWT's output when it exits early due to too many failures
545
546         Fixes <http://webkit.org/b/64456> TestFailures page reports way too many failures when NRWT
547         exits early
548
549         Reviewed by Dimitri Glazkov.
550
551         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
552         (Builder.prototype.getNumberOfFailingTests): Relaxed the "Exiting early" test not to require
553         it to be at the beginning of the line, since NRWT prints a bunch of junk earlier in the
554         line. Tightened up the regex that's used to parse the number of failing tests to require the
555         leading number to be followed by whitespace so that we won't parse the "2011" in
556         "2011-07-13" as a number of failures.
557
558         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
559         Added a test for the above.
560
561 2011-07-12  Dimitri Glazkov  <dglazkov@chromium.org>
562
563         Extract model-like TestExpectationLine and TestExpectationFile from TestExpectations.
564         https://bugs.webkit.org/show_bug.cgi?id=64386
565
566         This is the first step in converting TestExpectations to a real model.
567         * TestExpectationsLine represents a line in the test_expectations.txt file, and
568         * TestExpectationsFile represents the file, which is a collection of lines.
569
570         Reviewed by Adam Barth.
571
572         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
573         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
574         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
575
576 2011-07-13  Xan Lopez  <xlopez@igalia.com>
577
578         [GTK] Do not grab focus too early in DRT.
579
580         Reviewed by Gustavo Noronha.
581
582         It causes a layout to happen and a progress signal to be emitted
583         since r90900, but at this point we don't have a
584         LayoutTestController object and we'll eventually crash. Since we
585         already grab focus at the beginning of runTest() this is
586         redundant, so get rid of it to fix the crash.
587
588         * DumpRenderTree/gtk/DumpRenderTree.cpp:
589         (main): remove call to grab_focus
590
591 2011-07-13  Sheriff Bot  <webkit.review.bot@gmail.com>
592
593         Unreviewed, rolling out r90893 and r90894.
594         http://trac.webkit.org/changeset/90893
595         http://trac.webkit.org/changeset/90894
596         https://bugs.webkit.org/show_bug.cgi?id=64441
597
598         NRWT still doesn't work on qt-mac platform (Requested by Ossy
599         on #webkit).
600
601         * Scripts/run-webkit-tests:
602         (useNewRunWebKitTests):
603
604 2011-07-13  Csaba Osztrogonác  <ossy@webkit.org>
605
606         [Qt] NRWT should pick up the right httpd config file
607         https://bugs.webkit.org/show_bug.cgi?id=64086
608
609         * Scripts/run-webkit-tests: Enable NRWT on qt-mac platform after r90810.
610         (useNewRunWebKitTests):
611
612 2011-07-12  Mark Rowe  <mrowe@apple.com>
613
614         Fix the 32-bit build.
615
616         * DumpRenderTree/mac/UIDelegate.mm:
617         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:totalSpaceNeeded:]):
618         Cast the NSUInteger value to unsigned long to match the format specifier.
619
620 2011-07-12  Adam Barth  <abarth@webkit.org>
621
622         Improve garden-o-matic UI when the bots fail to report revision numbers
623         https://bugs.webkit.org/show_bug.cgi?id=64427
624
625         Reviewed by Dimitri Glazkov.
626
627         At least Win (dbg)(2) seems to fail to report the SVN revision number
628         in full_results.json.  This patch makes garden-o-matic more robust to
629         missing revision numbers.
630
631         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
632         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
633         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
634         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
635         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
636
637 2011-07-12  Adam Barth  <abarth@webkit.org>
638
639         Adjust garden-o-matic layout to use fewer lines
640         https://bugs.webkit.org/show_bug.cgi?id=64422
641
642         Reviewed by Ojan Vafai.
643
644         This patch effectively merges the test name line with the list of
645         builders on which the test fails.  Each test failure now occupies two
646         lines instead of three.
647
648         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
649         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
650         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
651
652 2011-07-12  Adam Barth  <abarth@webkit.org>
653
654         garden-o-matic should label tests failures of new tests as such
655         https://bugs.webkit.org/show_bug.cgi?id=64421
656
657         Reviewed by Ojan Vafai.
658
659         Also, we shouldn't dim failures of new tests, even if we've only seen
660         them once because they're likely to be real problems that need
661         attention.
662
663         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
664         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js: Added.
665         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout_unittests.js: Added.
666         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
667         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
668         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
669         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
670         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
671         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
672         * Scripts/webkitpy/tool/servers/gardeningserver.py:
673
674 2011-07-12  Adam Barth  <abarth@webkit.org>
675
676         garden-o-matic should display how many times we've seen a failure
677         https://bugs.webkit.org/show_bug.cgi?id=64417
678
679         Reviewed by Ojan Vafai.
680
681         This patch adds some UI to display how many times we've seen a given
682         failure, which can be helpful for determining whether that failure is a
683         real failure or a flaky test.
684
685         When a failure has only been seen once (i.e., only a single run on a
686         single bot), we set the opacity of to 50% to avoid distracting the
687         gardener.
688
689         This patch also refactors the failure walker to have a simpler API
690         internally by moving from an object-oriented paradigm to a functional
691         paradigm.
692
693         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
694         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
695         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
696         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
697         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
698         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
699
700 2011-07-12  Chris Rogers  <crogers@google.com>
701
702         Enable Web Audio for chromium DRT
703         https://bugs.webkit.org/show_bug.cgi?id=64409
704
705         Reviewed by James Robinson.
706
707         * DumpRenderTree/chromium/TestShell.cpp:
708         (TestShell::TestShell):
709
710 2011-07-12  Adam Barth  <abarth@webkit.org>
711
712         Tweak some UI in garden-o-matic now that the tree actually has a
713         failure and I can see what this all looks like.
714
715         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
716         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
717         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
718
719 2011-07-12  Joseph Pecoraro  <joepeck@webkit.org>
720
721         Unreviewed. Skipping a few tests which fail due to differing output
722         in recent ApplicationCache quota tests. Reenabling the tests is
723         tracked by <http://webkit.org/b/64410>.
724
725         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
726         (WebCore::DumpRenderTree::dumpApplicationCacheQuota): This should
727         have been a space, not a newline.
728
729 2011-07-12  Adam Barth  <abarth@webkit.org>
730
731         garden-o-matic should display regression ranges
732         https://bugs.webkit.org/show_bug.cgi?id=64407
733
734         Reviewed by Dimitri Glazkov.
735
736         This patch computes a regression range for a failure by intersecting
737         the regression ranges seen by the various bots.  We make the underlying
738         assumption that a test is only failing due to one revision at any given
739         moment.  If that's not true, this code probably explodes.
740
741         The regression ranges appear asynchronously, which might be a jarring
742         UI.  We'll have to experiment to see.
743
744         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
745         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
746         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
747         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
748         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
749         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
750         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
751
752 2011-07-12  Adam Barth  <abarth@webkit.org>
753
754         cr-linux-ews should run pixel tests
755         https://bugs.webkit.org/show_bug.cgi?id=64394
756
757         Reviewed by Ojan Vafai.
758
759         This patch enabled pixel test on the cr-linux-ews.  Running a quick
760         test on the bots, it looks like we have five pixel failures, related to
761         some missing fonts.  I've also made the bots skip failing tests to
762         avoid uploading huge zip files to bugs.webkit.org (with all the exected
763         IMAGE failures).
764
765         * Scripts/webkitpy/common/config/ports.py:
766         * Scripts/webkitpy/common/config/ports_unittest.py:
767
768 2011-07-12  Joseph Pecoraro  <joepeck@webkit.org>
769
770         ApplicationCache update should not immediately fail when reaching per-origin quota
771         https://bugs.webkit.org/show_bug.cgi?id=64177
772
773         Reviewed by Alexey Proskuryakov.
774
775         Add layoutTestController.disallowIncreaseForApplicationCacheQuota()
776         to disable the default behavior of raising the default per-origin
777         quota to 5MB when reached. This allows us to test what happens
778         when the quota is reached and not increased.
779
780         * DumpRenderTree/LayoutTestController.cpp:
781         (LayoutTestController::LayoutTestController):
782         (disallowIncreaseForApplicationCacheQuotaCallback):
783         (LayoutTestController::staticFunctions):
784         * DumpRenderTree/LayoutTestController.h:
785         (LayoutTestController::disallowIncreaseForApplicationCacheQuota):
786         (LayoutTestController::setDisallowIncreaseForApplicationCacheQuota):
787         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
788         (LayoutTestController::reset):
789         * DumpRenderTree/qt/LayoutTestControllerQt.h:
790         (LayoutTestController::disallowIncreaseForApplicationCacheQuota):
791         (LayoutTestController::dumpApplicationCacheDelegateCallbacks):
792         Expose a function to prevent automatically increasing the per-origin
793         quota when the quota is reached. This allows us to test what happens
794         when a user would "disallow" a quota increase.
795
796         * DumpRenderTree/mac/UIDelegate.mm:
797         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:spaceNeeded:]):
798         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
799         (WebCore::DumpRenderTree::dumpApplicationCacheQuota):
800         Respect the new disallowIncrease flag and dump the spaceNeeded values
801         for ports that implement application cache quotas.
802
803 2011-07-12  Ojan Vafai  <ojan@chromium.org>
804
805         fix flakiness dashboard to work with new crash log filenames
806         https://bugs.webkit.org/show_bug.cgi?id=64393
807
808         Reviewed by Adam Barth.
809
810         * TestResultServer/static-dashboards/flakiness_dashboard.html:
811
812 2011-07-12  Adam Barth  <abarth@webkit.org>
813
814         Add a happy animation to garden-o-matic when there are no failures
815         https://bugs.webkit.org/show_bug.cgi?id=64382
816
817         Reviewed by Ojan Vafai.
818
819         As requested by Ojan.
820
821         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
822         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
823         * Scripts/webkitpy/tool/servers/data/gardeningserver/partytime.gif: Added.
824         * Scripts/webkitpy/tool/servers/gardeningserver.py:
825
826 2011-07-12  Adam Barth  <abarth@webkit.org>
827
828         sheriffbot can't perform rollouts
829         https://bugs.webkit.org/show_bug.cgi?id=64370
830
831         Reviewed by Eric Seidel.
832
833         Now that we're not updating the working copy before processing every
834         IRC message, the rollout command needs to update it manually.
835
836         * Scripts/webkitpy/tool/bot/irc_command.py:
837         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
838
839 2011-07-12  Adam Barth  <abarth@webkit.org>
840
841         Fix the grammar in the FIXME comment.
842
843         * Scripts/webkitpy/common/net/resultsjsonparser.py:
844
845 2011-07-12  Eric Seidel  <eric@webkit.org>
846
847         NRWT should open test results page with Safari trunk, not the system provided one on Mac
848         https://bugs.webkit.org/show_bug.cgi?id=64346
849
850         Reviewed by Adam Barth.
851
852         To fix this I implemented Port.show_results_html_file in Mac, Gtk and Qt ports with
853         implementations (mostly) matching those found in old-run-webkit-tests.
854         There are still some minor differences for Qt which Qt hackers may wish to tweak.
855
856         I had to add a WebKitPort._port_flag_for_scripts method (similar to flag() in common.config.ports.py)
857         for the Qt/Gtk ports which always require a flag passed to scripts.
858
859         While trying to test this, I found that FactoryTest.test_chromium_gpu_linux
860         was using a real filesystem (due to assert_platform) which due to
861         global variables in config.py was causing set-webkit-configuration to have
862         an affect on unit test results!  So I fixed this by making FactoryTest.assert_port
863         pass a mock file system whenever calling factory.get.
864
865         Unfortunately TestPort was depending on always being passed a None filesystem
866         and asserting filesystem._tests (only true for unit_test_filesystem()).
867         So I just removed the FactoryTest.test_test and FactoryTest.test_dryrun tests
868         deciding that they were pretty much useless anyway.  If others feel strongly
869         I'm happy to fix this in a different way.
870
871         * Scripts/webkitpy/common/system/executive.py:
872          - default arguments in python are screwy.  They use a single shared
873            instance, so it's better to use argument=None and then argument = argument or Default()
874            if you have any chance of mutating (or returning) the default argument.
875         * Scripts/webkitpy/layout_tests/port/config.py:
876          - This code is wrong.  We don't need to use a global variable here (as far as I can tell).
877            I'm not fixing it in this patch, but I've marked it with a FIXME and we can convert to
878            storing the results of the read on the Config object (which should only be created once during normal operation).
879            Unit tests shouldn't be hitting the disk anyway.  It's possible Config should move off of Port and onto Tool/Host directly.
880         * Scripts/webkitpy/layout_tests/port/factory.py:
881         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
882         * Scripts/webkitpy/layout_tests/port/gtk.py:
883         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py: Added.
884         * Scripts/webkitpy/layout_tests/port/mac.py:
885         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
886         * Scripts/webkitpy/layout_tests/port/qt.py:
887         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
888         * Scripts/webkitpy/layout_tests/port/webkit.py:
889
890 2011-07-12  Adam Roben  <aroben@apple.com>
891
892         Teach TestFailures to recognize when run-webkit-tests gets killed by buildbot
893
894         Fixes <http://webkit.org/b/64358> TestFailures page thinks all tests passed in
895         http://build.webkit.org/builders/Windows%207%20Release%20(Tests)/builds/14672
896
897         Reviewed by Daniel Bates.
898
899         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
900         (Builder.prototype.getNumberOfFailingTests): If run-webkit-tests exited with a non-zero
901         exit status but we didn't find any failure counts, assume that there was some error that
902         caused run-webkit-tests to die early (like being killed by buildbot due to a timeout).
903
904         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
905         Added a new test that shows that we get a failingTestCount of -1 when run-webkit-tests dies
906         early.
907         (runGetNumberOfFailingTestsTest): Moved most code here from the only pre-existing test in this
908         file.
909
910         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
911         (LayoutTestResultsLoader.prototype.start): Bump the cache number so old cached data that was
912         tainted by the bug fixed in this patch will be evicted.
913
914 2011-07-12  Adam Barth  <abarth@webkit.org>
915
916         cr-linux-ews complains about tests that aren't actually failing
917         https://bugs.webkit.org/show_bug.cgi?id=64367
918
919         The underlying problem here is that full_results.json doesn't have
920         enough information to determine whether a given test result was
921         expected because whether an actual result is expected depends on
922         whether full_results.json was generated duing a test run that included
923         pixel tests.
924
925         The right long-term solution is to make full_results.json a complete
926         description of what happened durning a test run.  In this patch
927         however, to stop the spam, this patch makes the jsonresultsparser
928         ignore pixel failures.
929
930         (I'm landing this patch unreviewed in order to stop the bots from
931         spamming.  I'll happily address any review comments in a follow-up
932         patch.)
933
934         * Scripts/webkitpy/common/net/resultsjsonparser.py:
935         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
936
937 2011-07-12  Adam Barth  <abarth@webkit.org>
938
939         Re-work garden-o-matic UI to begin to look like the mocks
940         https://bugs.webkit.org/show_bug.cgi?id=64334
941
942         Reviewed by Dimitri Glazkov.
943
944         This patch changes how we display failures to be more like the mock the
945         dglazkov made. We now can display the same information much more
946         compactly.
947
948         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
949         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
950         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
951         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css: Added.
952         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
953         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
954         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
955         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
956         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
957         * Scripts/webkitpy/tool/servers/gardeningserver.py:
958
959 2011-07-11  Adam Roben  <aroben@apple.com>
960
961         Extract TestFailures's bug-filing code into two new classes
962
963         Fixes <http://webkit.org/b/64300> TestFailures page's new-bug-filing code is a mess!
964
965         Reviewed by Darin Adler and Adam Barth.
966
967         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js:
968         (Buildbot.prototype.resultsDirectoryURL): Changed to return a URI-encoded URL. Otherwise the
969         URL isn't valid (and it's harder to mock this function).
970
971         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot_unittests.js:
972         Added. Just tests the above change (for now).
973
974         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
975         Added a license header, enclosed everything in a closure to avoid polluting the global
976         namespace, changed the test name to actually describe the passing condition, and replaced
977         uses of equals() with equal(). (The latter is the actual name of the function, and matches
978         deepEqual, etc.)
979
980         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm.js: Added.
981         (NewBugForm): This class knows how to construct a <form> used for filing new bugs in
982         Bugzilla based on some parameters.
983         (NewBugForm.prototype.domElement): Creates and returns the <form> element. Code came from
984         ViewController.prototype._domForNewAndExistingBugs.
985
986         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm_unittests.js:
987         Added. Tests the above code.
988
989         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm.js: Added.
990         (TestFailureBugForm): This class knows how to construct a <form> element used for filing new
991         bugs specifically about test failures. Code came from
992         ViewController.prototype._domForNewAndExistingBugs.
993         (TestFailureBugForm.prototype.domElement): Slightly customizes the <form> element returned
994         by NewBugForm.
995         (TestFailureBugForm.prototype._computeOperatingSystem):
996         (TestFailureBugForm.prototype._computePlatform):
997         (TestFailureBugForm.prototype._createBugTitle):
998         (TestFailureBugForm.prototype._failingResultsHTMLURL):
999         (TestFailureBugForm.prototype._failingRevision):
1000         (TestFailureBugForm.prototype._passingRevision):
1001         (TestFailureBugForm.prototype._regressionRangeString):
1002         Code came from ViewController.prototype._domForNewAndExistingBugs. I broke it out into
1003         separate functions to break up the rat's nest a bit.
1004
1005         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm_unittests.js:
1006         Added. Tests the above code.
1007
1008         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1009         (ViewController.prototype._domForNewAndExistingBugs): Moved code from here to
1010         TestFailureBugForm/NewBugForm, and changed this code to use a TestFailureBugForm.
1011
1012         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
1013         NewBugForm/TestFailureBugForm.js.
1014
1015         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
1016         Added new tests and required files.
1017
1018 2011-07-12  Adam Roben  <aroben@apple.com>
1019
1020         Test that no intermediate WTF::Strings are created when concatenating with string literals
1021
1022         Test for <http://webkit.org/b/63330> Concatenating string literals and WTF::Strings using
1023         operator+ is suboptimal
1024
1025         Reviewed by Darin Adler.
1026
1027         * TestWebKitAPI/Tests/WTF/StringOperators.cpp: Added.
1028         (TestWebKitAPI::TEST): Test that a bunch of different string concatenation expressions don't
1029         create any intermediate WTF::Strings while they're being evaluated.
1030
1031         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1032         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
1033         Added new file.
1034
1035 2011-07-12  Eric Seidel  <eric@webkit.org>
1036
1037         [Qt] NRWT should pick up the right httpd config file
1038         https://bugs.webkit.org/show_bug.cgi?id=64086
1039
1040         Reviewed by Adam Barth.
1041
1042         This is more fall-out from the Port class inappropriately encapsulating
1043         both platform and port knowledge.  We need to split out some of this
1044         platform knowledge into a separate class which can be better shared
1045         between ports.
1046
1047         The fix was to move all the _path_to_apache_config_file logic down
1048         into the WebKitPort baseclass so that all of the ports can find the
1049         right apache config file, regardless of what platform they may be running on.
1050
1051         I did not move this down into base.Port (even though I considered it).
1052         Chromium duplicates some of this logic, but since they have separate
1053         subclasses for each port-OS combination (e.g. ChromiumMac, ChromiumWin)
1054         they wouldn't notice the change.  Eventually we'll move this logic
1055         out of Port entirely, and then it will be shared by all ports.
1056
1057         I also cleaned up the http-config logic for ORWT while I was there,
1058         although since we're killing that code, I'm happy to revert that part
1059         of the change if changing it is viewed as needlessly risky.
1060
1061         During this cleanup, I noticed that no ports use the "default" httpd.conf
1062         which is left over from Mac OS X Tiger (and old linux distros) which used
1063         Apache 1.3.  I've removed httpd.conf (and associated support in ORWT) since
1064         we no longer support any ports which use this httpd.conf.
1065
1066         * Scripts/webkitperl/httpd.pm:
1067         (getHTTPDConfigPathForTestDirectory):
1068         * Scripts/webkitpy/layout_tests/port/base.py:
1069         * Scripts/webkitpy/layout_tests/port/gtk.py:
1070         * Scripts/webkitpy/layout_tests/port/mac.py:
1071         * Scripts/webkitpy/layout_tests/port/qt.py:
1072         * Scripts/webkitpy/layout_tests/port/webkit.py:
1073         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1074         * Scripts/webkitpy/layout_tests/port/win.py:
1075         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
1076
1077 2011-07-11  Hironori Bono  <hbono@chromium.org>
1078
1079         Reviewed by Adam Roben.
1080
1081         Implement layoutTestController.setTextDirection for WebKit2 and Windows.
1082         https://bugs.webkit.org/show_bug.cgi?id=61931
1083
1084         This change implements layoutTextController.setTextDirection for WebKit2
1085         and Windows so we can run a layout test added by r87770 on them. Each
1086         implementation adds a wrapper function for Editor::setBaseWritingDirection()
1087         so LayoutTestController can call it.
1088
1089         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1090         (LayoutTestController::setTextDirection): Implemented the binding function
1091         for layoutTestController.setTextDirection (Windows).
1092         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
1093         Added a binding function for layoutTestController.setTextDirection (WebKit2).
1094         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp: ditto.
1095         (WTR::LayoutTestController::setTextDirection):
1096         * WebKitTestRunner/InjectedBundle/LayoutTestController.h: ditto.
1097
1098 2011-07-11  Raphael Kubo da Costa  <kubo@profusion.mobi>
1099
1100         [EFL] ImageDiff: Make sure gEcoreEvas is destroyed at the right time.
1101         https://bugs.webkit.org/show_bug.cgi?id=64293
1102
1103         Reviewed by Kent Tamura.
1104
1105         gEcoreEvas was being destroyed after shutdownEfl() was called, so the
1106         call to ecore_evas_free() failed.
1107
1108         We now shut it down manually before shutdownEfl() is called.
1109
1110         * DumpRenderTree/efl/ImageDiff.cpp:
1111         (main): Destroy gEcoreEvas before shutting down the EFL.
1112
1113 2011-07-11  Ryosuke Niwa  <rniwa@webkit.org>
1114
1115         Make Chromium Mac and Linux tests bot core builders
1116         https://bugs.webkit.org/show_bug.cgi?id=63196
1117
1118         Reviewed by Adam Barth.
1119
1120         Added Chromium Mac and Linux Release Tests bots core builders.
1121
1122         Also rearranged the order of bots so that test bots show up right next to
1123         respective builders.
1124
1125         * BuildSlaveSupport/build.webkit.org-config/config.json:
1126         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
1127         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
1128
1129 2011-07-11  Raphael Kubo da Costa  <kubo@profusion.mobi>
1130
1131         [EFL] DRT: Add code related to font management.
1132         https://bugs.webkit.org/show_bug.cgi?id=63989
1133
1134         Reviewed by Kent Tamura.
1135
1136         The added files are responsible for managing fontconfig and adding the
1137         fonts required by some layout tests to fontconfig's path.
1138
1139         * DumpRenderTree/efl/FontManagement.cpp: Added.
1140         (getFontDirectories):
1141         (getFontFiles):
1142         (addFontDirectories):
1143         (addFontFiles):
1144         (addFontsToEnvironment):
1145         * DumpRenderTree/efl/FontManagement.h: Added.
1146
1147 2011-07-08  Dirk Pranke  <dpranke@chromium.org>
1148
1149         nrwt: linting fixes
1150         https://bugs.webkit.org/show_bug.cgi?id=64225
1151
1152         Reviewed by Eric Siedel.
1153
1154         Miscellaneous linting fixes. The most notable change is that
1155         we add public attributes for user, executive, filesystem, and
1156         options on the Port object, so we don't have to refer to the
1157         "protected" versions all over the place".
1158
1159         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1160         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
1161         * Scripts/webkitpy/layout_tests/controllers/message_broker.py:
1162         * Scripts/webkitpy/layout_tests/controllers/message_broker_unittest.py:
1163         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1164         * Scripts/webkitpy/layout_tests/port/base.py:
1165         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1166
1167 2011-07-11  Ryosuke Niwa  <rniwa@webkit.org>
1168
1169         webkit-patch roll-chromium-deps no longer works
1170         https://bugs.webkit.org/show_bug.cgi?id=64324
1171
1172         Reviewed by James Robinson.
1173
1174         Lower the minimum commit message length to 10.
1175
1176         * Scripts/webkitpy/tool/steps/commit.py:
1177
1178 2011-07-11  Alice Boxhall  <aboxhall@chromium.org>
1179
1180         Convert json_results_generator.py to output version 4 JSON.
1181         https://bugs.webkit.org/show_bug.cgi?id=60869
1182
1183         Reviewed by Ojan Vafai.
1184
1185         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
1186         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
1187         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
1188
1189 2011-07-11  Martin Robinson  <mrobinson@igalia.com>
1190
1191         [GTK] [NRWT] Pixel tests do not work
1192         https://bugs.webkit.org/show_bug.cgi?id=64091
1193
1194         Reviewed by Eric Seidel.
1195
1196         Fix pixel test runs for GTK+ new-run-webkit-tests by exposing the
1197         path to the ImageDiff binary.
1198
1199         * Scripts/webkitpy/layout_tests/port/gtk.py: Implement _path_to_image_diff.
1200
1201 2011-07-11  Dirk Pranke  <dpranke@chromium.org>
1202
1203         Fix hang on win32 in a multiprocessing test that shouldn't have been running.
1204
1205         Unreviewed, build fix.
1206
1207         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1208
1209 2011-07-11  Adam Roben  <aroben@apple.com>
1210
1211         Don't count leaks as test failures on TestFailures's front page
1212
1213         As a bonus, this patch adds our first unit test.
1214
1215         Fixes <http://webkit.org/b/64303> REGRESSION (r90489): TestFailures page says far more tests
1216         are failing on the Leaks bot than actually are
1217
1218         Reviewed by Adam Barth.
1219
1220         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
1221         (Builder.prototype.getNumberOfFailingTests): Exclude lines containing "leak" when summing
1222         failure counts.
1223
1224         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
1225         Added. Contains a simple test to show that leaks aren't counted as test failures.
1226
1227         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
1228         Added. This is the test harness.
1229
1230 2011-06-14  Mike Stegeman  <mrstegeman@gmail.com>
1231
1232         [GTK] Add API to allow setting local storage database path
1233         https://bugs.webkit.org/show_bug.cgi?id=62091
1234
1235         Reviewed by Martin Robinson and Gustavo Noronha.
1236
1237         Expose the path of the localStorage databases through a setting
1238         to allow HTML5 localStorage to be persistent. New setting is:
1239         html5-local-storage-database-path
1240
1241         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1242         (resetDefaultsToConsistentValues):
1243
1244 2011-07-11  Adam Barth  <abarth@webkit.org>
1245
1246         garden-o-matic should be able to rebaseline tests
1247         https://bugs.webkit.org/show_bug.cgi?id=64186
1248
1249         Reviewed by Eric Seidel.
1250
1251         This patch contains a simple HTTP server binding for the
1252         rebaseline-test command.  The frontend will appear in another patch.
1253
1254         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1255         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
1256         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
1257
1258 2011-07-11  Adam Barth  <abarth@webkit.org>
1259
1260         Add some basic UI for showing regression ranges
1261         https://bugs.webkit.org/show_bug.cgi?id=64243
1262
1263         Reviewed by Eric Seidel.
1264
1265         This UI is all temporary.  It exists mostly just to have some buttons
1266         to click to exercise the code.
1267
1268         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
1269         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
1270         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
1271         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
1272         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
1273
1274 2011-07-11  Adam Barth  <abarth@webkit.org>
1275
1276         garden-o-matic should be able to determine which revisions caused a given failure
1277         https://bugs.webkit.org/show_bug.cgi?id=64189
1278
1279         Reviewed by Adam Roben.
1280
1281         Walking the failure history looking for failures turns out to be
1282         slightly tricky because the network requests are asynchronous.
1283         Currently we do all the fetches serially and our cache is unbounded.
1284         We'll probably optimize both those parameters eventually.
1285
1286         This patch also generalizes some functionality in the unit testing
1287         framework to make testing this sort of code easier.
1288
1289         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
1290         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
1291         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
1292
1293 2011-07-11  Adam Barth  <abarth@webkit.org>
1294
1295         Add a webkit-patch command for rebaselining an individual test
1296         https://bugs.webkit.org/show_bug.cgi?id=64246
1297
1298         Reviewed by Eric Seidel.
1299
1300         This patch introduces a command that's able to rebaseline a single
1301         test.  Currently, the command works only with the build.chromium.org
1302         buildbots, but extending it to work with the build.webkit.org bots
1303         shouldn't be that hard.
1304
1305         A complete rebaseling tool should also include an "optimize baselines"
1306         command (which moves/deletes baselines in order to reduce the number of
1307         expected results files), but that will come in a future patch.
1308
1309         Really BuilderToPort should be merged into builders.py, but I'm going
1310         to save that for a future patch as well.  (We need to stop shaving yaks
1311         at some point.)
1312
1313         * Scripts/webkitpy/tool/commands/rebaseline.py:
1314
1315 2011-07-11  Adam Barth  <abarth@webkit.org>
1316
1317         garden-o-matic should be able to roll out patches
1318         https://bugs.webkit.org/show_bug.cgi?id=64185
1319
1320         Reviewed by Eric Seidel.
1321
1322         This gardening server API is simply a binding to the rollout machinery
1323         already present in webkit-patch.
1324
1325         * Scripts/webkitpy/tool/commands/gardenomatic.py:
1326         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1327         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Added.
1328
1329 2011-07-11  Csaba Osztrogonác  <ossy@webkit.org>
1330
1331         [Qt][Mac] Unreviewed fix after r90746.
1332
1333         * Scripts/run-webkit-tests:
1334         (useNewRunWebKitTests): Really disable NRWT for qt-mac platform.
1335
1336 2011-07-11  Adam Roben  <aroben@apple.com>
1337
1338         Teach TestFailures that ORWT's results.html file might be missing due to all tests passing
1339
1340         There are three reasons why we might fail to fetch ORWT's results.html:
1341           - All tests passed, so no results.html was generated
1342           - Some error during the test run caused results.html not to be generated (e.g., ORWT
1343             timed out)
1344           - Some network error occurred when fetching results.html
1345
1346         We were failing to account for the first possibility in some cases. For test runs before
1347         r89610, we first check build.webkit.org/json to determine how many tests failed and whether
1348         ORWT exited early due to too many failures; if all tests passed then we don't fetch
1349         results.html at all. r89610 changed ORWT to put information in results.html about exiting
1350         early due to too many failures, so we no longer needed to check build.webkit.org/json to get
1351         that information, and in r89619 I changed TestFailures to do just that. But I forgot that we
1352         still needed to check build.webkit.org/json to find out if all tests passed!
1353
1354         Now, for test runs after r89610, we check results.html first, and then check
1355         build.webkit.org/json if we fail to fetch results.html. This lets us distinguish between all
1356         tests passing and the error cases.
1357
1358         Fixes <http://webkit.org/b/64280> TestFailures page can't pinpoint that r90699 caused 13
1359         tests to fail on Windows 7 Release (WebKit2 Tests)
1360
1361         Reviewed by David Kilzer.
1362
1363         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
1364         (LayoutTestResultsLoader.prototype.start): Bumped the cache version so that old, buggy
1365         cached data will get evicted. We were marking builds where all tests passed as errors!
1366         (LayoutTestResultsLoader.prototype._fetchAndParseORWTResults): Added success/error callback
1367         parameters to the fetchAndParseResultsHTML helper function, and added a similar
1368         fetchNumberOfFailingTests function that fetches data from build.webkit.org/json (code came
1369         from later in the function). For test runs before r89610, we first check
1370         build.webkit.org/json then check results.html. For builds after r89610, we first check
1371         results.html and then check build.webkit.org/json if we couldn't fetch results.html.
1372
1373 2011-07-11  Csaba Osztrogonác  <ossy@webkit.org>
1374
1375         new-run-webkit-tests does not support qt-arm or qt-4.8 results
1376         https://bugs.webkit.org/show_bug.cgi?id=64071
1377
1378         [Qt] NRWT should pick up the right httpd config file
1379         https://bugs.webkit.org/show_bug.cgi?id=64086
1380
1381         Reviewed by Zoltán Herczeg.
1382
1383         * Scripts/run-webkit-tests: Disable NRWT for qt-mac, qt-arm and qt-4.8 platforms until fix.
1384         (useNewRunWebKitTests):
1385
1386 2011-07-11  Kenichi Ishibashi  <bashi@chromium.org>
1387
1388         Add TestNetscapePlugIn/Tests/FormValue.cpp to DRT build files
1389         https://bugs.webkit.org/show_bug.cgi?id=64248
1390
1391         Reviewed by Kent Tamura.
1392
1393         * DumpRenderTree/DumpRenderTree.gypi: Added FormValue.cpp.
1394         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Ditto.
1395         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro: Ditto.
1396         * GNUmakefile.am: Ditto.
1397
1398 2011-07-10  Kenichi Ishibashi  <bashi@chromium.org>
1399
1400         Let plugins participate in form submission
1401         https://bugs.webkit.org/show_bug.cgi?id=13061
1402
1403         Adds NPPVformValue support to the testing plugin.
1404
1405         Reviewed by Anders Carlsson.
1406
1407         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added FormValue.cpp
1408         * DumpRenderTree/TestNetScapePlugIn/Tests/FormValue.cpp: Added.
1409         (FormValue::FormValue): Ditto.
1410         (FormValue::NPP_GetValue): Ditto.
1411
1412 2011-07-11  Roland Steiner  <rolandsteiner@chromium.org>
1413
1414         Reviewed by Tony Chang.
1415
1416         Complete functions in filesystem.py
1417         https://bugs.webkit.org/show_bug.cgi?id=63528
1418
1419         - missing text file functions added
1420         - functions sorted
1421         - removed 'append' optional parameters (were unused)
1422         - adapted filesystem_mock in the same way
1423
1424         * Scripts/webkitpy/common/system/filesystem.py:
1425         * Scripts/webkitpy/common/system/filesystem_mock.py:
1426         * Scripts/webkitpy/common/system/filesystem_unittest.py:
1427
1428 2011-07-10  Adam Barth  <abarth@webkit.org>
1429
1430         Clean up style in fallback path calculation
1431         https://bugs.webkit.org/show_bug.cgi?id=64239
1432
1433         Reviewed by Daniel Bates.
1434
1435         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1436         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1437         * Scripts/webkitpy/layout_tests/port/mac.py:
1438         * Scripts/webkitpy/layout_tests/port/win.py:
1439
1440 2011-07-10  Balazs Kelemen  <kbalazs@webkit.org>
1441
1442         [Qt][WK2] Implement the rest of WTR::PlatformWebView
1443         https://bugs.webkit.org/show_bug.cgi?id=63630
1444
1445         Reviewed by Kenneth Rohde Christiansen.
1446
1447         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
1448         (WTR::PlatformWebView::windowFrame):
1449         (WTR::PlatformWebView::setWindowFrame):
1450
1451 2011-07-10  Mark Rowe  <mrowe@apple.com>
1452
1453         Teach build-webkit how to find the latest WebKitSystemInterface binary.
1454
1455         * Scripts/build-webkit:
1456
1457 2011-07-09  Adam Roben  <aroben@apple.com>
1458
1459         Teach TestFailures to abbreviate the examples of test flakiness
1460
1461         These lists can get quite long, and it's not really helpful in most cases to have soooooo
1462         many examples of flakiness.
1463
1464         Fixes <http://webkit.org/b/64203> Lists of flaky revisions on TestFailures page can get so
1465         long they're hard to navigate
1466
1467         Reviewed by Dan Bates.
1468
1469         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
1470         (FlakyLayoutTestDetector.prototype.flakinessExamples): If we have more than a certain number
1471         of examples, replace the middle items with a separator.
1472
1473         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
1474         (.flakiness-example-separator): Added styles for the separator.
1475
1476         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1477         (ViewController.prototype._domForPossiblyFlakyTests): Use a vertical ellipsis to represent
1478         the separator.
1479
1480 2011-07-09  Dirk Pranke  <dpranke@chromium.org>
1481
1482         nrwt: stack traces from worker-side exceptions aren't very useful inside test-webkitpy
1483         https://bugs.webkit.org/show_bug.cgi?id=64218
1484
1485         Reviewed by Eric Seidel.
1486
1487         Exceptions aren't picklable and can't be sent across the
1488         manager/worker message queue without losing information. NRWT
1489         handles this by turning the stack trace into a set of strings,
1490         and logging the strings when we receive an exception from the
1491         worker. However, when you are running tests and something
1492         crashes on the worker side, test-webkitpy prints the
1493         manager-side stack trace, which is just confusing and useless.
1494
1495         This patch changes the logic so that exceptions are passed
1496         through as-is when the worker and manager are in the same
1497         process (the --worker-model=inline option). This increases the
1498         code paths slightly but makes crashes much more useful.
1499
1500         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1501         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
1502         * Scripts/webkitpy/layout_tests/controllers/message_broker.py:
1503         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1504         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1505
1506 2011-07-09  Darin Fisher  <darin@chromium.org>
1507
1508         Eliminate bad dependency on gfx::Point.
1509         https://bugs.webkit.org/show_bug.cgi?id=64228
1510
1511         Reviewed by Kent Tamura.
1512
1513         * DumpRenderTree/chromium/EventSender.cpp:
1514         (initMouseEvent):
1515
1516 2011-07-08  Ojan Vafai  <ojan@chromium.org>
1517
1518         pull static dashboard files into the appengine server from the chromium repository
1519         https://bugs.webkit.org/show_bug.cgi?id=64208
1520
1521         Reviewed by Dirk Pranke.
1522
1523         These files belong in the WebKit repo since they are tied primarily to the webkit tests.
1524         They have some extra bits to support chromium's gtests, but that seems fine.
1525         Mainly, this will allow other WebKit hackers to hack on the dashboards.
1526
1527         As a nice side-effect, we can now push the dashboard files when we do appengine pushes
1528         instead of the weird thing we used to do of pulling them from the Chromium repository
1529         and storing them in the appengine datastore. This allows for cleaning up a lot of code
1530         and will likely make the dashboards load a bit faster.
1531
1532         The new JS files don't fully match WebKit style, but I'd like to clean that up in a
1533         followup patch if possible to maintain my sanity with this patch.
1534
1535         * TestResultServer/app.yaml:
1536         * TestResultServer/handlers/dashboardhandler.py: Removed.
1537         * TestResultServer/handlers/menu.py:
1538         * TestResultServer/main.py:
1539         * TestResultServer/model/dashboardfile.py: Removed.
1540         * TestResultServer/static-dashboards/LICENSE.dygraph.txt: Added.
1541         * TestResultServer/static-dashboards/README.dygraph.txt: Added.
1542         * TestResultServer/static-dashboards/README.webtreemap.txt: Added.
1543         * TestResultServer/static-dashboards/aggregate_results.html: Added.
1544         * TestResultServer/static-dashboards/builders.js: Added.
1545         * TestResultServer/static-dashboards/dashboard_base.js: Added.
1546         * TestResultServer/static-dashboards/dygraph-combined.js: Added.
1547         * TestResultServer/static-dashboards/flakiness_dashboard.html: Added.
1548         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js: Added.
1549         * TestResultServer/static-dashboards/timeline_explorer.html: Added.
1550         * TestResultServer/static-dashboards/treemap.html: Added.
1551         * TestResultServer/static-dashboards/webtreemap.css: Added.
1552         * TestResultServer/static-dashboards/webtreemap.js: Added.
1553         * TestResultServer/stylesheets/dashboardfile.css: Removed.
1554         * TestResultServer/stylesheets/menu.css:
1555         * TestResultServer/templates/dashboardfilelist.html: Removed.
1556         * TestResultServer/templates/menu.html:
1557
1558 2011-07-08  Adam Roben  <aroben@apple.com>
1559
1560         Make TestFailures's list of flaky tests look more like the list of non-flaky tests
1561
1562         Fixes <http://webkit.org/b/64204> TestFailures page's flaky tests list is ugly!
1563
1564         Reviewed by Daniel Bates.
1565
1566         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
1567         (#failure-history, #possibly-flaky-tests): Expanded this rule to apply to the list of flaky
1568         tests.
1569         (#failure-history > li, #possibly-flaky-tests > li): Ditto, but moved the 50px left padding
1570         from here...
1571         (#failure-history > li): ...to here.
1572
1573         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1574         (ViewController.prototype._domForPossiblyFlakyTests): Give the list an id attribute for
1575         styling purposes.
1576
1577 2011-07-08  Jeffrey Pfau  <jpfau@apple.com>
1578
1579         Unreviewed, add myself as committer.
1580
1581         * Scripts/webkitpy/common/config/committers.py:
1582
1583 2011-07-08  Adam Barth  <abarth@webkit.org>
1584
1585         TestResultsServer should keep old test results
1586         https://bugs.webkit.org/show_bug.cgi?id=64199
1587
1588         Reviewed by Ojan Vafai.
1589
1590         Having historical data will help us do failure archeology.
1591
1592         * TestResultServer/handlers/testfilehandler.py:
1593         * TestResultServer/model/testfile.py:
1594         * TestResultServer/templates/showfilelist.jsonp: Added.
1595
1596 2011-07-08  Adam Roben  <aroben@apple.com>
1597
1598         Remove commit-log-editor's dependency on Module::Load::Conditional
1599
1600         This module isn't available in Perl 5.8.8 (the version used on Leopard).
1601
1602         Fixes <http://webkit.org/b/64198> REGRESSION (r90583):
1603         webkitpy.common.checkout.checkout_unittest failing on Leopard
1604
1605         Reviewed by Daniel Bates.
1606
1607         * Scripts/commit-log-editor: Use the new loadTermReadKey() function instead of
1608         Module::Load::Conditional::can_load.
1609         (loadTermReadKey): Added. Tries to load Term::ReadKey and returns true if it's successful.
1610
1611 2011-07-08  Adam Roben  <aroben@apple.com>
1612
1613         Make checkout_unittest more robust against files moving around
1614
1615         Fixes <http://webkit.org/b/64197> checkout_unittest contains ugly, fragile code to find the
1616         Scripts directory
1617
1618         Reviewed by Adam Barth.
1619
1620         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1621         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Instantiate a real SCM
1622         object and use it to get the path to the Scripts directory, rather than hard-coding the
1623         relative path from this file to Scripts.
1624
1625 2011-07-08  Adam Roben  <aroben@apple.com>
1626
1627         Make TestFailures remember that run-webkit-tests timed out, even across reloads
1628
1629         Fixes <http://webkit.org/b/64193> TestFailures page incorrectly thinks all tests passed in
1630         http://build.webkit.org/builders/Windows%207%20Release%20(Tests)/builds/14589 after a reload
1631
1632         Reviewed by Daniel Bates.
1633
1634         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
1635         (LayoutTestResultsLoader.prototype.start): Store an "error" attribute in the cached data.
1636         When true, it indicates that there was an error retrieving the results for this build and
1637         that the errorCallback should be called.
1638
1639 2011-07-08  Adam Barth  <abarth@webkit.org>
1640
1641         Teach garden-o-matic how to display test results
1642         https://bugs.webkit.org/show_bug.cgi?id=64141
1643
1644         Reviewed by Ojan Vafai.
1645
1646         This patch includes basic infrastructure for probing build.chromium.org
1647         for test results.  We only handle text and image tests, not anything
1648         complicated like reftests.  Also, we're using the revision/build
1649         independent results store on the server, so we're avoiding that
1650         complication for now.
1651
1652         It's slightly hacky that we need to probe the server to see what kinds
1653         of results exist.  A better solution would be to add CORS support to
1654         the server or to use the local server to help.
1655
1656         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
1657         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
1658         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
1659         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
1660         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
1661         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
1662
1663 2011-07-08  Dirk Pranke  <dpranke@chromium.org>
1664
1665         REGRESSION(90419) NRWT's httpd locking is broken for --child-processes=1
1666         https://bugs.webkit.org/show_bug.cgi?id=64092
1667
1668         Reviewed by Tony Chang.
1669
1670         The code was incorrectly creating empty shards in the
1671         shard_in_two case.
1672
1673         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1674         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1675
1676 2011-07-08  Adam Roben  <aroben@apple.com>
1677
1678         Teach buildbot to figure out how many webkitpy/webkitperl tests failed
1679
1680         Fixes <http://webkit.org/b/64192> It's hard to tell how many test-webkitpy/test-webkitperl
1681         tests failed when looking at build.webkit.org
1682
1683         Reviewed by Eric Seidel.
1684
1685         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1686         (TestWithFailureCount): New class that represents a test build step which has an associated
1687         failure count. Eventually we should move more of our test classes to deriving from this.
1688         (TestWithFailureCount.countFailures): Method for subclasses to override to say how many
1689         failures occurred.
1690
1691         (TestWithFailureCount.commandComplete):
1692         (TestWithFailureCount.evaluateCommand):
1693         (TestWithFailureCount.getText):
1694         (TestWithFailureCount.getText2):
1695         These were all based on RunGtkAPITests.
1696
1697         (RunPythonTests): Changed to inherit from TestWithFailureCount.
1698         (RunPythonTests.countFailures): Parses the test-webkitpy output looking for the count of
1699         failures.
1700         (RunPerlTests): Changed to inherit from TestWithFailureCount.
1701         (RunPerlTests.countFailures): Parses the test-webkitperl output looking for the count of
1702         failures.
1703
1704 2011-07-08  Adam Roben  <aroben@apple.com>
1705
1706         Ensure $CHANGE_LOG_EMAIL_ADDRESS is set when testing webkitpy's commit-log-editor integration
1707
1708         Fixes <http://webkit.org/b/64180> REGRESSION (r90564): test-webkitpy failing on multiple
1709         bots due to commit-log-editor errors
1710
1711         Reviewed by Adam Barth.
1712
1713         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1714         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Set
1715         $CHANGE_LOG_EMAIL_ADDRESS to the patch author's email address. This ensures that
1716         commit-log-editor can find a value for the committer's email, and that the committer and
1717         author email addresses match, which will prevent commit-log-editor from inserting a "Patch
1718         by" line in the commit message.
1719
1720 2011-07-08  Andreas Kling  <kling@webkit.org>
1721
1722         [Qt][WK2] Views should expose QActions for basic navigation.
1723         https://bugs.webkit.org/show_bug.cgi?id=64174
1724
1725         Reviewed by Benjamin Poulain.
1726
1727         Bring the toolbar in Qt's MiniBrowser back to life.
1728
1729         * MiniBrowser/qt/BrowserView.cpp:
1730         (BrowserView::navigationAction):
1731         * MiniBrowser/qt/BrowserView.h:
1732         * MiniBrowser/qt/BrowserWindow.cpp:
1733         (BrowserWindow::BrowserWindow):
1734
1735 2011-07-08  Adam Roben  <aroben@apple.com>
1736
1737         Teach TestFailures how to find test names in commit-log-editor-style commit messages
1738
1739         TestFailures was relying on Trac turning the list of modified files in our commit messages
1740         into an HTML list. But Trac only does this when the list of modified files is indented.
1741         commit-log-editor doesn't indent the file list, so the list wasn't being turned into an HTML
1742         list, which was confusing TestFailures.
1743
1744         TestFailures now does much simpler parsing of the commit message (i.e., just a substring
1745         search) without relying at all on its structure.
1746
1747         Fixes <http://webkit.org/b/64173> TestFailures page fails to blame r90608 for breaking
1748         fast/dom/HTMLProgressElement/progress-element-markup.html on Windows
1749
1750         Reviewed by David Kilzer.
1751
1752         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
1753         (Trac.prototype.getCommitDataForRevisionRange): Instead of trying to parse the commit
1754         message, just return its text.
1755
1756         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1757         (ViewController.prototype._domForRegressionRange): Instead of searching for test names in
1758         each commit's list of modified files, just search for test names anywhere in the commit's
1759         message.
1760
1761 2011-07-08  Adam Barth  <abarth@webkit.org>
1762
1763         sheriffbot is too spammy in IRC
1764         https://bugs.webkit.org/show_bug.cgi?id=64153
1765
1766         Reviewed by Eric Seidel.
1767
1768         Reporting failures in IRC worked well when the tree was greener than it
1769         is today.  Nowadays, this feature mostly just results in spam about
1770         false positives.  If we reach a greener state, we can bring this code
1771         back.
1772
1773         (Another possibility is to restrict this feature to builder bots,
1774         rather than including testers, as we did before this patch.)
1775
1776         * Scripts/webkitpy/tool/commands/sheriffbot.py:
1777         * Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
1778
1779 2011-07-08  Adam Roben  <aroben@apple.com>
1780
1781         Ensure commit-log-editor adds a "Patch by" line when the author and committer are different
1782
1783         Previously we were only adding a "Patch by" line when the ChangeLog contained a "Reviewed
1784         by" line. But some patches (like rollout patches) don't contain that line. Now we always add
1785         "Patch by" to the commit log regardless of the ChangeLog's contents.
1786
1787         Fixes <http://webkit.org/b/64127> Committer for r90588 is commit-queue@webkit.org, but
1788         should have been abarth@webkit.org
1789
1790         Reviewed by Anders Carlsson.
1791
1792         * Scripts/commit-log-editor:
1793         (createCommitMessage): Try to put the "Patch by" line just above the "Reviewed by" line, as
1794         before. If there is no "Reviewed by" line, try to put it just above the first modified file.
1795         If all else fails, put it at the end of the commit message.
1796         (patchAuthorshipString): Added. Code came from createCommitMessage.
1797
1798 2011-07-08  Herczeg Zoltan  <zherczeg@webkit.org>
1799
1800         Reviewed by Andreas Kling.
1801
1802         Adding myself as a reviewer.
1803
1804         * Scripts/webkitpy/common/config/committers.py:
1805
1806 2011-07-07  Kevin Ollivier  <kevino@theolliviers.com>
1807
1808         [wx] Unreviewed build fix, remove old files from the tree and update paths to new ones.
1809
1810         * wx/build: Removed.
1811         * wx/build/build_utils.py: Removed.
1812         * wx/build/settings.py: Removed.
1813         * wx/build/waf_extensions.py: Removed.
1814         * wx/build/wxpresets.py: Removed.
1815         * wx/packaging/build-mac-installer.py:
1816
1817 2011-07-07  Kevin Ollivier  <kevino@theolliviers.com>
1818
1819         [wx] Unreviewed build fix, more fixes for install name issues.
1820
1821         * wx/packaging/build-mac-installer.py:
1822
1823 2011-07-07  Dirk Pranke  <dpranke@chromium.org>
1824
1825         test-webkitpy fails on chromium win
1826         https://bugs.webkit.org/show_bug.cgi?id=64137
1827
1828         Reviewed by Eric Seidel.
1829
1830         test-webkitpy doesn't play nicely with the multiprocessing
1831         module on win32. An earlier change actually reenabled the tests
1832         on win32 by mistake.
1833
1834         This patch also fixes a few cases where path names will trip
1835         things up on win32 (e.g., by testing unix paths on windows).
1836         We do not lose any real coverage here with those fixes.
1837
1838         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
1839         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
1840         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
1841         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1842
1843 2011-07-07  Leandro Pereira  <leandro@profusion.mobi>
1844
1845         [ImageDiff] Calculate/print difference right after reading baseline image.
1846         https://bugs.webkit.org/show_bug.cgi?id=64117
1847
1848         Reviewed by Kent Tamura.
1849
1850         Otherwise, ImageDiff will block on fgets() until the universe ends or you ^C
1851         it.  Whichever happens first.
1852
1853         * DumpRenderTree/efl/ImageDiff.cpp:
1854         (main):
1855
1856 2011-07-07  Raphael Kubo da Costa  <kubo@profusion.mobi>
1857
1858         [EFL] DRT: Leak cairo_t to make the pixel tests stop crashing
1859         https://bugs.webkit.org/show_bug.cgi?id=64107
1860
1861         Reviewed by Kent Tamura.
1862
1863         The pointer is later managed by BitmapContext, but it was being
1864         de-refed and deleted earlier by our RefPtr, causing crashes in all
1865         pixel tests.
1866
1867         We now call leakRef() to make sure it is not removed when our
1868         createBitmapContextFromWebView goes out of context.
1869
1870         * DumpRenderTree/efl/PixelDumpSupportEfl.cpp:
1871         (createBitmapContextFromWebView):
1872
1873 2011-07-07  Adam Roben  <aroben@apple.com>
1874
1875         Update TestFailures's title and header immediately upon navigation
1876
1877         Fixes <http://webkit.org/b/64125> TestFailures page seems slow to react on link clicks
1878
1879         Reviewed by Daniel Bates.
1880
1881         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1882         (ViewController.prototype.loaded): Create and store an h1 element for displaying the page's
1883         title.
1884         (ViewController.prototype._displayBuilder): Set the page's title and clear out any old
1885         content immediately instead of waiting for the first set of results for the history
1886         analyzer. As we receive new results from the analyzer we'll just clear out the main content
1887         area and replace it with the new info.
1888         (ViewController.prototype._displayTesters): Set the page's title and clear out any old
1889         content immediately instead of waiting for the list of testers from the buildbot. Sprinkled
1890         in a little use of our removeAllChildren helper, too.
1891         (ViewController.prototype._setTitle): Added. Set's the page's title and the header text.
1892
1893 2011-07-07  Adam Roben  <aroben@apple.com>
1894
1895         Make Checkout use SCM's Executive instead of conjuring up its own
1896
1897         This will improve integration with the rest of webkitpy, particularly when invoked via
1898         webkit-patch.
1899
1900         Fixes <http://webkit.org/b/64115> REGRESSION (r90564): webkitpy's Checkout class uses
1901         Executive inappropriately
1902
1903         Reviewed by Adam Barth.
1904
1905         * Scripts/webkitpy/common/checkout/checkout.py:
1906         (Checkout.commit_message_for_this_commit): Use SCM.run instead of creating an Executive for
1907         our own use. SCM might have some super-special Executive that it uses under the covers, and
1908         we want to use it, too!
1909
1910         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1911         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Mock the SCM.run
1912         method to call through to Executive.run_command.
1913
1914 2011-07-07  Adam Roben  <aroben@apple.com>
1915
1916         Make Term/ReadKey.pm an optional dependency of commit-log-editor
1917
1918         When Term/ReadKey.pm isn't present, the --regenerate-log switch will be non-functional.
1919
1920         Fixes <http://webkit.org/b/64113> REGRESSION (r90564):
1921         webkitpy.common.checkout.checkout_unittest failing on GTK bots due to missing
1922         Term/ReadKey.pm module
1923
1924         Reviewed by Xan Lopez.
1925
1926         * Scripts/commit-log-editor: Use Module::Load::Conditional::can_load to only load
1927         Term/ReadKey.pm if it's present. If it isn't present, ignore the --regenerate-log switch so
1928         that existing commit messages will be preserved. (Users can manually remove existing commit
1929         messages using git-reset.)
1930
1931 2011-07-07  Adam Roben  <aroben@apple.com>
1932
1933         Keep testing that commit messages containing Unicode are handled properly by webkitpy
1934
1935         Fixes <http://webkit.org/b/64109> REGRESSION (r90571): test-webkitpy no longer tests that we
1936         correctly handle Unicode in commit messages
1937
1938         Reviewed by Eric Seidel.
1939
1940         * Scripts/webkitpy/common/checkout/checkout_unittest.py: Put some Unicode characters in
1941         places in the ChangeLog that will end up being included in the commit message.
1942
1943 2011-07-07  Andreas Kling  <kling@webkit.org>
1944
1945         [Qt][WK2] Remove Symbian code.
1946         https://bugs.webkit.org/show_bug.cgi?id=64101
1947
1948         Reviewed by Benjamin Poulain.
1949
1950         * MiniBrowser/qt/BrowserWindow.cpp:
1951         (BrowserWindow::BrowserWindow):
1952         (BrowserWindow::screenshot):
1953         * MiniBrowser/qt/MiniBrowser.pro:
1954         * MiniBrowser/qt/MiniBrowserApplication.h:
1955         (WindowOptions::WindowOptions):
1956
1957 2011-07-07  Andreas Kling  <kling@webkit.org>
1958
1959         [Qt][WK2] Don't support ridiculous matrix of QT_NO_FEATURE combinations.
1960         https://bugs.webkit.org/show_bug.cgi?id=64099
1961
1962         Reviewed by Benjamin Poulain.
1963
1964         * MiniBrowser/qt/BrowserWindow.cpp:
1965         (BrowserWindow::openFile):
1966         (BrowserWindow::screenshot):
1967         (BrowserWindow::loadURLListFromFile):
1968
1969 2011-07-07  Adam Roben  <aroben@apple.com>
1970
1971         Completely ignore too-many-failures builds in TestFailures in most circumstances
1972
1973         Because a semi-arbitrary subset of tests are run in a too-many-failures build, we can't
1974         really use them to perform regression analysis. The only time we want to pay attention to
1975         too-many-failures builds is when we're trying to explain when the current bout of
1976         too-many-failures started.
1977
1978         Fixes <http://webkit.org/b/64106> TestFailures page sometimes claims a test started failing
1979         in a build that didn't even run it (because it exited early due to too many failues)
1980
1981         Reviewed by David Kilzer.
1982
1983         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
1984         (LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): Removed old, broken
1985         too-many-failures handling that would cause us to blame builds that didn't even run a given
1986         test for breaking it. Instead, skip over all too-many-failures builds unless the most recent
1987         build was itself a too-many-failures build.
1988
1989 2011-07-07  Adam Roben  <aroben@apple.com>
1990
1991         Teach webkitpy's Checkout class to use commit-log-editor to create commit messages
1992
1993         Fixes <http://webkit.org/b/26755> webkit-patch's commit messages are less readable than
1994         commit-log-editor's
1995
1996         Reviewed by David Kilzer.
1997
1998         * Scripts/webkitpy/common/checkout/checkout.py:
1999         (Checkout.commit_message_for_this_commit): Run commit-log-editor, passing it the paths of
2000         the modified ChangeLogs, to generate the commit message, rather than trying to generate one
2001         ourselves.
2002
2003         * Scripts/webkitpy/common/checkout/checkout_unittest.py: Updated the expected commit message
2004         to match commit-log-editor's format.
2005         (CommitMessageForThisCommitTest.setUp): Write the ChangeLogs into Tools and LayoutTests
2006         directories so we can see how the various entries get labeled in the commit message.
2007         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Create a mock SCM
2008         instance that knows how to find commit-log-editor and pass it to our Checkout instance.
2009         Don't bother capturing output, since there shouldn't be any.
2010
2011 2011-07-07  Adam Roben  <aroben@apple.com>
2012
2013         Add a --print-log option to commit-log-editor
2014
2015         When specified, commit-log-editor takes a set of ChangeLog paths on the command line,
2016         generates a commit message from those ChangeLogs, prints it to stdout, and exits.
2017
2018         Prep work for fixing <http://webkit.org/b/26755> webkit-patch's commit messages are less
2019         readable than commit-log-editor's
2020
2021         Reviewed by David Kilzer.
2022
2023         * Scripts/commit-log-editor: Changed to use Getopt::Long to parse options. Added --print-log
2024         option, which calls through to createCommitMessage, prints the result, and exits.
2025         (printUsageAndExit): Renamed from "usage" for clarity. Beefed up the usage statement to
2026         explain commit-log-editor's 2.5 (normal, --print-log, and --help) modes of operation.
2027
2028 2011-07-07  Adam Roben  <aroben@apple.com>
2029
2030         Extract commit-log-editor's commit-message-generation code into a separate function
2031
2032         As a bonus, we now skip a bunch of work in the case where we are generating a commit message
2033         for staged git changes.
2034
2035         Prep work for fixing <http://webkit.org/b/26755> webkit-patch's commit messages are less
2036         readable than commit-log-editor's
2037
2038         Reviewed by David Kilzer.
2039
2040         * Scripts/commit-log-editor: Moved the default definition of $endl up toward the top of the
2041         file so that it will be defined even if we aren't operating on an existing log message.
2042         Moved a bunch of top-level code to generate the commit message from here...
2043         (createCommitMessage): ...to here.
2044
2045 2011-07-07  Carlos Garcia Campos  <cgarcia@igalia.com>
2046
2047         [GTK] Open links in a new window when clicking with the middle button in MiniBrowser
2048         https://bugs.webkit.org/show_bug.cgi?id=63988
2049
2050         Reviewed by Martin Robinson.
2051
2052         * MiniBrowser/gtk/BrowserWindow.c:
2053         (browserWindowConstructed):
2054         (decidePolicyForNavigationAction): Ignore the action if a link is
2055         clicked with the middle buttonm and open the link in a new window.
2056         (browserWindowPolicyClientInit):
2057
2058 2011-07-07  Adam Barth  <abarth@webkit.org>
2059
2060         Remove style_references.py
2061         https://bugs.webkit.org/show_bug.cgi?id=64070
2062
2063         Reviewed by Eric Seidel.
2064
2065         We never really adopted the _references design, and this file is a bit
2066         of an orphan at this point.
2067
2068         * Scripts/check-webkit-style:
2069         * Scripts/webkitpy/style/checker.py:
2070         * Scripts/webkitpy/style/checker_unittest.py:
2071         * Scripts/webkitpy/style/checkers/python.py:
2072         * Scripts/webkitpy/style/checkers/test_expectations.py:
2073         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2074         * Scripts/webkitpy/style/main_unittest.py:
2075         * Scripts/webkitpy/style_references.py: Removed.
2076
2077 2011-07-07  Eric Seidel  <eric@webkit.org>
2078
2079         Fix WebKit2 expected results search paths for Mac and Qt under new-run-webkit-tests
2080         https://bugs.webkit.org/show_bug.cgi?id=64056
2081
2082         Unreviewed.
2083
2084         I had written this unittest prior to landing my change, but
2085         had named it qt_unitest.py, so it wasn't running (nor was
2086         it added to my git repo).
2087
2088         Once I fixe the name of the unittest file, it was very easy
2089         to fix the error in webkit.py (which already had a FIXME).
2090
2091         * Scripts/webkitpy/layout_tests/port/qt_unittest.py: Added.
2092         * Scripts/webkitpy/layout_tests/port/webkit.py:
2093
2094 2011-07-07  Eric Seidel  <eric@webkit.org>
2095
2096         Unreviewed.  Add more debug logging about which test expectations we're using.
2097
2098         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2099         * Scripts/webkitpy/layout_tests/port/webkit.py:
2100
2101 2011-07-06  Adam Barth  <abarth@webkit.org>
2102
2103         Remove unexpected_results.json
2104         https://bugs.webkit.org/show_bug.cgi?id=64065
2105
2106         Reviewed by Eric Seidel.
2107
2108         No one should be using this file any more.  This patch removes it from
2109         the face of the earth.  This patch also removes two NRWT-specific
2110         command line options that were implemented using
2111         unexpected_results.json.
2112
2113         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
2114         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2115         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2116         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2117         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2118         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
2119         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
2120
2121 2011-07-07  Adam Barth  <abarth@webkit.org>
2122
2123         REGRESSION(90520): chromium GPU tests produce output in the LayoutTests directory
2124         https://bugs.webkit.org/show_bug.cgi?id=64067
2125
2126         Reviewed by Eric Seidel.
2127
2128         The author of r90520 forgot to change the chromium_gpu.py implemenation
2129         of this method as well.
2130
2131         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2132
2133 2011-07-06  Eric Seidel  <eric@webkit.org>
2134
2135         Fix WebKit2 expected results search paths for Mac and Qt under new-run-webkit-tests
2136         https://bugs.webkit.org/show_bug.cgi?id=64056
2137
2138         Reviewed by Adam Barth.
2139
2140         This fixes the MacPort's version of baseline_search_path to include
2141         wk2 fallback like WebKitPort's does.  This also re-works the QtPort
2142         to set self._name (as other ports do) so that WebKitPort's default
2143         implementations will understand Qt operating system flavors correctly
2144         (including when running wk2 tests).
2145
2146         In trying to test this, I realized that nearly all of our port-tests
2147         use the default FileSystem(), User() and Executive() objects, which
2148         is really bad, because they expose details of your system in the unittest results!
2149
2150         Once I fixed many of the Port() calls to pass MockFileSystem(), then
2151         MacPort.test_tests_for_other_platforms started failing, and I had to
2152         teach MockFileSystem.glob how to handle directories to make it work again.
2153
2154         I removed the useless PortTestCase.test_baseline_search_path and replaced
2155         it with more useful test_baseline_search_path tests in MacPortTest and QtPortTest.
2156
2157         * Scripts/webkitpy/common/config/ports.py:
2158         * Scripts/webkitpy/common/system/filesystem_mock.py:
2159         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2160         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2161         * Scripts/webkitpy/layout_tests/port/mac.py:
2162         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2163         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2164         * Scripts/webkitpy/layout_tests/port/qt.py:
2165         * Scripts/webkitpy/layout_tests/port/test_files.py:
2166         * Scripts/webkitpy/layout_tests/port/win.py:
2167         * Scripts/webkitpy/to_be_moved/deduplicate_tests.py:
2168         * Scripts/webkitpy/tool/mocktool.py:
2169
2170 2011-07-06  Adam Barth  <abarth@webkit.org>
2171
2172         Wean rebaseline-server off unexpected_results.json
2173         https://bugs.webkit.org/show_bug.cgi?id=64063
2174
2175         Reviewed by Eric Seidel.
2176
2177         In the process of changing this code to use full_results.json, I
2178         noticed that the code was broken (because it wasn't tested).  This
2179         patch also adds test coverage for the broken code.
2180
2181         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
2182         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
2183         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
2184
2185 2011-07-06  Adam Barth  <abarth@webkit.org>
2186
2187         Wean resultsjsonparser off unexpected_results.json
2188         https://bugs.webkit.org/show_bug.cgi?id=64061
2189
2190         Reviewed by Eric Seidel.
2191
2192         We're about to delete unexpected_results.json, so we need to remove all
2193         the users.
2194
2195         * Scripts/webkitpy/common/config/ports.py:
2196         * Scripts/webkitpy/common/net/resultsjsonparser.py:
2197         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
2198         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2199
2200 2011-07-06  Adam Barth  <abarth@webkit.org>
2201
2202         Cleanup result_summary.py
2203         https://bugs.webkit.org/show_bug.cgi?id=64057
2204
2205         Reviewed by Eric Seidel.
2206
2207         This class had a bunch of out-of-date docstrings that no longer make
2208         any sense.
2209
2210         * Scripts/webkitpy/layout_tests/models/result_summary.py:
2211
2212 2011-07-06  Adam Barth  <abarth@webkit.org>
2213
2214         Tweak the ChangeLog for DEPS rolls to say "Unreviewed" so that the
2215         patches can be landed by the commit-queue.
2216
2217         * Scripts/webkitpy/tool/steps/preparechangelogfordepsroll.py:
2218
2219 2011-07-06  Adam Barth  <abarth@webkit.org>
2220
2221         Normalize callers of logging.getLogger to use __name__
2222         https://bugs.webkit.org/show_bug.cgi?id=64054
2223
2224         Reviewed by Eric Seidel.
2225
2226         Most callers were already using __name__.  This patch makes everyone
2227         consistent.
2228
2229         * Scripts/webkitpy/common/checkout/diff_parser.py:
2230         * Scripts/webkitpy/common/checkout/scm/svn.py:
2231         * Scripts/webkitpy/common/net/statusserver.py:
2232         * Scripts/webkitpy/common/system/executive.py:
2233         * Scripts/webkitpy/common/system/file_lock.py:
2234         * Scripts/webkitpy/common/system/user.py:
2235         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2236         * Scripts/webkitpy/layout_tests/models/result_summary.py:
2237         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2238         * Scripts/webkitpy/layout_tests/port/chromium.py:
2239         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2240         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
2241         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2242         * Scripts/webkitpy/layout_tests/port/gtk.py:
2243         * Scripts/webkitpy/layout_tests/port/http_lock.py:
2244         * Scripts/webkitpy/layout_tests/port/mac.py:
2245         * Scripts/webkitpy/layout_tests/port/qt.py:
2246         * Scripts/webkitpy/layout_tests/port/server_process.py:
2247         * Scripts/webkitpy/layout_tests/port/webkit.py:
2248         * Scripts/webkitpy/layout_tests/port/win.py:
2249         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
2250         * Scripts/webkitpy/layout_tests/servers/http_server.py:
2251         * Scripts/webkitpy/layout_tests/servers/http_server_base.py:
2252         * Scripts/webkitpy/layout_tests/servers/websocket_server.py:
2253         * Scripts/webkitpy/layout_tests/views/printing.py:
2254         * Scripts/webkitpy/python24/versioning.py:
2255         * Scripts/webkitpy/style/checker.py:
2256         * Scripts/webkitpy/style/checkers/test_expectations.py:
2257         * Scripts/webkitpy/style/patchreader.py:
2258         * Scripts/webkitpy/to_be_moved/update_webgl_conformance_tests.py:
2259
2260 2011-07-06  Adam Barth  <abarth@webkit.org>
2261
2262         Move model classes out of layout_package and into models
2263         https://bugs.webkit.org/show_bug.cgi?id=64053
2264
2265         Reviewed by Eric Seidel.
2266
2267         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2268         * Scripts/webkitpy/common/net/layouttestresults.py:
2269         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
2270         * Scripts/webkitpy/common/net/resultsjsonparser.py:
2271         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
2272         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2273         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2274         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
2275         * Scripts/webkitpy/layout_tests/controllers/worker.py:
2276         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
2277         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2278         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2279         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py: Removed.
2280         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py: Removed.
2281         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py: Removed.
2282         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py: Removed.
2283         * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py: Removed.
2284         * Scripts/webkitpy/layout_tests/layout_package/test_input.py: Removed.
2285         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
2286         * Scripts/webkitpy/layout_tests/layout_package/test_results.py: Removed.
2287         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py: Removed.
2288         * Scripts/webkitpy/layout_tests/models: Added.
2289         * Scripts/webkitpy/layout_tests/models/__init__.py: Added.
2290         * Scripts/webkitpy/layout_tests/models/result_summary.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/result_summary.py.
2291         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_expectations.py.
2292         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py.
2293         * Scripts/webkitpy/layout_tests/models/test_failures.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_failures.py.
2294         * Scripts/webkitpy/layout_tests/models/test_failures_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py.
2295         * Scripts/webkitpy/layout_tests/models/test_input.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_input.py.
2296         * Scripts/webkitpy/layout_tests/models/test_results.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_results.py.
2297         * Scripts/webkitpy/layout_tests/models/test_results_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py.
2298         * Scripts/webkitpy/layout_tests/port/chromium.py:
2299         * Scripts/webkitpy/layout_tests/views/printing.py:
2300         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
2301         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2302         * Scripts/webkitpy/style_references.py:
2303         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
2304         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
2305         * Scripts/webkitpy/tool/bot/flakytestreporter_unittest.py:
2306         * Scripts/webkitpy/tool/commands/queues_unittest.py:
2307         * Scripts/webkitpy/tool/commands/rebaseline.py:
2308
2309 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
2310
2311         nrwt: fix typo preventing http startup on windows
2312         https://bugs.webkit.org/show_bug.cgi?id=64050
2313
2314         Reviewed by Eric Siedel.
2315
2316         NRWT checks whether it needs to start the servers by looking
2317         for "/http/" in the test names. After r90520, the leading slash
2318         is not present, so the check isn't quite right.
2319
2320         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2321         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2322
2323 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
2324
2325         nrwt: remove --use-apache from the command line
2326         https://bugs.webkit.org/show_bug.cgi?id=63358
2327
2328         Reviewed by Adam Barth.
2329
2330         Re-land the fix in bug 63358 - the prior fix was flawed in that
2331         we were trying to start the http server during check_sys_deps()
2332         but not actually setting up the environment properly. This
2333         broken prior to this change, but exposed by it, since the patch
2334         changes the chromium port to call the base class's checks.
2335
2336         * Scripts/webkitpy/common/system/executive.py:
2337         * Scripts/webkitpy/layout_tests/port/base.py:
2338         * Scripts/webkitpy/layout_tests/port/chromium.py:
2339         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2340         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2341         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2342         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
2343         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2344         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2345
2346 2011-07-06  Adam Barth  <abarth@webkit.org>
2347
2348         Move view-related code out of layout_package and into views
2349         https://bugs.webkit.org/show_bug.cgi?id=64051
2350
2351         Reviewed by Eric Seidel.
2352
2353         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2354         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2355         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
2356         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
2357         * Scripts/webkitpy/layout_tests/layout_package/metered_stream.py: Removed.
2358         * Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py: Removed.
2359         * Scripts/webkitpy/layout_tests/layout_package/printing.py: Removed.
2360         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py: Removed.
2361         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2362         * Scripts/webkitpy/layout_tests/views: Added.
2363         * Scripts/webkitpy/layout_tests/views/__init__.py: Added.
2364         * Scripts/webkitpy/layout_tests/views/metered_stream.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/metered_stream.py.
2365         * Scripts/webkitpy/layout_tests/views/metered_stream_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py.
2366         * Scripts/webkitpy/layout_tests/views/printing.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/printing.py.
2367         * Scripts/webkitpy/layout_tests/views/printing_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py.
2368
2369 2011-07-06  Adam Barth  <abarth@webkit.org>
2370
2371         Add goofy comments to explain why we have these files.
2372
2373         * Scripts/webkitpy/layout_tests/controllers/__init__.py:
2374         * Scripts/webkitpy/layout_tests/layout_package/__init__.py:
2375
2376 2011-07-06  Adam Barth  <abarth@webkit.org>
2377
2378         Move a bunch of classes out of layout_tests.layout_package and into layout_tests.controllers
2379         https://bugs.webkit.org/show_bug.cgi?id=64048
2380
2381         Reviewed by Eric Seidel.
2382
2383         * Scripts/webkitpy/layout_tests/controllers: Added.
2384         * Scripts/webkitpy/layout_tests/controllers/__init__.py: Added.
2385         * Scripts/webkitpy/layout_tests/controllers/manager.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager.py.
2386         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py.
2387         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py.
2388         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py.
2389         * Scripts/webkitpy/layout_tests/controllers/message_broker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker.py.
2390         * Scripts/webkitpy/layout_tests/controllers/message_broker_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py.
2391         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py.
2392         * Scripts/webkitpy/layout_tests/controllers/worker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/worker.py.
2393         * Scripts/webkitpy/layout_tests/layout_package/manager.py: Removed.
2394         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py: Removed.
2395         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py: Removed.
2396         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py: Removed.
2397         * Scripts/webkitpy/layout_tests/layout_package/message_broker.py: Removed.
2398         * Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py: Removed.
2399         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
2400         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: Removed.
2401         * Scripts/webkitpy/layout_tests/layout_package/worker.py: Removed.
2402         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2403
2404 2011-07-06  Adam Barth  <abarth@webkit.org>
2405
2406         Attempt to fix unittest on Leopard.  Let's enumerate the keys in sorted
2407         order so the test results are predictable.
2408
2409         * Scripts/webkitpy/tool/bot/irc_command.py:
2410         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
2411
2412 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
2413
2414         Reviewed by Eric Siedel.
2415
2416         nrwt: make the code be consistent about using test names instead of filenames or paths
2417
2418         https://bugs.webkit.org/show_bug.cgi?id=63597
2419
2420         It can be hard to figure out where in the code we're assuming
2421         test names are listed as unix-style relative filenames and where
2422         they are either absolute paths or relative paths following the
2423         host filesystem convention.
2424
2425         This patch changes things so that everything outside of the
2426         Port object uses (and must assume) unix-style relative
2427         filenames (with one exception, which is specifying host-local
2428         filenames as a list of test arguments on the command line).
2429
2430         This should make things clearer and more consistent, and also
2431         removes a lot of path conversion calls.
2432
2433         The changes in this patch outside of port/base.py are largely
2434         cosmetic changes of "filename" to "test". We add a few routines
2435         to the base Port implementation to manipulate test names to
2436         split them, figure out if we're referring to a directory of
2437         tests, etc.
2438
2439         * Scripts/webkitpy/common/net/layouttestresults.py:
2440         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
2441         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2442         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2443         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
2444         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
2445         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
2446         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
2447         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
2448         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
2449         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
2450         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
2451         * Scripts/webkitpy/layout_tests/layout_package/test_input.py:
2452         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
2453         * Scripts/webkitpy/layout_tests/layout_package/test_results.py:
2454         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py:
2455         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
2456         * Scripts/webkitpy/layout_tests/port/base.py:
2457         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2458         * Scripts/webkitpy/layout_tests/port/chromium.py:
2459         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2460         * Scripts/webkitpy/layout_tests/port/dryrun.py:
2461         * Scripts/webkitpy/layout_tests/port/test.py:
2462         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2463         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2464         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
2465         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
2466         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
2467         * Scripts/webkitpy/tool/commands/rebaseline.py:
2468         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
2469
2470 2011-07-06  Chris Guillory  <ctguil@chromium.org>
2471
2472         No review necessary.
2473
2474         Adding myself to committers list in committers.py.
2475         https://bugs.webkit.org/show_bug.cgi?id=63478
2476
2477         * Scripts/webkitpy/common/config/committers.py:
2478
2479 2011-07-06  Johnny Ding  <jnd@chromium.org>
2480
2481         https://bugs.webkit.org/show_bug.cgi?id=63857
2482         Implement createPopupMenu for Chromium DRT.
2483
2484         Reviewed by Tony Chang.
2485
2486         * DumpRenderTree/chromium/WebViewHost.cpp:
2487         (WebViewHost::createPopupMenu):
2488         (WebViewHost::~WebViewHost):
2489         * DumpRenderTree/chromium/WebViewHost.h:
2490
2491 2011-07-06  Mihai Parparita  <mihaip@chromium.org>
2492
2493         Unreviewed, rolling out r90503.
2494         http://trac.webkit.org/changeset/90503
2495         https://bugs.webkit.org/show_bug.cgi?id=63358
2496
2497         Makes Chromium Windows canary bots hang.
2498
2499         * Scripts/webkitpy/layout_tests/port/base.py:
2500         * Scripts/webkitpy/layout_tests/port/chromium.py:
2501         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2502         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2503         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2504         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
2505         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2506         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2507
2508 2011-07-06  Adam Barth  <abarth@webkit.org>
2509
2510         Move buildbot URL in webkitpy to config.urls
2511         https://bugs.webkit.org/show_bug.cgi?id=64041
2512
2513         Reviewed by Eric Seidel.
2514
2515         This patch prepares for using the buildbot class to talk to the
2516         Chromium buildbots.
2517
2518         * Scripts/webkitpy/common/config/urls.py:
2519         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
2520         * Scripts/webkitpy/tool/commands/download_unittest.py:
2521         * Scripts/webkitpy/tool/commands/queries.py:
2522         * Scripts/webkitpy/tool/mocktool.py:
2523         * Scripts/webkitpy/tool/steps/ensurebuildersaregreen.py:
2524
2525 2011-07-06  Kevin Ollivier  <kevino@theolliviers.com>
2526
2527         [wx] Unreviewed build fix, add the correct path to the waf build settings.
2528         
2529         * wx/packaging/build-win-installer.py:
2530
2531 2011-07-06  Adam Barth  <abarth@webkit.org>
2532
2533         Add missing include.
2534
2535         * Scripts/webkitpy/tool/bot/irc_command.py:
2536
2537 2011-07-06  Adam Barth  <abarth@webkit.org>
2538
2539         Add roll-chromium-deps command to sheriff-bot
2540         https://bugs.webkit.org/show_bug.cgi?id=64037
2541
2542         Reviewed by Eric Seidel.
2543
2544         Lo, and the DEPS were rolled on command.
2545
2546         * Scripts/webkitpy/tool/bot/irc_command.py:
2547         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
2548         * Scripts/webkitpy/tool/bot/sheriff.py:
2549         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
2550         * Scripts/webkitpy/tool/commands/roll.py:
2551         * Scripts/webkitpy/tool/commands/roll_unittest.py:
2552         * Scripts/webkitpy/tool/steps/updatechromiumdeps.py:
2553
2554 2011-07-06  Eric Seidel  <eric@webkit.org>
2555
2556         Split Driver, DriverOutput and DriverInput out into a new driver.py file
2557         https://bugs.webkit.org/show_bug.cgi?id=64031
2558
2559         Reviewed by Adam Barth.
2560
2561         This ended up kinda large and I'm happy to break it up if desired.
2562
2563         As part of cleaning up the Port module, I split Driver, DriverInput and DriverOutput
2564         out into its own driver.py file.
2565
2566         I also cleaned up how we import Port objects, by making port/__init__.py
2567         expose Port, Driver, DriverInput, and DriverOutput.  I removed exposure of
2568         some various test objects as they were causing circular dependencies.
2569         base.py is now an implementation detail of port.  No longer is "base" referenced
2570         anywhere in the code.
2571
2572         I also made Driver.__init__ non-virtual so that sub-classers could all share
2573         the data storage on Driver.  This more closely matches how other subclasses
2574         work throughout our python code.
2575
2576         Standard line-unwrapping and doc-string removal as I went through the code.
2577
2578         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2579         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
2580         * Scripts/webkitpy/layout_tests/port/__init__.py:
2581         * Scripts/webkitpy/layout_tests/port/base.py:
2582         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2583         * Scripts/webkitpy/layout_tests/port/chromium.py:
2584         * Scripts/webkitpy/layout_tests/port/driver.py: Added.
2585         * Scripts/webkitpy/layout_tests/port/dryrun.py:
2586         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
2587         * Scripts/webkitpy/layout_tests/port/test.py:
2588         * Scripts/webkitpy/layout_tests/port/webkit.py:
2589         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2590         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
2591         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
2592
2593 2011-07-06  Adam Roben  <aroben@apple.com>
2594
2595         Clear TestFailures's getNumberOfFailingTests cache
2596
2597         Now that we support NRWT, we need to clear out old zero values for the NRWT bots that might
2598         have been cached by old versions of TestFailures.
2599
2600         Fixes <http://webkit.org/b/64029> TestFailures page still doesn't show NRWT bots if you've
2601         used it recently
2602
2603         Reviewed by Adam Barth.
2604
2605         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
2606         (Builder.prototype.getNumberOfFailingTests): Added a .version property to the cache data
2607         that must match the current version. Otherwise we'll ignore the cached data.
2608
2609 2011-06-24  Dirk Pranke  <dpranke@chromium.org>
2610
2611         Reviewed by Adam Barth.
2612
2613         nrwt: remove --use-apache from the command line
2614         https://bugs.webkit.org/show_bug.cgi?id=63358
2615
2616         This change removes the --use-apache command line argument. It
2617         was initially put in when we were trying to get the cygwin
2618         apache instance to work with Chromium win, but that code has
2619         bitrotted and doesn't work at all now.
2620
2621         Arguably we should remove all of the code to allow for a choice
2622         of web servers, but since we may still want to switch off of
2623         LigHTTPd at some point on Windows, I'll leave the rest in for
2624         now (or at least yank it in a different change.)
2625
2626         * Scripts/webkitpy/layout_tests/port/base.py:
2627         * Scripts/webkitpy/layout_tests/port/chromium.py:
2628         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2629         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2630         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
2631         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2632
2633 2011-07-06  Dimitri Glazkov  <dglazkov@chromium.org>
2634
2635         garden-o-matic should use a favicon to indicate current state.
2636         https://bugs.webkit.org/show_bug.cgi?id=64027
2637
2638         Reviewed by Adam Barth.
2639
2640         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-green.png: Added.
2641         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-red.png: Added.
2642         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html: Added blurb about icons.
2643         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js: Added setting icon state.
2644         * Scripts/webkitpy/tool/servers/gardeningserver.py: Allowed the use of favicons.
2645
2646 2011-07-06  Daniel Bates  <dbates@rim.com>
2647
2648         Fix webkitpy error after changeset 90483 <http://trac.webkit.org/changeset/90483>; Add missing
2649         parameter to _fetch_file_from_results lamdba function.
2650
2651         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2652
2653 2011-07-06  Kevin Ollivier  <kevino@theolliviers.com>
2654
2655         [wx] Unreviewed build fix. Generate the proper install_name for dylibs on Mac.
2656
2657         * wx/packaging/build-mac-installer.py:
2658
2659 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2660
2661         [GTK] Only set env. variables in start when strictly needed
2662         https://bugs.webkit.org/show_bug.cgi?id=64026
2663
2664         Reviewed by Eric Seidel.
2665
2666         * Scripts/webkitpy/layout_tests/port/gtk.py: add a
2667         setup_environ_for_server method in GtkPort and set as many
2668         env. variables as possible there.
2669
2670 2011-07-06  Adam Roben  <aroben@apple.com>
2671
2672         Teach TestFailures how to load, parse, and interpret NRWT test results
2673
2674         Fixes <http://webkit.org/b/61877> TestFailures page doesn't show testers that use
2675         new-run-webkit-tests
2676
2677         Reviewed by Adam Barth.
2678
2679         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
2680         (Builder.prototype.failureDiagnosisTextAndURL): Added support for the new 'flaky' failure
2681         type. For now we don't account for image-only flakes (but TestFailures doesn't deal with
2682         pixel tests at all currently).
2683         (Builder.prototype.getNumberOfFailingTests): Relaxed the regex used for parsing the number
2684         of failing tests from the buildbot output. Make sure not to count "new passes" (a new
2685         category introduced by NRWT) as failures.
2686
2687         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
2688         (LayoutTestResultsLoader.prototype.start): Moved a bunch of code from here to
2689         _fetchAndParseORWTResults. This function now attempts to load NRWT results, then falls back
2690         to loading ORWT results.
2691         (LayoutTestResultsLoader.prototype._fetchAndParseNRWTResults): Added. Tries to load and
2692         parse the full_results.json file for the given build.
2693         (LayoutTestResultsLoader.prototype._fetchAndParseORWTResults): Added. Code came from start.
2694         Fixed a bug along the way where we were sometimes calling the error callback instead of the
2695         success callback when all tests passed.
2696
2697         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NRWTResultsParser.js: Added.
2698         (NRWTResultsParser): Do-nothing constructor.
2699         (NRWTResultsParser.prototype.parse): Uses eval() (eww!) to get the test results out of the
2700         JS string, then iterates over all the tests in the results data and builds up a data
2701         structure matching what ORWTResultsParser returns.
2702
2703         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
2704         (Array.prototype.contains):
2705         (String.prototype.contains):
2706         Added these simple helper functions.
2707
2708         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
2709         NRWTResultsParser.js.
2710
2711 2011-07-06  Adam Roben  <aroben@apple.com>
2712
2713         Extract code to parse ORWT's results.html file into its own class
2714
2715         Prep work for fixing <http://webkit.org/b/61877> TestFailures page doesn't show testers that
2716         use new-run-webkit-tests
2717
2718         Reviewed by Adam Barth.
2719
2720         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
2721         (LayoutTestResultsLoader.prototype.start): Moved a bunch of code from here...
2722
2723         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ORWTResultsParser.js: Added.
2724         (ORWTResultsParser):
2725         (ORWTResultsParser.prototype.parse):
2726         ...to here.
2727
2728         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
2729         ORWTResultsParser.js.
2730
2731 2011-07-06  Eric Seidel  <eric@webkit.org>
2732
2733         Unreviewed.  Fix a unittest from hitting the network. :)
2734
2735         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2736
2737 2011-07-06  Eric Seidel  <eric@webkit.org>
2738
2739         Unreviewed.  Fixing an exception in test-webkitpy, caused by bug 64006.
2740
2741         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2742
2743 2011-07-06  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
2744
2745         Run an Xvfb instance to run the API tests on, like we do for the
2746         layout tests.
2747
2748         Reviewed by Xan Lopez.
2749
2750         * Scripts/run-gtk-tests:
2751
2752 2011-07-06  Daniel Bates  <dbates@webkit.org>
2753
2754         Make SCM unit tests faster
2755         https://bugs.webkit.org/show_bug.cgi?id=63883
2756
2757         Speeds up the SCM unit tests by a factor of 4. Currently, we create a
2758         mock SVN repo for each test_ method in SVNTest and GitTest and creating
2759         this repo is expensive.
2760
2761         Instead, it is significantly faster to create the mock SVN repo once
2762         and then perform a filesystem copy of it for each test_ method.
2763
2764         Note, Python 2.7's unittest module implements support for per class and
2765         per module setup and tear down methods which could be used to implement
2766         similar functionality. At the time of writing, test-webkitpy is designed
2767         to support Python 2.5. So, we can't take advantage of these Python 2.7
2768         features :(
2769
2770         Reviewed by Eric Seidel.
2771
2772         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
2773
2774 2011-07-06  Adam Barth  <abarth@webkit.org>
2775
2776         Remove unused parameter from _add_test in test_expectations.py
2777         https://bugs.webkit.org/show_bug.cgi?id=64008
2778
2779         Reviewed by Eric Seidel.
2780
2781         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
2782
2783 2011-07-06  Adam Barth  <abarth@webkit.org>
2784
2785         Use old-run-webkit-tests when --leaks is present
2786         https://bugs.webkit.org/show_bug.cgi?id=64012
2787
2788         Reviewed by Adam Roben.
2789
2790         We'll remove this branch once support of --leaks is added to
2791         new-run-webkit-tests.
2792
2793         * Scripts/run-webkit-tests:
2794
2795 2011-07-06  Benjamin Poulain  <benjamin@webkit.org>
2796
2797         [Qt] [WK2] Add touch mocking to Qt's MiniBrowser
2798         https://bugs.webkit.org/show_bug.cgi?id=63995
2799
2800         Reviewed by Andreas Kling.
2801
2802         The mouse events are intercepted through QCoreApplication::notify(),
2803         and fake touch events are generated for mouse events when necessary.
2804
2805         If touch events are received from the platform, we assume the current
2806         hardware has a touch screen and we stop generating fake events.
2807
2808         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2809         (isTouchEvent):
2810         (isMouseEvent):
2811         (MiniBrowserApplication::MiniBrowserApplication):
2812         (MiniBrowserApplication::notify):
2813         * MiniBrowser/qt/MiniBrowserApplication.h:
2814
2815 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2816
2817         Reviewed by Gustavo Noronha.
2818
2819         * Scripts/webkitpy/layout_tests/port/gtk.py: use Popen correctly.
2820
2821 2011-07-06  Eric Seidel  <eric@webkit.org>
2822
2823         webkit-patch failure-reason does not understand NRWT results
2824         https://bugs.webkit.org/show_bug.cgi?id=64006
2825
2826         Reviewed by Adam Barth.
2827
2828         Adam Barth tells me failure-reason should be deleted soon,
2829         but aroben's fancy new TestFailures/ page doesn't work for
2830         NRWT yet, and I needed to know when
2831         third-party-cookie-relaxing started failing.
2832
2833         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
2834         * Scripts/webkitpy/tool/commands/queries.py:
2835
2836 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2837
2838         [GTK] Add missing environment variables in NWRT
2839         https://bugs.webkit.org/show_bug.cgi?id=64004
2840
2841         Reviewed by Gustavo Noronha Silva.
2842
2843         Add missing environment variables needed by the GTK+ port.
2844
2845         * Scripts/webkitpy/layout_tests/port/gtk.py: ditto.
2846
2847 2011-07-06  Adam Roben  <aroben@apple.com>
2848
2849         Make run-api-tests output intelligible on the bots
2850
2851         run-api-tests's output on the bots was very hard to read for two reasons:
2852           1) It was mixing its own output with gtest's
2853           2) It was using ANSI escape sequences to print colored text, but the bots don't support
2854              that
2855
2856         Now, the --verbose flag turns off almost all of run-api-tests's own output so that gtest's
2857         output will not be obscured. We still print "Timeout" messages even in verbose mode, since
2858         gtest doesn't have any native support for timeouts. Also, when our output is being
2859         redirected to a file, we don't print ANSI escape sequences.
2860
2861         Fixes <http://webkit.org/b/63996> It's very hard to read run-api-tests output on the bots
2862
2863         Reviewed by Sam Weinig.
2864
2865         * Scripts/run-api-tests: Don't use the :constants interface of Term::ANSIColor. We want to
2866         use the colored function instead.
2867         (runAllTestsInSuite): Don't print out suite names in verbose mode; gtest will give us enough
2868         context that they aren't needed.
2869         (runTest): Don't print out test names or pass/fail messages in verbose mode; gtest will do
2870         that for us. Also, use the new possiblyColored function instead of always coloring output.
2871         (possiblyColored): Added. When printing to a tty, returns the string with the appropriate
2872         ANSI color escape sequences added. Otherwise just returns the string unmodified.
2873
2874 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2875
2876         Reviewed by Gustavo Noronha.
2877
2878         Launch Xvfb (mostly) the same way we were launching it in the
2879         bots, for maximum compatibility.
2880
2881         * Scripts/webkitpy/layout_tests/port/gtk.py:
2882
2883 2011-07-06  Andreas Kling  <kling@webkit.org>
2884
2885         [Qt][WK2] Split Qt API into two different web views (touch and desktop)
2886         https://bugs.webkit.org/show_bug.cgi?id=63950
2887
2888         Co-authored with Benjamin Poulain <benjamin@webkit.org>.
2889
2890         Reviewed by Kenneth Rohde Christiansen.
2891
2892         Port WTR/Qt to use QDesktopWebView and force a 800x600 viewport.
2893
2894         The Qt MiniBrowser now defaults to QDesktopWebView but can be started
2895         in QTouchWebView mode by passing -touch on the command line.
2896
2897         A lot of functionality has been removed and/or disabled from
2898         MiniBrowser until we reintroduce the necessary API's.
2899
2900         * MiniBrowser/qt/BrowserView.cpp:
2901         (BrowserView::BrowserView):
2902         (BrowserView::~BrowserView):
2903         (BrowserView::load):
2904         (BrowserView::view):
2905         (BrowserView::touchWebView):
2906         (BrowserView::desktopWebView):
2907         * MiniBrowser/qt/BrowserView.h:
2908         * MiniBrowser/qt/BrowserWindow.cpp:
2909         (newPageFunction):
2910         (BrowserWindow::BrowserWindow):
2911         (BrowserWindow::webView):
2912         (BrowserWindow::newWindow):
2913         (BrowserWindow::showUserAgentDialog):
2914         (BrowserWindow::updateUserAgentList):
2915         * MiniBrowser/qt/BrowserWindow.h:
2916         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2917         (MiniBrowserApplication::handleUserOptions):
2918         * MiniBrowser/qt/MiniBrowserApplication.h:
2919         (WindowOptions::WindowOptions):
2920         * MiniBrowser/qt/UrlLoader.cpp:
2921         (UrlLoader::UrlLoader):
2922         * MiniBrowser/qt/main.cpp:
2923         (main):
2924         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
2925         (WTR::WebView::wkView):
2926         (WTR::WebView::pageRef):
2927         (WTR::WebView::WebView):
2928         (WTR::PlatformWebView::page):
2929         (WTR::PlatformWebView::windowFrame):
2930         (WTR::PlatformWebView::setWindowFrame):
2931
2932 2011-07-06  Eric Seidel  <eric@webkit.org>
2933
2934         Unreviewed..
2935
2936         * Scripts/run-webkit-tests:
2937         (useNewRunWebKitTests):
2938          - Fix the pass-thru of --gtk and --qt for old-run-webkit-tests,
2939            it was previously broken by the existence of isGtk/isQt checks which
2940            removed the --gtk/--qt args from @ARGV.
2941
2942 2011-07-06  Eric Seidel  <eric@webkit.org>
2943
2944         new-run-webkit-tests uses a 35s timeout for Mac and 6s for all other ports
2945         https://bugs.webkit.org/show_bug.cgi?id=63983
2946
2947         Reviewed by Xan Lopez.
2948
2949         All WebKit ports want to use a 35s timeout.  Only chromium currently
2950         uses a 6s timeout.  Eventually we'll move all WebKit ports to 6s too
2951         to further speed up new-run-webkit-tests runs, but that's going to be a while.
2952
2953         * Scripts/webkitpy/layout_tests/port/mac.py:
2954         * Scripts/webkitpy/layout_tests/port/webkit.py:
2955         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2956
2957 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2958
2959         Unreviewed.
2960
2961         NWRT actually seems to work locally, so it should work in the bots
2962         as long as we take care of the TIMEOUT issue (which is being fixed
2963         as I write this). Let's give this another shot and fix whatever
2964         breaks.
2965
2966         * Scripts/run-webkit-tests:
2967         (useNewRunWebKitTests):
2968
2969 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2970
2971         Unreviewed.
2972
2973         Disable NWRT on GTK. At the very least it needs an updated exected
2974         results file since NWRT is more sensitive to timeouts, not sure
2975         what the exact problem is.
2976
2977         * Scripts/run-webkit-tests:
2978         (useNewRunWebKitTests):
2979
2980 2011-07-06  Adam Barth  <abarth@webkit.org>
2981
2982         Fix global variable leak detected by noglobals.
2983
2984         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
2985
2986 2011-07-06  Adam Barth  <abarth@webkit.org>
2987
2988         garden-o-matic shouldn't show flaky tests by default
2989         https://bugs.webkit.org/show_bug.cgi?id=63979
2990
2991         Reviewed by Ojan Vafai.
2992
2993         If any of the actual test results are PASS, the test is flaky and we
2994         can ignore it (at least for now).  If we get awesome at dealing with
2995         failing tests, we might want to tackle flaky tests by refering the user
2996         to the flakiness dashboard, which is way more awesome at dealing with
2997         flaky tests that this tool ever will be.
2998
2999         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
3000         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
3001
3002 2011-07-05  Adam Barth  <abarth@webkit.org>
3003
3004         garden-o-matic should list the failing tests
3005         https://bugs.webkit.org/show_bug.cgi?id=63978
3006
3007         Reviewed by Eric Seidel.
3008
3009         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js: Added.
3010         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
3011         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
3012         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
3013         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
3014         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
3015         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js: Added.
3016         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js: Added.
3017         * Scripts/webkitpy/tool/servers/gardeningserver.py:
3018
3019
3020 2011-07-05  Eric Seidel  <eric@webkit.org>
3021
3022         Re-enable NRWT on Leopard now that bug 63973 is fixed.
3023
3024         Reviewed by Adam Barth.
3025
3026         I also made the should-enable expression a black-list instead
3027         of a white-list since every bot on build.webkit.org is
3028         now using NRWT except windows and WebKit2 bots.
3029
3030         * Scripts/run-webkit-tests:
3031         (useNewRunWebKitTests):
3032
3033 2011-07-05  Eric Seidel  <eric@webkit.org>
3034
3035         Perl string concat failure on Leopard (causes NRWT to fail)
3036         https://bugs.webkit.org/show_bug.cgi?id=63973
3037
3038         Reviewed by Daniel Bates.
3039
3040         Leopard's "xcodebuild -version" has a different format from Snow Leopard/Lion
3041         thus the version parsing logic in webkitdirs.pm seems to fail.
3042
3043         Default to "3.0" instead of undef when we can't parse the version number.
3044         This should send us down the correct path for Leopard.
3045
3046         * Scripts/webkitdirs.pm:
3047         (determineBaseProductDir):
3048
3049 2011-07-05  Kevin Ollivier  <kevino@theolliviers.com>
3050
3051         [wx] Unreviewed build fix after addition of Source/WebCore/xml/parser dir.
3052
3053         * waf/build/settings.py:
3054
3055 2011-07-05  Eric Seidel  <eric@webkit.org>
3056
3057         Disable NRWT for Leopard until https://bugs.webkit.org/show_bug.cgi?id=63973 can be fixed.
3058
3059         Reviewed by Adam Barth.
3060
3061         * Scripts/run-webkit-tests:
3062         (useNewRunWebKitTests):
3063
3064 2011-07-05  Eric Seidel  <eric@webkit.org>
3065
3066         Make run-webkit-tests pass --verbose to new-run-webkit-tests on qt/gtk bots.
3067
3068         Reviewed by Adam Barth.
3069
3070         * Scripts/run-webkit-tests:
3071         (runningOnBuildBot):
3072
3073 2011-07-05  Adam Barth  <abarth@webkit.org>
3074
3075         Add results_unittests.js to garden-o-matic
3076         https://bugs.webkit.org/show_bug.cgi?id=63971
3077
3078         Reviewed by Eric Seidel.
3079
3080         Adding these unit tests forced me to commit to an API for the results
3081         module.  I'm not sure if this is the API we'll end up with, but it's
3082         something not entirely unreasonable.
3083
3084         Also, I've changed results.js over to WebKit style, as discussed previously.
3085
3086         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
3087         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
3088         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js: Added.
3089         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
3090
3091 2011-07-05  Eric Seidel  <eric@webkit.org>
3092
3093         Teach run-webkit-tests how to translate --qt and --gtk for new-run-webkit-tests.
3094
3095         Reviewed by Adam Barth.
3096
3097         * Scripts/run-webkit-tests:
3098
3099 2011-07-05  Adam Barth  <abarth@webkit.org>
3100
3101         Add first unit tests for garden-o-matic
3102         https://bugs.webkit.org/show_bug.cgi?id=63969
3103
3104         Reviewed by Eric Seidel.
3105
3106         I've also reformated base.js to match WebKit style, as discussed in the
3107         previous bug.
3108
3109         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
3110         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js: Added.
3111         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html: Added.
3112
3113 2011-07-05  Eric Seidel  <eric@webkit.org>
3114
3115         Move Qt, Gtk and Leopard to NRWT.
3116
3117         We'll roll out any of these if we see problems.
3118
3119         Reviewed by Adam Barth.
3120
3121         * Scripts/run-webkit-tests:
3122         (useNewRunWebKitTests):
3123
3124 2011-07-05  Adam Barth  <abarth@webkit.org>
3125
3126         garden-o-matic should understand which tests have unexpected failures
3127         https://bugs.webkit.org/show_bug.cgi?id=63965
3128
3129         Reviewed by Eric Seidel.
3130
3131         Currently, this code just logs the list of unexpected failures to the
3132         console.  A future patch will do something useful with this
3133         information.  A testing harness will also come in the next patch.
3134
3135         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js: Added.
3136         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
3137         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
3138         * Scripts/webkitpy/tool/servers/gardeningserver.py:
3139
3140 2011-07-05  Adam Barth  <abarth@webkit.org>
3141
3142         garden-o-matic should know how to fetch test results from the (NRWT) bots
3143         https://bugs.webkit.org/show_bug.cgi?id=63959
3144
3145         Reviewed by Eric Seidel.
3146
3147         Currently, this code just fetches the results and expectations JSON
3148         blob for Linux from the server and logs it to the console.
3149
3150         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
3151         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js: Added.
3152         * Scripts/webkitpy/tool/servers/gardeningserver.py:
3153
3154 2011-07-05  Dirk Pranke  <dpranke@chromium.org>
3155
3156         Reviewed by Eric Seidel.
3157
3158         Python tests are failing on leopard
3159         https://bugs.webkit.org/show_bug.cgi?id=63842
3160
3161         Fix the missing flush() call that appears to be needed by the
3162         logging package in python 2.5. Also fix the regressions
3163         introduced in the run_webkit_test tests when we switched from
3164         thread to processes (since processes aren't available on 2.5).
3165
3166         Also fix a missing "from __future__ import with_statement' in
3167         testfilehandler (unreviewed).
3168
3169         * Scripts/webkitpy/layout_tests/layout_package/metered_stream.py:
3170         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3171         * TestResultServer/handlers/testfilehandler.py:
3172
3173 2011-07-05  Eric Seidel  <eric@webkit.org>
3174
3175         Reviewed by Adam Barth.
3176
3177         re-enabling new-run-webkit-tests on the Snow Leopard WebKit1 bots
3178         now that we believe the stray http-server issue to be solved.
3179
3180         * Scripts/run-webkit-tests:
3181         (useNewRunWebKitTests):
3182
3183 2011-07-05  Adam Barth  <abarth@webkit.org>
3184
3185         Turns out we need to support numerals too because that's what jQuery
3186         uses by default.
3187
3188         * TestResultServer/handlers/testfilehandler.py:
3189
3190 2011-07-05  Dirk Pranke  <dpranke@chromium.org>
3191
3192         nrwt: allow for multiple http shards
3193         https://bugs.webkit.org/show_bug.cgi?id=63116
3194
3195         Reviewed by Tony Chang.
3196
3197         This modifies the sharding logic to support multiple http
3198         shards, but for now we clamp to one shard until we can test
3199         perf impact and flakiness impact.
3200
3201         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3202         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
3203
3204 2011-07-05  Dirk Pranke  <dpranke@chromium.org>
3205
3206         Re-land nrwt: make sharding tests needing locks less hard-coded
3207         https://bugs.webkit.org/show_bug.cgi?id=63112
3208
3209         Reviewed by Ojan Vafai.
3210
3211         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3212         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
3213
3214 2011-07-05  Adam Barth  <abarth@webkit.org>
3215
3216         TestResultServer should support callback parameter for JSON
3217         https://bugs.webkit.org/show_bug.cgi?id=63961
3218
3219         Reviewed by Ojan Vafai.
3220
3221         This makes using jQuery.ajax so much more pleasant.
3222
3223         * TestResultServer/handlers/testfilehandler.py:
3224
3225 2011-07-05  Eric Seidel  <eric@webkit.org>
3226
3227         buildbot needs to understand whether NRWT exited early after having too many failures
3228         https://bugs.webkit.org/show_bug.cgi?id=63839
3229
3230         Reviewed by Adam Barth.
3231
3232         Fix-up the exited early messages printed by NRWT so that
3233         the buildbot can parse them as expected.
3234         It looks for lines using "if line.find('Exiting early') >= 0:"
3235
3236         I also plumbed the "early exit" status through to results.json
3237         in the form of an "interrupted" bool.  It was unclear to me
3238         if results.json already had enough information to compute this bool
3239         itself.  It's possible Ojan could come up with a better fix.
3240
3241         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3242         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
3243         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
3244         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
3245
3246 2011-07-05  Eric Seidel  <eric@webkit.org>
3247
3248         new-run-webkit-tests fails to start http server if one is already running
3249         https://bugs.webkit.org/show_bug.cgi?id=63956
3250
3251         Reviewed by Adam Barth.
3252
3253         * BuildSlaveSupport/kill-old-processes:
3254          - Removed ^M line endings and sorted the process names.
3255          - There are no functional changes to this file.
3256            (except that now that its using unix line endings it can be executed directly!)
3257         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
3258          - Change the name to 'httpd' to match old-run-webkit-test paths.
3259
3260 2011-07-05  Adam Barth  <abarth@webkit.org>
3261
3262         Add basic ajax support to garden-o-matic
3263         https://bugs.webkit.org/show_bug.cgi?id=63874
3264
3265         Reviewed by Eric Seidel.
3266
3267         This patch adds some basic infrastructure to garden-o-matic.
3268         Currently, the infrastructure is just used to make the "quit" command
3269         use Ajax, but in the future, this infrastructure will be used to do
3270         more sophistocated remote proceedure calls.
3271
3272         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
3273         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js: Added.
3274         * Scripts/webkitpy/tool/servers/gardeningserver.py:
3275         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
3276
3277 2011-07-05  Adam Barth  <abarth@webkit.org>
3278
3279         Add trivial garden-o-matic command (with server)
3280         https://bugs.webkit.org/show_bug.cgi?id=63872
3281
3282         Reviewed by Eric Seidel.
3283
3284         This patch adds a trivial "Hello, world" garden-o-matic command,
3285         complete with an HTTP server.  This command re-uses a bunch of code
3286         from the existing rebaseline-server command.  Over time, this command
3287         will grow to be a tool useful for gardening the WebKit tree.
3288
3289         * Scripts/webkitpy/tool/commands/__init__.py:
3290         * Scripts/webkitpy/tool/commands/abstractservercommand.py: Added.
3291         * Scripts/webkitpy/tool/commands/gardenomatic.py: Added.
3292         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
3293         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html: Added.
3294         * Scripts/webkitpy/tool/servers/gardeningserver.py: Added.
3295         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
3296
3297 2011-07-05  Adam Roben  <aroben@apple.com>
3298
3299         Make prepare-ChangeLog include modified Perl functions in its ChangeLog template
3300
3301         This is a very simple first cut. Functions must start with a line that starts with "sub "
3302         and end with a line that starts with a closing brace. No leading whitespace is allowed.
3303         Package names aren't parsed at all.
3304
3305         Fixes <http://webkit.org/b/21591> prepare-ChangeLog should know how to find functions in
3306         Perl files
3307
3308         Reviewed by David Kilzer.
3309
3310         * Scripts/prepare-ChangeLog: Removed redundant code that ignored certain files when
3311         searching for function line ranges. This is already done inside the get_function_line_ranges
3312         function.
3313         (get_function_line_ranges): Cleaned up coding style a little bit. Call
3314         get_function_line_ranges_for_perl for files with .pl and .pm extensions. For files with an
3315         unknown extension or no extension, read the shebang line to try to determine the script
3316         interpreter. Call get_function_line_ranges_for_perl if the interpreter seems to be Perl.
3317         (get_function_line_ranges_for_perl): Added. Does extremely basic parsing of the file to find
3318         lines starting with "sub " or "}".
3319
3320 2011-07-05  Adam Roben  <aroben@apple.com>
3321
3322         Clean up run-api-tests output on Windows
3323
3324         We were mixing run-api-tests output with gtest output, and the result was a mess.
3325
3326         Fixes <http://webkit.org/b/63954> run-api-tests output is very confusing on Windows
3327
3328         Reviewed by Dan Bates.
3329
3330         * Scripts/run-api-tests:
3331         (runTest):
3332         (populateTests):
3333         Made Windows use the formerly-Mac-only codepaths which correctly suppress gtest output
3334         except in verbose mode. Most changes are due to indentation. Use of the arch utility is now
3335         the only Mac-specific thing in these functions. Some other platform-specific code was moved
3336         from here...
3337
3338         (prepareEnvironmentForRunningTestTool):
3339         (testToolPath):
3340         ...to here.
3341
3342 2011-07-04  Gavin Barraclough  <barraclough@apple.com>
3343
3344         Reviewed by Sam Weinig.
3345
3346         https://bugs.webkit.org/show_bug.cgi?id=16652
3347         Firefox and JavaScriptCore differ in Number.toString(integer)
3348
3349         Added forwarding header.
3350
3351         * DumpRenderTree/ForwardingHeaders/wtf/StdLibExtras.h: Added.
3352
3353 2011-07-04  Anders Carlsson  <andersca@apple.com>
3354
3355         Try to fix Qt build.
3356
3357         * DumpRenderTree/TestNetscapePlugIn/Tests/NPRuntimeRemoveProperty.cpp:
3358
3359 2011-07-04  Anders Carlsson  <andersca@apple.com>
3360
3361         NP_RemoveProperty is not called back by Safari when delete npObject.prop is encountered in JavaScript
3362         https://bugs.webkit.org/show_bug.cgi?id=63915
3363         <rdar://problem/7124300>
3364
3365         Reviewed by Sam Weinig.
3366
3367         Extend the NPRuntimeRemoveProperty to handle delete object.property from JavaScript.
3368
3369         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
3370         (PluginTest::NPN_IdentifierIsString):
3371         (PluginTest::NPN_UTF8FromIdentifier):
3372         (PluginTest::NPN_IntFromIdentifier):
3373         (PluginTest::NPN_RetainObject):
3374         (PluginTest::NPN_ReleaseObject):
3375         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
3376         (PluginTest::Object::removeProperty):
3377         (PluginTest::Object::identifierIs):
3378         (PluginTest::Object::NP_RemoveProperty):
3379         (PluginTest::Object::npClass):
3380         * DumpRenderTree/TestNetscapePlugIn/Tests/NPRuntimeRemoveProperty.cpp:
3381         (NPRuntimeRemoveProperty::TestObject::TestObject):
3382         (NPRuntimeRemoveProperty::TestObject::hasProperty):
3383         (NPRuntimeRemoveProperty::TestObject::getProperty):
3384         (NPRuntimeRemoveProperty::TestObject::removeProperty):
3385         (NPRuntimeRemoveProperty::PluginObject::PluginObject):
3386         (NPRuntimeRemoveProperty::PluginObject::~PluginObject):
3387         (NPRuntimeRemoveProperty::PluginObject::hasMethod):
3388         (NPRuntimeRemoveProperty::PluginObject::invoke):
3389         (NPRuntimeRemoveProperty::PluginObject::hasProperty):
3390         (NPRuntimeRemoveProperty::PluginObject::getProperty):
3391         (NPRuntimeRemoveProperty::NPP_GetValue):
3392
3393 2011-07-04  Leandro Pereira  <leandro@profusion.mobi>
3394
3395         Reviewed by Kent Tamura.
3396
3397         [EFL] Add LayoutTestControllerEfl.
3398         https://bugs.webkit.org/show_bug.cgi?id=61974
3399
3400         Only the core features are implemented. Most functions are actually
3401         stubs and will be implemented as soon as its features are needed.
3402         Build system changes will be made as soon as DRT is completely
3403         upstreamed.
3404
3405         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp: Added.
3406         (LayoutTestController::~LayoutTestController):
3407         (LayoutTestController::addDisallowedURL):
3408         (LayoutTestController::clearBackForwardList):
3409         (LayoutTestController::copyDecodedHostName):
3410         (LayoutTestController::copyEncodedHostName):
3411         (LayoutTestController::dispatchPendingLoadRequests):
3412         (LayoutTestController::display):
3413         (LayoutTestController::counterValueForElementById):
3414         (LayoutTestController::keepWebHistory):
3415         (LayoutTestController::computedStyleIncludingVisitedInfo):
3416         (LayoutTestController::nodesFromRect):
3417         (LayoutTestController::layerTreeAsText):
3418         (LayoutTestController::pageNumberForElementById):
3419         (LayoutTestController::numberOfPages):
3420         (LayoutTestController::pageProperty):
3421         (LayoutTestController::isPageBoxVisible):
3422         (LayoutTestController::pageSizeAndMarginsInPixels):
3423         (LayoutTestController::webHistoryItemCount):
3424         (LayoutTestController::workerThreadCount):
3425         (LayoutTestController::notifyDone):
3426         (LayoutTestController::pathToLocalResource):
3427         (LayoutTestController::queueLoad):
3428         (LayoutTestController::setAcceptsEditing):
3429         (LayoutTestController::setAlwaysAcceptCookies):
3430         (LayoutTestController::setCustomPolicyDelegate):
3431         (LayoutTestController::waitForPolicyDelegate):
3432         (LayoutTestController::setScrollbarPolicy):
3433         (LayoutTestController::addOriginAccessWhitelistEntry):
3434         (LayoutTestController::removeOriginAccessWhitelistEntry):
3435         (LayoutTestController::setMainFrameIsFirstResponder):
3436         (LayoutTestController::setTabKeyCyclesThroughElements):
3437         (LayoutTestController::setUseDashboardCompatibilityMode):
3438         (LayoutTestController::setUserStyleSheetEnabled):
3439         (LayoutTestController::setUserStyleSheetLocation):
3440         (LayoutTestController::setValueForUser):
3441         (LayoutTestController::setViewModeMediaFeature):
3442         (LayoutTestController::setWindowIsKey):
3443         (LayoutTestController::setSmartInsertDeleteEnabled):
3444         (waitToDumpWatchdogFired):
3445         (LayoutTestController::setWaitToDump):
3446         (LayoutTestController::windowCount):
3447         (LayoutTestController::setPrivateBrowsingEnabled):
3448         (LayoutTestController::setJavaScriptCanAccessClipboard):
3449         (LayoutTestController::setXSSAuditorEnabled):
3450         (LayoutTestController::setFrameFlatteningEnabled):
3451         (LayoutTestController::setSpatialNavigationEnabled):
3452         (LayoutTestController::setAllowUniversalAccessFromFileURLs):
3453         (LayoutTestController::setAllowFileAccessFromFileURLs):
3454         (LayoutTestController::setAuthorAndUserStylesEnabled):
3455         (LayoutTestController::setAutofilled):
3456         (LayoutTestController::disableImageLoading):
3457         (LayoutTestController::setMockDeviceOrientation):
3458         (LayoutTestController::setMockGeolocationPosition):
3459         (LayoutTestController::setMockGeolocationError):
3460         (LayoutTestController::setGeolocationPermission):
3461         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
3462         (LayoutTestController::addMockSpeechInputResult):
3463         (LayoutTestController::setIconDatabaseEnabled):
3464         (LayoutTestController::setJavaScriptProfilingEnabled):
3465         (LayoutTestController::setSelectTrailingWhitespaceEnabled):
3466         (LayoutTestController::setPopupBlockingEnabled):
3467         (LayoutTestController::setPluginsEnabled):
3468         (LayoutTestController::elementDoesAutoCompleteForElementWithId):
3469         (LayoutTestController::execCommand):
3470         (LayoutTestController::findString):
3471         (LayoutTestController::isCommandEnabled):
3472         (LayoutTestController::setCacheModel):
3473         (LayoutTestController::setPersistentUserStyleSheetLocation):
3474         (LayoutTestController::clearPersistentUserStyleSheet):
3475         (LayoutTestController::clearAllApplicationCaches):
3476         (LayoutTestController::setApplicationCacheOriginQuota):
3477         (LayoutTestController::clearApplicationCacheForOrigin):
3478         (LayoutTestController::localStorageDiskUsageForOrigin):
3479         (LayoutTestController::originsWithApplicationCache):
3480         (LayoutTestController::applicationCacheDiskUsageForOrigin):
3481         (LayoutTestController::clearAllDatabases):
3482         (LayoutTestController::setDatabaseQuota):
3483         (LayoutTestController::originsWithLocalStorage):
3484         (LayoutTestController::deleteAllLocalStorage):
3485         (LayoutTestController::deleteLocalStorageForOrigin):
3486         (LayoutTestController::observeStorageTrackerNotifications):
3487         (LayoutTestController::syncLocalStorage):
3488         (LayoutTestController::setDomainRelaxationForbiddenForURLScheme):
3489         (LayoutTestController::goBack):
3490         (LayoutTestController::setDefersLoading):
3491         (LayoutTestController::setAppCacheMaximumSize):
3492         (LayoutTestController::pauseAnimationAtTimeOnElementWithId):
3493         (LayoutTestController::pauseTransitionAtTimeOnElementWithId):
3494         (LayoutTestController::sampleSVGAnimationForElementAtTime):
3495         (LayoutTestController::numberOfActiveAnimations):
3496         (LayoutTestController::suspendAnimations):
3497         (LayoutTestController::resumeAnimations):
3498         (LayoutTestController::overridePreference):
3499         (LayoutTestController::addUserScript):
3500         (LayoutTestController::addUserStyleSheet):
3501         (LayoutTestController::setDeveloperExtrasEnabled):
3502         (LayoutTestController::setAsynchronousSpellCheckingEnabled):
3503         (LayoutTestController::showWebInspector):
3504         (LayoutTestController::closeWebInspector):
3505         (LayoutTestController::evaluateInWebInspector):
3506         (LayoutTestController::evaluateScriptInIsolatedWorld):
3507         (LayoutTestController::removeAllVisitedLinks):
3508         (LayoutTestController::callShouldCloseOnWebView):
3509         (LayoutTestController::apiTestNewWindowDataLoadBaseURL):
3510         (LayoutTestController::apiTestGoToCurrentBackForwardItem):
3511         (LayoutTestController::setWebViewEditable):
3512         (LayoutTestController::markerTextForListItem):
3513         (LayoutTestController::authenticateSession):
3514         (LayoutTestController::setEditingBehavior):
3515         (LayoutTestController::abortModal):
3516         (LayoutTestController::hasSpellingMarker):
3517         (LayoutTestController::hasGrammarMarker):
3518         (LayoutTestController::dumpConfigurationForViewport):
3519         (LayoutTestController::setSerializeHTTPLoads):
3520         (LayoutTestController::setMinimumTimerInterval):
3521         (LayoutTestController::setTextDirection):
3522         (LayoutTestController::allowRoundingHacks):
3523
3524 2011-06-24  Vsevolod Vlasov  <vsevik@chromium.org>
3525
3526         Unreviewed, add myself as committer.
3527
3528         * Scripts/webkitpy/common/config/committers.py:
3529
3530 2011-07-03  Sheriff Bot  <webkit.review.bot@gmail.com>
3531
3532         Unreviewed, rolling out r90347.
3533         http://trac.webkit.org/changeset/90347
3534         https://bugs.webkit.org/show_bug.cgi?id=63886
3535
3536         Build breaks on Leopard, Chromium-win, WinCairo, and WinCE.
3537         (Requested by tkent on #webkit).
3538
3539         * DumpRenderTree/ForwardingHeaders/wtf/StdLibExtras.h: Removed.
3540
3541 2011-07-03  Daniel Bates  <dbates@webkit.org>
3542
3543         Import Checkout in scm_unittests.py and logging in svn.py to fix
3544         Python missing global variable errors
3545
3546         Rubber-stamped by Adam Barth.
3547
3548         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py: Import Checkout.
3549         * Scripts/webkitpy/common/checkout/scm/svn.py: Import logging.
3550
3551 2011-07-02  Gavin Barraclough  <barraclough@apple.com>
3552
3553         Reviewed by Sam Weinig.
3554
3555         https://bugs.webkit.org/show_bug.cgi?id=16652
3556         Firefox and JavaScriptCore differ in Number.toString(integer)
3557
3558         Added forwarding header.
3559
3560         * DumpRenderTree/ForwardingHeaders/wtf/StdLibExtras.h: Added.
3561
3562 2011-04-02  Robert Hogan  <robert@webkit.org>
3563
3564         Reviewed by Benjamin Poulain.
3565
3566         [Qt] Support third-party cookie policy for Qt clients