Standardize WinCairo conditionalized code under PLATFORM macro.
[WebKit-https.git] / Tools / ChangeLog
1 2011-07-12  Brent Fulgham  <bfulgham@webkit.org>
2
3         Standardize WinCairo conditionalized code under PLATFORM macro.
4         https://bugs.webkit.org/show_bug.cgi?id=64377
5
6         Reviewed by Maciej Stachowiak.
7
8         Update compile-time conditions that had previously used #ifdef
9         of WIN_CAIRO to consistently use PLATFORM(WIN_CAIRO).
10
11         * DumpRenderTree/config.h: Switch to PLATFORM(WIN_CAIRO)
12         * DumpRenderTree/win/DumpRenderTreeCairo.vsprops: Remove redundant
13           define of WIN_CAIRO=1.
14         * DumpRenderTree/win/DumpRenderTreeDebugCairoCFLite.vsprops: Add
15           WinCairo.vsprops to set (to match Release build).
16
17 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
18
19         Eliminate TestExpectationsFile.
20         https://bugs.webkit.org/show_bug.cgi?id=64458
21
22         Turns out, we can just use a Python list.
23
24         Reviewed by Adam Barth.
25
26         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Folded TestExpectationsFile.append into TestExpectationParser.parse_list,
27             removed TestExpectationsFile.
28         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Moved tests to better reflect new names, removed iterator test,
29             since there's no more custom iterator machinery.
30
31 2011-07-13  Philippe Normand  <pnormand@igalia.com>
32
33         Unreviewed, added my other email addresses.
34
35         * Scripts/webkitpy/common/config/committers.py:
36
37 2011-07-13  Adam Barth  <abarth@webkit.org>
38
39         Folks should only be listed once.
40
41         * Scripts/webkitpy/common/config/committers.py:
42
43 2011-07-13  Philippe Normand  <pnormand@igalia.com>
44
45         Unreviewed, adding myself as Reviewer.
46
47         * Scripts/webkitpy/common/config/committers.py:
48
49 2011-07-13  Adam Barth  <abarth@webkit.org>
50
51         Fix flickering bug introduced by my previous patch.  I forgot to change
52         the name of the class everywhere.
53
54         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
55
56 2011-07-12  Philippe Normand  <pnormand@igalia.com>
57
58         [GTK] media/media-blocked-by-willsendrequest.html fails
59         https://bugs.webkit.org/show_bug.cgi?id=63699
60
61         Reviewed by Martin Robinson.
62
63         * DumpRenderTree/gtk/DumpRenderTree.cpp:
64         (willSendRequestCallback): Abort the request if explicitely asked
65         by the LayoutTestController's willSendRequestReturnsNull() function.
66
67 2011-07-13  Adam Barth  <abarth@webkit.org>
68
69         Remove "Dismiss" button from garden-o-matic butterbar
70         https://bugs.webkit.org/show_bug.cgi?id=64443
71
72         Reviewed by Dimitri Glazkov.
73
74         We don't have any persistent butter bar messages yet, so the dismiss
75         button is premature (and fairly heavy, visually).
76
77         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
78         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
79
80 2011-07-13  Adam Barth  <abarth@webkit.org>
81
82         Refine garden-o-matic status pane
83         https://bugs.webkit.org/show_bug.cgi?id=64442
84
85         Reviewed by Dimitri Glazkov.
86
87         This patch makes two improvements to the status pane:
88
89         1) We only query the server for failure types that we're expecting.
90            This dramatically reduces the number of HTTP requests, making
91            loading the status pane faster.
92
93         2) The status pane now displays which test and which builder it is
94            showing results for.
95
96         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js:
97         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
98         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
99         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
100         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
101         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
102         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
103
104 2011-07-13  Adam Barth  <abarth@webkit.org>
105
106         garden-o-matic should show test results
107         https://bugs.webkit.org/show_bug.cgi?id=64440
108
109         Reviewed by Adam Roben.
110
111         This patch introduces the results detail pane, which appears at the
112         bottom of the window and contains test results from the bots.
113         Currently, you can activate the pane by mousing over one of the builder
114         names associated with a failing test.
115
116         This is just a first iteration of the UI.  There's no way to resize or
117         zoom in on elements of the details pane, and images likely aren't sized
118         correctly, but it's a place to start.
119
120         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js:
121             - Add constants for our data attributes so we don't typo them!
122         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
123             - Add DOM structure for the details pane.
124         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
125         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
126             - Bind events for showing and hiding the details pane.
127         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
128             - Reduce the number of result types that we query for to avoid
129               overfilling the details pane.  We'll probably need another
130               solution here in the long-term.  For example, we could use tabs
131               to pack more results into the pane.
132         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
133         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
134             - Now that we're using the config package during testing, we need
135               to include it in the testing HTML.
136         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
137         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
138             - Change the DOM structure of the results detail to make it fit
139               nicely in the details pane (rather than flowing freely in the
140               body, as it did before).
141
142 2011-07-13  Adam Roben  <aroben@apple.com>
143
144         Make TestFailures show every time a possibly-flaky test failed, but hide it by default
145
146         It's useful to be able to see every time a flaky test failed to see whether it failed the
147         same way every time. But doing so takes a lot of space, so the list of failures is now
148         collapsed by default and can be revealed using a disclosure triangle.
149
150         Fixes <http://webkit.org/b/64455> TestFailures page doesn't show as much information for
151         flaky tests as I would like, even though the page is already so long
152
153         Reviewed by Dimitri Glazkov.
154
155         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
156         (FlakyLayoutTestDetector.prototype.allFailures): Replaced flakinessExamples with this
157         function. Now returns all failures for the given test.
158
159         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector_unittests.js:
160         Added. This just contains some simple tests of the FlakyLayoutTestDetector class. We'll add
161         more over time.
162
163         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
164         (LayoutTestHistoryAnalyzer.prototype.start): Updated the documentation comment to reflect
165         that we no longer return passing builds for possibly-flaky tests.
166
167         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
168         (.existing-bugs-list, .suspect-revisions-list, .flakiness-examples-list): Make the list of
169         flakiness examples small, too, since it can get quite long.
170
171         (.disclosure-triangle):
172         (.expanded > .disclosure-triangle):
173         Simple styles for the disclosure triangle.
174
175         (.flakiness-examples-list): Collapse the list by default.
176
177         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
178         (ViewController.prototype._displayBuilder): Pass the total number of builds analyzed to
179         _domForPossiblyFlakyTests.
180         (ViewController.prototype._domForPossiblyFlakyTests): Put a disclosure triangle to the left
181         of each test name, and the number of failures to the right. When the disclosure triangle is
182         clicked for the first time, we build up the list of failures and expand the element. After
183         that we just collapse or expand the element on subsequent clicks.
184
185         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
186         Pulled in new tests.
187
188 2011-07-13  Adam Roben  <aroben@apple.com>
189         
190         Teach TestFailures to understand NRWT's output when it exits early due to too many failures
191
192         Fixes <http://webkit.org/b/64456> TestFailures page reports way too many failures when NRWT
193         exits early
194
195         Reviewed by Dimitri Glazkov.
196
197         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
198         (Builder.prototype.getNumberOfFailingTests): Relaxed the "Exiting early" test not to require
199         it to be at the beginning of the line, since NRWT prints a bunch of junk earlier in the
200         line. Tightened up the regex that's used to parse the number of failing tests to require the
201         leading number to be followed by whitespace so that we won't parse the "2011" in
202         "2011-07-13" as a number of failures.
203
204         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
205         Added a test for the above.
206
207 2011-07-12  Dimitri Glazkov  <dglazkov@chromium.org>
208
209         Extract model-like TestExpectationLine and TestExpectationFile from TestExpectations.
210         https://bugs.webkit.org/show_bug.cgi?id=64386
211
212         This is the first step in converting TestExpectations to a real model.
213         * TestExpectationsLine represents a line in the test_expectations.txt file, and
214         * TestExpectationsFile represents the file, which is a collection of lines.
215
216         Reviewed by Adam Barth.
217
218         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
219         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
220         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
221
222 2011-07-13  Xan Lopez  <xlopez@igalia.com>
223
224         [GTK] Do not grab focus too early in DRT.
225
226         Reviewed by Gustavo Noronha.
227
228         It causes a layout to happen and a progress signal to be emitted
229         since r90900, but at this point we don't have a
230         LayoutTestController object and we'll eventually crash. Since we
231         already grab focus at the beginning of runTest() this is
232         redundant, so get rid of it to fix the crash.
233
234         * DumpRenderTree/gtk/DumpRenderTree.cpp:
235         (main): remove call to grab_focus
236
237 2011-07-13  Sheriff Bot  <webkit.review.bot@gmail.com>
238
239         Unreviewed, rolling out r90893 and r90894.
240         http://trac.webkit.org/changeset/90893
241         http://trac.webkit.org/changeset/90894
242         https://bugs.webkit.org/show_bug.cgi?id=64441
243
244         NRWT still doesn't work on qt-mac platform (Requested by Ossy
245         on #webkit).
246
247         * Scripts/run-webkit-tests:
248         (useNewRunWebKitTests):
249
250 2011-07-13  Csaba Osztrogonác  <ossy@webkit.org>
251
252         [Qt] NRWT should pick up the right httpd config file
253         https://bugs.webkit.org/show_bug.cgi?id=64086
254
255         * Scripts/run-webkit-tests: Enable NRWT on qt-mac platform after r90810.
256         (useNewRunWebKitTests):
257
258 2011-07-12  Mark Rowe  <mrowe@apple.com>
259
260         Fix the 32-bit build.
261
262         * DumpRenderTree/mac/UIDelegate.mm:
263         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:totalSpaceNeeded:]):
264         Cast the NSUInteger value to unsigned long to match the format specifier.
265
266 2011-07-12  Adam Barth  <abarth@webkit.org>
267
268         Improve garden-o-matic UI when the bots fail to report revision numbers
269         https://bugs.webkit.org/show_bug.cgi?id=64427
270
271         Reviewed by Dimitri Glazkov.
272
273         At least Win (dbg)(2) seems to fail to report the SVN revision number
274         in full_results.json.  This patch makes garden-o-matic more robust to
275         missing revision numbers.
276
277         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
278         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
279         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
280         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
281         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
282
283 2011-07-12  Adam Barth  <abarth@webkit.org>
284
285         Adjust garden-o-matic layout to use fewer lines
286         https://bugs.webkit.org/show_bug.cgi?id=64422
287
288         Reviewed by Ojan Vafai.
289
290         This patch effectively merges the test name line with the list of
291         builders on which the test fails.  Each test failure now occupies two
292         lines instead of three.
293
294         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
295         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
296         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
297
298 2011-07-12  Adam Barth  <abarth@webkit.org>
299
300         garden-o-matic should label tests failures of new tests as such
301         https://bugs.webkit.org/show_bug.cgi?id=64421
302
303         Reviewed by Ojan Vafai.
304
305         Also, we shouldn't dim failures of new tests, even if we've only seen
306         them once because they're likely to be real problems that need
307         attention.
308
309         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
310         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js: Added.
311         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout_unittests.js: Added.
312         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
313         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
314         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
315         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
316         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
317         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
318         * Scripts/webkitpy/tool/servers/gardeningserver.py:
319
320 2011-07-12  Adam Barth  <abarth@webkit.org>
321
322         garden-o-matic should display how many times we've seen a failure
323         https://bugs.webkit.org/show_bug.cgi?id=64417
324
325         Reviewed by Ojan Vafai.
326
327         This patch adds some UI to display how many times we've seen a given
328         failure, which can be helpful for determining whether that failure is a
329         real failure or a flaky test.
330
331         When a failure has only been seen once (i.e., only a single run on a
332         single bot), we set the opacity of to 50% to avoid distracting the
333         gardener.
334
335         This patch also refactors the failure walker to have a simpler API
336         internally by moving from an object-oriented paradigm to a functional
337         paradigm.
338
339         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
340         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
341         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
342         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
343         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
344         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
345
346 2011-07-12  Chris Rogers  <crogers@google.com>
347
348         Enable Web Audio for chromium DRT
349         https://bugs.webkit.org/show_bug.cgi?id=64409
350
351         Reviewed by James Robinson.
352
353         * DumpRenderTree/chromium/TestShell.cpp:
354         (TestShell::TestShell):
355
356 2011-07-12  Adam Barth  <abarth@webkit.org>
357
358         Tweak some UI in garden-o-matic now that the tree actually has a
359         failure and I can see what this all looks like.
360
361         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
362         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
363         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
364
365 2011-07-12  Joseph Pecoraro  <joepeck@webkit.org>
366
367         Unreviewed. Skipping a few tests which fail due to differing output
368         in recent ApplicationCache quota tests. Reenabling the tests is
369         tracked by <http://webkit.org/b/64410>.
370
371         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
372         (WebCore::DumpRenderTree::dumpApplicationCacheQuota): This should
373         have been a space, not a newline.
374
375 2011-07-12  Adam Barth  <abarth@webkit.org>
376
377         garden-o-matic should display regression ranges
378         https://bugs.webkit.org/show_bug.cgi?id=64407
379
380         Reviewed by Dimitri Glazkov.
381
382         This patch computes a regression range for a failure by intersecting
383         the regression ranges seen by the various bots.  We make the underlying
384         assumption that a test is only failing due to one revision at any given
385         moment.  If that's not true, this code probably explodes.
386
387         The regression ranges appear asynchronously, which might be a jarring
388         UI.  We'll have to experiment to see.
389
390         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
391         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
392         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
393         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
394         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
395         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
396         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
397
398 2011-07-12  Adam Barth  <abarth@webkit.org>
399
400         cr-linux-ews should run pixel tests
401         https://bugs.webkit.org/show_bug.cgi?id=64394
402
403         Reviewed by Ojan Vafai.
404
405         This patch enabled pixel test on the cr-linux-ews.  Running a quick
406         test on the bots, it looks like we have five pixel failures, related to
407         some missing fonts.  I've also made the bots skip failing tests to
408         avoid uploading huge zip files to bugs.webkit.org (with all the exected
409         IMAGE failures).
410
411         * Scripts/webkitpy/common/config/ports.py:
412         * Scripts/webkitpy/common/config/ports_unittest.py:
413
414 2011-07-12  Joseph Pecoraro  <joepeck@webkit.org>
415
416         ApplicationCache update should not immediately fail when reaching per-origin quota
417         https://bugs.webkit.org/show_bug.cgi?id=64177
418
419         Reviewed by Alexey Proskuryakov.
420
421         Add layoutTestController.disallowIncreaseForApplicationCacheQuota()
422         to disable the default behavior of raising the default per-origin
423         quota to 5MB when reached. This allows us to test what happens
424         when the quota is reached and not increased.
425
426         * DumpRenderTree/LayoutTestController.cpp:
427         (LayoutTestController::LayoutTestController):
428         (disallowIncreaseForApplicationCacheQuotaCallback):
429         (LayoutTestController::staticFunctions):
430         * DumpRenderTree/LayoutTestController.h:
431         (LayoutTestController::disallowIncreaseForApplicationCacheQuota):
432         (LayoutTestController::setDisallowIncreaseForApplicationCacheQuota):
433         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
434         (LayoutTestController::reset):
435         * DumpRenderTree/qt/LayoutTestControllerQt.h:
436         (LayoutTestController::disallowIncreaseForApplicationCacheQuota):
437         (LayoutTestController::dumpApplicationCacheDelegateCallbacks):
438         Expose a function to prevent automatically increasing the per-origin
439         quota when the quota is reached. This allows us to test what happens
440         when a user would "disallow" a quota increase.
441
442         * DumpRenderTree/mac/UIDelegate.mm:
443         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:spaceNeeded:]):
444         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
445         (WebCore::DumpRenderTree::dumpApplicationCacheQuota):
446         Respect the new disallowIncrease flag and dump the spaceNeeded values
447         for ports that implement application cache quotas.
448
449 2011-07-12  Ojan Vafai  <ojan@chromium.org>
450
451         fix flakiness dashboard to work with new crash log filenames
452         https://bugs.webkit.org/show_bug.cgi?id=64393
453
454         Reviewed by Adam Barth.
455
456         * TestResultServer/static-dashboards/flakiness_dashboard.html:
457
458 2011-07-12  Adam Barth  <abarth@webkit.org>
459
460         Add a happy animation to garden-o-matic when there are no failures
461         https://bugs.webkit.org/show_bug.cgi?id=64382
462
463         Reviewed by Ojan Vafai.
464
465         As requested by Ojan.
466
467         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
468         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
469         * Scripts/webkitpy/tool/servers/data/gardeningserver/partytime.gif: Added.
470         * Scripts/webkitpy/tool/servers/gardeningserver.py:
471
472 2011-07-12  Adam Barth  <abarth@webkit.org>
473
474         sheriffbot can't perform rollouts
475         https://bugs.webkit.org/show_bug.cgi?id=64370
476
477         Reviewed by Eric Seidel.
478
479         Now that we're not updating the working copy before processing every
480         IRC message, the rollout command needs to update it manually.
481
482         * Scripts/webkitpy/tool/bot/irc_command.py:
483         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
484
485 2011-07-12  Adam Barth  <abarth@webkit.org>
486
487         Fix the grammar in the FIXME comment.
488
489         * Scripts/webkitpy/common/net/resultsjsonparser.py:
490
491 2011-07-12  Eric Seidel  <eric@webkit.org>
492
493         NRWT should open test results page with Safari trunk, not the system provided one on Mac
494         https://bugs.webkit.org/show_bug.cgi?id=64346
495
496         Reviewed by Adam Barth.
497
498         To fix this I implemented Port.show_results_html_file in Mac, Gtk and Qt ports with
499         implementations (mostly) matching those found in old-run-webkit-tests.
500         There are still some minor differences for Qt which Qt hackers may wish to tweak.
501
502         I had to add a WebKitPort._port_flag_for_scripts method (similar to flag() in common.config.ports.py)
503         for the Qt/Gtk ports which always require a flag passed to scripts.
504
505         While trying to test this, I found that FactoryTest.test_chromium_gpu_linux
506         was using a real filesystem (due to assert_platform) which due to
507         global variables in config.py was causing set-webkit-configuration to have
508         an affect on unit test results!  So I fixed this by making FactoryTest.assert_port
509         pass a mock file system whenever calling factory.get.
510
511         Unfortunately TestPort was depending on always being passed a None filesystem
512         and asserting filesystem._tests (only true for unit_test_filesystem()).
513         So I just removed the FactoryTest.test_test and FactoryTest.test_dryrun tests
514         deciding that they were pretty much useless anyway.  If others feel strongly
515         I'm happy to fix this in a different way.
516
517         * Scripts/webkitpy/common/system/executive.py:
518          - default arguments in python are screwy.  They use a single shared
519            instance, so it's better to use argument=None and then argument = argument or Default()
520            if you have any chance of mutating (or returning) the default argument.
521         * Scripts/webkitpy/layout_tests/port/config.py:
522          - This code is wrong.  We don't need to use a global variable here (as far as I can tell).
523            I'm not fixing it in this patch, but I've marked it with a FIXME and we can convert to
524            storing the results of the read on the Config object (which should only be created once during normal operation).
525            Unit tests shouldn't be hitting the disk anyway.  It's possible Config should move off of Port and onto Tool/Host directly.
526         * Scripts/webkitpy/layout_tests/port/factory.py:
527         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
528         * Scripts/webkitpy/layout_tests/port/gtk.py:
529         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py: Added.
530         * Scripts/webkitpy/layout_tests/port/mac.py:
531         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
532         * Scripts/webkitpy/layout_tests/port/qt.py:
533         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
534         * Scripts/webkitpy/layout_tests/port/webkit.py:
535
536 2011-07-12  Adam Roben  <aroben@apple.com>
537
538         Teach TestFailures to recognize when run-webkit-tests gets killed by buildbot
539
540         Fixes <http://webkit.org/b/64358> TestFailures page thinks all tests passed in
541         http://build.webkit.org/builders/Windows%207%20Release%20(Tests)/builds/14672
542
543         Reviewed by Daniel Bates.
544
545         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
546         (Builder.prototype.getNumberOfFailingTests): If run-webkit-tests exited with a non-zero
547         exit status but we didn't find any failure counts, assume that there was some error that
548         caused run-webkit-tests to die early (like being killed by buildbot due to a timeout).
549
550         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
551         Added a new test that shows that we get a failingTestCount of -1 when run-webkit-tests dies
552         early.
553         (runGetNumberOfFailingTestsTest): Moved most code here from the only pre-existing test in this
554         file.
555
556         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
557         (LayoutTestResultsLoader.prototype.start): Bump the cache number so old cached data that was
558         tainted by the bug fixed in this patch will be evicted.
559
560 2011-07-12  Adam Barth  <abarth@webkit.org>
561
562         cr-linux-ews complains about tests that aren't actually failing
563         https://bugs.webkit.org/show_bug.cgi?id=64367
564
565         The underlying problem here is that full_results.json doesn't have
566         enough information to determine whether a given test result was
567         expected because whether an actual result is expected depends on
568         whether full_results.json was generated duing a test run that included
569         pixel tests.
570
571         The right long-term solution is to make full_results.json a complete
572         description of what happened durning a test run.  In this patch
573         however, to stop the spam, this patch makes the jsonresultsparser
574         ignore pixel failures.
575
576         (I'm landing this patch unreviewed in order to stop the bots from
577         spamming.  I'll happily address any review comments in a follow-up
578         patch.)
579
580         * Scripts/webkitpy/common/net/resultsjsonparser.py:
581         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
582
583 2011-07-12  Adam Barth  <abarth@webkit.org>
584
585         Re-work garden-o-matic UI to begin to look like the mocks
586         https://bugs.webkit.org/show_bug.cgi?id=64334
587
588         Reviewed by Dimitri Glazkov.
589
590         This patch changes how we display failures to be more like the mock the
591         dglazkov made. We now can display the same information much more
592         compactly.
593
594         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
595         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
596         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
597         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css: Added.
598         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
599         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
600         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
601         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
602         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
603         * Scripts/webkitpy/tool/servers/gardeningserver.py:
604
605 2011-07-11  Adam Roben  <aroben@apple.com>
606
607         Extract TestFailures's bug-filing code into two new classes
608
609         Fixes <http://webkit.org/b/64300> TestFailures page's new-bug-filing code is a mess!
610
611         Reviewed by Darin Adler and Adam Barth.
612
613         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js:
614         (Buildbot.prototype.resultsDirectoryURL): Changed to return a URI-encoded URL. Otherwise the
615         URL isn't valid (and it's harder to mock this function).
616
617         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot_unittests.js:
618         Added. Just tests the above change (for now).
619
620         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
621         Added a license header, enclosed everything in a closure to avoid polluting the global
622         namespace, changed the test name to actually describe the passing condition, and replaced
623         uses of equals() with equal(). (The latter is the actual name of the function, and matches
624         deepEqual, etc.)
625
626         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm.js: Added.
627         (NewBugForm): This class knows how to construct a <form> used for filing new bugs in
628         Bugzilla based on some parameters.
629         (NewBugForm.prototype.domElement): Creates and returns the <form> element. Code came from
630         ViewController.prototype._domForNewAndExistingBugs.
631
632         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm_unittests.js:
633         Added. Tests the above code.
634
635         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm.js: Added.
636         (TestFailureBugForm): This class knows how to construct a <form> element used for filing new
637         bugs specifically about test failures. Code came from
638         ViewController.prototype._domForNewAndExistingBugs.
639         (TestFailureBugForm.prototype.domElement): Slightly customizes the <form> element returned
640         by NewBugForm.
641         (TestFailureBugForm.prototype._computeOperatingSystem):
642         (TestFailureBugForm.prototype._computePlatform):
643         (TestFailureBugForm.prototype._createBugTitle):
644         (TestFailureBugForm.prototype._failingResultsHTMLURL):
645         (TestFailureBugForm.prototype._failingRevision):
646         (TestFailureBugForm.prototype._passingRevision):
647         (TestFailureBugForm.prototype._regressionRangeString):
648         Code came from ViewController.prototype._domForNewAndExistingBugs. I broke it out into
649         separate functions to break up the rat's nest a bit.
650
651         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm_unittests.js:
652         Added. Tests the above code.
653
654         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
655         (ViewController.prototype._domForNewAndExistingBugs): Moved code from here to
656         TestFailureBugForm/NewBugForm, and changed this code to use a TestFailureBugForm.
657
658         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
659         NewBugForm/TestFailureBugForm.js.
660
661         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
662         Added new tests and required files.
663
664 2011-07-12  Adam Roben  <aroben@apple.com>
665
666         Test that no intermediate WTF::Strings are created when concatenating with string literals
667
668         Test for <http://webkit.org/b/63330> Concatenating string literals and WTF::Strings using
669         operator+ is suboptimal
670
671         Reviewed by Darin Adler.
672
673         * TestWebKitAPI/Tests/WTF/StringOperators.cpp: Added.
674         (TestWebKitAPI::TEST): Test that a bunch of different string concatenation expressions don't
675         create any intermediate WTF::Strings while they're being evaluated.
676
677         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
678         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
679         Added new file.
680
681 2011-07-12  Eric Seidel  <eric@webkit.org>
682
683         [Qt] NRWT should pick up the right httpd config file
684         https://bugs.webkit.org/show_bug.cgi?id=64086
685
686         Reviewed by Adam Barth.
687
688         This is more fall-out from the Port class inappropriately encapsulating
689         both platform and port knowledge.  We need to split out some of this
690         platform knowledge into a separate class which can be better shared
691         between ports.
692
693         The fix was to move all the _path_to_apache_config_file logic down
694         into the WebKitPort baseclass so that all of the ports can find the
695         right apache config file, regardless of what platform they may be running on.
696
697         I did not move this down into base.Port (even though I considered it).
698         Chromium duplicates some of this logic, but since they have separate
699         subclasses for each port-OS combination (e.g. ChromiumMac, ChromiumWin)
700         they wouldn't notice the change.  Eventually we'll move this logic
701         out of Port entirely, and then it will be shared by all ports.
702
703         I also cleaned up the http-config logic for ORWT while I was there,
704         although since we're killing that code, I'm happy to revert that part
705         of the change if changing it is viewed as needlessly risky.
706
707         During this cleanup, I noticed that no ports use the "default" httpd.conf
708         which is left over from Mac OS X Tiger (and old linux distros) which used
709         Apache 1.3.  I've removed httpd.conf (and associated support in ORWT) since
710         we no longer support any ports which use this httpd.conf.
711
712         * Scripts/webkitperl/httpd.pm:
713         (getHTTPDConfigPathForTestDirectory):
714         * Scripts/webkitpy/layout_tests/port/base.py:
715         * Scripts/webkitpy/layout_tests/port/gtk.py:
716         * Scripts/webkitpy/layout_tests/port/mac.py:
717         * Scripts/webkitpy/layout_tests/port/qt.py:
718         * Scripts/webkitpy/layout_tests/port/webkit.py:
719         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
720         * Scripts/webkitpy/layout_tests/port/win.py:
721         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
722
723 2011-07-11  Hironori Bono  <hbono@chromium.org>
724
725         Reviewed by Adam Roben.
726
727         Implement layoutTestController.setTextDirection for WebKit2 and Windows.
728         https://bugs.webkit.org/show_bug.cgi?id=61931
729
730         This change implements layoutTextController.setTextDirection for WebKit2
731         and Windows so we can run a layout test added by r87770 on them. Each
732         implementation adds a wrapper function for Editor::setBaseWritingDirection()
733         so LayoutTestController can call it.
734
735         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
736         (LayoutTestController::setTextDirection): Implemented the binding function
737         for layoutTestController.setTextDirection (Windows).
738         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
739         Added a binding function for layoutTestController.setTextDirection (WebKit2).
740         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp: ditto.
741         (WTR::LayoutTestController::setTextDirection):
742         * WebKitTestRunner/InjectedBundle/LayoutTestController.h: ditto.
743
744 2011-07-11  Raphael Kubo da Costa  <kubo@profusion.mobi>
745
746         [EFL] ImageDiff: Make sure gEcoreEvas is destroyed at the right time.
747         https://bugs.webkit.org/show_bug.cgi?id=64293
748
749         Reviewed by Kent Tamura.
750
751         gEcoreEvas was being destroyed after shutdownEfl() was called, so the
752         call to ecore_evas_free() failed.
753
754         We now shut it down manually before shutdownEfl() is called.
755
756         * DumpRenderTree/efl/ImageDiff.cpp:
757         (main): Destroy gEcoreEvas before shutting down the EFL.
758
759 2011-07-11  Ryosuke Niwa  <rniwa@webkit.org>
760
761         Make Chromium Mac and Linux tests bot core builders
762         https://bugs.webkit.org/show_bug.cgi?id=63196
763
764         Reviewed by Adam Barth.
765
766         Added Chromium Mac and Linux Release Tests bots core builders.
767
768         Also rearranged the order of bots so that test bots show up right next to
769         respective builders.
770
771         * BuildSlaveSupport/build.webkit.org-config/config.json:
772         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
773         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
774
775 2011-07-11  Raphael Kubo da Costa  <kubo@profusion.mobi>
776
777         [EFL] DRT: Add code related to font management.
778         https://bugs.webkit.org/show_bug.cgi?id=63989
779
780         Reviewed by Kent Tamura.
781
782         The added files are responsible for managing fontconfig and adding the
783         fonts required by some layout tests to fontconfig's path.
784
785         * DumpRenderTree/efl/FontManagement.cpp: Added.
786         (getFontDirectories):
787         (getFontFiles):
788         (addFontDirectories):
789         (addFontFiles):
790         (addFontsToEnvironment):
791         * DumpRenderTree/efl/FontManagement.h: Added.
792
793 2011-07-08  Dirk Pranke  <dpranke@chromium.org>
794
795         nrwt: linting fixes
796         https://bugs.webkit.org/show_bug.cgi?id=64225
797
798         Reviewed by Eric Siedel.
799
800         Miscellaneous linting fixes. The most notable change is that
801         we add public attributes for user, executive, filesystem, and
802         options on the Port object, so we don't have to refer to the
803         "protected" versions all over the place".
804
805         * Scripts/webkitpy/layout_tests/controllers/manager.py:
806         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
807         * Scripts/webkitpy/layout_tests/controllers/message_broker.py:
808         * Scripts/webkitpy/layout_tests/controllers/message_broker_unittest.py:
809         * Scripts/webkitpy/layout_tests/controllers/worker.py:
810         * Scripts/webkitpy/layout_tests/port/base.py:
811         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
812
813 2011-07-11  Ryosuke Niwa  <rniwa@webkit.org>
814
815         webkit-patch roll-chromium-deps no longer works
816         https://bugs.webkit.org/show_bug.cgi?id=64324
817
818         Reviewed by James Robinson.
819
820         Lower the minimum commit message length to 10.
821
822         * Scripts/webkitpy/tool/steps/commit.py:
823
824 2011-07-11  Alice Boxhall  <aboxhall@chromium.org>
825
826         Convert json_results_generator.py to output version 4 JSON.
827         https://bugs.webkit.org/show_bug.cgi?id=60869
828
829         Reviewed by Ojan Vafai.
830
831         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
832         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
833         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
834
835 2011-07-11  Martin Robinson  <mrobinson@igalia.com>
836
837         [GTK] [NRWT] Pixel tests do not work
838         https://bugs.webkit.org/show_bug.cgi?id=64091
839
840         Reviewed by Eric Seidel.
841
842         Fix pixel test runs for GTK+ new-run-webkit-tests by exposing the
843         path to the ImageDiff binary.
844
845         * Scripts/webkitpy/layout_tests/port/gtk.py: Implement _path_to_image_diff.
846
847 2011-07-11  Dirk Pranke  <dpranke@chromium.org>
848
849         Fix hang on win32 in a multiprocessing test that shouldn't have been running.
850
851         Unreviewed, build fix.
852
853         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
854
855 2011-07-11  Adam Roben  <aroben@apple.com>
856
857         Don't count leaks as test failures on TestFailures's front page
858
859         As a bonus, this patch adds our first unit test.
860
861         Fixes <http://webkit.org/b/64303> REGRESSION (r90489): TestFailures page says far more tests
862         are failing on the Leaks bot than actually are
863
864         Reviewed by Adam Barth.
865
866         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
867         (Builder.prototype.getNumberOfFailingTests): Exclude lines containing "leak" when summing
868         failure counts.
869
870         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
871         Added. Contains a simple test to show that leaks aren't counted as test failures.
872
873         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
874         Added. This is the test harness.
875
876 2011-06-14  Mike Stegeman  <mrstegeman@gmail.com>
877
878         [GTK] Add API to allow setting local storage database path
879         https://bugs.webkit.org/show_bug.cgi?id=62091
880
881         Reviewed by Martin Robinson and Gustavo Noronha.
882
883         Expose the path of the localStorage databases through a setting
884         to allow HTML5 localStorage to be persistent. New setting is:
885         html5-local-storage-database-path
886
887         * DumpRenderTree/gtk/DumpRenderTree.cpp:
888         (resetDefaultsToConsistentValues):
889
890 2011-07-11  Adam Barth  <abarth@webkit.org>
891
892         garden-o-matic should be able to rebaseline tests
893         https://bugs.webkit.org/show_bug.cgi?id=64186
894
895         Reviewed by Eric Seidel.
896
897         This patch contains a simple HTTP server binding for the
898         rebaseline-test command.  The frontend will appear in another patch.
899
900         * Scripts/webkitpy/tool/servers/gardeningserver.py:
901         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
902         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
903
904 2011-07-11  Adam Barth  <abarth@webkit.org>
905
906         Add some basic UI for showing regression ranges
907         https://bugs.webkit.org/show_bug.cgi?id=64243
908
909         Reviewed by Eric Seidel.
910
911         This UI is all temporary.  It exists mostly just to have some buttons
912         to click to exercise the code.
913
914         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
915         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
916         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
917         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
918         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
919
920 2011-07-11  Adam Barth  <abarth@webkit.org>
921
922         garden-o-matic should be able to determine which revisions caused a given failure
923         https://bugs.webkit.org/show_bug.cgi?id=64189
924
925         Reviewed by Adam Roben.
926
927         Walking the failure history looking for failures turns out to be
928         slightly tricky because the network requests are asynchronous.
929         Currently we do all the fetches serially and our cache is unbounded.
930         We'll probably optimize both those parameters eventually.
931
932         This patch also generalizes some functionality in the unit testing
933         framework to make testing this sort of code easier.
934
935         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
936         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
937         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
938
939 2011-07-11  Adam Barth  <abarth@webkit.org>
940
941         Add a webkit-patch command for rebaselining an individual test
942         https://bugs.webkit.org/show_bug.cgi?id=64246
943
944         Reviewed by Eric Seidel.
945
946         This patch introduces a command that's able to rebaseline a single
947         test.  Currently, the command works only with the build.chromium.org
948         buildbots, but extending it to work with the build.webkit.org bots
949         shouldn't be that hard.
950
951         A complete rebaseling tool should also include an "optimize baselines"
952         command (which moves/deletes baselines in order to reduce the number of
953         expected results files), but that will come in a future patch.
954
955         Really BuilderToPort should be merged into builders.py, but I'm going
956         to save that for a future patch as well.  (We need to stop shaving yaks
957         at some point.)
958
959         * Scripts/webkitpy/tool/commands/rebaseline.py:
960
961 2011-07-11  Adam Barth  <abarth@webkit.org>
962
963         garden-o-matic should be able to roll out patches
964         https://bugs.webkit.org/show_bug.cgi?id=64185
965
966         Reviewed by Eric Seidel.
967
968         This gardening server API is simply a binding to the rollout machinery
969         already present in webkit-patch.
970
971         * Scripts/webkitpy/tool/commands/gardenomatic.py:
972         * Scripts/webkitpy/tool/servers/gardeningserver.py:
973         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Added.
974
975 2011-07-11  Csaba Osztrogonác  <ossy@webkit.org>
976
977         [Qt][Mac] Unreviewed fix after r90746.
978
979         * Scripts/run-webkit-tests:
980         (useNewRunWebKitTests): Really disable NRWT for qt-mac platform.
981
982 2011-07-11  Adam Roben  <aroben@apple.com>
983
984         Teach TestFailures that ORWT's results.html file might be missing due to all tests passing
985
986         There are three reasons why we might fail to fetch ORWT's results.html:
987           - All tests passed, so no results.html was generated
988           - Some error during the test run caused results.html not to be generated (e.g., ORWT
989             timed out)
990           - Some network error occurred when fetching results.html
991
992         We were failing to account for the first possibility in some cases. For test runs before
993         r89610, we first check build.webkit.org/json to determine how many tests failed and whether
994         ORWT exited early due to too many failures; if all tests passed then we don't fetch
995         results.html at all. r89610 changed ORWT to put information in results.html about exiting
996         early due to too many failures, so we no longer needed to check build.webkit.org/json to get
997         that information, and in r89619 I changed TestFailures to do just that. But I forgot that we
998         still needed to check build.webkit.org/json to find out if all tests passed!
999
1000         Now, for test runs after r89610, we check results.html first, and then check
1001         build.webkit.org/json if we fail to fetch results.html. This lets us distinguish between all
1002         tests passing and the error cases.
1003
1004         Fixes <http://webkit.org/b/64280> TestFailures page can't pinpoint that r90699 caused 13
1005         tests to fail on Windows 7 Release (WebKit2 Tests)
1006
1007         Reviewed by David Kilzer.
1008
1009         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
1010         (LayoutTestResultsLoader.prototype.start): Bumped the cache version so that old, buggy
1011         cached data will get evicted. We were marking builds where all tests passed as errors!
1012         (LayoutTestResultsLoader.prototype._fetchAndParseORWTResults): Added success/error callback
1013         parameters to the fetchAndParseResultsHTML helper function, and added a similar
1014         fetchNumberOfFailingTests function that fetches data from build.webkit.org/json (code came
1015         from later in the function). For test runs before r89610, we first check
1016         build.webkit.org/json then check results.html. For builds after r89610, we first check
1017         results.html and then check build.webkit.org/json if we couldn't fetch results.html.
1018
1019 2011-07-11  Csaba Osztrogonác  <ossy@webkit.org>
1020
1021         new-run-webkit-tests does not support qt-arm or qt-4.8 results
1022         https://bugs.webkit.org/show_bug.cgi?id=64071
1023
1024         [Qt] NRWT should pick up the right httpd config file
1025         https://bugs.webkit.org/show_bug.cgi?id=64086
1026
1027         Reviewed by Zoltán Herczeg.
1028
1029         * Scripts/run-webkit-tests: Disable NRWT for qt-mac, qt-arm and qt-4.8 platforms until fix.
1030         (useNewRunWebKitTests):
1031
1032 2011-07-11  Kenichi Ishibashi  <bashi@chromium.org>
1033
1034         Add TestNetscapePlugIn/Tests/FormValue.cpp to DRT build files
1035         https://bugs.webkit.org/show_bug.cgi?id=64248
1036
1037         Reviewed by Kent Tamura.
1038
1039         * DumpRenderTree/DumpRenderTree.gypi: Added FormValue.cpp.
1040         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Ditto.
1041         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro: Ditto.
1042         * GNUmakefile.am: Ditto.
1043
1044 2011-07-10  Kenichi Ishibashi  <bashi@chromium.org>
1045
1046         Let plugins participate in form submission
1047         https://bugs.webkit.org/show_bug.cgi?id=13061
1048
1049         Adds NPPVformValue support to the testing plugin.
1050
1051         Reviewed by Anders Carlsson.
1052
1053         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added FormValue.cpp
1054         * DumpRenderTree/TestNetScapePlugIn/Tests/FormValue.cpp: Added.
1055         (FormValue::FormValue): Ditto.
1056         (FormValue::NPP_GetValue): Ditto.
1057
1058 2011-07-11  Roland Steiner  <rolandsteiner@chromium.org>
1059
1060         Reviewed by Tony Chang.
1061
1062         Complete functions in filesystem.py
1063         https://bugs.webkit.org/show_bug.cgi?id=63528
1064
1065         - missing text file functions added
1066         - functions sorted
1067         - removed 'append' optional parameters (were unused)
1068         - adapted filesystem_mock in the same way
1069
1070         * Scripts/webkitpy/common/system/filesystem.py:
1071         * Scripts/webkitpy/common/system/filesystem_mock.py:
1072         * Scripts/webkitpy/common/system/filesystem_unittest.py:
1073
1074 2011-07-10  Adam Barth  <abarth@webkit.org>
1075
1076         Clean up style in fallback path calculation
1077         https://bugs.webkit.org/show_bug.cgi?id=64239
1078
1079         Reviewed by Daniel Bates.
1080
1081         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1082         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1083         * Scripts/webkitpy/layout_tests/port/mac.py:
1084         * Scripts/webkitpy/layout_tests/port/win.py:
1085
1086 2011-07-10  Balazs Kelemen  <kbalazs@webkit.org>
1087
1088         [Qt][WK2] Implement the rest of WTR::PlatformWebView
1089         https://bugs.webkit.org/show_bug.cgi?id=63630
1090
1091         Reviewed by Kenneth Rohde Christiansen.
1092
1093         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
1094         (WTR::PlatformWebView::windowFrame):
1095         (WTR::PlatformWebView::setWindowFrame):
1096
1097 2011-07-10  Mark Rowe  <mrowe@apple.com>
1098
1099         Teach build-webkit how to find the latest WebKitSystemInterface binary.
1100
1101         * Scripts/build-webkit:
1102
1103 2011-07-09  Adam Roben  <aroben@apple.com>
1104
1105         Teach TestFailures to abbreviate the examples of test flakiness
1106
1107         These lists can get quite long, and it's not really helpful in most cases to have soooooo
1108         many examples of flakiness.
1109
1110         Fixes <http://webkit.org/b/64203> Lists of flaky revisions on TestFailures page can get so
1111         long they're hard to navigate
1112
1113         Reviewed by Dan Bates.
1114
1115         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
1116         (FlakyLayoutTestDetector.prototype.flakinessExamples): If we have more than a certain number
1117         of examples, replace the middle items with a separator.
1118
1119         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
1120         (.flakiness-example-separator): Added styles for the separator.
1121
1122         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1123         (ViewController.prototype._domForPossiblyFlakyTests): Use a vertical ellipsis to represent
1124         the separator.
1125
1126 2011-07-09  Dirk Pranke  <dpranke@chromium.org>
1127
1128         nrwt: stack traces from worker-side exceptions aren't very useful inside test-webkitpy
1129         https://bugs.webkit.org/show_bug.cgi?id=64218
1130
1131         Reviewed by Eric Seidel.
1132
1133         Exceptions aren't picklable and can't be sent across the
1134         manager/worker message queue without losing information. NRWT
1135         handles this by turning the stack trace into a set of strings,
1136         and logging the strings when we receive an exception from the
1137         worker. However, when you are running tests and something
1138         crashes on the worker side, test-webkitpy prints the
1139         manager-side stack trace, which is just confusing and useless.
1140
1141         This patch changes the logic so that exceptions are passed
1142         through as-is when the worker and manager are in the same
1143         process (the --worker-model=inline option). This increases the
1144         code paths slightly but makes crashes much more useful.
1145
1146         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1147         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
1148         * Scripts/webkitpy/layout_tests/controllers/message_broker.py:
1149         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1150         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1151
1152 2011-07-09  Darin Fisher  <darin@chromium.org>
1153
1154         Eliminate bad dependency on gfx::Point.
1155         https://bugs.webkit.org/show_bug.cgi?id=64228
1156
1157         Reviewed by Kent Tamura.
1158
1159         * DumpRenderTree/chromium/EventSender.cpp:
1160         (initMouseEvent):
1161
1162 2011-07-08  Ojan Vafai  <ojan@chromium.org>
1163
1164         pull static dashboard files into the appengine server from the chromium repository
1165         https://bugs.webkit.org/show_bug.cgi?id=64208
1166
1167         Reviewed by Dirk Pranke.
1168
1169         These files belong in the WebKit repo since they are tied primarily to the webkit tests.
1170         They have some extra bits to support chromium's gtests, but that seems fine.
1171         Mainly, this will allow other WebKit hackers to hack on the dashboards.
1172
1173         As a nice side-effect, we can now push the dashboard files when we do appengine pushes
1174         instead of the weird thing we used to do of pulling them from the Chromium repository
1175         and storing them in the appengine datastore. This allows for cleaning up a lot of code
1176         and will likely make the dashboards load a bit faster.
1177
1178         The new JS files don't fully match WebKit style, but I'd like to clean that up in a
1179         followup patch if possible to maintain my sanity with this patch.
1180
1181         * TestResultServer/app.yaml:
1182         * TestResultServer/handlers/dashboardhandler.py: Removed.
1183         * TestResultServer/handlers/menu.py:
1184         * TestResultServer/main.py:
1185         * TestResultServer/model/dashboardfile.py: Removed.
1186         * TestResultServer/static-dashboards/LICENSE.dygraph.txt: Added.
1187         * TestResultServer/static-dashboards/README.dygraph.txt: Added.
1188         * TestResultServer/static-dashboards/README.webtreemap.txt: Added.
1189         * TestResultServer/static-dashboards/aggregate_results.html: Added.
1190         * TestResultServer/static-dashboards/builders.js: Added.
1191         * TestResultServer/static-dashboards/dashboard_base.js: Added.
1192         * TestResultServer/static-dashboards/dygraph-combined.js: Added.
1193         * TestResultServer/static-dashboards/flakiness_dashboard.html: Added.
1194         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js: Added.
1195         * TestResultServer/static-dashboards/timeline_explorer.html: Added.
1196         * TestResultServer/static-dashboards/treemap.html: Added.
1197         * TestResultServer/static-dashboards/webtreemap.css: Added.
1198         * TestResultServer/static-dashboards/webtreemap.js: Added.
1199         * TestResultServer/stylesheets/dashboardfile.css: Removed.
1200         * TestResultServer/stylesheets/menu.css:
1201         * TestResultServer/templates/dashboardfilelist.html: Removed.
1202         * TestResultServer/templates/menu.html:
1203
1204 2011-07-08  Adam Roben  <aroben@apple.com>
1205
1206         Make TestFailures's list of flaky tests look more like the list of non-flaky tests
1207
1208         Fixes <http://webkit.org/b/64204> TestFailures page's flaky tests list is ugly!
1209
1210         Reviewed by Daniel Bates.
1211
1212         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
1213         (#failure-history, #possibly-flaky-tests): Expanded this rule to apply to the list of flaky
1214         tests.
1215         (#failure-history > li, #possibly-flaky-tests > li): Ditto, but moved the 50px left padding
1216         from here...
1217         (#failure-history > li): ...to here.
1218
1219         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1220         (ViewController.prototype._domForPossiblyFlakyTests): Give the list an id attribute for
1221         styling purposes.
1222
1223 2011-07-08  Jeffrey Pfau  <jpfau@apple.com>
1224
1225         Unreviewed, add myself as committer.
1226
1227         * Scripts/webkitpy/common/config/committers.py:
1228
1229 2011-07-08  Adam Barth  <abarth@webkit.org>
1230
1231         TestResultsServer should keep old test results
1232         https://bugs.webkit.org/show_bug.cgi?id=64199
1233
1234         Reviewed by Ojan Vafai.
1235
1236         Having historical data will help us do failure archeology.
1237
1238         * TestResultServer/handlers/testfilehandler.py:
1239         * TestResultServer/model/testfile.py:
1240         * TestResultServer/templates/showfilelist.jsonp: Added.
1241
1242 2011-07-08  Adam Roben  <aroben@apple.com>
1243
1244         Remove commit-log-editor's dependency on Module::Load::Conditional
1245
1246         This module isn't available in Perl 5.8.8 (the version used on Leopard).
1247
1248         Fixes <http://webkit.org/b/64198> REGRESSION (r90583):
1249         webkitpy.common.checkout.checkout_unittest failing on Leopard
1250
1251         Reviewed by Daniel Bates.
1252
1253         * Scripts/commit-log-editor: Use the new loadTermReadKey() function instead of
1254         Module::Load::Conditional::can_load.
1255         (loadTermReadKey): Added. Tries to load Term::ReadKey and returns true if it's successful.
1256
1257 2011-07-08  Adam Roben  <aroben@apple.com>
1258
1259         Make checkout_unittest more robust against files moving around
1260
1261         Fixes <http://webkit.org/b/64197> checkout_unittest contains ugly, fragile code to find the
1262         Scripts directory
1263
1264         Reviewed by Adam Barth.
1265
1266         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1267         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Instantiate a real SCM
1268         object and use it to get the path to the Scripts directory, rather than hard-coding the
1269         relative path from this file to Scripts.
1270
1271 2011-07-08  Adam Roben  <aroben@apple.com>
1272
1273         Make TestFailures remember that run-webkit-tests timed out, even across reloads
1274
1275         Fixes <http://webkit.org/b/64193> TestFailures page incorrectly thinks all tests passed in
1276         http://build.webkit.org/builders/Windows%207%20Release%20(Tests)/builds/14589 after a reload
1277
1278         Reviewed by Daniel Bates.
1279
1280         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
1281         (LayoutTestResultsLoader.prototype.start): Store an "error" attribute in the cached data.
1282         When true, it indicates that there was an error retrieving the results for this build and
1283         that the errorCallback should be called.
1284
1285 2011-07-08  Adam Barth  <abarth@webkit.org>
1286
1287         Teach garden-o-matic how to display test results
1288         https://bugs.webkit.org/show_bug.cgi?id=64141
1289
1290         Reviewed by Ojan Vafai.
1291
1292         This patch includes basic infrastructure for probing build.chromium.org
1293         for test results.  We only handle text and image tests, not anything
1294         complicated like reftests.  Also, we're using the revision/build
1295         independent results store on the server, so we're avoiding that
1296         complication for now.
1297
1298         It's slightly hacky that we need to probe the server to see what kinds
1299         of results exist.  A better solution would be to add CORS support to
1300         the server or to use the local server to help.
1301
1302         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
1303         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
1304         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
1305         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
1306         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
1307         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
1308
1309 2011-07-08  Dirk Pranke  <dpranke@chromium.org>
1310
1311         REGRESSION(90419) NRWT's httpd locking is broken for --child-processes=1
1312         https://bugs.webkit.org/show_bug.cgi?id=64092
1313
1314         Reviewed by Tony Chang.
1315
1316         The code was incorrectly creating empty shards in the
1317         shard_in_two case.
1318
1319         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1320         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1321
1322 2011-07-08  Adam Roben  <aroben@apple.com>
1323
1324         Teach buildbot to figure out how many webkitpy/webkitperl tests failed
1325
1326         Fixes <http://webkit.org/b/64192> It's hard to tell how many test-webkitpy/test-webkitperl
1327         tests failed when looking at build.webkit.org
1328
1329         Reviewed by Eric Seidel.
1330
1331         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1332         (TestWithFailureCount): New class that represents a test build step which has an associated
1333         failure count. Eventually we should move more of our test classes to deriving from this.
1334         (TestWithFailureCount.countFailures): Method for subclasses to override to say how many
1335         failures occurred.
1336
1337         (TestWithFailureCount.commandComplete):
1338         (TestWithFailureCount.evaluateCommand):
1339         (TestWithFailureCount.getText):
1340         (TestWithFailureCount.getText2):
1341         These were all based on RunGtkAPITests.
1342
1343         (RunPythonTests): Changed to inherit from TestWithFailureCount.
1344         (RunPythonTests.countFailures): Parses the test-webkitpy output looking for the count of
1345         failures.
1346         (RunPerlTests): Changed to inherit from TestWithFailureCount.
1347         (RunPerlTests.countFailures): Parses the test-webkitperl output looking for the count of
1348         failures.
1349
1350 2011-07-08  Adam Roben  <aroben@apple.com>
1351
1352         Ensure $CHANGE_LOG_EMAIL_ADDRESS is set when testing webkitpy's commit-log-editor integration
1353
1354         Fixes <http://webkit.org/b/64180> REGRESSION (r90564): test-webkitpy failing on multiple
1355         bots due to commit-log-editor errors
1356
1357         Reviewed by Adam Barth.
1358
1359         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1360         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Set
1361         $CHANGE_LOG_EMAIL_ADDRESS to the patch author's email address. This ensures that
1362         commit-log-editor can find a value for the committer's email, and that the committer and
1363         author email addresses match, which will prevent commit-log-editor from inserting a "Patch
1364         by" line in the commit message.
1365
1366 2011-07-08  Andreas Kling  <kling@webkit.org>
1367
1368         [Qt][WK2] Views should expose QActions for basic navigation.
1369         https://bugs.webkit.org/show_bug.cgi?id=64174
1370
1371         Reviewed by Benjamin Poulain.
1372
1373         Bring the toolbar in Qt's MiniBrowser back to life.
1374
1375         * MiniBrowser/qt/BrowserView.cpp:
1376         (BrowserView::navigationAction):
1377         * MiniBrowser/qt/BrowserView.h:
1378         * MiniBrowser/qt/BrowserWindow.cpp:
1379         (BrowserWindow::BrowserWindow):
1380
1381 2011-07-08  Adam Roben  <aroben@apple.com>
1382
1383         Teach TestFailures how to find test names in commit-log-editor-style commit messages
1384
1385         TestFailures was relying on Trac turning the list of modified files in our commit messages
1386         into an HTML list. But Trac only does this when the list of modified files is indented.
1387         commit-log-editor doesn't indent the file list, so the list wasn't being turned into an HTML
1388         list, which was confusing TestFailures.
1389
1390         TestFailures now does much simpler parsing of the commit message (i.e., just a substring
1391         search) without relying at all on its structure.
1392
1393         Fixes <http://webkit.org/b/64173> TestFailures page fails to blame r90608 for breaking
1394         fast/dom/HTMLProgressElement/progress-element-markup.html on Windows
1395
1396         Reviewed by David Kilzer.
1397
1398         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
1399         (Trac.prototype.getCommitDataForRevisionRange): Instead of trying to parse the commit
1400         message, just return its text.
1401
1402         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1403         (ViewController.prototype._domForRegressionRange): Instead of searching for test names in
1404         each commit's list of modified files, just search for test names anywhere in the commit's
1405         message.
1406
1407 2011-07-08  Adam Barth  <abarth@webkit.org>
1408
1409         sheriffbot is too spammy in IRC
1410         https://bugs.webkit.org/show_bug.cgi?id=64153
1411
1412         Reviewed by Eric Seidel.
1413
1414         Reporting failures in IRC worked well when the tree was greener than it
1415         is today.  Nowadays, this feature mostly just results in spam about
1416         false positives.  If we reach a greener state, we can bring this code
1417         back.
1418
1419         (Another possibility is to restrict this feature to builder bots,
1420         rather than including testers, as we did before this patch.)
1421
1422         * Scripts/webkitpy/tool/commands/sheriffbot.py:
1423         * Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
1424
1425 2011-07-08  Adam Roben  <aroben@apple.com>
1426
1427         Ensure commit-log-editor adds a "Patch by" line when the author and committer are different
1428
1429         Previously we were only adding a "Patch by" line when the ChangeLog contained a "Reviewed
1430         by" line. But some patches (like rollout patches) don't contain that line. Now we always add
1431         "Patch by" to the commit log regardless of the ChangeLog's contents.
1432
1433         Fixes <http://webkit.org/b/64127> Committer for r90588 is commit-queue@webkit.org, but
1434         should have been abarth@webkit.org
1435
1436         Reviewed by Anders Carlsson.
1437
1438         * Scripts/commit-log-editor:
1439         (createCommitMessage): Try to put the "Patch by" line just above the "Reviewed by" line, as
1440         before. If there is no "Reviewed by" line, try to put it just above the first modified file.
1441         If all else fails, put it at the end of the commit message.
1442         (patchAuthorshipString): Added. Code came from createCommitMessage.
1443
1444 2011-07-08  Herczeg Zoltan  <zherczeg@webkit.org>
1445
1446         Reviewed by Andreas Kling.
1447
1448         Adding myself as a reviewer.
1449
1450         * Scripts/webkitpy/common/config/committers.py:
1451
1452 2011-07-07  Kevin Ollivier  <kevino@theolliviers.com>
1453
1454         [wx] Unreviewed build fix, remove old files from the tree and update paths to new ones.
1455
1456         * wx/build: Removed.
1457         * wx/build/build_utils.py: Removed.
1458         * wx/build/settings.py: Removed.
1459         * wx/build/waf_extensions.py: Removed.
1460         * wx/build/wxpresets.py: Removed.
1461         * wx/packaging/build-mac-installer.py:
1462
1463 2011-07-07  Kevin Ollivier  <kevino@theolliviers.com>
1464
1465         [wx] Unreviewed build fix, more fixes for install name issues.
1466
1467         * wx/packaging/build-mac-installer.py:
1468
1469 2011-07-07  Dirk Pranke  <dpranke@chromium.org>
1470
1471         test-webkitpy fails on chromium win
1472         https://bugs.webkit.org/show_bug.cgi?id=64137
1473
1474         Reviewed by Eric Seidel.
1475
1476         test-webkitpy doesn't play nicely with the multiprocessing
1477         module on win32. An earlier change actually reenabled the tests
1478         on win32 by mistake.
1479
1480         This patch also fixes a few cases where path names will trip
1481         things up on win32 (e.g., by testing unix paths on windows).
1482         We do not lose any real coverage here with those fixes.
1483
1484         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
1485         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
1486         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
1487         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1488
1489 2011-07-07  Leandro Pereira  <leandro@profusion.mobi>
1490
1491         [ImageDiff] Calculate/print difference right after reading baseline image.
1492         https://bugs.webkit.org/show_bug.cgi?id=64117
1493
1494         Reviewed by Kent Tamura.
1495
1496         Otherwise, ImageDiff will block on fgets() until the universe ends or you ^C
1497         it.  Whichever happens first.
1498
1499         * DumpRenderTree/efl/ImageDiff.cpp:
1500         (main):
1501
1502 2011-07-07  Raphael Kubo da Costa  <kubo@profusion.mobi>
1503
1504         [EFL] DRT: Leak cairo_t to make the pixel tests stop crashing
1505         https://bugs.webkit.org/show_bug.cgi?id=64107
1506
1507         Reviewed by Kent Tamura.
1508
1509         The pointer is later managed by BitmapContext, but it was being
1510         de-refed and deleted earlier by our RefPtr, causing crashes in all
1511         pixel tests.
1512
1513         We now call leakRef() to make sure it is not removed when our
1514         createBitmapContextFromWebView goes out of context.
1515
1516         * DumpRenderTree/efl/PixelDumpSupportEfl.cpp:
1517         (createBitmapContextFromWebView):
1518
1519 2011-07-07  Adam Roben  <aroben@apple.com>
1520
1521         Update TestFailures's title and header immediately upon navigation
1522
1523         Fixes <http://webkit.org/b/64125> TestFailures page seems slow to react on link clicks
1524
1525         Reviewed by Daniel Bates.
1526
1527         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1528         (ViewController.prototype.loaded): Create and store an h1 element for displaying the page's
1529         title.
1530         (ViewController.prototype._displayBuilder): Set the page's title and clear out any old
1531         content immediately instead of waiting for the first set of results for the history
1532         analyzer. As we receive new results from the analyzer we'll just clear out the main content
1533         area and replace it with the new info.
1534         (ViewController.prototype._displayTesters): Set the page's title and clear out any old
1535         content immediately instead of waiting for the list of testers from the buildbot. Sprinkled
1536         in a little use of our removeAllChildren helper, too.
1537         (ViewController.prototype._setTitle): Added. Set's the page's title and the header text.
1538
1539 2011-07-07  Adam Roben  <aroben@apple.com>
1540
1541         Make Checkout use SCM's Executive instead of conjuring up its own
1542
1543         This will improve integration with the rest of webkitpy, particularly when invoked via
1544         webkit-patch.
1545
1546         Fixes <http://webkit.org/b/64115> REGRESSION (r90564): webkitpy's Checkout class uses
1547         Executive inappropriately
1548
1549         Reviewed by Adam Barth.
1550
1551         * Scripts/webkitpy/common/checkout/checkout.py:
1552         (Checkout.commit_message_for_this_commit): Use SCM.run instead of creating an Executive for
1553         our own use. SCM might have some super-special Executive that it uses under the covers, and
1554         we want to use it, too!
1555
1556         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1557         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Mock the SCM.run
1558         method to call through to Executive.run_command.
1559
1560 2011-07-07  Adam Roben  <aroben@apple.com>
1561
1562         Make Term/ReadKey.pm an optional dependency of commit-log-editor
1563
1564         When Term/ReadKey.pm isn't present, the --regenerate-log switch will be non-functional.
1565
1566         Fixes <http://webkit.org/b/64113> REGRESSION (r90564):
1567         webkitpy.common.checkout.checkout_unittest failing on GTK bots due to missing
1568         Term/ReadKey.pm module
1569
1570         Reviewed by Xan Lopez.
1571
1572         * Scripts/commit-log-editor: Use Module::Load::Conditional::can_load to only load
1573         Term/ReadKey.pm if it's present. If it isn't present, ignore the --regenerate-log switch so
1574         that existing commit messages will be preserved. (Users can manually remove existing commit
1575         messages using git-reset.)
1576
1577 2011-07-07  Adam Roben  <aroben@apple.com>
1578
1579         Keep testing that commit messages containing Unicode are handled properly by webkitpy
1580
1581         Fixes <http://webkit.org/b/64109> REGRESSION (r90571): test-webkitpy no longer tests that we
1582         correctly handle Unicode in commit messages
1583
1584         Reviewed by Eric Seidel.
1585
1586         * Scripts/webkitpy/common/checkout/checkout_unittest.py: Put some Unicode characters in
1587         places in the ChangeLog that will end up being included in the commit message.
1588
1589 2011-07-07  Andreas Kling  <kling@webkit.org>
1590
1591         [Qt][WK2] Remove Symbian code.
1592         https://bugs.webkit.org/show_bug.cgi?id=64101
1593
1594         Reviewed by Benjamin Poulain.
1595
1596         * MiniBrowser/qt/BrowserWindow.cpp:
1597         (BrowserWindow::BrowserWindow):
1598         (BrowserWindow::screenshot):
1599         * MiniBrowser/qt/MiniBrowser.pro:
1600         * MiniBrowser/qt/MiniBrowserApplication.h:
1601         (WindowOptions::WindowOptions):
1602
1603 2011-07-07  Andreas Kling  <kling@webkit.org>
1604
1605         [Qt][WK2] Don't support ridiculous matrix of QT_NO_FEATURE combinations.
1606         https://bugs.webkit.org/show_bug.cgi?id=64099
1607
1608         Reviewed by Benjamin Poulain.
1609
1610         * MiniBrowser/qt/BrowserWindow.cpp:
1611         (BrowserWindow::openFile):
1612         (BrowserWindow::screenshot):
1613         (BrowserWindow::loadURLListFromFile):
1614
1615 2011-07-07  Adam Roben  <aroben@apple.com>
1616
1617         Completely ignore too-many-failures builds in TestFailures in most circumstances
1618
1619         Because a semi-arbitrary subset of tests are run in a too-many-failures build, we can't
1620         really use them to perform regression analysis. The only time we want to pay attention to
1621         too-many-failures builds is when we're trying to explain when the current bout of
1622         too-many-failures started.
1623
1624         Fixes <http://webkit.org/b/64106> TestFailures page sometimes claims a test started failing
1625         in a build that didn't even run it (because it exited early due to too many failues)
1626
1627         Reviewed by David Kilzer.
1628
1629         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
1630         (LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): Removed old, broken
1631         too-many-failures handling that would cause us to blame builds that didn't even run a given
1632         test for breaking it. Instead, skip over all too-many-failures builds unless the most recent
1633         build was itself a too-many-failures build.
1634
1635 2011-07-07  Adam Roben  <aroben@apple.com>
1636
1637         Teach webkitpy's Checkout class to use commit-log-editor to create commit messages
1638
1639         Fixes <http://webkit.org/b/26755> webkit-patch's commit messages are less readable than
1640         commit-log-editor's
1641
1642         Reviewed by David Kilzer.
1643
1644         * Scripts/webkitpy/common/checkout/checkout.py:
1645         (Checkout.commit_message_for_this_commit): Run commit-log-editor, passing it the paths of
1646         the modified ChangeLogs, to generate the commit message, rather than trying to generate one
1647         ourselves.
1648
1649         * Scripts/webkitpy/common/checkout/checkout_unittest.py: Updated the expected commit message
1650         to match commit-log-editor's format.
1651         (CommitMessageForThisCommitTest.setUp): Write the ChangeLogs into Tools and LayoutTests
1652         directories so we can see how the various entries get labeled in the commit message.
1653         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Create a mock SCM
1654         instance that knows how to find commit-log-editor and pass it to our Checkout instance.
1655         Don't bother capturing output, since there shouldn't be any.
1656
1657 2011-07-07  Adam Roben  <aroben@apple.com>
1658
1659         Add a --print-log option to commit-log-editor
1660
1661         When specified, commit-log-editor takes a set of ChangeLog paths on the command line,
1662         generates a commit message from those ChangeLogs, prints it to stdout, and exits.
1663
1664         Prep work for fixing <http://webkit.org/b/26755> webkit-patch's commit messages are less
1665         readable than commit-log-editor's
1666
1667         Reviewed by David Kilzer.
1668
1669         * Scripts/commit-log-editor: Changed to use Getopt::Long to parse options. Added --print-log
1670         option, which calls through to createCommitMessage, prints the result, and exits.
1671         (printUsageAndExit): Renamed from "usage" for clarity. Beefed up the usage statement to
1672         explain commit-log-editor's 2.5 (normal, --print-log, and --help) modes of operation.
1673
1674 2011-07-07  Adam Roben  <aroben@apple.com>
1675
1676         Extract commit-log-editor's commit-message-generation code into a separate function
1677
1678         As a bonus, we now skip a bunch of work in the case where we are generating a commit message
1679         for staged git changes.
1680
1681         Prep work for fixing <http://webkit.org/b/26755> webkit-patch's commit messages are less
1682         readable than commit-log-editor's
1683
1684         Reviewed by David Kilzer.
1685
1686         * Scripts/commit-log-editor: Moved the default definition of $endl up toward the top of the
1687         file so that it will be defined even if we aren't operating on an existing log message.
1688         Moved a bunch of top-level code to generate the commit message from here...
1689         (createCommitMessage): ...to here.
1690
1691 2011-07-07  Carlos Garcia Campos  <cgarcia@igalia.com>
1692
1693         [GTK] Open links in a new window when clicking with the middle button in MiniBrowser
1694         https://bugs.webkit.org/show_bug.cgi?id=63988
1695
1696         Reviewed by Martin Robinson.
1697
1698         * MiniBrowser/gtk/BrowserWindow.c:
1699         (browserWindowConstructed):
1700         (decidePolicyForNavigationAction): Ignore the action if a link is
1701         clicked with the middle buttonm and open the link in a new window.
1702         (browserWindowPolicyClientInit):
1703
1704 2011-07-07  Adam Barth  <abarth@webkit.org>
1705
1706         Remove style_references.py
1707         https://bugs.webkit.org/show_bug.cgi?id=64070
1708
1709         Reviewed by Eric Seidel.
1710
1711         We never really adopted the _references design, and this file is a bit
1712         of an orphan at this point.
1713
1714         * Scripts/check-webkit-style:
1715         * Scripts/webkitpy/style/checker.py:
1716         * Scripts/webkitpy/style/checker_unittest.py:
1717         * Scripts/webkitpy/style/checkers/python.py:
1718         * Scripts/webkitpy/style/checkers/test_expectations.py:
1719         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
1720         * Scripts/webkitpy/style/main_unittest.py:
1721         * Scripts/webkitpy/style_references.py: Removed.
1722
1723 2011-07-07  Eric Seidel  <eric@webkit.org>
1724
1725         Fix WebKit2 expected results search paths for Mac and Qt under new-run-webkit-tests
1726         https://bugs.webkit.org/show_bug.cgi?id=64056
1727
1728         Unreviewed.
1729
1730         I had written this unittest prior to landing my change, but
1731         had named it qt_unitest.py, so it wasn't running (nor was
1732         it added to my git repo).
1733
1734         Once I fixe the name of the unittest file, it was very easy
1735         to fix the error in webkit.py (which already had a FIXME).
1736
1737         * Scripts/webkitpy/layout_tests/port/qt_unittest.py: Added.
1738         * Scripts/webkitpy/layout_tests/port/webkit.py:
1739
1740 2011-07-07  Eric Seidel  <eric@webkit.org>
1741
1742         Unreviewed.  Add more debug logging about which test expectations we're using.
1743
1744         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1745         * Scripts/webkitpy/layout_tests/port/webkit.py:
1746
1747 2011-07-06  Adam Barth  <abarth@webkit.org>
1748
1749         Remove unexpected_results.json
1750         https://bugs.webkit.org/show_bug.cgi?id=64065
1751
1752         Reviewed by Eric Seidel.
1753
1754         No one should be using this file any more.  This patch removes it from
1755         the face of the earth.  This patch also removes two NRWT-specific
1756         command line options that were implemented using
1757         unexpected_results.json.
1758
1759         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
1760         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1761         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1762         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1763         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1764         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
1765         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
1766
1767 2011-07-07  Adam Barth  <abarth@webkit.org>
1768
1769         REGRESSION(90520): chromium GPU tests produce output in the LayoutTests directory
1770         https://bugs.webkit.org/show_bug.cgi?id=64067
1771
1772         Reviewed by Eric Seidel.
1773
1774         The author of r90520 forgot to change the chromium_gpu.py implemenation
1775         of this method as well.
1776
1777         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
1778
1779 2011-07-06  Eric Seidel  <eric@webkit.org>
1780
1781         Fix WebKit2 expected results search paths for Mac and Qt under new-run-webkit-tests
1782         https://bugs.webkit.org/show_bug.cgi?id=64056
1783
1784         Reviewed by Adam Barth.
1785
1786         This fixes the MacPort's version of baseline_search_path to include
1787         wk2 fallback like WebKitPort's does.  This also re-works the QtPort
1788         to set self._name (as other ports do) so that WebKitPort's default
1789         implementations will understand Qt operating system flavors correctly
1790         (including when running wk2 tests).
1791
1792         In trying to test this, I realized that nearly all of our port-tests
1793         use the default FileSystem(), User() and Executive() objects, which
1794         is really bad, because they expose details of your system in the unittest results!
1795
1796         Once I fixed many of the Port() calls to pass MockFileSystem(), then
1797         MacPort.test_tests_for_other_platforms started failing, and I had to
1798         teach MockFileSystem.glob how to handle directories to make it work again.
1799
1800         I removed the useless PortTestCase.test_baseline_search_path and replaced
1801         it with more useful test_baseline_search_path tests in MacPortTest and QtPortTest.
1802
1803         * Scripts/webkitpy/common/config/ports.py:
1804         * Scripts/webkitpy/common/system/filesystem_mock.py:
1805         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1806         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
1807         * Scripts/webkitpy/layout_tests/port/mac.py:
1808         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1809         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1810         * Scripts/webkitpy/layout_tests/port/qt.py:
1811         * Scripts/webkitpy/layout_tests/port/test_files.py:
1812         * Scripts/webkitpy/layout_tests/port/win.py:
1813         * Scripts/webkitpy/to_be_moved/deduplicate_tests.py:
1814         * Scripts/webkitpy/tool/mocktool.py:
1815
1816 2011-07-06  Adam Barth  <abarth@webkit.org>
1817
1818         Wean rebaseline-server off unexpected_results.json
1819         https://bugs.webkit.org/show_bug.cgi?id=64063
1820
1821         Reviewed by Eric Seidel.
1822
1823         In the process of changing this code to use full_results.json, I
1824         noticed that the code was broken (because it wasn't tested).  This
1825         patch also adds test coverage for the broken code.
1826
1827         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
1828         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
1829         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
1830
1831 2011-07-06  Adam Barth  <abarth@webkit.org>
1832
1833         Wean resultsjsonparser off unexpected_results.json
1834         https://bugs.webkit.org/show_bug.cgi?id=64061
1835
1836         Reviewed by Eric Seidel.
1837
1838         We're about to delete unexpected_results.json, so we need to remove all
1839         the users.
1840
1841         * Scripts/webkitpy/common/config/ports.py:
1842         * Scripts/webkitpy/common/net/resultsjsonparser.py:
1843         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
1844         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1845
1846 2011-07-06  Adam Barth  <abarth@webkit.org>
1847
1848         Cleanup result_summary.py
1849         https://bugs.webkit.org/show_bug.cgi?id=64057
1850
1851         Reviewed by Eric Seidel.
1852
1853         This class had a bunch of out-of-date docstrings that no longer make
1854         any sense.
1855
1856         * Scripts/webkitpy/layout_tests/models/result_summary.py:
1857
1858 2011-07-06  Adam Barth  <abarth@webkit.org>
1859
1860         Tweak the ChangeLog for DEPS rolls to say "Unreviewed" so that the
1861         patches can be landed by the commit-queue.
1862
1863         * Scripts/webkitpy/tool/steps/preparechangelogfordepsroll.py:
1864
1865 2011-07-06  Adam Barth  <abarth@webkit.org>
1866
1867         Normalize callers of logging.getLogger to use __name__
1868         https://bugs.webkit.org/show_bug.cgi?id=64054
1869
1870         Reviewed by Eric Seidel.
1871
1872         Most callers were already using __name__.  This patch makes everyone
1873         consistent.
1874
1875         * Scripts/webkitpy/common/checkout/diff_parser.py:
1876         * Scripts/webkitpy/common/checkout/scm/svn.py:
1877         * Scripts/webkitpy/common/net/statusserver.py:
1878         * Scripts/webkitpy/common/system/executive.py:
1879         * Scripts/webkitpy/common/system/file_lock.py:
1880         * Scripts/webkitpy/common/system/user.py:
1881         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
1882         * Scripts/webkitpy/layout_tests/models/result_summary.py:
1883         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1884         * Scripts/webkitpy/layout_tests/port/chromium.py:
1885         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1886         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1887         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1888         * Scripts/webkitpy/layout_tests/port/gtk.py:
1889         * Scripts/webkitpy/layout_tests/port/http_lock.py:
1890         * Scripts/webkitpy/layout_tests/port/mac.py:
1891         * Scripts/webkitpy/layout_tests/port/qt.py:
1892         * Scripts/webkitpy/layout_tests/port/server_process.py:
1893         * Scripts/webkitpy/layout_tests/port/webkit.py:
1894         * Scripts/webkitpy/layout_tests/port/win.py:
1895         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
1896         * Scripts/webkitpy/layout_tests/servers/http_server.py:
1897         * Scripts/webkitpy/layout_tests/servers/http_server_base.py:
1898         * Scripts/webkitpy/layout_tests/servers/websocket_server.py:
1899         * Scripts/webkitpy/layout_tests/views/printing.py:
1900         * Scripts/webkitpy/python24/versioning.py:
1901         * Scripts/webkitpy/style/checker.py:
1902         * Scripts/webkitpy/style/checkers/test_expectations.py:
1903         * Scripts/webkitpy/style/patchreader.py:
1904         * Scripts/webkitpy/to_be_moved/update_webgl_conformance_tests.py:
1905
1906 2011-07-06  Adam Barth  <abarth@webkit.org>
1907
1908         Move model classes out of layout_package and into models
1909         https://bugs.webkit.org/show_bug.cgi?id=64053
1910
1911         Reviewed by Eric Seidel.
1912
1913         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
1914         * Scripts/webkitpy/common/net/layouttestresults.py:
1915         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
1916         * Scripts/webkitpy/common/net/resultsjsonparser.py:
1917         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
1918         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1919         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1920         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
1921         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1922         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
1923         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
1924         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
1925         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py: Removed.
1926         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py: Removed.
1927         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py: Removed.
1928         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py: Removed.
1929         * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py: Removed.
1930         * Scripts/webkitpy/layout_tests/layout_package/test_input.py: Removed.
1931         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
1932         * Scripts/webkitpy/layout_tests/layout_package/test_results.py: Removed.
1933         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py: Removed.
1934         * Scripts/webkitpy/layout_tests/models: Added.
1935         * Scripts/webkitpy/layout_tests/models/__init__.py: Added.
1936         * Scripts/webkitpy/layout_tests/models/result_summary.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/result_summary.py.
1937         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_expectations.py.
1938         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py.
1939         * Scripts/webkitpy/layout_tests/models/test_failures.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_failures.py.
1940         * Scripts/webkitpy/layout_tests/models/test_failures_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py.
1941         * Scripts/webkitpy/layout_tests/models/test_input.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_input.py.
1942         * Scripts/webkitpy/layout_tests/models/test_results.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_results.py.
1943         * Scripts/webkitpy/layout_tests/models/test_results_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py.
1944         * Scripts/webkitpy/layout_tests/port/chromium.py:
1945         * Scripts/webkitpy/layout_tests/views/printing.py:
1946         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
1947         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
1948         * Scripts/webkitpy/style_references.py:
1949         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
1950         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
1951         * Scripts/webkitpy/tool/bot/flakytestreporter_unittest.py:
1952         * Scripts/webkitpy/tool/commands/queues_unittest.py:
1953         * Scripts/webkitpy/tool/commands/rebaseline.py:
1954
1955 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
1956
1957         nrwt: fix typo preventing http startup on windows
1958         https://bugs.webkit.org/show_bug.cgi?id=64050
1959
1960         Reviewed by Eric Siedel.
1961
1962         NRWT checks whether it needs to start the servers by looking
1963         for "/http/" in the test names. After r90520, the leading slash
1964         is not present, so the check isn't quite right.
1965
1966         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1967         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1968
1969 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
1970
1971         nrwt: remove --use-apache from the command line
1972         https://bugs.webkit.org/show_bug.cgi?id=63358
1973
1974         Reviewed by Adam Barth.
1975
1976         Re-land the fix in bug 63358 - the prior fix was flawed in that
1977         we were trying to start the http server during check_sys_deps()
1978         but not actually setting up the environment properly. This
1979         broken prior to this change, but exposed by it, since the patch
1980         changes the chromium port to call the base class's checks.
1981
1982         * Scripts/webkitpy/common/system/executive.py:
1983         * Scripts/webkitpy/layout_tests/port/base.py:
1984         * Scripts/webkitpy/layout_tests/port/chromium.py:
1985         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1986         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1987         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1988         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
1989         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1990         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1991
1992 2011-07-06  Adam Barth  <abarth@webkit.org>
1993
1994         Move view-related code out of layout_package and into views
1995         https://bugs.webkit.org/show_bug.cgi?id=64051
1996
1997         Reviewed by Eric Seidel.
1998
1999         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2000         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
2001         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
2002         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
2003         * Scripts/webkitpy/layout_tests/layout_package/metered_stream.py: Removed.
2004         * Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py: Removed.
2005         * Scripts/webkitpy/layout_tests/layout_package/printing.py: Removed.
2006         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py: Removed.
2007         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2008         * Scripts/webkitpy/layout_tests/views: Added.
2009         * Scripts/webkitpy/layout_tests/views/__init__.py: Added.
2010         * Scripts/webkitpy/layout_tests/views/metered_stream.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/metered_stream.py.
2011         * Scripts/webkitpy/layout_tests/views/metered_stream_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py.
2012         * Scripts/webkitpy/layout_tests/views/printing.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/printing.py.
2013         * Scripts/webkitpy/layout_tests/views/printing_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py.
2014
2015 2011-07-06  Adam Barth  <abarth@webkit.org>
2016
2017         Add goofy comments to explain why we have these files.
2018
2019         * Scripts/webkitpy/layout_tests/controllers/__init__.py:
2020         * Scripts/webkitpy/layout_tests/layout_package/__init__.py:
2021
2022 2011-07-06  Adam Barth  <abarth@webkit.org>
2023
2024         Move a bunch of classes out of layout_tests.layout_package and into layout_tests.controllers
2025         https://bugs.webkit.org/show_bug.cgi?id=64048
2026
2027         Reviewed by Eric Seidel.
2028
2029         * Scripts/webkitpy/layout_tests/controllers: Added.
2030         * Scripts/webkitpy/layout_tests/controllers/__init__.py: Added.
2031         * Scripts/webkitpy/layout_tests/controllers/manager.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager.py.
2032         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py.
2033         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py.
2034         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py.
2035         * Scripts/webkitpy/layout_tests/controllers/message_broker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker.py.
2036         * Scripts/webkitpy/layout_tests/controllers/message_broker_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py.
2037         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py.
2038         * Scripts/webkitpy/layout_tests/controllers/worker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/worker.py.
2039         * Scripts/webkitpy/layout_tests/layout_package/manager.py: Removed.
2040         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py: Removed.
2041         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py: Removed.
2042         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py: Removed.
2043         * Scripts/webkitpy/layout_tests/layout_package/message_broker.py: Removed.
2044         * Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py: Removed.
2045         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
2046         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: Removed.
2047         * Scripts/webkitpy/layout_tests/layout_package/worker.py: Removed.
2048         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2049
2050 2011-07-06  Adam Barth  <abarth@webkit.org>
2051
2052         Attempt to fix unittest on Leopard.  Let's enumerate the keys in sorted
2053         order so the test results are predictable.
2054
2055         * Scripts/webkitpy/tool/bot/irc_command.py:
2056         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
2057
2058 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
2059
2060         Reviewed by Eric Siedel.
2061
2062         nrwt: make the code be consistent about using test names instead of filenames or paths
2063
2064         https://bugs.webkit.org/show_bug.cgi?id=63597
2065
2066         It can be hard to figure out where in the code we're assuming
2067         test names are listed as unix-style relative filenames and where
2068         they are either absolute paths or relative paths following the
2069         host filesystem convention.
2070
2071         This patch changes things so that everything outside of the
2072         Port object uses (and must assume) unix-style relative
2073         filenames (with one exception, which is specifying host-local
2074         filenames as a list of test arguments on the command line).
2075
2076         This should make things clearer and more consistent, and also
2077         removes a lot of path conversion calls.
2078
2079         The changes in this patch outside of port/base.py are largely
2080         cosmetic changes of "filename" to "test". We add a few routines
2081         to the base Port implementation to manipulate test names to
2082         split them, figure out if we're referring to a directory of
2083         tests, etc.
2084
2085         * Scripts/webkitpy/common/net/layouttestresults.py:
2086         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
2087         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2088         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2089         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
2090         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
2091         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
2092         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
2093         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
2094         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
2095         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
2096         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
2097         * Scripts/webkitpy/layout_tests/layout_package/test_input.py:
2098         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
2099         * Scripts/webkitpy/layout_tests/layout_package/test_results.py:
2100         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py:
2101         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
2102         * Scripts/webkitpy/layout_tests/port/base.py:
2103         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2104         * Scripts/webkitpy/layout_tests/port/chromium.py:
2105         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2106         * Scripts/webkitpy/layout_tests/port/dryrun.py:
2107         * Scripts/webkitpy/layout_tests/port/test.py:
2108         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2109         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2110         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
2111         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
2112         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
2113         * Scripts/webkitpy/tool/commands/rebaseline.py:
2114         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
2115
2116 2011-07-06  Chris Guillory  <ctguil@chromium.org>
2117
2118         No review necessary.
2119
2120         Adding myself to committers list in committers.py.
2121         https://bugs.webkit.org/show_bug.cgi?id=63478
2122
2123         * Scripts/webkitpy/common/config/committers.py:
2124
2125 2011-07-06  Johnny Ding  <jnd@chromium.org>
2126
2127         https://bugs.webkit.org/show_bug.cgi?id=63857
2128         Implement createPopupMenu for Chromium DRT.
2129
2130         Reviewed by Tony Chang.
2131
2132         * DumpRenderTree/chromium/WebViewHost.cpp:
2133         (WebViewHost::createPopupMenu):
2134         (WebViewHost::~WebViewHost):
2135         * DumpRenderTree/chromium/WebViewHost.h:
2136
2137 2011-07-06  Mihai Parparita  <mihaip@chromium.org>
2138
2139         Unreviewed, rolling out r90503.
2140         http://trac.webkit.org/changeset/90503
2141         https://bugs.webkit.org/show_bug.cgi?id=63358
2142
2143         Makes Chromium Windows canary bots hang.
2144
2145         * Scripts/webkitpy/layout_tests/port/base.py:
2146         * Scripts/webkitpy/layout_tests/port/chromium.py:
2147         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2148         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2149         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
2150         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
2151         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
2152         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2153
2154 2011-07-06  Adam Barth  <abarth@webkit.org>
2155
2156         Move buildbot URL in webkitpy to config.urls
2157         https://bugs.webkit.org/show_bug.cgi?id=64041
2158
2159         Reviewed by Eric Seidel.
2160
2161         This patch prepares for using the buildbot class to talk to the
2162         Chromium buildbots.
2163
2164         * Scripts/webkitpy/common/config/urls.py:
2165         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
2166         * Scripts/webkitpy/tool/commands/download_unittest.py:
2167         * Scripts/webkitpy/tool/commands/queries.py:
2168         * Scripts/webkitpy/tool/mocktool.py:
2169         * Scripts/webkitpy/tool/steps/ensurebuildersaregreen.py:
2170
2171 2011-07-06  Kevin Ollivier  <kevino@theolliviers.com>
2172
2173         [wx] Unreviewed build fix, add the correct path to the waf build settings.
2174         
2175         * wx/packaging/build-win-installer.py:
2176
2177 2011-07-06  Adam Barth  <abarth@webkit.org>
2178
2179         Add missing include.
2180
2181         * Scripts/webkitpy/tool/bot/irc_command.py:
2182
2183 2011-07-06  Adam Barth  <abarth@webkit.org>
2184
2185         Add roll-chromium-deps command to sheriff-bot
2186         https://bugs.webkit.org/show_bug.cgi?id=64037
2187
2188         Reviewed by Eric Seidel.
2189
2190         Lo, and the DEPS were rolled on command.
2191
2192         * Scripts/webkitpy/tool/bot/irc_command.py:
2193         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
2194         * Scripts/webkitpy/tool/bot/sheriff.py:
2195         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
2196         * Scripts/webkitpy/tool/commands/roll.py:
2197         * Scripts/webkitpy/tool/commands/roll_unittest.py:
2198         * Scripts/webkitpy/tool/steps/updatechromiumdeps.py:
2199
2200 2011-07-06  Eric Seidel  <eric@webkit.org>
2201
2202         Split Driver, DriverOutput and DriverInput out into a new driver.py file
2203         https://bugs.webkit.org/show_bug.cgi?id=64031
2204
2205         Reviewed by Adam Barth.
2206
2207         This ended up kinda large and I'm happy to break it up if desired.
2208
2209         As part of cleaning up the Port module, I split Driver, DriverInput and DriverOutput
2210         out into its own driver.py file.
2211
2212         I also cleaned up how we import Port objects, by making port/__init__.py
2213         expose Port, Driver, DriverInput, and DriverOutput.  I removed exposure of
2214         some various test objects as they were causing circular dependencies.
2215         base.py is now an implementation detail of port.  No longer is "base" referenced
2216         anywhere in the code.
2217
2218         I also made Driver.__init__ non-virtual so that sub-classers could all share
2219         the data storage on Driver.  This more closely matches how other subclasses
2220         work throughout our python code.
2221
2222         Standard line-unwrapping and doc-string removal as I went through the code.
2223
2224         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2225         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
2226         * Scripts/webkitpy/layout_tests/port/__init__.py:
2227         * Scripts/webkitpy/layout_tests/port/base.py:
2228         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2229         * Scripts/webkitpy/layout_tests/port/chromium.py:
2230         * Scripts/webkitpy/layout_tests/port/driver.py: Added.
2231         * Scripts/webkitpy/layout_tests/port/dryrun.py:
2232         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
2233         * Scripts/webkitpy/layout_tests/port/test.py:
2234         * Scripts/webkitpy/layout_tests/port/webkit.py:
2235         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2236         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
2237         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
2238
2239 2011-07-06  Adam Roben  <aroben@apple.com>
2240
2241         Clear TestFailures's getNumberOfFailingTests cache
2242
2243         Now that we support NRWT, we need to clear out old zero values for the NRWT bots that might
2244         have been cached by old versions of TestFailures.
2245
2246         Fixes <http://webkit.org/b/64029> TestFailures page still doesn't show NRWT bots if you've
2247         used it recently
2248
2249         Reviewed by Adam Barth.
2250
2251         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
2252         (Builder.prototype.getNumberOfFailingTests): Added a .version property to the cache data
2253         that must match the current version. Otherwise we'll ignore the cached data.
2254
2255 2011-06-24  Dirk Pranke  <dpranke@chromium.org>
2256
2257         Reviewed by Adam Barth.
2258
2259         nrwt: remove --use-apache from the command line
2260         https://bugs.webkit.org/show_bug.cgi?id=63358
2261
2262         This change removes the --use-apache command line argument. It
2263         was initially put in when we were trying to get the cygwin
2264         apache instance to work with Chromium win, but that code has
2265         bitrotted and doesn't work at all now.
2266
2267         Arguably we should remove all of the code to allow for a choice
2268         of web servers, but since we may still want to switch off of
2269         LigHTTPd at some point on Windows, I'll leave the rest in for
2270         now (or at least yank it in a different change.)
2271
2272         * Scripts/webkitpy/layout_tests/port/base.py:
2273         * Scripts/webkitpy/layout_tests/port/chromium.py:
2274         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2275         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2276         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
2277         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2278
2279 2011-07-06  Dimitri Glazkov  <dglazkov@chromium.org>
2280
2281         garden-o-matic should use a favicon to indicate current state.
2282         https://bugs.webkit.org/show_bug.cgi?id=64027
2283
2284         Reviewed by Adam Barth.
2285
2286         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-green.png: Added.
2287         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-red.png: Added.
2288         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html: Added blurb about icons.
2289         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js: Added setting icon state.
2290         * Scripts/webkitpy/tool/servers/gardeningserver.py: Allowed the use of favicons.
2291
2292 2011-07-06  Daniel Bates  <dbates@rim.com>
2293
2294         Fix webkitpy error after changeset 90483 <http://trac.webkit.org/changeset/90483>; Add missing
2295         parameter to _fetch_file_from_results lamdba function.
2296
2297         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2298
2299 2011-07-06  Kevin Ollivier  <kevino@theolliviers.com>
2300
2301         [wx] Unreviewed build fix. Generate the proper install_name for dylibs on Mac.
2302
2303         * wx/packaging/build-mac-installer.py:
2304
2305 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2306
2307         [GTK] Only set env. variables in start when strictly needed
2308         https://bugs.webkit.org/show_bug.cgi?id=64026
2309
2310         Reviewed by Eric Seidel.
2311
2312         * Scripts/webkitpy/layout_tests/port/gtk.py: add a
2313         setup_environ_for_server method in GtkPort and set as many
2314         env. variables as possible there.
2315
2316 2011-07-06  Adam Roben  <aroben@apple.com>
2317
2318         Teach TestFailures how to load, parse, and interpret NRWT test results
2319
2320         Fixes <http://webkit.org/b/61877> TestFailures page doesn't show testers that use
2321         new-run-webkit-tests
2322
2323         Reviewed by Adam Barth.
2324
2325         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
2326         (Builder.prototype.failureDiagnosisTextAndURL): Added support for the new 'flaky' failure
2327         type. For now we don't account for image-only flakes (but TestFailures doesn't deal with
2328         pixel tests at all currently).
2329         (Builder.prototype.getNumberOfFailingTests): Relaxed the regex used for parsing the number
2330         of failing tests from the buildbot output. Make sure not to count "new passes" (a new
2331         category introduced by NRWT) as failures.
2332
2333         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
2334         (LayoutTestResultsLoader.prototype.start): Moved a bunch of code from here to
2335         _fetchAndParseORWTResults. This function now attempts to load NRWT results, then falls back
2336         to loading ORWT results.
2337         (LayoutTestResultsLoader.prototype._fetchAndParseNRWTResults): Added. Tries to load and
2338         parse the full_results.json file for the given build.
2339         (LayoutTestResultsLoader.prototype._fetchAndParseORWTResults): Added. Code came from start.
2340         Fixed a bug along the way where we were sometimes calling the error callback instead of the
2341         success callback when all tests passed.
2342
2343         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NRWTResultsParser.js: Added.
2344         (NRWTResultsParser): Do-nothing constructor.
2345         (NRWTResultsParser.prototype.parse): Uses eval() (eww!) to get the test results out of the
2346         JS string, then iterates over all the tests in the results data and builds up a data
2347         structure matching what ORWTResultsParser returns.
2348
2349         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
2350         (Array.prototype.contains):
2351         (String.prototype.contains):
2352         Added these simple helper functions.
2353
2354         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
2355         NRWTResultsParser.js.
2356
2357 2011-07-06  Adam Roben  <aroben@apple.com>
2358
2359         Extract code to parse ORWT's results.html file into its own class
2360
2361         Prep work for fixing <http://webkit.org/b/61877> TestFailures page doesn't show testers that
2362         use new-run-webkit-tests
2363
2364         Reviewed by Adam Barth.
2365
2366         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
2367         (LayoutTestResultsLoader.prototype.start): Moved a bunch of code from here...
2368
2369         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ORWTResultsParser.js: Added.
2370         (ORWTResultsParser):
2371         (ORWTResultsParser.prototype.parse):
2372         ...to here.
2373
2374         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
2375         ORWTResultsParser.js.
2376
2377 2011-07-06  Eric Seidel  <eric@webkit.org>
2378
2379         Unreviewed.  Fix a unittest from hitting the network. :)
2380
2381         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2382
2383 2011-07-06  Eric Seidel  <eric@webkit.org>
2384
2385         Unreviewed.  Fixing an exception in test-webkitpy, caused by bug 64006.
2386
2387         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2388
2389 2011-07-06  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
2390
2391         Run an Xvfb instance to run the API tests on, like we do for the
2392         layout tests.
2393
2394         Reviewed by Xan Lopez.
2395
2396         * Scripts/run-gtk-tests:
2397
2398 2011-07-06  Daniel Bates  <dbates@webkit.org>
2399
2400         Make SCM unit tests faster
2401         https://bugs.webkit.org/show_bug.cgi?id=63883
2402
2403         Speeds up the SCM unit tests by a factor of 4. Currently, we create a
2404         mock SVN repo for each test_ method in SVNTest and GitTest and creating
2405         this repo is expensive.
2406
2407         Instead, it is significantly faster to create the mock SVN repo once
2408         and then perform a filesystem copy of it for each test_ method.
2409
2410         Note, Python 2.7's unittest module implements support for per class and
2411         per module setup and tear down methods which could be used to implement
2412         similar functionality. At the time of writing, test-webkitpy is designed
2413         to support Python 2.5. So, we can't take advantage of these Python 2.7
2414         features :(
2415
2416         Reviewed by Eric Seidel.
2417
2418         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
2419
2420 2011-07-06  Adam Barth  <abarth@webkit.org>
2421
2422         Remove unused parameter from _add_test in test_expectations.py
2423         https://bugs.webkit.org/show_bug.cgi?id=64008
2424
2425         Reviewed by Eric Seidel.
2426
2427         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
2428
2429 2011-07-06  Adam Barth  <abarth@webkit.org>
2430
2431         Use old-run-webkit-tests when --leaks is present
2432         https://bugs.webkit.org/show_bug.cgi?id=64012
2433
2434         Reviewed by Adam Roben.
2435
2436         We'll remove this branch once support of --leaks is added to
2437         new-run-webkit-tests.
2438
2439         * Scripts/run-webkit-tests:
2440
2441 2011-07-06  Benjamin Poulain  <benjamin@webkit.org>
2442
2443         [Qt] [WK2] Add touch mocking to Qt's MiniBrowser
2444         https://bugs.webkit.org/show_bug.cgi?id=63995
2445
2446         Reviewed by Andreas Kling.
2447
2448         The mouse events are intercepted through QCoreApplication::notify(),
2449         and fake touch events are generated for mouse events when necessary.
2450
2451         If touch events are received from the platform, we assume the current
2452         hardware has a touch screen and we stop generating fake events.
2453
2454         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2455         (isTouchEvent):
2456         (isMouseEvent):
2457         (MiniBrowserApplication::MiniBrowserApplication):
2458         (MiniBrowserApplication::notify):
2459         * MiniBrowser/qt/MiniBrowserApplication.h:
2460
2461 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2462
2463         Reviewed by Gustavo Noronha.
2464
2465         * Scripts/webkitpy/layout_tests/port/gtk.py: use Popen correctly.
2466
2467 2011-07-06  Eric Seidel  <eric@webkit.org>
2468
2469         webkit-patch failure-reason does not understand NRWT results
2470         https://bugs.webkit.org/show_bug.cgi?id=64006
2471
2472         Reviewed by Adam Barth.
2473
2474         Adam Barth tells me failure-reason should be deleted soon,
2475         but aroben's fancy new TestFailures/ page doesn't work for
2476         NRWT yet, and I needed to know when
2477         third-party-cookie-relaxing started failing.
2478
2479         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
2480         * Scripts/webkitpy/tool/commands/queries.py:
2481
2482 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2483
2484         [GTK] Add missing environment variables in NWRT
2485         https://bugs.webkit.org/show_bug.cgi?id=64004
2486
2487         Reviewed by Gustavo Noronha Silva.
2488
2489         Add missing environment variables needed by the GTK+ port.
2490
2491         * Scripts/webkitpy/layout_tests/port/gtk.py: ditto.
2492
2493 2011-07-06  Adam Roben  <aroben@apple.com>
2494
2495         Make run-api-tests output intelligible on the bots
2496
2497         run-api-tests's output on the bots was very hard to read for two reasons:
2498           1) It was mixing its own output with gtest's
2499           2) It was using ANSI escape sequences to print colored text, but the bots don't support
2500              that
2501
2502         Now, the --verbose flag turns off almost all of run-api-tests's own output so that gtest's
2503         output will not be obscured. We still print "Timeout" messages even in verbose mode, since
2504         gtest doesn't have any native support for timeouts. Also, when our output is being
2505         redirected to a file, we don't print ANSI escape sequences.
2506
2507         Fixes <http://webkit.org/b/63996> It's very hard to read run-api-tests output on the bots
2508
2509         Reviewed by Sam Weinig.
2510
2511         * Scripts/run-api-tests: Don't use the :constants interface of Term::ANSIColor. We want to
2512         use the colored function instead.
2513         (runAllTestsInSuite): Don't print out suite names in verbose mode; gtest will give us enough
2514         context that they aren't needed.
2515         (runTest): Don't print out test names or pass/fail messages in verbose mode; gtest will do
2516         that for us. Also, use the new possiblyColored function instead of always coloring output.
2517         (possiblyColored): Added. When printing to a tty, returns the string with the appropriate
2518         ANSI color escape sequences added. Otherwise just returns the string unmodified.
2519
2520 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2521
2522         Reviewed by Gustavo Noronha.
2523
2524         Launch Xvfb (mostly) the same way we were launching it in the
2525         bots, for maximum compatibility.
2526
2527         * Scripts/webkitpy/layout_tests/port/gtk.py:
2528
2529 2011-07-06  Andreas Kling  <kling@webkit.org>
2530
2531         [Qt][WK2] Split Qt API into two different web views (touch and desktop)
2532         https://bugs.webkit.org/show_bug.cgi?id=63950
2533
2534         Co-authored with Benjamin Poulain <benjamin@webkit.org>.
2535
2536         Reviewed by Kenneth Rohde Christiansen.
2537
2538         Port WTR/Qt to use QDesktopWebView and force a 800x600 viewport.
2539
2540         The Qt MiniBrowser now defaults to QDesktopWebView but can be started
2541         in QTouchWebView mode by passing -touch on the command line.
2542
2543         A lot of functionality has been removed and/or disabled from
2544         MiniBrowser until we reintroduce the necessary API's.
2545
2546         * MiniBrowser/qt/BrowserView.cpp:
2547         (BrowserView::BrowserView):
2548         (BrowserView::~BrowserView):
2549         (BrowserView::load):
2550         (BrowserView::view):
2551         (BrowserView::touchWebView):
2552         (BrowserView::desktopWebView):
2553         * MiniBrowser/qt/BrowserView.h:
2554         * MiniBrowser/qt/BrowserWindow.cpp:
2555         (newPageFunction):
2556         (BrowserWindow::BrowserWindow):
2557         (BrowserWindow::webView):
2558         (BrowserWindow::newWindow):
2559         (BrowserWindow::showUserAgentDialog):
2560         (BrowserWindow::updateUserAgentList):
2561         * MiniBrowser/qt/BrowserWindow.h:
2562         * MiniBrowser/qt/MiniBrowserApplication.cpp:
2563         (MiniBrowserApplication::handleUserOptions):
2564         * MiniBrowser/qt/MiniBrowserApplication.h:
2565         (WindowOptions::WindowOptions):
2566         * MiniBrowser/qt/UrlLoader.cpp:
2567         (UrlLoader::UrlLoader):
2568         * MiniBrowser/qt/main.cpp:
2569         (main):
2570         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
2571         (WTR::WebView::wkView):
2572         (WTR::WebView::pageRef):
2573         (WTR::WebView::WebView):
2574         (WTR::PlatformWebView::page):
2575         (WTR::PlatformWebView::windowFrame):
2576         (WTR::PlatformWebView::setWindowFrame):
2577
2578 2011-07-06  Eric Seidel  <eric@webkit.org>
2579
2580         Unreviewed..
2581
2582         * Scripts/run-webkit-tests:
2583         (useNewRunWebKitTests):
2584          - Fix the pass-thru of --gtk and --qt for old-run-webkit-tests,
2585            it was previously broken by the existence of isGtk/isQt checks which
2586            removed the --gtk/--qt args from @ARGV.
2587
2588 2011-07-06  Eric Seidel  <eric@webkit.org>
2589
2590         new-run-webkit-tests uses a 35s timeout for Mac and 6s for all other ports
2591         https://bugs.webkit.org/show_bug.cgi?id=63983
2592
2593         Reviewed by Xan Lopez.
2594
2595         All WebKit ports want to use a 35s timeout.  Only chromium currently
2596         uses a 6s timeout.  Eventually we'll move all WebKit ports to 6s too
2597         to further speed up new-run-webkit-tests runs, but that's going to be a while.
2598
2599         * Scripts/webkitpy/layout_tests/port/mac.py:
2600         * Scripts/webkitpy/layout_tests/port/webkit.py:
2601         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2602
2603 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2604
2605         Unreviewed.
2606
2607         NWRT actually seems to work locally, so it should work in the bots
2608         as long as we take care of the TIMEOUT issue (which is being fixed
2609         as I write this). Let's give this another shot and fix whatever
2610         breaks.
2611
2612         * Scripts/run-webkit-tests:
2613         (useNewRunWebKitTests):
2614
2615 2011-07-06  Xan Lopez  <xlopez@igalia.com>
2616
2617         Unreviewed.
2618
2619         Disable NWRT on GTK. At the very least it needs an updated exected
2620         results file since NWRT is more sensitive to timeouts, not sure
2621         what the exact problem is.
2622
2623         * Scripts/run-webkit-tests:
2624         (useNewRunWebKitTests):
2625
2626 2011-07-06  Adam Barth  <abarth@webkit.org>
2627
2628         Fix global variable leak detected by noglobals.
2629
2630         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
2631
2632 2011-07-06  Adam Barth  <abarth@webkit.org>
2633
2634         garden-o-matic shouldn't show flaky tests by default
2635         https://bugs.webkit.org/show_bug.cgi?id=63979
2636
2637         Reviewed by Ojan Vafai.
2638
2639         If any of the actual test results are PASS, the test is flaky and we
2640         can ignore it (at least for now).  If we get awesome at dealing with
2641         failing tests, we might want to tackle flaky tests by refering the user
2642         to the flakiness dashboard, which is way more awesome at dealing with
2643         flaky tests that this tool ever will be.
2644
2645         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
2646         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
2647
2648 2011-07-05  Adam Barth  <abarth@webkit.org>
2649
2650         garden-o-matic should list the failing tests
2651         https://bugs.webkit.org/show_bug.cgi?id=63978
2652
2653         Reviewed by Eric Seidel.
2654
2655         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js: Added.
2656         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
2657         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
2658         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
2659         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
2660         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
2661         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js: Added.
2662         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js: Added.
2663         * Scripts/webkitpy/tool/servers/gardeningserver.py:
2664
2665
2666 2011-07-05  Eric Seidel  <eric@webkit.org>
2667
2668         Re-enable NRWT on Leopard now that bug 63973 is fixed.
2669
2670         Reviewed by Adam Barth.
2671
2672         I also made the should-enable expression a black-list instead
2673         of a white-list since every bot on build.webkit.org is
2674         now using NRWT except windows and WebKit2 bots.
2675
2676         * Scripts/run-webkit-tests:
2677         (useNewRunWebKitTests):
2678
2679 2011-07-05  Eric Seidel  <eric@webkit.org>
2680
2681         Perl string concat failure on Leopard (causes NRWT to fail)
2682         https://bugs.webkit.org/show_bug.cgi?id=63973
2683
2684         Reviewed by Daniel Bates.
2685
2686         Leopard's "xcodebuild -version" has a different format from Snow Leopard/Lion
2687         thus the version parsing logic in webkitdirs.pm seems to fail.
2688
2689         Default to "3.0" instead of undef when we can't parse the version number.
2690         This should send us down the correct path for Leopard.
2691
2692         * Scripts/webkitdirs.pm:
2693         (determineBaseProductDir):
2694
2695 2011-07-05  Kevin Ollivier  <kevino@theolliviers.com>
2696
2697         [wx] Unreviewed build fix after addition of Source/WebCore/xml/parser dir.
2698
2699         * waf/build/settings.py:
2700
2701 2011-07-05  Eric Seidel  <eric@webkit.org>
2702
2703         Disable NRWT for Leopard until https://bugs.webkit.org/show_bug.cgi?id=63973 can be fixed.
2704
2705         Reviewed by Adam Barth.
2706
2707         * Scripts/run-webkit-tests:
2708         (useNewRunWebKitTests):
2709
2710 2011-07-05  Eric Seidel  <eric@webkit.org>
2711
2712         Make run-webkit-tests pass --verbose to new-run-webkit-tests on qt/gtk bots.
2713
2714         Reviewed by Adam Barth.
2715
2716         * Scripts/run-webkit-tests:
2717         (runningOnBuildBot):
2718
2719 2011-07-05  Adam Barth  <abarth@webkit.org>
2720
2721         Add results_unittests.js to garden-o-matic
2722         https://bugs.webkit.org/show_bug.cgi?id=63971
2723
2724         Reviewed by Eric Seidel.
2725
2726         Adding these unit tests forced me to commit to an API for the results
2727         module.  I'm not sure if this is the API we'll end up with, but it's
2728         something not entirely unreasonable.
2729
2730         Also, I've changed results.js over to WebKit style, as discussed previously.
2731
2732         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
2733         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
2734         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js: Added.
2735         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
2736
2737 2011-07-05  Eric Seidel  <eric@webkit.org>
2738
2739         Teach run-webkit-tests how to translate --qt and --gtk for new-run-webkit-tests.
2740
2741         Reviewed by Adam Barth.
2742
2743         * Scripts/run-webkit-tests:
2744
2745 2011-07-05  Adam Barth  <abarth@webkit.org>
2746
2747         Add first unit tests for garden-o-matic
2748         https://bugs.webkit.org/show_bug.cgi?id=63969
2749
2750         Reviewed by Eric Seidel.
2751
2752         I've also reformated base.js to match WebKit style, as discussed in the
2753         previous bug.
2754
2755         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
2756         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js: Added.
2757         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html: Added.
2758
2759 2011-07-05  Eric Seidel  <eric@webkit.org>
2760
2761         Move Qt, Gtk and Leopard to NRWT.
2762
2763         We'll roll out any of these if we see problems.
2764
2765         Reviewed by Adam Barth.
2766
2767         * Scripts/run-webkit-tests:
2768         (useNewRunWebKitTests):
2769
2770 2011-07-05  Adam Barth  <abarth@webkit.org>
2771
2772         garden-o-matic should understand which tests have unexpected failures
2773         https://bugs.webkit.org/show_bug.cgi?id=63965
2774
2775         Reviewed by Eric Seidel.
2776
2777         Currently, this code just logs the list of unexpected failures to the
2778         console.  A future patch will do something useful with this
2779         information.  A testing harness will also come in the next patch.
2780
2781         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js: Added.
2782         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
2783         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
2784         * Scripts/webkitpy/tool/servers/gardeningserver.py:
2785
2786 2011-07-05  Adam Barth  <abarth@webkit.org>
2787
2788         garden-o-matic should know how to fetch test results from the (NRWT) bots
2789         https://bugs.webkit.org/show_bug.cgi?id=63959
2790
2791         Reviewed by Eric Seidel.
2792
2793         Currently, this code just fetches the results and expectations JSON
2794         blob for Linux from the server and logs it to the console.
2795
2796         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
2797         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js: Added.
2798         * Scripts/webkitpy/tool/servers/gardeningserver.py:
2799
2800 2011-07-05  Dirk Pranke  <dpranke@chromium.org>
2801
2802         Reviewed by Eric Seidel.
2803
2804         Python tests are failing on leopard
2805         https://bugs.webkit.org/show_bug.cgi?id=63842
2806
2807         Fix the missing flush() call that appears to be needed by the
2808         logging package in python 2.5. Also fix the regressions
2809         introduced in the run_webkit_test tests when we switched from
2810         thread to processes (since processes aren't available on 2.5).
2811
2812         Also fix a missing "from __future__ import with_statement' in
2813         testfilehandler (unreviewed).
2814
2815         * Scripts/webkitpy/layout_tests/layout_package/metered_stream.py:
2816         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2817         * TestResultServer/handlers/testfilehandler.py:
2818
2819 2011-07-05  Eric Seidel  <eric@webkit.org>
2820
2821         Reviewed by Adam Barth.
2822
2823         re-enabling new-run-webkit-tests on the Snow Leopard WebKit1 bots
2824         now that we believe the stray http-server issue to be solved.
2825
2826         * Scripts/run-webkit-tests:
2827         (useNewRunWebKitTests):
2828
2829 2011-07-05  Adam Barth  <abarth@webkit.org>
2830
2831         Turns out we need to support numerals too because that's what jQuery
2832         uses by default.
2833
2834         * TestResultServer/handlers/testfilehandler.py:
2835
2836 2011-07-05  Dirk Pranke  <dpranke@chromium.org>
2837
2838         nrwt: allow for multiple http shards
2839         https://bugs.webkit.org/show_bug.cgi?id=63116
2840
2841         Reviewed by Tony Chang.
2842
2843         This modifies the sharding logic to support multiple http
2844         shards, but for now we clamp to one shard until we can test
2845         perf impact and flakiness impact.
2846
2847         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
2848         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
2849
2850 2011-07-05  Dirk Pranke  <dpranke@chromium.org>
2851
2852         Re-land nrwt: make sharding tests needing locks less hard-coded
2853         https://bugs.webkit.org/show_bug.cgi?id=63112
2854
2855         Reviewed by Ojan Vafai.
2856
2857         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
2858         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
2859
2860 2011-07-05  Adam Barth  <abarth@webkit.org>
2861
2862         TestResultServer should support callback parameter for JSON
2863         https://bugs.webkit.org/show_bug.cgi?id=63961
2864
2865         Reviewed by Ojan Vafai.
2866
2867         This makes using jQuery.ajax so much more pleasant.
2868
2869         * TestResultServer/handlers/testfilehandler.py:
2870
2871 2011-07-05  Eric Seidel  <eric@webkit.org>
2872
2873         buildbot needs to understand whether NRWT exited early after having too many failures
2874         https://bugs.webkit.org/show_bug.cgi?id=63839
2875
2876         Reviewed by Adam Barth.
2877
2878         Fix-up the exited early messages printed by NRWT so that
2879         the buildbot can parse them as expected.
2880         It looks for lines using "if line.find('Exiting early') >= 0:"
2881
2882         I also plumbed the "early exit" status through to results.json
2883         in the form of an "interrupted" bool.  It was unclear to me
2884         if results.json already had enough information to compute this bool
2885         itself.  It's possible Ojan could come up with a better fix.
2886
2887         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
2888         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
2889         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
2890         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
2891
2892 2011-07-05  Eric Seidel  <eric@webkit.org>
2893
2894         new-run-webkit-tests fails to start http server if one is already running
2895         https://bugs.webkit.org/show_bug.cgi?id=63956
2896
2897         Reviewed by Adam Barth.
2898
2899         * BuildSlaveSupport/kill-old-processes:
2900          - Removed ^M line endings and sorted the process names.
2901          - There are no functional changes to this file.
2902            (except that now that its using unix line endings it can be executed directly!)
2903         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
2904          - Change the name to 'httpd' to match old-run-webkit-test paths.
2905
2906 2011-07-05  Adam Barth  <abarth@webkit.org>
2907
2908         Add basic ajax support to garden-o-matic
2909         https://bugs.webkit.org/show_bug.cgi?id=63874
2910
2911         Reviewed by Eric Seidel.
2912
2913         This patch adds some basic infrastructure to garden-o-matic.
2914         Currently, the infrastructure is just used to make the "quit" command
2915         use Ajax, but in the future, this infrastructure will be used to do
2916         more sophistocated remote proceedure calls.
2917
2918         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
2919         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js: Added.
2920         * Scripts/webkitpy/tool/servers/gardeningserver.py:
2921         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
2922
2923 2011-07-05  Adam Barth  <abarth@webkit.org>
2924
2925         Add trivial garden-o-matic command (with server)
2926         https://bugs.webkit.org/show_bug.cgi?id=63872
2927
2928         Reviewed by Eric Seidel.
2929
2930         This patch adds a trivial "Hello, world" garden-o-matic command,
2931         complete with an HTTP server.  This command re-uses a bunch of code
2932         from the existing rebaseline-server command.  Over time, this command
2933         will grow to be a tool useful for gardening the WebKit tree.
2934
2935         * Scripts/webkitpy/tool/commands/__init__.py:
2936         * Scripts/webkitpy/tool/commands/abstractservercommand.py: Added.
2937         * Scripts/webkitpy/tool/commands/gardenomatic.py: Added.
2938         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
2939         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html: Added.
2940         * Scripts/webkitpy/tool/servers/gardeningserver.py: Added.
2941         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
2942
2943 2011-07-05  Adam Roben  <aroben@apple.com>
2944
2945         Make prepare-ChangeLog include modified Perl functions in its ChangeLog template
2946
2947         This is a very simple first cut. Functions must start with a line that starts with "sub "
2948         and end with a line that starts with a closing brace. No leading whitespace is allowed.
2949         Package names aren't parsed at all.
2950
2951         Fixes <http://webkit.org/b/21591> prepare-ChangeLog should know how to find functions in
2952         Perl files
2953
2954         Reviewed by David Kilzer.
2955
2956         * Scripts/prepare-ChangeLog: Removed redundant code that ignored certain files when
2957         searching for function line ranges. This is already done inside the get_function_line_ranges
2958         function.
2959         (get_function_line_ranges): Cleaned up coding style a little bit. Call
2960         get_function_line_ranges_for_perl for files with .pl and .pm extensions. For files with an
2961         unknown extension or no extension, read the shebang line to try to determine the script
2962         interpreter. Call get_function_line_ranges_for_perl if the interpreter seems to be Perl.
2963         (get_function_line_ranges_for_perl): Added. Does extremely basic parsing of the file to find
2964         lines starting with "sub " or "}".
2965
2966 2011-07-05  Adam Roben  <aroben@apple.com>
2967
2968         Clean up run-api-tests output on Windows
2969
2970         We were mixing run-api-tests output with gtest output, and the result was a mess.
2971
2972         Fixes <http://webkit.org/b/63954> run-api-tests output is very confusing on Windows
2973
2974         Reviewed by Dan Bates.
2975
2976         * Scripts/run-api-tests:
2977         (runTest):
2978         (populateTests):
2979         Made Windows use the formerly-Mac-only codepaths which correctly suppress gtest output
2980         except in verbose mode. Most changes are due to indentation. Use of the arch utility is now
2981         the only Mac-specific thing in these functions. Some other platform-specific code was moved
2982         from here...
2983
2984         (prepareEnvironmentForRunningTestTool):
2985         (testToolPath):
2986         ...to here.
2987
2988 2011-07-04  Gavin Barraclough  <barraclough@apple.com>
2989
2990         Reviewed by Sam Weinig.
2991
2992         https://bugs.webkit.org/show_bug.cgi?id=16652
2993         Firefox and JavaScriptCore differ in Number.toString(integer)
2994
2995         Added forwarding header.
2996
2997         * DumpRenderTree/ForwardingHeaders/wtf/StdLibExtras.h: Added.
2998
2999 2011-07-04  Anders Carlsson  <andersca@apple.com>
3000
3001         Try to fix Qt build.
3002
3003         * DumpRenderTree/TestNetscapePlugIn/Tests/NPRuntimeRemoveProperty.cpp:
3004
3005 2011-07-04  Anders Carlsson  <andersca@apple.com>
3006
3007         NP_RemoveProperty is not called back by Safari when delete npObject.prop is encountered in JavaScript
3008         https://bugs.webkit.org/show_bug.cgi?id=63915
3009         <rdar://problem/7124300>
3010
3011         Reviewed by Sam Weinig.
3012
3013         Extend the NPRuntimeRemoveProperty to handle delete object.property from JavaScript.
3014
3015         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
3016         (PluginTest::NPN_IdentifierIsString):
3017         (PluginTest::NPN_UTF8FromIdentifier):
3018         (PluginTest::NPN_IntFromIdentifier):
3019         (PluginTest::NPN_RetainObject):
3020         (PluginTest::NPN_ReleaseObject):
3021         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
3022         (PluginTest::Object::removeProperty):
3023         (PluginTest::Object::identifierIs):
3024         (PluginTest::Object::NP_RemoveProperty):
3025         (PluginTest::Object::npClass):
3026         * DumpRenderTree/TestNetscapePlugIn/Tests/NPRuntimeRemoveProperty.cpp:
3027         (NPRuntimeRemoveProperty::TestObject::TestObject):
3028         (NPRuntimeRemoveProperty::TestObject::hasProperty):
3029         (NPRuntimeRemoveProperty::TestObject::getProperty):
3030         (NPRuntimeRemoveProperty::TestObject::removeProperty):
3031         (NPRuntimeRemoveProperty::PluginObject::PluginObject):
3032         (NPRuntimeRemoveProperty::PluginObject::~PluginObject):
3033         (NPRuntimeRemoveProperty::PluginObject::hasMethod):
3034         (NPRuntimeRemoveProperty::PluginObject::invoke):
3035         (NPRuntimeRemoveProperty::PluginObject::hasProperty):
3036         (NPRuntimeRemoveProperty::PluginObject::getProperty):
3037         (NPRuntimeRemoveProperty::NPP_GetValue):
3038
3039 2011-07-04  Leandro Pereira  <leandro@profusion.mobi>
3040
3041         Reviewed by Kent Tamura.
3042
3043         [EFL] Add LayoutTestControllerEfl.
3044         https://bugs.webkit.org/show_bug.cgi?id=61974
3045
3046         Only the core features are implemented. Most functions are actually
3047         stubs and will be implemented as soon as its features are needed.
3048         Build system changes will be made as soon as DRT is completely
3049         upstreamed.
3050
3051         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp: Added.
3052         (LayoutTestController::~LayoutTestController):
3053         (LayoutTestController::addDisallowedURL):
3054         (LayoutTestController::clearBackForwardList):
3055         (LayoutTestController::copyDecodedHostName):
3056         (LayoutTestController::copyEncodedHostName):
3057         (LayoutTestController::dispatchPendingLoadRequests):
3058         (LayoutTestController::display):
3059         (LayoutTestController::counterValueForElementById):
3060         (LayoutTestController::keepWebHistory):
3061         (LayoutTestController::computedStyleIncludingVisitedInfo):
3062         (LayoutTestController::nodesFromRect):
3063         (LayoutTestController::layerTreeAsText):
3064         (LayoutTestController::pageNumberForElementById):
3065         (LayoutTestController::numberOfPages):
3066         (LayoutTestController::pageProperty):
3067         (LayoutTestController::isPageBoxVisible):
3068         (LayoutTestController::pageSizeAndMarginsInPixels):
3069         (LayoutTestController::webHistoryItemCount):
3070         (LayoutTestController::workerThreadCount):
3071         (LayoutTestController::notifyDone):
3072         (LayoutTestController::pathToLocalResource):
3073         (LayoutTestController::queueLoad):
3074         (LayoutTestController::setAcceptsEditing):
3075         (LayoutTestController::setAlwaysAcceptCookies):
3076         (LayoutTestController::setCustomPolicyDelegate):
3077         (LayoutTestController::waitForPolicyDelegate):
3078         (LayoutTestController::setScrollbarPolicy):
3079         (LayoutTestController::addOriginAccessWhitelistEntry):
3080         (LayoutTestController::removeOriginAccessWhitelistEntry):
3081         (LayoutTestController::setMainFrameIsFirstResponder):
3082         (LayoutTestController::setTabKeyCyclesThroughElements):
3083         (LayoutTestController::setUseDashboardCompatibilityMode):
3084         (LayoutTestController::setUserStyleSheetEnabled):
3085         (LayoutTestController::setUserStyleSheetLocation):
3086         (LayoutTestController::setValueForUser):
3087         (LayoutTestController::setViewModeMediaFeature):
3088         (LayoutTestController::setWindowIsKey):
3089         (LayoutTestController::setSmartInsertDeleteEnabled):
3090         (waitToDumpWatchdogFired):
3091         (LayoutTestController::setWaitToDump):
3092         (LayoutTestController::windowCount):
3093         (LayoutTestController::setPrivateBrowsingEnabled):
3094         (LayoutTestController::setJavaScriptCanAccessClipboard):
3095         (LayoutTestController::setXSSAuditorEnabled):
3096         (LayoutTestController::setFrameFlatteningEnabled):
3097         (LayoutTestController::setSpatialNavigationEnabled):
3098         (LayoutTestController::setAllowUniversalAccessFromFileURLs):
3099         (LayoutTestController::setAllowFileAccessFromFileURLs):
3100         (LayoutTestController::setAuthorAndUserStylesEnabled):
3101         (LayoutTestController::setAutofilled):
3102         (LayoutTestController::disableImageLoading):
3103         (LayoutTestController::setMockDeviceOrientation):
3104         (LayoutTestController::setMockGeolocationPosition):
3105         (LayoutTestController::setMockGeolocationError):
3106         (LayoutTestController::setGeolocationPermission):
3107         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
3108         (LayoutTestController::addMockSpeechInputResult):
3109         (LayoutTestController::setIconDatabaseEnabled):
3110         (LayoutTestController::setJavaScriptProfilingEnabled):
3111         (LayoutTestController::setSelectTrailingWhitespaceEnabled):
3112         (LayoutTestController::setPopupBlockingEnabled):
3113         (LayoutTestController::setPluginsEnabled):
3114         (LayoutTestController::elementDoesAutoCompleteForElementWithId):
3115         (LayoutTestController::execCommand):
3116         (LayoutTestController::findString):
3117         (LayoutTestController::isCommandEnabled):
3118         (LayoutTestController::setCacheModel):
3119         (LayoutTestController::setPersistentUserStyleSheetLocation):
3120         (LayoutTestController::clearPersistentUserStyleSheet):
3121         (LayoutTestController::clearAllApplicationCaches):
3122         (LayoutTestController::setApplicationCacheOriginQuota):
3123         (LayoutTestController::clearApplicationCacheForOrigin):
3124         (LayoutTestController::localStorageDiskUsageForOrigin):
3125         (LayoutTestController::originsWithApplicationCache):
3126         (LayoutTestController::applicationCacheDiskUsageForOrigin):
3127         (LayoutTestController::clearAllDatabases):
3128         (LayoutTestController::setDatabaseQuota):
3129         (LayoutTestController::originsWithLocalStorage):
3130         (LayoutTestController::deleteAllLocalStorage):
3131         (LayoutTestController::deleteLocalStorageForOrigin):
3132         (LayoutTestController::observeStorageTrackerNotifications):
3133         (LayoutTestController::syncLocalStorage):
3134         (LayoutTestController::setDomainRelaxationForbiddenForURLScheme):
3135         (LayoutTestController::goBack):
3136         (LayoutTestController::setDefersLoading):
3137         (LayoutTestController::setAppCacheMaximumSize):
3138         (LayoutTestController::pauseAnimationAtTimeOnElementWithId):
3139         (LayoutTestController::pauseTransitionAtTimeOnElementWithId):
3140         (LayoutTestController::sampleSVGAnimationForElementAtTime):
3141         (LayoutTestController::numberOfActiveAnimations):
3142         (LayoutTestController::suspendAnimations):
3143         (LayoutTestController::resumeAnimations):
3144         (LayoutTestController::overridePreference):
3145         (LayoutTestController::addUserScript):
3146         (LayoutTestController::addUserStyleSheet):
3147         (LayoutTestController::setDeveloperExtrasEnabled):
3148         (LayoutTestController::setAsynchronousSpellCheckingEnabled):
3149         (LayoutTestController::showWebInspector):
3150         (LayoutTestController::closeWebInspector):
3151         (LayoutTestController::evaluateInWebInspector):
3152         (LayoutTestController::evaluateScriptInIsolatedWorld):
3153         (LayoutTestController::removeAllVisitedLinks):
3154         (LayoutTestController::callShouldCloseOnWebView):
3155         (LayoutTestController::apiTestNewWindowDataLoadBaseURL):
3156         (LayoutTestController::apiTestGoToCurrentBackForwardItem):
3157         (LayoutTestController::setWebViewEditable):
3158         (LayoutTestController::markerTextForListItem):
3159         (LayoutTestController::authenticateSession):
3160         (LayoutTestController::setEditingBehavior):
3161         (LayoutTestController::abortModal):
3162         (LayoutTestController::hasSpellingMarker):
3163         (LayoutTestController::hasGrammarMarker):
3164         (LayoutTestController::dumpConfigurationForViewport):
3165         (LayoutTestController::setSerializeHTTPLoads):
3166         (LayoutTestController::setMinimumTimerInterval):
3167         (LayoutTestController::setTextDirection):
3168         (LayoutTestController::allowRoundingHacks):
3169
3170 2011-06-24  Vsevolod Vlasov  <vsevik@chromium.org>
3171
3172         Unreviewed, add myself as committer.
3173
3174         * Scripts/webkitpy/common/config/committers.py:
3175
3176 2011-07-03  Sheriff Bot  <webkit.review.bot@gmail.com>
3177
3178         Unreviewed, rolling out r90347.
3179         http://trac.webkit.org/changeset/90347
3180         https://bugs.webkit.org/show_bug.cgi?id=63886
3181
3182         Build breaks on Leopard, Chromium-win, WinCairo, and WinCE.
3183         (Requested by tkent on #webkit).
3184
3185         * DumpRenderTree/ForwardingHeaders/wtf/StdLibExtras.h: Removed.
3186
3187 2011-07-03  Daniel Bates  <dbates@webkit.org>
3188
3189         Import Checkout in scm_unittests.py and logging in svn.py to fix
3190         Python missing global variable errors
3191
3192         Rubber-stamped by Adam Barth.
3193
3194         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py: Import Checkout.
3195         * Scripts/webkitpy/common/checkout/scm/svn.py: Import logging.
3196
3197 2011-07-02  Gavin Barraclough  <barraclough@apple.com>
3198
3199         Reviewed by Sam Weinig.
3200
3201         https://bugs.webkit.org/show_bug.cgi?id=16652
3202         Firefox and JavaScriptCore differ in Number.toString(integer)
3203
3204         Added forwarding header.
3205
3206         * DumpRenderTree/ForwardingHeaders/wtf/StdLibExtras.h: Added.
3207
3208 2011-04-02  Robert Hogan  <robert@webkit.org>
3209
3210         Reviewed by Benjamin Poulain.
3211
3212         [Qt] Support third-party cookie policy for Qt clients
3213         https://bugs.webkit.org/show_bug.cgi?id=45455
3214
3215         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
3216         (LayoutTestController::reset):
3217         (LayoutTestController::setAlwaysAcceptCookies):
3218         (LayoutTestController::setAlwaysBlockCookies):
3219         * DumpRenderTree/qt/LayoutTestControllerQt.h:
3220
3221 2011-07-03  Adam Barth  <abarth@webkit.org>
3222
3223         Factor ReflectionHandler out of the rebaseline server.  Just code motion.
3224
3225         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
3226         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
3227         * Scripts/webkitpy/tool/servers/reflectionhandler.py: Added.
3228
3229 2011-07-03  Adam Barth  <abarth@webkit.org>
3230
3231         Move the rebaseline server out of the commands package and into the
3232         (new) servers package.  This patch prepares for refactoring this code
3233         to share server infrastructure with the (forthcoming) garden-o-matic
3234         command.
3235
3236         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
3237         * Scripts/webkitpy/tool/servers/__init__.py: Added.
3238         * Scripts/webkitpy/tool/servers/data/rebaselineserver/loupe.js:
3239         * Scripts/webkitpy/tool/servers/data/rebaselineserver/main.css:
3240         * Scripts/webkitpy/tool/servers/data/rebaselineserver/queue.js:
3241         * Scripts/webkitpy/tool/servers/rebaselineserver.py: Added.
3242
3243 2011-07-01  Adam Barth  <abarth@webkit.org>
3244
3245         Fix typo in master.cfg, hoping to trigger a master restart.
3246
3247         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
3248
3249 2011-07-01  Adam Barth  <abarth@webkit.org>
3250
3251         Turn off new-run-webkit-tests and remove some debug logging.
3252
3253         * Scripts/run-webkit-tests:
3254         * Scripts/webkitpy/layout_tests/port/http_lock.py:
3255
3256 2011-07-01  Adam Barth  <abarth@webkit.org>
3257
3258         Add some more debug logging to try to understand why we're getting
3259         stuck acquring the HTTP lock.
3260
3261         * Scripts/webkitpy/layout_tests/port/http_lock.py:
3262
3263 2011-07-01  Adam Barth  <abarth@webkit.org>
3264
3265         Refactor NRWT parsing logic in master.cfg to understand that run-webkit-tests might run NRWT
3266         https://bugs.webkit.org/show_bug.cgi?id=63854
3267
3268         Reviewed by Dirk Pranke.
3269
3270         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
3271
3272 2011-07-01  Eric Seidel  <eric@webkit.org>
3273
3274         new-run-webkit-tests results does not understand that mac uses test_expectations files
3275         https://bugs.webkit.org/show_bug.cgi?id=63838
3276
3277         Reviewed by Adam Barth.
3278
3279         We've decided to "enable" test_expectations for the Mac port for now.
3280         It makes the results.html results much more confusing to read, but at least they're
3281         no longer lying to us.
3282
3283         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3284         * Scripts/webkitpy/layout_tests/port/base.py:
3285         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3286
3287 2011-07-01  Adam Barth  <abarth@webkit.org>
3288
3289         Temporarily disable waiting for ReportCrash to see how this code
3290         affects performance on the bots.
3291
3292         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3293         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
3294
3295 2011-07-01  Dmitry Lomov  <dslomov@google.com>
3296
3297         Reviewed by Adam Roben.
3298
3299         https://bugs.webkit.org/show_bug.cgi?id=63818
3300         Test bots shouldn't attempt to build TestWebKitAPI on Windows.
3301         TestWebKitAPI builds as a part of general webkit build and is a
3302         part of build products archive.
3303
3304         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
3305         * Scripts/run-api-tests:
3306
3307 2011-07-01  Adam Barth  <abarth@webkit.org>
3308
3309         We shouldn't wait for ReportCrash while the timeout is running
3310         https://bugs.webkit.org/show_bug.cgi?id=63837
3311
3312         Reviewed by Eric Seidel.
3313
3314         Instead, we should move this check to worker.py.
3315
3316         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
3317         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
3318
3319 2011-07-01  Adam Roben  <aroben@apple.com>
3320
3321         Stop setting unsafe headers on POSTs via XHR from TestFailures page
3322
3323         Fixes <http://webkit.org/b/63833> TestFailures page causes a bunch of spew in the
3324         Inspector's console about unsafe headers
3325
3326         Reviewed by Dan Bates.
3327
3328         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
3329         (fetchResource): Removed unnecessary code to set Content-Length and Connection headers.
3330
3331 2011-07-01  Eric Seidel  <eric@webkit.org>
3332
3333         Reviewed by Adam Barth.
3334
3335         Switch Snow Leopard over to new-run-webkit-tests (except for wk2)
3336         https://bugs.webkit.org/show_bug.cgi?id=63786
3337
3338         We're down to one bug blocking the "move to NRWT" bug:
3339         https://bugs.webkit.org/show_bug.cgi?id=34984
3340         so we're trying transitioning Snow Leopard to NRWT as an experiment.
3341
3342         If we run into any blocking issues we'll of course roll this back out
3343         (or anyone reading this should feel encouraged to).
3344
3345         * Scripts/run-webkit-tests:
3346
3347 2011-07-01  Adam Barth  <abarth@webkit.org>
3348
3349         Add a compat shim for --leaks.  Support for --leaks will be added in
3350         https://bugs.webkit.org/show_bug.cgi?id=63832.
3351
3352         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3353
3354 2011-07-01  Adam Barth  <abarth@webkit.org>
3355
3356         Reviewed by Darin Adler.
3357
3358         ReportCrash destabilizes new-run-webkit-tests
3359         https://bugs.webkit.org/show_bug.cgi?id=63795
3360
3361         ReportCrash chews up too many system resources that it destablizes
3362         tests run concurrently.  This patch causes us to spin down all the
3363         worker processes when ReportCrash is running.
3364
3365         Also, this patch causes the master process to tell the user we're
3366         waiting for the crash reporter rather than having that be the worker's
3367         job.  This stops the user from getting spammed with too many "waiting
3368         for crash reporter" messages and also cleans up some minor UI issues
3369         with the meter.
3370
3371         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3372         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
3373         * Scripts/webkitpy/layout_tests/port/base.py:
3374         * Scripts/webkitpy/layout_tests/port/server_process.py:
3375
3376 2011-07-01  Adam Roben  <aroben@apple.com>
3377
3378         Teach webkitpy about the new format of our ChangeLog template
3379
3380         r90229 moved the "Need a short description and bug URL" line to the top of the ChangeLog
3381         template. But webkitpy didn't know this, so the various rollout-related commands were
3382         leaving that line in the ChangeLog.
3383
3384         Fixes <http://webkit.org/b/63815> REGRESSION (r90229): webkit-patch rollout and sheriffbot
3385         rollout are broken
3386
3387         Reviewed by Dan Bates.
3388
3389         * Scripts/webkitpy/common/checkout/changelog.py:
3390         (ChangeLog.update_with_unreviewed_message): Changed to look for the "Need a short
3391         description and bug URL" line instead of the "Reviewed by" line as the first line of the
3392         ChangeLog template.
3393
3394         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
3395         (ChangeLogTest): Updated _new_entry_boilerplate to match what prepare-ChangeLog now
3396         generates.
3397
3398 2011-07-01  Adam Roben  <aroben@apple.com>
3399
3400         Make TestFailures stop loading old builds once we've explained all current failures
3401
3402         We used to do this, but some refactoring in r90165 broke this logic.
3403
3404         Fixes <http://webkit.org/b/63808> REGRESSION (r90165): TestFailures page always loads all
3405         builds back to the beginning of time
3406
3407         Reviewed by Dan Bates.
3408
3409         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
3410         (LayoutTestHistoryAnalyzer.prototype.start): Stop fetching old builds once we're told that
3411         all current failures have been explained.
3412
3413 2011-07-01  Adam Roben  <aroben@apple.com>
3414
3415         commit-log-editor reorders ChangeLog entries in unexpected ways
3416         https://bugs.webkit.org/show_bug.cgi?id=63804
3417
3418         Reviewed by Darin Adler.
3419
3420         commit-log-editor was changed in r46899 to move the bug title and URL to the top of the
3421         commit message in order to make git-based tools, which expect the first line of the commit
3422         message to be a summary of the change, to work better. But this reordering of the ChangeLog
3423         entry was unpredictable and unexpected.
3424
3425         I changed prepare-ChangeLog to put the bug title and URL at the top of the ChangeLog entry
3426         so that commit-log-editor doesn't have to move them. Then I changed commit-log-editor not to
3427         try to move them.
3428
3429         * Scripts/commit-log-editor: Partially reverted r46899. We no longer save the "Reviewed by"
3430         line and try to move it down below the bug title and URL. The order of the text in the
3431         ChangeLog is now preserved, except that we insert a "Patch by" line just above the "Reviewed
3432         by" line in cases where the committer and author are not the same person.
3433
3434         * Scripts/prepare-ChangeLog: Moved the "Reviewed by" line down below the bug title and URL.
3435
3436 2011-07-01  Mihnea Ovidenie  <mihnea@adobe.com>
3437
3438         Reviewed by Adam Roben.
3439
3440         [CSSRegions] Add build slave
3441         https://bugs.webkit.org/show_bug.cgi?id=62372
3442
3443         * BuildSlaveSupport/build.webkit.org-config/config.json:
3444         Add an entry for a release build of Mac.
3445
3446 2011-07-01  Sheriff Bot  <webkit.review.bot@gmail.com>
3447
3448         Unreviewed, rolling out r90213.
3449         http://trac.webkit.org/changeset/90213
3450         https://bugs.webkit.org/show_bug.cgi?id=63792
3451
3452         The SL bot has an errant HTTP server holding port 8080
3453         (Requested by abarth on #webkit).
3454
3455         * Scripts/run-webkit-tests:
3456
3457 2011-07-01  Carlos Garcia Campos  <cgarcia@igalia.com>
3458
3459         Unreviewed. Fix WebKit2 GTK build after r90163.
3460
3461         * MiniBrowser/gtk/WebBundle/WebBundleMain.c:
3462         (didCreatePage):
3463
3464 2011-07-01  Adam Barth  <abarth@webkit.org>
3465
3466         Exit faster when we get a keyboard interrupt.  Previously, we'd open
3467         Safari if you ^C NRWT after it found a failure.
3468
3469         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3470
3471 2011-07-01  Eric Seidel  <eric@webkit.org>
3472
3473         Reviewed by Adam Barth.
3474
3475         Switch Snow Leopard over to new-run-webkit-tests (except for wk2)
3476         https://bugs.webkit.org/show_bug.cgi?id=63786
3477
3478         We're down to one bug blocking the "move to NRWT" bug:
3479         https://bugs.webkit.org/show_bug.cgi?id=34984
3480         so we're trying transitioning Snow Leopard to NRWT as an experiment.
3481
3482         If we run into any blocking issues we'll of course roll this back out
3483         (or anyone reading this should feel encouraged to).
3484
3485         * Scripts/run-webkit-tests:
3486
3487 2011-07-01  Sheriff Bot  <webkit.review.bot@gmail.com>
3488
3489         Unreviewed, rolling out r90192.
3490         http://trac.webkit.org/changeset/90192
3491         https://bugs.webkit.org/show_bug.cgi?id=63788
3492
3493         Appears to have caused NRWT on Chromium WebKit Vista to hang
3494         (Requested by abarth on #webkit).
3495
3496         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3497         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
3498
3499 2011-06-30  Adam Barth  <abarth@webkit.org>
3500
3501         Reviewed by Eric Seidel.
3502
3503         Remove the concept of "being wedged" from new-run-webkit-tests
3504         https://bugs.webkit.org/show_bug.cgi?id=63767
3505
3506         Worker processes shouldn't ever become wedged.  My understanding is
3507         that this code was originally motivated by the old threading-based
3508         design but no longer servers any purpose.
3509
3510         Note: If we actually have a problem with the test harness getting
3511         stuck, buildbot will kill us.
3512
3513         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3514         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
3515         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
3516
3517 2011-06-30  Adam Barth  <abarth@webkit.org>
3518
3519         The win32 implementation of fork is sad about forking test-webkitpy.
3520         Hopefully this will heal the bot.
3521
3522         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3523
3524 2011-06-30  Eric Seidel  <eric@webkit.org>
3525
3526         Reviewed by Adam Barth.
3527
3528         Update new-run-webkit-tests now that the wk2 shared skipped list has moved
3529         https://bugs.webkit.org/show_bug.cgi?id=63780
3530
3531         The shared list was moved from mac-wk2 to wk2 as part of
3532         http://trac.webkit.org/changeset/90125.
3533
3534         * Scripts/webkitpy/layout_tests/port/webkit.py:
3535         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3536
3537 2011-06-30  Kenichi Ishibashi  <bashi@chromium.org>
3538
3539         Reviewed by Darin Adler.
3540
3541         Style Checker should flag removal of "developmentRegion = English;" from project.pbxproj.
3542         https://bugs.webkit.org/show_bug.cgi?id=62022
3543
3544         Adds XcodeProjectFileChecker to check removal of "developmentRegion".
3545
3546         * Scripts/webkitpy/style/checker.py: Added XcodeProjectFileChecker as a checker for *.pbxproj.
3547         * Scripts/webkitpy/style/checkers/xcodeproj.py: Added.
3548         * Scripts/webkitpy/style/checkers/xcodeproj_unittest.py: Added.
3549
3550 2011-06-30  Dirk Pranke  <dpranke@chromium.org>
3551
3552         Reviewed by Ojan Vafai.
3553
3554         nrwt: make sharding tests needing locks less hard-coded
3555         https://bugs.webkit.org/show_bug.cgi?id=63112
3556
3557         This change also changes the manager logic so that it will
3558         drop the server lock as soon as all of the shards requiring
3559         the lock have completed.
3560
3561         This change includes some minor namespace/import changes in the
3562         unit tests, and also makes the Manager a new-style object, which
3563         it should've been all along.
3564
3565         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3566         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
3567