369940de9fe93bcf9127f03ac3652b32f0e306c8
[WebKit-https.git] / Tools / ChangeLog
1 2011-06-07  Martin Robinson  <mrobinson@igalia.com>
2
3         Reviewed by Gustavo Noronha Silva.
4
5         [GTK] Use WebCore scrollbars for the DRT instead of GtkScrolledWindow
6         https://bugs.webkit.org/show_bug.cgi?id=62252
7
8         Add a new subclass of WebKitWebView that overrides the size request to
9         return 1 pixel by 1 pixel. This ensure that it operates properly when
10         not packed into a GtkScrolledWindow. Also construct the WebKitWebView
11         with self-scrolling set to true.
12
13         * DumpRenderTree/gtk/DumpRenderTree.cpp:
14         (webInspectorInspectWebView): Use the new SelfScrollingWebKitWebview class here.
15         (createWebView): Ditto.
16         (main): Pack the widget into a GtkBox instead of a GtkScrolledWindow.
17         * DumpRenderTree/gtk/SelfScrollingWebKitWebView.cpp: Added.
18         (self_scrolling_webkit_web_view_class_init):
19         (self_scrolling_webkit_web_view_init):
20         (self_scrolling_webkit_web_view_new):
21         (sizeRequestMethod):
22         (getPreferredSizeMethod):
23         * DumpRenderTree/gtk/SelfScrollingWebKitWebView.h: Added.
24         * GNUmakefile.am: Added SelfScrollingWebKitWebView files to the source list.
25
26 2011-07-25  Ojan Vafai  <ojan@chromium.org>
27
28         by default hide tests that have passes for all recorded runs
29         https://bugs.webkit.org/show_bug.cgi?id=65127
30
31         Reviewed by Adam Barth.
32
33         This is part of simplifying the default views of the dashboard to
34         make it show less information for the common use cases.
35
36         * TestResultServer/static-dashboards/flakiness_dashboard.html:
37         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
38
39 2011-07-25  Ojan Vafai  <ojan@chromium.org>
40
41         stop generating expectations.json now that it's unused
42         https://bugs.webkit.org/show_bug.cgi?id=65130
43
44         Reviewed by Adam Barth.
45
46         * Scripts/webkitpy/layout_tests/controllers/manager.py:
47         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
48         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
49         * TestResultServer/handlers/menu.py:
50         * TestResultServer/static-dashboards/dashboard_base.js:
51         * TestResultServer/static-dashboards/flakiness_dashboard.html:
52
53 2011-07-25  Ojan Vafai  <ojan@chromium.org>
54
55         don't show builders in the flakiness dashboard that have stopped running a given test suite
56         https://bugs.webkit.org/show_bug.cgi?id=65126
57
58         Reviewed by Mihai Parparita.
59
60         * TestResultServer/static-dashboards/dashboard_base.js:
61
62 2011-07-25  Adam Barth  <abarth@webkit.org>
63
64         Overhaul garden-o-matic frontend to integrate revision history and failures
65         https://bugs.webkit.org/show_bug.cgi?id=65089
66
67         Reviewed by Dimitri Glazkov.
68
69         This is a major overhaul of the garden-o-matic frontend.  The new UI
70         integrates the revision history information with failures.  Failures
71         appear in context adjacent to the revisions that caused them (which are
72         highlighted).  Each buildbot receives a collumn to the left, which
73         indicates which revisions that bot has tested and whether that bot
74         showed the indicated failures.
75
76         Failures are selectable with checkboxes, and can then either be
77         examined or rebaselined.
78
79         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
80         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js:
81         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
82         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
83         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
84         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
85         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js:
86         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
87         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
88
89 2011-07-25  Holger Hans Peter Freyther  <zecke@selfish.org>
90
91         [buildbot] Add Qt Linux MIPSEL Configuration
92         https://bugs.webkit.org/show_bug.cgi?id=65116
93
94         Reviewed by Adam Roben.
95
96         Introduce a Qt Linux MIPS EL build slave. The configuration
97         is a build only configuration.
98
99         * BuildSlaveSupport/build.webkit.org-config/config.json: Add MIPS buildbot.
100
101 2011-07-25  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
102
103         [Qt][WK2] Add multi-point touch mocking to MiniBrowser
104         https://bugs.webkit.org/show_bug.cgi?id=64374
105
106         Reviewed by Benjamin Poulain.
107
108         Each mouse button generate a touch point. The touch point stay
109         virtually on screen if the mouse release happened while the Ctrl
110         modifier was pressed.
111
112         * MiniBrowser/qt/MiniBrowserApplication.cpp:
113         (MiniBrowserApplication::notify):
114         * MiniBrowser/qt/MiniBrowserApplication.h:
115
116 2011-07-25  Balazs Kelemen  <kbalazs@webkit.org>
117
118         [Qt][WK2] Use NRWT for Qt-WK2
119         https://bugs.webkit.org/show_bug.cgi?id=65097
120
121         Reviewed by Andreas Kling.
122
123         * Scripts/run-webkit-tests:
124         (useNewRunWebKitTests):
125
126 2011-07-24  Sheriff Bot  <webkit.review.bot@gmail.com>
127
128         Unreviewed, rolling out r91635.
129         http://trac.webkit.org/changeset/91635
130         https://bugs.webkit.org/show_bug.cgi?id=65083
131
132         It broke the minimal build (Requested by Ossy___OFFLINE on
133         #webkit).
134
135         * QtTestBrowser/launcherwindow.cpp:
136         (LauncherWindow::initializeView):
137         (LauncherWindow::createChrome):
138         * QtTestBrowser/launcherwindow.h:
139         (WindowOptions::WindowOptions):
140         * QtTestBrowser/main.cpp:
141         (LauncherApplication::handleUserOptions):
142
143 2011-07-24  Adam Barth  <abarth@webkit.org>
144
145         Rename garden-o-matic "alert" to "infobar" and increase the width to 100%.
146
147         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
148         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
149         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
150         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
151         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
152         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
153
154 2011-07-23  Dimitri Glazkov  <dglazkov@chromium.org>
155
156         TestConfiguration should be hashable and comparable for equality.
157         https://bugs.webkit.org/show_bug.cgi?id=65070
158
159         Reviewed by Adam Barth.
160
161         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added __hash__ and __eq__ methods.
162         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added unit tests.
163
164 2011-07-23  Dimitri Glazkov  <dglazkov@chromium.org>
165
166         Extract TestConfiguration from base and write a few unit tests to capture current behavior.
167         https://bugs.webkit.org/show_bug.cgi?id=65066
168
169         Reviewed by Adam Barth.
170
171         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added.
172         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added.
173         * Scripts/webkitpy/layout_tests/port/base.py: Moved code to test_configuration.py
174
175 2011-07-23  Alok Priyadarshi  <alokp@chromium.org>
176
177         Switching off acceleration for small canvas broke gpu tests
178         https://bugs.webkit.org/show_bug.cgi?id=65053
179
180         Reviewed by Stephen White.
181
182         * DumpRenderTree/chromium/WebPreferences.cpp:
183         (WebPreferences::reset):
184         (WebPreferences::applyTo):
185         * DumpRenderTree/chromium/WebPreferences.h:
186
187 2011-07-23  Keith Kyzivat  <keith.kyzivat@nokia.com>
188
189         [Qt] Add option to turn on disk caching in QtTestBrowser
190         https://bugs.webkit.org/show_bug.cgi?id=65007
191
192         Reviewed by Noam Rosenthal.
193
194         Add menu item and command line option to turn on disk caching in
195         QtTestBrowser. Simple QNetworkDiskCache added to the page's
196         QNetworkAccssManager.
197
198         * QtTestBrowser/launcherwindow.cpp:
199         (LauncherWindow::initializeView):
200         (LauncherWindow::createChrome):
201         (LauncherWindow::setDiskCache):
202         * QtTestBrowser/launcherwindow.h:
203         (WindowOptions::WindowOptions):
204         * QtTestBrowser/main.cpp:
205         (LauncherApplication::handleUserOptions):
206
207 2011-07-22  Ojan Vafai  <ojan@chromium.org>
208
209         Don't match 'DEBUG' when trying to filter BUG modifiers.
210         * TestResultServer/static-dashboards/flakiness_dashboard.html:
211         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
212
213 2011-07-22  Ojan Vafai  <ojan@chromium.org>
214
215         Fix issue loading gtests since there's no g_expectations.
216         Also, make filing a bug file a webkit.org bug, not a crbug since
217         the link is only shown for layout tests.
218         * TestResultServer/static-dashboards/dashboard_base.js:
219         * TestResultServer/static-dashboards/flakiness_dashboard.html:
220
221 2011-07-22  Ojan Vafai  <ojan@chromium.org>
222
223         pull test expectations into the dashboard from wvn.webkit.org
224         https://bugs.webkit.org/show_bug.cgi?id=65054
225
226         Reviewed by Adam Barth.
227
228         Now that svn.webkit.org supports cors headers, we can just grab the txt
229         file from there instead of generating a JSON file that we upload
230         to appengine. We'll stop generating the JSON file in a followup patch.
231
232         * TestResultServer/static-dashboards/dashboard_base.js:
233         * TestResultServer/static-dashboards/flakiness_dashboard.html:
234         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
235
236 2011-07-22  Adam Barth  <abarth@webkit.org>
237
238         Fix typo so the tool works on a case-sensitive file system.
239
240         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
241
242 2011-07-22  Adam Barth  <abarth@webkit.org>
243
244         garden-o-matic should know what patches are landing
245         https://bugs.webkit.org/show_bug.cgi?id=64978
246
247         Reviewed by Dimitri Glazkov.
248
249         This patch adds a display of the recently committed patches to
250         garden-o-matic.  The UI needs iteration, but this patch establishes the
251         plumbing.
252
253         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
254             - Teach trac how to extract more information from the commit log.
255         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
256             - Add a data attribute for storing the revision.
257         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
258             - Add a container for the recent commit history.
259         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
260             - CSS for displaying non-ugly commit history.
261         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
262             - Wire up trac to the UI container.
263         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
264             - A basic, compact display for a list of commits.
265
266 2011-07-22  Ojan Vafai  <ojan@chromium.org>
267
268         Fix bug when switching from layout-tests to another test-type if
269         ToT webkit.org is the selected group.
270         * TestResultServer/static-dashboards/flakiness_dashboard.html:
271
272 2011-07-22  Ojan Vafai  <ojan@chromium.org>
273
274         Add back in line accidentally deleted in http://trac.webkit.org/changeset/91612.
275         * TestResultServer/static-dashboards/flakiness_dashboard.html:
276
277 2011-07-22  Ojan Vafai  <ojan@chromium.org>
278
279         simplify the flakiness dashboard to only show modifiers/expectations for the current builder
280         https://bugs.webkit.org/show_bug.cgi?id=65049
281
282         Reviewed by Tony Chang.
283
284         This communicates less information, but probably makes the dashboard more usable.
285         Also make it show that bug lines that don't point to bug numbers still
286         show up in the bugs column (e.g. BUG_TONY) and stop showing the extra/missing
287         columns in any view. The data is still used, but only in the view that updates
288         the test_expectations.txt file.
289
290         * TestResultServer/static-dashboards/dashboard_base.js:
291         * TestResultServer/static-dashboards/flakiness_dashboard.html:
292         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
293
294 2011-07-21  Ojan Vafai  <ojan@chromium.org>
295
296         hook in the rest of the bots to the dashboard now they run NRWT
297         https://bugs.webkit.org/show_bug.cgi?id=64985
298
299         Reviewed by Adam Barth.
300
301         * TestResultServer/static-dashboards/builders.js:
302
303 2011-07-22  Alexey Proskuryakov  <ap@apple.com>
304
305         Adding Tim Horton to contributor list for Bugzilla auto-completion. He is CC'ed quite frequently.
306
307         Unreviewed.
308
309         * Scripts/webkitpy/common/config/committers.py:
310
311 2011-07-22  Michal Pakula vel Rutka  <m.pakula@samsung.com>
312
313         [EFL] ewk_frame_hit_test_new enchancement
314         https://bugs.webkit.org/show_bug.cgi?id=64260
315
316         Apply changes done in ewk_frame_hit_test_new in EWebLauncher:
317         Replacing Ewk_Hit_Test_Result_Context structure 'flags'
318         in hit test output to by enum 'context'.
319
320         Reviewed by Antonio Gomes.
321
322         * EWebLauncher/main.c:
323         (on_key_down):
324
325 2011-07-21  Adam Roben  <aroben@apple.com>
326
327         Fix typo in TestFailures
328
329         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
330
331 2011-07-21  Ojan Vafai  <ojan@chromium.org>
332
333         simplify gtest display now that we strip modifiers from the JSON
334         https://bugs.webkit.org/show_bug.cgi?id=64990
335
336         Reviewed by Adam Barth.
337
338         Now that names are normalized, we can remove all the code that
339         handles gtest name changes (e.g. for adding modifiers like DISABLED_).
340         Instead, if you try to list a test with a modifier in it, we need
341         to strip the modifier so we get the normalized value.
342
343         We also get rid of the concept of extra/missing expectations for gtests.
344         In a patch soon, we'll stop showing extra/missing expectations from the UI
345         entirely and only leave it for the special updating test_expectations.txt
346         view of the dashboard, which doesn't apply to gtests.
347
348         * TestResultServer/static-dashboards/flakiness_dashboard.html:
349         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
350
351 2011-07-21  Adam Barth  <abarth@webkit.org>
352
353         Refactor Trac.js for use in garden-o-matic
354         https://bugs.webkit.org/show_bug.cgi?id=64998
355
356         Reviewed by Adam Roben.
357
358         This patch refactors Trac.js to match the architecture and style of
359         garden-o-matic.  I've also factored AsynchronousCache out of Trac (and
360         moved it to base) because it was redundant with some other code we
361         already had.
362
363         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js:
364         (FailingTestsBugForm):
365         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js:
366         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
367         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
368         (ViewController.prototype._displayBuilder.start):
369         (ViewController.prototype._displayBuilder):
370         (ViewController.prototype._domForRegressionRange.trac.getCommitDataForRevisionRange):
371         (ViewController.prototype._domForRegressionRange):
372         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js:
373         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
374         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html:
375         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
376         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
377
378 2011-07-21  Ojan Vafai  <ojan@chromium.org>
379
380         Fix issue throwing a 500 to the bots. We were modifying
381         the dictionary during an iteration over it's keys and would get
382         RuntimeError: dictionary changed size during iteration
383
384         * TestResultServer/model/jsonresults.py:
385
386 2011-07-21  Adam Barth  <abarth@webkit.org>
387
388         ORWT fails to consume some NRWT arguments
389         https://bugs.webkit.org/show_bug.cgi?id=64988
390
391         Reviewed by Adam Roben.
392
393         * Scripts/old-run-webkit-tests:
394
395 2011-07-21  Adam Barth  <abarth@webkit.org>
396
397         Merge TestFailures and garden-o-matic unit tests harness
398         https://bugs.webkit.org/show_bug.cgi?id=64984
399
400         Reviewed by Adam Roben.
401
402         One unit testing framework is better than two.
403
404         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-more-unittests.html: Removed.
405         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
406
407 2011-07-21  Adam Barth  <abarth@webkit.org>
408
409         Rubber-stamped by Adam Roben.
410
411         Move garden-o-matic files into the same directory as TestFailures to
412         make it easier to integrate the codebases.
413
414         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/base.js.
415         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js.
416         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js.
417         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout_unittests.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/checkout_unittests.js.
418         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/config.js.
419         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon-green.png: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-green.png.
420         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon-red.png: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-red.png.
421         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/index.html.
422         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/main.css.
423         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/main.js.
424         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/partytime.gif: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/partytime.gif.
425         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/results.js.
426         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js.
427         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-more-unittests.html: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html.
428         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js.
429         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js.
430         * Scripts/webkitpy/tool/commands/abstractlocalservercommand.py:
431         * Scripts/webkitpy/tool/commands/gardenomatic.py:
432         * Scripts/webkitpy/tool/servers/data/gardeningserver: Removed.
433         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js: Removed.
434         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js: Removed.
435         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js: Removed.
436         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout_unittests.js: Removed.
437         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js: Removed.
438         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-green.png: Removed.
439         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-red.png: Removed.
440         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html: Removed.
441         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css: Removed.
442         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js: Removed.
443         * Scripts/webkitpy/tool/servers/data/gardeningserver/partytime.gif: Removed.
444         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js: Removed.
445         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js: Removed.
446         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html: Removed.
447         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js: Removed.
448         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js: Removed.
449         * Scripts/webkitpy/tool/servers/gardeningserver.py:
450
451 2011-07-19  Ojan Vafai  <ojan@chromium.org>
452
453         fix links on the flakiness dashboard
454         https://bugs.webkit.org/show_bug.cgi?id=64832
455
456         Reviewed by Tony Chang.
457
458         Make them actual anchor tags and linkify the test name instead of
459         adding an extra link. Also simplify the update view since we now
460         show the linkified test name.
461
462         * TestResultServer/static-dashboards/flakiness_dashboard.html:
463
464 2011-07-21  Adam Barth  <abarth@webkit.org>
465
466         webkit.org builders should upload results to test-results AppEngine instance
467         https://bugs.webkit.org/show_bug.cgi?id=64979
468
469         Reviewed by Adam Roben.
470
471         The master just needs to pass some flags to NRWT.  This patch also lets
472         ORWT consume the flags (ignoring them) so we can pass the flags
473         unconditionally.
474
475         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
476         * Scripts/old-run-webkit-tests:
477
478 2011-07-21  Ojan Vafai  <ojan@chromium.org>
479
480         normalize gtest names in the result json
481         https://bugs.webkit.org/show_bug.cgi?id=64971
482
483         Reviewed by Eric Seidel.
484
485         This will make the resulting JSON smaller and make the
486         flakiness dashboard much more sane (e.g. we won't need to
487         show you the FLAKY_ version and the normal version).
488
489         * TestResultServer/model/jsonresults.py:
490         * TestResultServer/model/jsonresults_unittest.py:
491
492 2011-07-21  Eric Seidel  <eric@webkit.org>
493
494         Unreviewed.  Reenable ORWT for the --leaks bot.
495         Something is still wrong with the leaks summarizing
496         but I'm about to head out on vacation for 2 weeks,
497         so it's best to just revert back to ORWT until I'm back.
498
499         * Scripts/run-webkit-tests:
500         (usingLeaks):
501         (useNewRunWebKitTests):
502
503 2011-07-21  Ojan Vafai  <ojan@chromium.org>
504
505         cleanup jsonresults.py style in preparation for making some changes
506         https://bugs.webkit.org/show_bug.cgi?id=64968
507
508         Reviewed by Adam Barth.
509
510         No code changes. Unittest still passes. Mostly just removing
511         useless comments and 80 character wrapping.
512
513         * TestResultServer/model/jsonresults.py:
514         * TestResultServer/model/jsonresults_unittest.py:
515
516 2011-07-21  Adam Barth  <abarth@webkit.org>
517
518         Add missing column header.
519
520         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
521
522 2011-07-21  Eric Seidel  <eric@webkit.org>
523
524         Unreviewed.  Just fixing (and unittesting) a previous typo.
525
526         Fix typo in print_leaks_summary regexp which was causing
527         leaks summarizing to fail on the --leaks bot.
528
529         * Scripts/webkitpy/layout_tests/port/mac.py:
530         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
531
532 2011-07-21  Adam Barth  <abarth@webkit.org>
533
534         garden-o-matic should be able to rebaseline many tests at once
535         https://bugs.webkit.org/show_bug.cgi?id=64963
536
537         Reviewed by Dimitri Glazkov.
538
539         This patch contains some UI for rebaselining tests in bulk.  I don't
540         expect this UI to be the final UI, but the underlying infrastructure
541         will be useful.
542
543         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js:
544         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
545         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
546         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
547         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
548
549 2011-07-21  Adam Barth  <abarth@webkit.org>
550
551         garden-o-matic has trouble when the same test fails in different ways on diffrent bots
552         https://bugs.webkit.org/show_bug.cgi?id=64830
553
554         Unreviewed.  (Missing the patch is blocking me from fixing the tree.)
555
556         This patch is just some minor bug fixes.  One of the bugs results from
557         the sadness of having two representations (a list and a space-separated
558         string) for the set of failure types.
559
560         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
561         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
562         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
563         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
564
565 2011-07-20  Ryosuke Niwa  <rniwa@webkit.org>
566
567         Buildbot marks a nrwt bot red when tests are missing results
568         https://bugs.webkit.org/show_bug.cgi?id=64812
569
570         Reviewed by Adam Barth.
571
572         The bug was caused by multiple expressions matching on the single output.
573         Fixed it by exiting the loop as soon as one expression matches.
574
575         Because the regular expression for 'failures' is most general,
576         moved it to the end of the list to avoid it catching other cases.
577
578         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
579
580 2011-07-20  Chang Shu  <cshu@webkit.org>
581
582         Adding myself to the reviewers list; No review needed.
583
584         * Scripts/webkitpy/common/config/committers.py:
585
586 2011-07-20  Eric Seidel  <eric@webkit.org>
587
588         new-run-webkit-tests --leaks does not restart DumpRenderTree often enough
589         https://bugs.webkit.org/show_bug.cgi?id=64912
590
591         Reviewed by Adam Barth.
592
593         The bug was that set_default_option only works if the option's value
594         is "None".  So I fixed --batch-size to default to None instead of 0
595         (which kinda makes sense).
596
597         * Scripts/webkitpy/layout_tests/controllers/worker.py:
598         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
599
600 2011-07-20  Eric Seidel  <eric@webkit.org>
601
602         Make kill-old-processes kill httpd on mac
603         https://bugs.webkit.org/show_bug.cgi?id=64898
604
605         Reviewed by Adam Barth.
606
607         * BuildSlaveSupport/kill-old-processes:
608          - The line endings seem to have flipped again,
609            Thus making this diff gigantic.  The only line
610            changed is the "httpd" line.  I suspect we'll need
611            to set the line ending style from an svn checkout.
612
613 2011-07-20  Ojan Vafai  <ojan@chromium.org>
614
615         show a list of average test times in the treemap
616         https://bugs.webkit.org/show_bug.cgi?id=64899
617
618         Reviewed by Adam Roben.
619
620         For now, you click a link and it replaces the treemap
621         with a reverse sorted list of directories based on
622         average test runtime within that directory.
623
624         To cut some of the noise, we don't show directories
625         that only have one test in them or where the average
626         runtime is less than 100ms.
627
628         * TestResultServer/static-dashboards/treemap.html:
629
630 2011-07-20  Eric Seidel  <eric@webkit.org>
631
632         Teach build.webkit.org how to identify leaks in NRWT output
633         https://bugs.webkit.org/show_bug.cgi?id=64891
634
635         Reviewed by Adam Barth.
636
637         This was the last remaining piece blocking turning on
638         NRWT for the leaks bot, so I did that in this change too.
639         If for some reason we see trouble on the bots, I'll
640         revert that part of this change.
641
642         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
643         * Scripts/run-webkit-tests:
644
645 2011-07-20  Ojan Vafai  <ojan@chromium.org>
646
647         make the treemap display more human friendly
648         https://bugs.webkit.org/show_bug.cgi?id=64896
649
650         Reviewed by Tony Chang.
651
652         * TestResultServer/static-dashboards/treemap.html:
653
654 2011-07-20  Eric Seidel  <eric@webkit.org>
655
656         Move HttpLock to using a FileSystem object
657         https://bugs.webkit.org/show_bug.cgi?id=64885
658
659         Reviewed by Adam Barth.
660
661         I made a typo in my previous change, taking the value
662         of read_text_file(pid_file) and setting it to lock_pid_file
663         instead of current_pid.  Fixed now. :)
664
665         In order to test my new change I had to overhaul the unittests
666         for this class to create a separate set of tests which work off
667         of Mock objects instead of the real filesystem.
668
669         Since Executive doesn't yet wrap os.getpid() I added a FIXME
670         in several places where we're currently calling os.getpid().  I
671         felt adding Executive.getpid was outside of the scope of this change
672         but once it exists some of this code will be much simpler to mock.
673
674         * Scripts/webkitpy/layout_tests/port/http_lock.py:
675         * Scripts/webkitpy/layout_tests/port/http_lock_unittest.py:
676         * Scripts/webkitpy/tool/mocktool.py:
677
678 2011-07-20  Adam Roben  <aroben@apple.com>
679
680         Fix typo in TestFailures's Bugzilla constants
681
682         This was preventing the "Version" field from being pre-filled when filing new bugs.
683
684         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBugzilla.js:
685
686 2011-07-20  Eric Seidel  <eric@webkit.org>
687
688         Move HttpLock to using a FileSystem object
689         https://bugs.webkit.org/show_bug.cgi?id=64885
690
691         Reviewed by Adam Barth.
692
693         There is no functional change here.  Just moving off
694         of direct os.foo calls to their filesystem equivalent.
695
696         This makes it possible to now unittest HttpLock using a
697         MockFileSystem instead of needing to talk to the filesystem directly.
698
699         This also will have changed any port test cases which were using a
700         mock file system to now have an HttpLock based on a mock filesystem
701         instead.  I saw no change in the unittests so it's possible
702         we're not testing that configuration (or more likely that it "just works").
703
704         * Scripts/webkitpy/layout_tests/port/base.py:
705         * Scripts/webkitpy/layout_tests/port/http_lock.py:
706         * Scripts/webkitpy/layout_tests/port/http_lock_unittest.py:
707
708 2011-07-20  Adam Barth  <abarth@webkit.org>
709
710         NRWT results uses too much space per run
711         https://bugs.webkit.org/show_bug.cgi?id=64845
712
713         Reviewed by Ojan Vafai.
714
715         Removing these files saves about 750K per build.
716
717         * Scripts/webkitpy/layout_tests/controllers/manager.py:
718
719 2011-07-20  Zan Dobersek  <zandobersek@gmail.com>
720
721         [Gtk] [NRWT] Xvfb produces a lot of stderr output
722         https://bugs.webkit.org/show_bug.cgi?id=64669
723
724         Reviewed by Philippe Normand.
725
726         Redirect the stderr output of the driver's Xvfb process
727         to /dev/null to suppress unneeded output.
728
729         * Scripts/webkitpy/layout_tests/port/gtk.py:
730
731 2011-07-19  Eric Seidel  <eric@webkit.org>
732
733         parse-malloc-history always exits 1, causing NRWT to fail
734         https://bugs.webkit.org/show_bug.cgi?id=64835
735
736         Reviewed by Adam Barth.
737
738         ORWT just never checked the return code. :)
739
740         * Scripts/parse-malloc-history:
741         (main):
742         * Scripts/webkitpy/layout_tests/port/mac.py:
743
744 2011-07-19  Kevin Ollivier  <kevino@theolliviers.com>
745
746         [wx] Unreviewed build fix, don't add the debug prefix as 2.9 no longer uses one.
747
748         * waf/build/wxpresets.py:
749
750 2011-07-19  Eric Seidel  <eric@webkit.org>
751
752         new-run-webkit-tests should only enable MallocStackLogging for DRT
753         https://bugs.webkit.org/show_bug.cgi?id=64792
754
755         Reviewed by Adam Barth.
756
757         The previous code would enable it for all servers launched
758         by the port, which included Apache, the python websocket server
759         as well as ImageDiff.  Now only DumpRenderTree will have
760         MallocStackLogging enabled or the GuardMalloc library injected.
761
762         I also cleaned up the websocket_server code to use filesystem
763         while I was in it.
764
765         I also made DRT restart every 1000 tests when running with
766         --leaks enabled.  I believe this made the --leaks run slightly
767         faster, but it still takes over an hour on my machine. :(
768
769         * Scripts/webkitpy/layout_tests/controllers/worker.py:
770         * Scripts/webkitpy/layout_tests/port/base.py:
771         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
772         * Scripts/webkitpy/layout_tests/port/gtk.py:
773         * Scripts/webkitpy/layout_tests/port/mac.py:
774         * Scripts/webkitpy/layout_tests/port/qt.py:
775         * Scripts/webkitpy/layout_tests/port/server_process.py:
776         * Scripts/webkitpy/layout_tests/port/webkit.py:
777         * Scripts/webkitpy/layout_tests/servers/http_server.py:
778         * Scripts/webkitpy/layout_tests/servers/websocket_server.py:
779
780 2011-07-19  Ojan Vafai  <ojan@chromium.org>
781
782         remove the concept of platform fallbacks
783         https://bugs.webkit.org/show_bug.cgi?id=64829
784
785         Reviewed by Adam Barth.
786         
787         In the process, wrote tests for this code and fixed a pretty major
788         bug (now covered by the test).
789
790         * TestResultServer/static-dashboards/flakiness_dashboard.html:
791         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
792
793 2011-07-19  Ojan Vafai  <ojan@chromium.org>
794
795         cleanup some of the PLATFORM logic in the flakiness dashboard
796         https://bugs.webkit.org/show_bug.cgi?id=64821
797
798         Reviewed by Adam Barth.
799
800         * TestResultServer/static-dashboards/flakiness_dashboard.html:
801         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
802
803 2011-07-19  Ojan Vafai  <ojan@chromium.org>
804
805         cleanup some of the PLATFORM logic in the flakiness dashboard
806         https://bugs.webkit.org/show_bug.cgi?id=64821
807
808         Reviewed by Adam Barth.
809
810         * TestResultServer/static-dashboards/flakiness_dashboard.html:
811         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
812
813 2011-07-19  Dimitri Glazkov  <dglazkov@chromium.org>
814
815         Store line number on TestExpectationLine.
816         https://bugs.webkit.org/show_bug.cgi?id=64800
817
818         Reviewed by Adam Barth.
819
820         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Refactored to store line number on TestExpectationLine instances.
821
822 2011-07-19  Ojan Vafai  <ojan@chromium.org>
823
824         fix flakiness dashboard for XP bots with XP in the name instead of WIN
825         https://bugs.webkit.org/show_bug.cgi?id=64826
826
827         Reviewed by Mihai Parparita.
828
829         * TestResultServer/static-dashboards/flakiness_dashboard.html:
830         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
831
832 2011-07-19  Cary Clark  <caryclark@google.com>
833
834         add myself to committers.py
835         https://bugs.webkit.org/show_bug.cgi?id=64816
836
837         No review needed.
838
839         * Scripts/webkitpy/common/config/committers.py:
840
841 2011-07-19  Daniel Bates  <dbates@webkit.org>
842
843         Rename variable isEmptyTextPatch in svn-{apply, unapply}
844         https://bugs.webkit.org/show_bug.cgi?id=64648
845
846         Reviewed by Adam Roben.
847
848         Rename the variable isEmptyTextPatch to hasTextChunks, which is
849         more descriptive and makes its usage read well in control statements.
850
851         * Scripts/svn-apply:
852         (patch):
853         * Scripts/svn-unapply:
854         (patch):
855
856 2011-07-19  Adam Roben  <aroben@apple.com>
857
858         Make TestFailures show existing bugs and a new bug link for flaky tests
859
860         Fixes <http://webkit.org/b/63728> TestFailures page should make it easy to file bugs about
861         flaky tests
862         and
863         <http://webkit.org/b/63830> TestFailures page doesn't show related bugs for possibly-flaky
864         tests, but should
865
866         Reviewed by Sam Weinig.
867
868         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm.js:
869         Added. This is what's used to file new bugs about flaky tests.
870         (FlakyTestBugForm): Calls up to the base class, stores the arguments, and sets our
871         title, description, and URL.
872
873         (FlakyTestBugForm.prototype._createBugDescription):
874         (FlakyTestBugForm.prototype._createBugTitle):
875         These do what they say.
876
877         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm_unittests.js:
878         Added. Tests of the above.
879
880         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
881         (#failure-history > li):
882         (#possibly-flaky-tests > li):
883         (#failure-history > li, #possibly-flaky-tests > li):
884         (#possibly-flaky-tests > li > :first-child):
885         (.flakiness-examples-list):
886         Styling to account for the new elements.
887
888         (.expandable):
889         (.expanded > .expandable):
890         Generalized the .flakiness-example-list code to a generic .expandable class. Removed the
891         transition properties because we can't transition to/from 'auto' (but really would like
892         to!).
893
894         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
895         (ViewController.prototype._displayBuilder): Moved code to create the FailingTestsBugForm
896         here from _domForNewAndExistingBugs. Updated for changes to _domForPossiblyFlakyTests.
897         (ViewController.prototype._domForNewAndExistingBugs): Changed to take a NewBugForm as an
898         argument instead of creating one ourselves.
899         (ViewController.prototype._domForPossiblyFlakyTests): Changed to take all analyzed builds as
900         an argument instead of just the number of analyzed builds. Put the examples list inside a
901         container <div>, which is also used to hold the new/existing bugs UI. Removed code to deal
902         with animating the height of the examples list; we don't do this anymore because it's hard
903         to make it work correctly with the asynchronous loading of existing bugs. When we populate
904         the examples list, also set up the new/existing bug UI.
905
906         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
907         FlakyTestBugForm.
908
909         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
910         Pull in FlakyTestBugForm and its tests.
911
912 2011-07-19  Ryosuke Niwa  <rniwa@webkit.org>
913
914         Buildbot marks a nrwt bot red when tests are missing results
915         https://bugs.webkit.org/show_bug.cgi?id=64812
916
917         Reviewed by Tony Chang.
918
919         Added "missing results" label for tests that are missing results.
920
921         buildbot Now reports shows a label "X missing results" on a run when X tests were missing
922         expected results on that run.
923
924         Also rephrased "X failed" and "X were flaky" to "X failures" and "X flakes".
925
926         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
927
928 2011-07-19  Adam Roben  <aroben@apple.com>
929
930         Extract some of FailingTestsBugForm's code into a base class
931
932         Prep work for fixing <http://webkit.org/b/63728> TestFailures page should make it easy to
933         file bugs about flaky tests
934
935         Reviewed by Sam Weinig.
936
937         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js:
938         Moved BugzillaConstants to new WebKitBugzilla file. Moved a bunch of other code from here to
939         TestRelatedBugForm.js.
940
941         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js:
942         Moved some tests to TestRelatedBugForm_unittests.js.
943
944         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm.js: Added.
945         (TestRelatedBugForm):
946         (TestRelatedBugForm.prototype.domElement):
947         (TestRelatedBugForm.prototype._computeOperatingSystem):
948         (TestRelatedBugForm.prototype._computePlatform):
949         Code came from FailingTestsBugForm.
950
951         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm_unittests.js:
952         Added. Tests came from FailingTestsBugForm_unittests.js
953
954         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBugzilla.js:
955         Added. Code came from FailingTestsBugForm.js.
956
957         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
958         TestRelatedBugForm.js and WebKitBugzilla.js. Moved Bugzilla.js out of the list of files that
959         need to be pulled in early for parsing reasons.
960
961         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
962         Pull in TestRelatedBugForm and tests and WebKitBugzilla.
963
964 2011-07-19  Sam Weinig  <sam@webkit.org>
965
966         Remove obsolete Mac NRWT test bots
967         https://bugs.webkit.org/show_bug.cgi?id=64806
968
969         Reviewed by Adam Roben.
970
971         Remove "Leopard Intel Release (NRWT)" and "SnowLeopard Intel Release (NRWT)"
972         which have been disconnected for some time and have been obsoleted by
973         the main testers using NRWT.
974
975         * BuildSlaveSupport/build.webkit.org-config/config.json:
976
977 2011-07-19  Sam Weinig  <sam@webkit.org>
978
979         webkit-patch rebaseline crashes on use
980         https://bugs.webkit.org/show_bug.cgi?id=64775
981
982         Reviewed by Adam Roben.
983
984         Fix incorrectly spelled attribute name.
985
986         * Scripts/webkitpy/common/net/layouttestresults.py:
987         The attribute is test_name, not filename.
988
989         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
990         Add test for tests_matching_failure_types.
991
992 2011-07-18  Eric Seidel  <eric@webkit.org>
993
994         new-run-webkit-tests should support --leaks
995         https://bugs.webkit.org/show_bug.cgi?id=63832
996
997         Reviewed by Dirk Pranke.
998
999         This may not be sufficient to actually transition over the leaks bot,
1000         but this is a huge step in the right direction.
1001
1002         I had to make parse-malloc-history understand being passed more than
1003         one file (to avoid the silly cat | nonsense in old-run-webkit-tests).
1004
1005         I removed some dead code relating to previous iterations of our crash detection.
1006
1007         I created a new class "LeakDetector" to encapsulate all this logic.
1008         Eventually we should consider pulling that class out of mac.py and
1009         sharing with other ports.  However given that ORWT has had
1010         --leaks support on Mac for almost 7 years and no other port has added
1011         it, leaves me to believe we're in no rush to move LeakDetector.
1012
1013         I've tested --leaks locally.  I suspect there are more bugs to shake out
1014         but it seems to work well enough to start.
1015
1016         I also added support for --guard-malloc, but have not tested it much.  It
1017         should be viewed as experimental at this time.
1018
1019         I also fixed various os.path uses to self._filesystem as I was reading
1020         through the various files to understand how best to fix this bug.
1021
1022         * Scripts/old-run-webkit-tests:
1023         (parseLeaksandPrintUniqueLeaks):
1024         * Scripts/parse-malloc-history:
1025         (main):
1026         * Scripts/webkitpy/common/system/crashlogs.py:
1027         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1028         * Scripts/webkitpy/layout_tests/controllers/worker.py:
1029         * Scripts/webkitpy/layout_tests/port/base.py:
1030         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1031         * Scripts/webkitpy/layout_tests/port/gtk.py:
1032         * Scripts/webkitpy/layout_tests/port/mac.py:
1033         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1034         * Scripts/webkitpy/layout_tests/port/server_process.py:
1035         * Scripts/webkitpy/layout_tests/port/webkit.py:
1036         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1037
1038 2011-07-19  Adam Barth  <abarth@webkit.org>
1039
1040         garden-o-matic should have a "Triage Failures" button for iterating over failures
1041         https://bugs.webkit.org/show_bug.cgi?id=64769
1042
1043         Reviewed by Eric Seidel.
1044
1045         This patch adds some global static state and refactors things a bit so
1046         that we can iterate over all the failures in the details pane.
1047
1048         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
1049             - Add a generic callback iterator to iterate through a series of
1050               callbacks. We use this to iterate through the failures we want to
1051               display the details of.
1052         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
1053             - Add a toolbar for the results summary and change the buttons to
1054               be real buttons.
1055         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
1056             - Make the detail pane 75% of the window.
1057             - Make the buttons pretty.
1058             - Allow the failure type badges to apply to the details pane as well.
1059         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
1060             - Restructure how we bring up the details pane so we can iterate
1061               through a bunch of failures.
1062             - Remove the transition between results details because it's
1063               annoying when you want to click through a bunch of failures.
1064         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
1065             - Add some more structure to the title bar for the results details
1066               so it's clearer which results we're currently displaying.
1067
1068 2011-07-19  Adam Barth  <abarth@webkit.org>
1069
1070         garden-o-matic should live update as conditions change on the buildbot
1071         https://bugs.webkit.org/show_bug.cgi?id=64646
1072
1073         Reviewed by Eric Seidel.
1074
1075         This patch causes the garden-o-matic display to update every 10 minutes
1076         to show the latest failures.  This patch required a moderately large
1077         refactoring of main.js to make the display incremental instead of
1078         all-at-once.
1079
1080         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
1081         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
1082         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
1083         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
1084
1085 2011-07-18  Tab Atkins  <jackalmage@gmail.com>
1086
1087         Tab Atkins is not in committers.py (as a non-committer)
1088         https://bugs.webkit.org/show_bug.cgi?id=64770
1089
1090         Reviewed by Eric Seidel.
1091
1092         * Scripts/webkitpy/common/config/committers.py:
1093
1094 2011-07-18  Ojan Vafai  <ojan@chromium.org>
1095
1096         Load ToT chromium bots by default in the flakiness dashboard
1097         https://bugs.webkit.org/show_bug.cgi?id=64756
1098
1099         Reviewed by Adam Barth.
1100
1101         For now, these are the most useful bots to look at. Eventually,
1102         we may want to load the build.webkit.org bots by default once
1103         most of the bots there upload results.
1104
1105         * TestResultServer/static-dashboards/dashboard_base.js:
1106
1107 2011-07-13  Jon Honeycutt  <jhoneycutt@apple.com>
1108
1109         Focus and selection events are not fired when a <select>'s selection
1110         changes
1111         https://bugs.webkit.org/show_bug.cgi?id=64504
1112         <rdar://problem/9319881>
1113
1114         Reviewed by Alice Liu.
1115
1116         * DumpRenderTree/AccessibilityController.h:
1117         Added m_notificationsEventHook for addNotificationListener().
1118         m_allEventsHook will now be used for setLogAccessibilityEvents().
1119
1120         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
1121         (AccessibilityController::AccessibilityController):
1122         Initialize m_notificationsEventHook.
1123         (AccessibilityController::~AccessibilityController):
1124         Turn off logging of all accessibility events. If
1125         m_notificationsEventHook is non-null, unhook it.
1126         (logEventProc):
1127         Add handling of EVENT_OBJECT_SELECTION.
1128         (AccessibilityController::setLogAccessibilityEvents):
1129         If the state of logging is not changing, return early. If we're turning
1130         off logging, unhook m_allEventsHook, and zero it out. Otherwise, add a
1131         hook for all events.
1132         (AccessibilityController::addNotificationListener):
1133         Use m_notificationsEventHook rather than m_allEventsHook.
1134
1135 2011-07-18  Eric Seidel  <eric@webkit.org>
1136
1137         webkit-patch apply-attachment does not work when not called from the root of the checkout
1138         https://bugs.webkit.org/show_bug.cgi?id=64751
1139
1140         Unreviewed.  Just updating unit test results after bug 64751.
1141
1142         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1143         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1144
1145 2011-07-18  Adam Barth  <abarth@webkit.org>
1146
1147         simplejson has trouble on chromium-linux
1148         https://bugs.webkit.org/show_bug.cgi?id=64757
1149
1150         Reviewed by Eric Seidel.
1151
1152         Use the native JSON, if available.
1153
1154         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
1155
1156 2011-07-18  Ojan Vafai  <ojan@chromium.org>
1157
1158         update the flakiness dashboard to understand the new platforms/formats in test_expectations
1159         https://bugs.webkit.org/show_bug.cgi?id=64743
1160
1161         Reviewed by Adam Barth.
1162
1163         The test expectations format changed a long time ago and the flakiness dashboard
1164         was never updated to match.
1165
1166         * TestResultServer/static-dashboards/dashboard_base.js:
1167         * TestResultServer/static-dashboards/flakiness_dashboard.html:
1168         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
1169         * TestResultServer/static-dashboards/timeline_explorer.html:
1170
1171 2011-07-18  Eric Seidel  <eric@webkit.org>
1172
1173         webkit-patch apply-attachment does not work when not called from the root of the checkout
1174         https://bugs.webkit.org/show_bug.cgi?id=64751
1175
1176         Reviewed by Adam Barth.
1177
1178         Last week I removed a os.chdir in EnsureWorkingDirectoryClean, which
1179         is a step that we run for most commands (as one of the earliest steps).
1180         EnsureWorkingDirectoryClean was incorrectly calling os.chdir to change
1181         the CWD to the checkout root.  This magically made a bunch of otherwise
1182         wrong code work.
1183
1184         When I realized that apply-attachment no longer worked as expected today,
1185         I went and fixed the bug, then realizing that we were not testing
1186         what the cwd was when running various commands.  I fixed our MockExecutive
1187         to always log what the cwd is and fixed a whole bunch of places
1188         where we needed to be setting the cwd.
1189
1190         Hopefully this will solve our cwd problems once and for-all, and webkit-patch
1191         will again correctly work when called from any directory (including outside
1192         of a webkit checkout).
1193
1194         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
1195         * Scripts/webkitpy/common/system/executive.py:
1196         * Scripts/webkitpy/common/system/workspace_unittest.py:
1197         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1198         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
1199         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1200         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
1201         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
1202         * Scripts/webkitpy/tool/bot/irc_command.py:
1203         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
1204         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
1205         * Scripts/webkitpy/tool/commands/download_unittest.py:
1206         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
1207         * Scripts/webkitpy/tool/commands/queues.py:
1208         * Scripts/webkitpy/tool/commands/queues_unittest.py:
1209         * Scripts/webkitpy/tool/commands/queuestest.py:
1210         * Scripts/webkitpy/tool/mocktool.py:
1211         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1212         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
1213         * Scripts/webkitpy/tool/steps/build.py:
1214         * Scripts/webkitpy/tool/steps/checkstyle.py:
1215         * Scripts/webkitpy/tool/steps/editchangelog.py:
1216         * Scripts/webkitpy/tool/steps/preparechangelog.py:
1217         * Scripts/webkitpy/tool/steps/preparechangelogforrevert.py:
1218         * Scripts/webkitpy/tool/steps/runtests.py:
1219         * Scripts/webkitpy/tool/steps/steps_unittest.py:
1220         * Scripts/webkitpy/tool/steps/update.py:
1221
1222 2011-07-18  Adam Barth  <abarth@webkit.org>
1223
1224         Increase information garden-o-matic information density by switching to a table
1225         https://bugs.webkit.org/show_bug.cgi?id=64642
1226
1227         Reviewed by Dimitri Glazkov.
1228
1229         The old layout couldn't handle the large number of failures we had
1230         today.  This patch switches garden-o-matic to a table-based layout,
1231         which can handle many more failures gracefully.  (I expect we'll need
1232         more UI iterations.)
1233
1234         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
1235         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
1236         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
1237         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
1238
1239 2011-07-18  Adam Barth  <abarth@webkit.org>
1240
1241         garden-o-matic should be able to determine when compile breaks
1242         https://bugs.webkit.org/show_bug.cgi?id=64190
1243
1244         Reviewed by Dimitri Glazkov.
1245
1246         This patch adds a red-ish box to the top of the page whenever there is
1247         a compile error on the bots.  The box automatically opens and closes as
1248         appropriate and links to the waterfall display.  In the future, we
1249         might want to compute a regression range.
1250
1251         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js:
1252             - Add the build-only bots to the config.  We use these to check
1253               whether the build failed, which is faster than waiting for the
1254               tester bots to cycle.
1255         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
1256             - Add DOM for the alert bar.
1257         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
1258             - CSS to support the alert bar.
1259         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
1260             - Wiring up events to poll the buildbot to see whether compile has
1261               failed.
1262         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
1263             - Infrastructure for fetching and parsing the buildbot status JSON
1264               blob.  This code could be better factored for testability.  :(
1265         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
1266             - UI for displaying compile errors.
1267         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
1268
1269 2011-07-18  Sheriff Bot  <webkit.review.bot@gmail.com>
1270
1271         Unreviewed, rolling out r91132 and r91135.
1272         http://trac.webkit.org/changeset/91132
1273         http://trac.webkit.org/changeset/91135
1274         https://bugs.webkit.org/show_bug.cgi?id=64681
1275
1276         Broke GTK and Chromium debug bots (Requested by rniwa on
1277         #webkit).
1278
1279         * DumpRenderTree/AccessibilityController.h:
1280         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
1281         (AccessibilityController::AccessibilityController):
1282         (AccessibilityController::~AccessibilityController):
1283         (logEventProc):
1284         (AccessibilityController::setLogAccessibilityEvents):
1285         (AccessibilityController::addNotificationListener):
1286
1287 2011-07-17  Philippe Normand  <pnormand@igalia.com>
1288
1289         test-webkitpy failing with Python 2.5
1290         https://bugs.webkit.org/show_bug.cgi?id=64594
1291
1292         Reviewed by Eric Seidel.
1293
1294         Updated simplejson to version 2.1.6 that fixes an issue when
1295         dumping slashes. Also use cgi.parse_qs instead of the urlparse
1296         version to be compatible with Python2.5. Verified the fix with
1297         test-webkitpy on python 2.5 and python 2.6.
1298
1299         * Scripts/webkitpy/thirdparty/simplejson/README.txt:
1300         * Scripts/webkitpy/thirdparty/simplejson/__init__.py:
1301         * Scripts/webkitpy/thirdparty/simplejson/_speedups.c:
1302         (json_PyOS_string_to_double):
1303         (_convertPyInt_AsSsize_t):
1304         (_convertPyInt_FromSsize_t):
1305         (ascii_escape_char):
1306         (ascii_escape_unicode):
1307         (ascii_escape_str):
1308         (raise_errmsg):
1309         (join_list_unicode):
1310         (join_list_string):
1311         (_build_rval_index_tuple):
1312         (scanstring_str):
1313         (scanstring_unicode):
1314         (py_scanstring):
1315         (py_encode_basestring_ascii):
1316         (scanner_dealloc):
1317         (scanner_traverse):
1318         (scanner_clear):
1319         (_parse_object_str):
1320         (_parse_object_unicode):
1321         (_parse_array_str):
1322         (_parse_array_unicode):
1323         (_parse_constant):
1324         (_match_number_str):
1325         (_match_number_unicode):
1326         (scan_once_str):
1327         (scan_once_unicode):
1328         (scanner_call):
1329         (scanner_new):
1330         (scanner_init):
1331         (encoder_new):
1332         (encoder_init):
1333         (encoder_call):
1334         (_encoded_const):
1335         (encoder_encode_float):
1336         (encoder_encode_string):
1337         (_steal_list_append):
1338         (encoder_listencode_obj):
1339         (encoder_listencode_dict):
1340         (encoder_listencode_list):
1341         (encoder_dealloc):
1342         (encoder_traverse):
1343         (encoder_clear):
1344         (init_speedups):
1345         * Scripts/webkitpy/thirdparty/simplejson/decoder.py:
1346         * Scripts/webkitpy/thirdparty/simplejson/encoder.py:
1347         * Scripts/webkitpy/thirdparty/simplejson/ordered_dict.py: Added.
1348         * Scripts/webkitpy/thirdparty/simplejson/scanner.py:
1349         * Scripts/webkitpy/thirdparty/simplejson/tool.py: Added.
1350         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
1351
1352 2011-07-17  Dimitri Glazkov  <dglazkov@chromium.org>
1353
1354         Rename ModifierMatcher to SpecificityCalculator.
1355         https://bugs.webkit.org/show_bug.cgi?id=64660
1356
1357         It's a little longer than before, but it is much clearer.
1358
1359         Reviewed by Adam Barth.
1360
1361         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Renamed and adjusted
1362             all callsites, also rewrote the comments.
1363         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Adjusted callsites
1364             and renamed tests.
1365
1366 2011-07-15  Adam Barth  <abarth@webkit.org>
1367
1368         gardening server should proxy buildbot status for garden-o-matic
1369         https://bugs.webkit.org/show_bug.cgi?id=64588
1370
1371         Reviewed by Eric Seidel.
1372
1373         This information will be used to detect build breaks.  I took this
1374         opportunity to make ChromiumBuildBot non-static.
1375
1376         * Scripts/webkitpy/common/host.py:
1377         * Scripts/webkitpy/tool/commands/rebaseline.py:
1378         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
1379         * Scripts/webkitpy/tool/mocktool.py:
1380         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1381         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
1382
1383 2011-07-15  Julien Chaffraix  <jchaffraix@webkit.org>
1384
1385         [NRWT] Add support for --no-http
1386         https://bugs.webkit.org/show_bug.cgi?id=64564
1387
1388         Reviewed by Dirk Pranke.
1389
1390         Added support for --no-http, which disables both HTTP and websockets tests.
1391         It also matches the old-run-webkit-tests behavior if --force is used.
1392
1393         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1394         Fixed HTTP_SUBDIR and WEBSOCKET_SUBDIR as tests do not start with a leading separator.
1395         We check if --no-http is set and add the HTTP / websockets tests to the skipped list prior to looking
1396         at the expectation file. Fixed the  _test_requires_lock function to use the same code path to determine
1397         what is worth have an HTTP lock as --no-http to avoid badness.
1398
1399         * Scripts/webkitpy/layout_tests/port/test.py: Added 2 new tests to our mock filesystem to validate that
1400         we do skip properly HTTP / websocket tests inside platform/.
1401
1402         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1403         Added tests that we properly skip all the tests.
1404
1405         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1406         Added checks for the command line arguments.
1407
1408 2011-07-13  Jon Honeycutt  <jhoneycutt@apple.com>
1409
1410         Focus and selection events are not fired when a <select>'s selection
1411         changes
1412         https://bugs.webkit.org/show_bug.cgi?id=64504
1413         <rdar://problem/9319881>
1414
1415         Reviewed by Alice Liu.
1416
1417         * DumpRenderTree/AccessibilityController.h:
1418         Added m_notificationsEventHook for addNotificationListener().
1419         m_allEventsHook will now be used for setLogAccessibilityEvents().
1420
1421         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
1422         (AccessibilityController::AccessibilityController):
1423         Initialize m_notificationsEventHook.
1424         (AccessibilityController::~AccessibilityController):
1425         Turn off logging of all accessibility events. If
1426         m_notificationsEventHook is non-null, unhook it.
1427         (logEventProc):
1428         Add handling of EVENT_OBJECT_SELECTION.
1429         (AccessibilityController::setLogAccessibilityEvents):
1430         If the state of logging is not changing, return early. If we're turning
1431         off logging, unhook m_allEventsHook, and zero it out. Otherwise, add a
1432         hook for all events.
1433         (AccessibilityController::addNotificationListener):
1434         Use m_notificationsEventHook rather than m_allEventsHook.
1435
1436 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
1437
1438         Refactor TestExpectationModel to use TestExpectationLine as data item.
1439         https://bugs.webkit.org/show_bug.cgi?id=64635
1440
1441         This is a bit largish in scope. Does the following things:
1442
1443         1) Adds "path" member to TestExpectationLine to hold normalized path to test, computed at parsing,
1444            and changes code that used Port.normalize_test_name to rely on TestExpectationLine.path. As a result, TestExpectationModel no longer
1445            needs to have any port knowledge.
1446
1447         2) Adds "create_passing_expectation" class method to TestExpectationLine to generate a pristine passing expectation out of a test name,
1448            and changes TestExpectations._process_tests_without_expectations to use it, thus eliminating the need for a special API entry point.
1449            Now all expectations are added to the model in the same way!
1450
1451         3) Changes TestExpectationModel's main test index to store a tuple consisting of line number and TestExpectationLine instance.
1452
1453         Reviewed by Adam Barth.
1454
1455         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Refactored code.
1456
1457 2011-07-15  Adam Roben  <aroben@apple.com>
1458
1459         Teach TestFailures how to detect interrupted build steps
1460
1461         Fixes <http://webkit.org/b/64619> TestFailures page thinks all tests passed in
1462         http://build.webkit.org/builders/SnowLeopard%20Intel%20Release%20(WebKit2%20Tests)/builds/13401
1463
1464         Reviewed by Daniel Bates.
1465
1466         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
1467         (Builder.prototype.getNumberOfFailingTests): If the build step has a result code of 4, the
1468         build step was interrupted. Treat it as an error (by returning a failureCount of -1).
1469
1470         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
1471         Added a test for the above.
1472
1473         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
1474         (LayoutTestResultsLoader.prototype.start): Bumped the cache version to evict old, buggy
1475         cache data that was afflicted by the above bug.
1476
1477 2011-07-15  Eric Seidel  <eric@webkit.org>
1478
1479         new-run-webkit-tests crashes on Apple's Windows port when trying to launch Apache
1480         https://bugs.webkit.org/show_bug.cgi?id=64533
1481
1482         Reviewed by Adam Roben.
1483
1484         Blind removal of Chromium-specific code (now that Chromium does not use this file).
1485
1486         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
1487
1488 2011-07-15  Adam Roben  <aroben@apple.com>
1489
1490         Make TestFailures's list of possibly-flaky tests not so tall
1491
1492         Fixes <http://webkit.org/b/64618> TestFailures page's list of flaky tests takes up way too
1493         much room
1494
1495         Reviewed by Daniel Bates.
1496
1497         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
1498         Removed a no-longer-used .flakiness-example-separator selector.
1499
1500         (#failure-history > li):
1501         (#possibly-flaky-tests > li):
1502         (#failure-history > li, #possibly-flaky-tests > li):
1503         Use a lot less padding for the possibly-flaky tests.
1504
1505 2011-07-15  Tony Chang  <tony@chromium.org>
1506
1507         [chromium] land linux selection color layout test
1508         https://bugs.webkit.org/show_bug.cgi?id=64631
1509
1510         Reviewed by Ojan Vafai.
1511
1512         I had removed this code in r76620 because I thought it wasn't used.
1513         Turns out there was a layout test in the chromium tree that never
1514         got migrated that used it.  Re-add the code and land the layout
1515         test (I will remove it from the chromium tree soon).
1516
1517         * DumpRenderTree/chromium/LayoutTestController.cpp:
1518         (LayoutTestController::LayoutTestController):
1519         (LayoutTestController::forceRedSelectionColors):
1520         * DumpRenderTree/chromium/LayoutTestController.h:
1521
1522 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
1523
1524         Revert http://trac.webkit.org/changeset/91091, since Python 2.5
1525         does not support enumerate() start argument.
1526
1527         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Reverted.
1528
1529 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
1530
1531         Clean up test_expectations.py after refactorings.
1532         https://bugs.webkit.org/show_bug.cgi?id=64620
1533
1534         * Renamed all variables holding TestExpectationLine to expectation_line to avoid confusion with
1535           its sub-part, the actual expectation.
1536         * Renamed all references to options to modifier to eliminate dual terminology.
1537         * Made a bunch of parser constants and changed all callsites to use them.
1538         * Various other minor clean-ups.
1539
1540         Reviewed by Adam Barth.
1541
1542         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Cleaned up stuff.
1543         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Changed callsites after cleanup.
1544         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Ditto.
1545
1546 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
1547
1548         Move expectation parsing code to TestExpectationParser.
1549         https://bugs.webkit.org/show_bug.cgi?id=64605
1550
1551         This is a somewhat mechanical move, with two interesting bits:
1552         
1553         1) TestExpectationParser.parse methods renamed to tokenize, to better
1554         reflect what they do
1555
1556         2) TestExpectationLine now carries all of its info, from tokens to parsed data,
1557         and even the list of tests that it matches (a line may refer to more than one test).
1558
1559         Reviewed by Adam Barth.
1560
1561         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Moved parsing-related TestExpectaions methods
1562             to TestExpectationParser, added more members to TestExpectationLine to carry parsed info, renamed existing parse methods
1563             to "tokenize", changed callsites to use new code.
1564         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Changed unit tests to reflect renames.
1565
1566 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
1567
1568         Switch to use Python enumerate function to enumerate line numbers.
1569         https://bugs.webkit.org/show_bug.cgi?id=64602
1570
1571         Reviewed by Adam Barth.
1572
1573         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Changed to use enumerate.
1574
1575 2011-07-15  Martin Robinson  <mrobinson@igalia.com>
1576
1577         Build fixes for WebKit2. Ensure that all generated sources are
1578         on nodist primaries, that they are on forward declared variables
1579         so that BUILT_SOURCES is calculated properly and that zlib is 
1580         included during linking (for WOFF support).
1581
1582         * WebKitTestRunner/GNUmakefile.am:
1583
1584 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
1585
1586         Store error and warning information on TestExpectationLine.
1587         https://bugs.webkit.org/show_bug.cgi?id=64565
1588
1589         Keeping errors and warnings on the TestExpectationLine instance allows us to decouple storing errors
1590         from various parsing and validation mechanisms and have more flexibility in reporting and understanding the origin of the errors.
1591
1592         Reviewed by Adam Barth.
1593
1594         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Added TestExpectationLine.warnings list to keep track of non-fatal errors,
1595             converted the code to add errors and warnings to corresponding TestExpectationLine instances, removed the code that used to store
1596             this info on TestExpectations. In the process, had to refactor ModifierMatcher a bit to take in a TestExpdectationLine instance.
1597         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Changed ModifierMatcher callsite.
1598
1599 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
1600
1601         Plumb the use of TestExpectationLine deeper, clean up.
1602         https://bugs.webkit.org/show_bug.cgi?id=64559
1603
1604         Instead of carrying various bits of TestExpectationLine, plumb it down to its consumers,
1605         also cleaning up names and remove an unused TestExpectations._get_options_list member.
1606
1607         Reviewed by Adam Barth.
1608
1609         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
1610
1611 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
1612
1613         Remove the notion of TestExpectationLine.valid, start storing parsing errors in expectations themselves.
1614         https://bugs.webkit.org/show_bug.cgi?id=64554
1615
1616         This moves us toward the world where errors are collected on the expectations, which allows us to
1617         easily enumerate them, keep association with the point of origin, and freely pass TestExpectationLine instances around.
1618
1619         Also eliminate the validator idea, since validation is a context-dependent concept and has to be decoupled from initial
1620         parsing.
1621
1622         Reviewed by Adam Barth.
1623
1624         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Removed TestExpectation.valid, validator,
1625             changed TestExpectationParser to collect errors in TestExpectationLine, refactored surrounding code.
1626         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Changed tests to accommodate changes.
1627         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Ditto.
1628
1629 2011-07-15  Adam Roben  <aroben@apple.com>
1630
1631         Rename TestFailureBugForm to FailingTestsBugForm
1632
1633         The new name will match better with a forthcoming FlakyTestBugForm class.
1634
1635         Fixes <http://webkit.org/b/64598> TestFailures page's TestFailureBugForm class has a bad
1636         name
1637
1638         Reviewed by Daniel Bates.
1639
1640         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js: Renamed from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm.js.
1641         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js: Renamed from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm_unittests.js.
1642
1643         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1644         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
1645         Updated for renames.
1646
1647         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Ditto, and
1648         reordered <script>s to put files which others depend on for parsing first.
1649
1650 2011-07-15  Adam Barth  <abarth@webkit.org>
1651
1652         NRWT stores the Chromium revision number in full_results.json
1653         https://bugs.webkit.org/show_bug.cgi?id=64586
1654
1655         I have no earthly idea how to test this change.
1656
1657         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1658         * Scripts/webkitpy/layout_tests/port/base.py:
1659
1660 2011-07-14  Ojan Vafai  <ojan@chromium.org>
1661
1662         fix remaining style issues in the static-dashboards directory
1663         https://bugs.webkit.org/show_bug.cgi?id=64561
1664
1665         Reviewed by Adam Barth.
1666
1667         * TestResultServer/static-dashboards/aggregate_results.html:
1668         * TestResultServer/static-dashboards/builders.js:
1669         * TestResultServer/static-dashboards/dashboard_base.js:
1670         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
1671         * TestResultServer/static-dashboards/timeline_explorer.html:
1672         * TestResultServer/static-dashboards/treemap.html:
1673
1674 2011-07-14  Yuta Kitamura  <yutak@chromium.org>
1675
1676         WebSocket: Introduce pywebsocket-0.6b2
1677         https://bugs.webkit.org/show_bug.cgi?id=64534
1678
1679         Reviewed by Kent Tamura.
1680
1681         * Scripts/webkitpy/thirdparty/__init__.py:
1682
1683 2011-07-14  Dimitri Glazkov  <dglazkov@chromium.org>
1684
1685         Introduce TestExpectationsModel, split out of TestExpectations.
1686         https://bugs.webkit.org/show_bug.cgi?id=64531
1687
1688         This is a simple split-and-make-work refactoring, a first step among many.
1689
1690         Reviewed by Adam Barth.
1691
1692         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Moved all model-related members
1693             out of TestExpectations and into TestExpectationsModel.
1694
1695 2011-07-14  Eric Seidel  <eric@webkit.org>
1696
1697         Move webkitpy off of loose mocks
1698         https://bugs.webkit.org/show_bug.cgi?id=64508
1699
1700         Unreviewed.  Fixing a test which fails under test-webkitpy --all (but not in a normal run).
1701
1702         * Scripts/webkitpy/common/checkout/scm/git.py:
1703
1704 2011-07-14  Adam Barth  <abarth@webkit.org>
1705
1706         garden-o-matic should have a "rebaseline" button
1707         https://bugs.webkit.org/show_bug.cgi?id=64446
1708
1709         Reviewed by Ojan Vafai.
1710
1711         This patch adds a basic Rebaseline button that copies the baselines
1712         displayed in the results pane into the appropriate directory in your
1713         working copy.
1714
1715         There are two main limitations:
1716
1717         1) There is no UI for actually committing the baselines.
1718
1719         2) The baselines are not optimized for redundancy (meaning you can have
1720            identical baselines in both chromium-mac and chromium-win).
1721
1722         * Scripts/webkitpy/tool/commands/rebaseline.py:
1723             - Turns out we need to create the directory for the baseline if it doesn't exist yet.
1724         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js:
1725             - Add an programatic API to call the server.
1726         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
1727             - Add the rebaseline button itself.
1728         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
1729             - Change the CSS so that the Rebaseline and Close buttons can
1730               appear in the normal order in the DOM.
1731         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
1732             - Bind the event and translate the parameters.
1733             - Hide/show the rebaseline button, as appropriate.
1734         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
1735             - Add some helpful utility functions for manipulating failure types.
1736         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
1737
1738 2011-07-14  Ryosuke Niwa  <rniwa@webkit.org>
1739
1740         REGRESSION: webkit-patch roll-chromium-deps is broken
1741         https://bugs.webkit.org/show_bug.cgi?id=64568
1742
1743         Reviewed by Dirk Pranke
1744
1745         Use _filesystem.join.
1746
1747         * Scripts/webkitpy/common/checkout/checkout.py:
1748         * Scripts/webkitpy/common/checkout/checkout_unittests.py:
1749
1750 2011-07-14  Noam Rosenthal  <noam.rosenthal@nokia.com>
1751
1752         Adding myself to the reviewers list; No review needed.
1753
1754         * Scripts/webkitpy/common/config/committers.py:
1755
1756 2011-07-14  Adam Roben  <aroben@apple.com>
1757
1758         Don't use Element.prototype.classList in TestFailures
1759
1760         Safari 5 doesn't support it.
1761
1762         Fixes <http://webkit.org/b/64550> Can't expand flaky tests on TestFailures page in Safari 5
1763
1764         Reviewed by Daniel Bates.
1765
1766         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
1767         (Element.prototype.hasStyleClass):
1768         (Element.prototype.addStyleClass):
1769         (Element.prototype.removeStyleClass):
1770         (Element.prototype.toggleStyleClass):
1771         Added these helper functions which simulate classList functionality.
1772
1773         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities_unittests.js:
1774         Added. Tests for the above.
1775
1776         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1777         (ViewController.prototype._domForPossiblyFlakyTests): Changed to use
1778         toggleStyleClass/hasStyleClass instead of classList.
1779
1780         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
1781         Added Utilities_unittests.js, and reordered the tested files to be in
1782         roughly dependency order (i.e., the lowest-level files are imported
1783         and tested first).
1784
1785 2011-07-14  Eric Seidel  <eric@webkit.org>
1786
1787         NRWT doesn't store the svn revision in full_results.json on chromium-win
1788         https://bugs.webkit.org/show_bug.cgi?id=64492
1789
1790         Unreviewed.  Just fixing my test-webkitpy regression.
1791
1792         Fix the unit tests.  Unfortunately scm does not use a filesystem
1793         object so we can't control the result of detect_scm_system.
1794         When detect_scm_system would fail, we would log, which would
1795         cause all passing_run integration tests to fail.
1796
1797         * Scripts/webkitpy/common/checkout/scm/scm.py:
1798         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1799         * Scripts/webkitpy/layout_tests/port/base.py:
1800         * Scripts/webkitpy/layout_tests/port/test.py:
1801         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1802
1803 2011-07-14  Eric Seidel  <eric@webkit.org>
1804
1805         NRWT doesn't store the svn revision in full_results.json on chromium-win
1806         https://bugs.webkit.org/show_bug.cgi?id=64492
1807
1808         Reviewed by Ojan Vafai.
1809
1810         This should fix the bug.  Unfortunately this code is currently
1811         impossible to test since it's impossible to mock detect_scm_system at the
1812         moment.  I started re-writing scm.detection.py to be mockable, but decided that
1813         was best left for another day.
1814
1815         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1816         * Scripts/webkitpy/layout_tests/port/base.py:
1817
1818 2011-07-14  Ojan Vafai  <ojan@chromium.org>
1819
1820         fix coding style of dashboard_base.js
1821         https://bugs.webkit.org/show_bug.cgi?id=64545
1822
1823         Reviewed by Adam Roben.
1824
1825         The code changes to the files other than dashboard_base.js
1826         are just fallout from renaming globals in dashboard_base.js.
1827
1828         * TestResultServer/static-dashboards/aggregate_results.html:
1829         * TestResultServer/static-dashboards/builders.js:
1830         * TestResultServer/static-dashboards/dashboard_base.js:
1831         * TestResultServer/static-dashboards/flakiness_dashboard.html:
1832         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
1833         * TestResultServer/static-dashboards/timeline_explorer.html:
1834         * TestResultServer/static-dashboards/treemap.html:
1835
1836 2011-07-14  Carlos Garcia Campos  <cgarcia@igalia.com>
1837
1838         Unreviewed. Fix WebKit2 GTK build after r90953.
1839
1840         * WebKitTestRunner/GNUmakefile.am:
1841
1842 2011-07-14  Adam Barth  <abarth@webkit.org>
1843
1844         garden-o-matic results pane should be more discoverable
1845         https://bugs.webkit.org/show_bug.cgi?id=64513
1846
1847         Reviewed by Eric Seidel.
1848
1849         This patch causes us to trigger the results pane on mouse clicks
1850         instead of mousein.  Also, handle the case of no results details more
1851         elegantly.
1852
1853         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
1854         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
1855             - Change the binding to "click".
1856         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
1857             - Handle the "no results URLs" case explicitly instead of never
1858               calling the callback.
1859         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
1860         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
1861             - Show a message when there are no result details to show.
1862         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
1863
1864 2011-07-14  Adam Barth  <abarth@webkit.org>
1865
1866         garden-o-matic should badge TIMEOUT failures
1867         https://bugs.webkit.org/show_bug.cgi?id=64435
1868
1869         Reviewed by Eric Seidel.
1870
1871         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
1872         (.regression .what a[draggable].TIMEOUT::before):
1873             - Add TIMEOUT.
1874
1875 2011-07-14  Adam Barth  <abarth@webkit.org>
1876
1877         Fix Python exception blocking the commit-queue.
1878
1879         * Scripts/webkitpy/common/checkout/scm/git.py:
1880
1881 2011-07-13  Eric Seidel  <eric@webkit.org>
1882
1883         Move webkitpy off of loose mocks
1884         https://bugs.webkit.org/show_bug.cgi?id=64508
1885
1886         Reviewed by Adam Barth.
1887
1888         Using Mock has caused us more pain than help.
1889         It's possible that there was a cleaner way to use it
1890         (maybe Mock(class) instead of inheriting from it?).
1891         But for now, I've removed all uses of Mock from mocktool.py.
1892
1893         I also moved run_command into the only 3 files which call it
1894         instead of leaving the deprecated method in executive.py.
1895
1896         I changed various direct calls to os.* to use filesystem instead.
1897
1898         * Scripts/webkitpy/common/checkout/checkout.py:
1899         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1900         * Scripts/webkitpy/common/checkout/scm/git.py:
1901         * Scripts/webkitpy/common/checkout/scm/scm.py:
1902         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
1903         * Scripts/webkitpy/common/checkout/scm/svn.py:
1904         * Scripts/webkitpy/common/system/executive.py:
1905         * Scripts/webkitpy/common/system/executive_unittest.py:
1906         * Scripts/webkitpy/tool/mocktool.py:
1907         * Scripts/webkitpy/tool/steps/cleanworkingdirectory_unittest.py:
1908         * Scripts/webkitpy/tool/steps/ensurelocalcommitifneeded.py:
1909
1910 2011-07-13  Eric Seidel  <eric@webkit.org>
1911
1912         NRWT doesn't store the svn revision in full_results.json on chromium-win
1913         https://bugs.webkit.org/show_bug.cgi?id=64492
1914
1915         Reviewed by Adam Barth.
1916
1917         Add a warning message in the revision='' case to help us diagnose
1918         what's going wrong on the bot.
1919
1920         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1921
1922 2011-07-13  Dirk Pranke  <dpranke@chromium.org>
1923
1924         new-run-webkit-tests: print baseline search path as part of config output
1925         https://bugs.webkit.org/show_bug.cgi?id=64499
1926
1927         Reviewed by Eric Seidel.
1928
1929         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1930         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
1931
1932 2011-07-13  Adam Barth  <abarth@webkit.org>
1933
1934         gardening server should have an API for parsing changelogs
1935         https://bugs.webkit.org/show_bug.cgi?id=64495
1936
1937         Reviewed by Eric Seidel.
1938
1939         This patch exposes much of the same information from CommitInfo in a
1940         dictionary form, which is easier to send over-the-wire as JSON to the
1941         frontend.
1942
1943         * Scripts/webkitpy/common/checkout/checkout.py:
1944         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1945         * Scripts/webkitpy/tool/mocktool.py:
1946         * Scripts/webkitpy/tool/servers/gardeningserver.py:
1947         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
1948         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
1949
1950 2011-07-13  Eric Seidel  <eric@webkit.org>
1951
1952         REGRESSION: GitTestWithMock.test_create_patch fails
1953         https://bugs.webkit.org/show_bug.cgi?id=62945
1954
1955         Reviewed by Daniel Bates.
1956
1957         I was not able to reproduce the exact failure seen in the bug,
1958         however this test was failing on my machine for other reasons.
1959
1960         I went through and did an audit of our run_command usage, it's
1961         entirely in scm classes after this change.  (Not surprising given
1962         that scm.py was the second file ever created in webkit.py.)
1963
1964         The real bug I'm fixing here is that we were setting executive.should_log
1965         when the value had changed to executive._should_log.  Now we set the right one
1966         and the test works again.
1967
1968         * Scripts/webkitpy/common/checkout/checkout.py:
1969         * Scripts/webkitpy/common/checkout/scm/git.py:
1970         * Scripts/webkitpy/common/checkout/scm/scm.py:
1971         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
1972         * Scripts/webkitpy/common/checkout/scm/svn.py:
1973
1974 2011-07-13  Ilya Sherman  <isherman@chromium.org>
1975
1976         Fix WTF header guard style check
1977         https://bugs.webkit.org/show_bug.cgi?id=64488
1978
1979         Reviewed by David Levin.
1980
1981         * Scripts/webkitpy/style/checkers/cpp.py:
1982         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Verify that we don't *always* suggest "WTF_" as a prefix
1983
1984 2011-07-13  Ojan Vafai  <ojan@chromium.org>
1985
1986         bring flakiness_dashboard.html closer to webkit style
1987         https://bugs.webkit.org/show_bug.cgi?id=64477
1988
1989         Reviewed by Adam Barth.
1990
1991         * TestResultServer/static-dashboards/dashboard_base.js:
1992         Fix bad variable name that would hit only when running the tests.
1993         * TestResultServer/static-dashboards/flakiness_dashboard.html:
1994         No code/logic changes. All moving/removing brackets, indents and moving things
1995         to one line.
1996
1997 2011-07-13  Xan Lopez  <xlopez@igalia.com>
1998
1999         [GTK] Fix distcheck
2000
2001         Reviewed by Martin Robinson.
2002
2003         * WebKitTestRunner/GNUmakefile.am: mark built sources as nodist.
2004
2005 2011-07-13  Joseph Pecoraro  <joepeck@webkit.org>
2006
2007         Unreviewed fix. Add a "\n" to a printf that somehow got lost.
2008
2009         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2010         (WebCore::DumpRenderTree::dumpApplicationCacheQuota):
2011
2012 2011-07-13  Joseph Pecoraro  <joepeck@webkit.org>
2013
2014         ApplicationCache Quota Output is Flakey
2015         https://bugs.webkit.org/show_bug.cgi?id=64410
2016
2017         Reviewed by Alexey Proskuryakov.
2018
2019         Unify the delegate logging for reaching application cache
2020         quotas. Truncate the space needed to the nearest 10000
2021         for less flakey test output. This also better supports
2022         printing NSUInteger on non-64 bit systems by casting to
2023         an unsigned long.
2024
2025         * DumpRenderTree/mac/UIDelegate.mm:
2026         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:totalSpaceNeeded:]):
2027         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2028         (WebCore::DumpRenderTree::dumpApplicationCacheQuota):
2029
2030 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
2031
2032         Remove unused TestExpectations._overrides.
2033         https://bugs.webkit.org/show_bug.cgi?id=64470
2034
2035         Reviewed by Adam Barth.
2036
2037         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Removed unused member.
2038
2039 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
2040
2041         Introduce TestExpectationSerializer.list_to_string.
2042         https://bugs.webkit.org/show_bug.cgi?id=64462
2043
2044         Reviewed by Adam Barth.
2045
2046         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Added list_to_string and change the relevant callsite to use it.
2047         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added tests for it.
2048
2049 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
2050
2051         Consolidate expectations parsing code.
2052         https://bugs.webkit.org/show_bug.cgi?id=64460
2053
2054         Reviewed by Adam Barth.
2055
2056         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Folded TestExpectationParser._split_expectation_string into TestExpectationParser.parse.
2057
2058 2011-07-12  Brent Fulgham  <bfulgham@webkit.org>
2059
2060         Standardize WinCairo conditionalized code under PLATFORM macro.
2061         https://bugs.webkit.org/show_bug.cgi?id=64377
2062
2063         Reviewed by Maciej Stachowiak.
2064
2065         Update compile-time conditions that had previously used #ifdef
2066         of WIN_CAIRO to consistently use PLATFORM(WIN_CAIRO).
2067
2068         * DumpRenderTree/config.h: Switch to PLATFORM(WIN_CAIRO)
2069         * DumpRenderTree/win/DumpRenderTreeCairo.vsprops: Remove redundant
2070           define of WIN_CAIRO=1.
2071         * DumpRenderTree/win/DumpRenderTreeDebugCairoCFLite.vsprops: Add
2072           WinCairo.vsprops to set (to match Release build).
2073
2074 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
2075
2076         Eliminate TestExpectationsFile.
2077         https://bugs.webkit.org/show_bug.cgi?id=64458
2078
2079         Turns out, we can just use a Python list.
2080
2081         Reviewed by Adam Barth.
2082
2083         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Folded TestExpectationsFile.append into TestExpectationParser.parse_list,
2084             removed TestExpectationsFile.
2085         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Moved tests to better reflect new names, removed iterator test,
2086             since there's no more custom iterator machinery.
2087
2088 2011-07-13  Philippe Normand  <pnormand@igalia.com>
2089
2090         Unreviewed, added my other email addresses.
2091
2092         * Scripts/webkitpy/common/config/committers.py:
2093
2094 2011-07-13  Adam Barth  <abarth@webkit.org>
2095
2096         Folks should only be listed once.
2097
2098         * Scripts/webkitpy/common/config/committers.py:
2099
2100 2011-07-13  Philippe Normand  <pnormand@igalia.com>
2101
2102         Unreviewed, adding myself as Reviewer.
2103
2104         * Scripts/webkitpy/common/config/committers.py:
2105
2106 2011-07-13  Adam Barth  <abarth@webkit.org>
2107
2108         Fix flickering bug introduced by my previous patch.  I forgot to change
2109         the name of the class everywhere.
2110
2111         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
2112
2113 2011-07-12  Philippe Normand  <pnormand@igalia.com>
2114
2115         [GTK] media/media-blocked-by-willsendrequest.html fails
2116         https://bugs.webkit.org/show_bug.cgi?id=63699
2117
2118         Reviewed by Martin Robinson.
2119
2120         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2121         (willSendRequestCallback): Abort the request if explicitely asked
2122         by the LayoutTestController's willSendRequestReturnsNull() function.
2123
2124 2011-07-13  Adam Barth  <abarth@webkit.org>
2125
2126         Remove "Dismiss" button from garden-o-matic butterbar
2127         https://bugs.webkit.org/show_bug.cgi?id=64443
2128
2129         Reviewed by Dimitri Glazkov.
2130
2131         We don't have any persistent butter bar messages yet, so the dismiss
2132         button is premature (and fairly heavy, visually).
2133
2134         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
2135         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
2136
2137 2011-07-13  Adam Barth  <abarth@webkit.org>
2138
2139         Refine garden-o-matic status pane
2140         https://bugs.webkit.org/show_bug.cgi?id=64442
2141
2142         Reviewed by Dimitri Glazkov.
2143
2144         This patch makes two improvements to the status pane:
2145
2146         1) We only query the server for failure types that we're expecting.
2147            This dramatically reduces the number of HTTP requests, making
2148            loading the status pane faster.
2149
2150         2) The status pane now displays which test and which builder it is
2151            showing results for.
2152
2153         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js:
2154         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
2155         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
2156         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
2157         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
2158         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
2159         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
2160
2161 2011-07-13  Adam Barth  <abarth@webkit.org>
2162
2163         garden-o-matic should show test results
2164         https://bugs.webkit.org/show_bug.cgi?id=64440
2165
2166         Reviewed by Adam Roben.
2167
2168         This patch introduces the results detail pane, which appears at the
2169         bottom of the window and contains test results from the bots.
2170         Currently, you can activate the pane by mousing over one of the builder
2171         names associated with a failing test.
2172
2173         This is just a first iteration of the UI.  There's no way to resize or
2174         zoom in on elements of the details pane, and images likely aren't sized
2175         correctly, but it's a place to start.
2176
2177         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js:
2178             - Add constants for our data attributes so we don't typo them!
2179         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
2180             - Add DOM structure for the details pane.
2181         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
2182         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
2183             - Bind events for showing and hiding the details pane.
2184         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
2185             - Reduce the number of result types that we query for to avoid
2186               overfilling the details pane.  We'll probably need another
2187               solution here in the long-term.  For example, we could use tabs
2188               to pack more results into the pane.
2189         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
2190         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
2191             - Now that we're using the config package during testing, we need
2192               to include it in the testing HTML.
2193         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
2194         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
2195             - Change the DOM structure of the results detail to make it fit
2196               nicely in the details pane (rather than flowing freely in the
2197               body, as it did before).
2198
2199 2011-07-13  Adam Roben  <aroben@apple.com>
2200
2201         Make TestFailures show every time a possibly-flaky test failed, but hide it by default
2202
2203         It's useful to be able to see every time a flaky test failed to see whether it failed the
2204         same way every time. But doing so takes a lot of space, so the list of failures is now
2205         collapsed by default and can be revealed using a disclosure triangle.
2206
2207         Fixes <http://webkit.org/b/64455> TestFailures page doesn't show as much information for
2208         flaky tests as I would like, even though the page is already so long
2209
2210         Reviewed by Dimitri Glazkov.
2211
2212         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
2213         (FlakyLayoutTestDetector.prototype.allFailures): Replaced flakinessExamples with this
2214         function. Now returns all failures for the given test.
2215
2216         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector_unittests.js:
2217         Added. This just contains some simple tests of the FlakyLayoutTestDetector class. We'll add
2218         more over time.
2219
2220         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
2221         (LayoutTestHistoryAnalyzer.prototype.start): Updated the documentation comment to reflect
2222         that we no longer return passing builds for possibly-flaky tests.
2223
2224         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
2225         (.existing-bugs-list, .suspect-revisions-list, .flakiness-examples-list): Make the list of
2226         flakiness examples small, too, since it can get quite long.
2227
2228         (.disclosure-triangle):
2229         (.expanded > .disclosure-triangle):
2230         Simple styles for the disclosure triangle.
2231
2232         (.flakiness-examples-list): Collapse the list by default.
2233
2234         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
2235         (ViewController.prototype._displayBuilder): Pass the total number of builds analyzed to
2236         _domForPossiblyFlakyTests.
2237         (ViewController.prototype._domForPossiblyFlakyTests): Put a disclosure triangle to the left
2238         of each test name, and the number of failures to the right. When the disclosure triangle is
2239         clicked for the first time, we build up the list of failures and expand the element. After
2240         that we just collapse or expand the element on subsequent clicks.
2241
2242         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
2243         Pulled in new tests.
2244
2245 2011-07-13  Adam Roben  <aroben@apple.com>
2246         
2247         Teach TestFailures to understand NRWT's output when it exits early due to too many failures
2248
2249         Fixes <http://webkit.org/b/64456> TestFailures page reports way too many failures when NRWT
2250         exits early
2251
2252         Reviewed by Dimitri Glazkov.
2253
2254         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
2255         (Builder.prototype.getNumberOfFailingTests): Relaxed the "Exiting early" test not to require
2256         it to be at the beginning of the line, since NRWT prints a bunch of junk earlier in the
2257         line. Tightened up the regex that's used to parse the number of failing tests to require the
2258         leading number to be followed by whitespace so that we won't parse the "2011" in
2259         "2011-07-13" as a number of failures.
2260
2261         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
2262         Added a test for the above.
2263
2264 2011-07-12  Dimitri Glazkov  <dglazkov@chromium.org>
2265
2266         Extract model-like TestExpectationLine and TestExpectationFile from TestExpectations.
2267         https://bugs.webkit.org/show_bug.cgi?id=64386
2268
2269         This is the first step in converting TestExpectations to a real model.
2270         * TestExpectationsLine represents a line in the test_expectations.txt file, and
2271         * TestExpectationsFile represents the file, which is a collection of lines.
2272
2273         Reviewed by Adam Barth.
2274
2275         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
2276         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
2277         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2278
2279 2011-07-13  Xan Lopez  <xlopez@igalia.com>
2280
2281         [GTK] Do not grab focus too early in DRT.
2282
2283         Reviewed by Gustavo Noronha.
2284
2285         It causes a layout to happen and a progress signal to be emitted
2286         since r90900, but at this point we don't have a
2287         LayoutTestController object and we'll eventually crash. Since we
2288         already grab focus at the beginning of runTest() this is
2289         redundant, so get rid of it to fix the crash.
2290
2291         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2292         (main): remove call to grab_focus
2293
2294 2011-07-13  Sheriff Bot  <webkit.review.bot@gmail.com>
2295
2296         Unreviewed, rolling out r90893 and r90894.
2297         http://trac.webkit.org/changeset/90893
2298         http://trac.webkit.org/changeset/90894
2299         https://bugs.webkit.org/show_bug.cgi?id=64441
2300
2301         NRWT still doesn't work on qt-mac platform (Requested by Ossy
2302         on #webkit).
2303
2304         * Scripts/run-webkit-tests:
2305         (useNewRunWebKitTests):
2306
2307 2011-07-13  Csaba Osztrogonác  <ossy@webkit.org>
2308
2309         [Qt] NRWT should pick up the right httpd config file
2310         https://bugs.webkit.org/show_bug.cgi?id=64086
2311
2312         * Scripts/run-webkit-tests: Enable NRWT on qt-mac platform after r90810.
2313         (useNewRunWebKitTests):
2314
2315 2011-07-12  Mark Rowe  <mrowe@apple.com>
2316
2317         Fix the 32-bit build.
2318
2319         * DumpRenderTree/mac/UIDelegate.mm:
2320         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:totalSpaceNeeded:]):
2321         Cast the NSUInteger value to unsigned long to match the format specifier.
2322
2323 2011-07-12  Adam Barth  <abarth@webkit.org>
2324
2325         Improve garden-o-matic UI when the bots fail to report revision numbers
2326         https://bugs.webkit.org/show_bug.cgi?id=64427
2327
2328         Reviewed by Dimitri Glazkov.
2329
2330         At least Win (dbg)(2) seems to fail to report the SVN revision number
2331         in full_results.json.  This patch makes garden-o-matic more robust to
2332         missing revision numbers.
2333
2334         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
2335         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
2336         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
2337         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
2338         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
2339
2340 2011-07-12  Adam Barth  <abarth@webkit.org>
2341
2342         Adjust garden-o-matic layout to use fewer lines
2343         https://bugs.webkit.org/show_bug.cgi?id=64422
2344
2345         Reviewed by Ojan Vafai.
2346
2347         This patch effectively merges the test name line with the list of
2348         builders on which the test fails.  Each test failure now occupies two
2349         lines instead of three.
2350
2351         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
2352         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
2353         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
2354
2355 2011-07-12  Adam Barth  <abarth@webkit.org>
2356
2357         garden-o-matic should label tests failures of new tests as such
2358         https://bugs.webkit.org/show_bug.cgi?id=64421
2359
2360         Reviewed by Ojan Vafai.
2361
2362         Also, we shouldn't dim failures of new tests, even if we've only seen
2363         them once because they're likely to be real problems that need
2364         attention.
2365
2366         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
2367         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js: Added.
2368         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout_unittests.js: Added.
2369         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
2370         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
2371         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
2372         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
2373         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
2374         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
2375         * Scripts/webkitpy/tool/servers/gardeningserver.py:
2376
2377 2011-07-12  Adam Barth  <abarth@webkit.org>
2378
2379         garden-o-matic should display how many times we've seen a failure
2380         https://bugs.webkit.org/show_bug.cgi?id=64417
2381
2382         Reviewed by Ojan Vafai.
2383
2384         This patch adds some UI to display how many times we've seen a given
2385         failure, which can be helpful for determining whether that failure is a
2386         real failure or a flaky test.
2387
2388         When a failure has only been seen once (i.e., only a single run on a
2389         single bot), we set the opacity of to 50% to avoid distracting the
2390         gardener.
2391
2392         This patch also refactors the failure walker to have a simpler API
2393         internally by moving from an object-oriented paradigm to a functional
2394         paradigm.
2395
2396         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
2397         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
2398         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
2399         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
2400         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
2401         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
2402
2403 2011-07-12  Chris Rogers  <crogers@google.com>
2404
2405         Enable Web Audio for chromium DRT
2406         https://bugs.webkit.org/show_bug.cgi?id=64409
2407
2408         Reviewed by James Robinson.
2409
2410         * DumpRenderTree/chromium/TestShell.cpp:
2411         (TestShell::TestShell):
2412
2413 2011-07-12  Adam Barth  <abarth@webkit.org>
2414
2415         Tweak some UI in garden-o-matic now that the tree actually has a
2416         failure and I can see what this all looks like.
2417
2418         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
2419         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
2420         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
2421
2422 2011-07-12  Joseph Pecoraro  <joepeck@webkit.org>
2423
2424         Unreviewed. Skipping a few tests which fail due to differing output
2425         in recent ApplicationCache quota tests. Reenabling the tests is
2426         tracked by <http://webkit.org/b/64410>.
2427
2428         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2429         (WebCore::DumpRenderTree::dumpApplicationCacheQuota): This should
2430         have been a space, not a newline.
2431
2432 2011-07-12  Adam Barth  <abarth@webkit.org>
2433
2434         garden-o-matic should display regression ranges
2435         https://bugs.webkit.org/show_bug.cgi?id=64407
2436
2437         Reviewed by Dimitri Glazkov.
2438
2439         This patch computes a regression range for a failure by intersecting
2440         the regression ranges seen by the various bots.  We make the underlying
2441         assumption that a test is only failing due to one revision at any given
2442         moment.  If that's not true, this code probably explodes.
2443
2444         The regression ranges appear asynchronously, which might be a jarring
2445         UI.  We'll have to experiment to see.
2446
2447         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
2448         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
2449         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
2450         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
2451         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
2452         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
2453         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
2454
2455 2011-07-12  Adam Barth  <abarth@webkit.org>
2456
2457         cr-linux-ews should run pixel tests
2458         https://bugs.webkit.org/show_bug.cgi?id=64394
2459
2460         Reviewed by Ojan Vafai.
2461
2462         This patch enabled pixel test on the cr-linux-ews.  Running a quick
2463         test on the bots, it looks like we have five pixel failures, related to
2464         some missing fonts.  I've also made the bots skip failing tests to
2465         avoid uploading huge zip files to bugs.webkit.org (with all the exected
2466         IMAGE failures).
2467
2468         * Scripts/webkitpy/common/config/ports.py:
2469         * Scripts/webkitpy/common/config/ports_unittest.py:
2470
2471 2011-07-12  Joseph Pecoraro  <joepeck@webkit.org>
2472
2473         ApplicationCache update should not immediately fail when reaching per-origin quota
2474         https://bugs.webkit.org/show_bug.cgi?id=64177
2475
2476         Reviewed by Alexey Proskuryakov.
2477
2478         Add layoutTestController.disallowIncreaseForApplicationCacheQuota()
2479         to disable the default behavior of raising the default per-origin
2480         quota to 5MB when reached. This allows us to test what happens
2481         when the quota is reached and not increased.
2482
2483         * DumpRenderTree/LayoutTestController.cpp:
2484         (LayoutTestController::LayoutTestController):
2485         (disallowIncreaseForApplicationCacheQuotaCallback):
2486         (LayoutTestController::staticFunctions):
2487         * DumpRenderTree/LayoutTestController.h:
2488         (LayoutTestController::disallowIncreaseForApplicationCacheQuota):
2489         (LayoutTestController::setDisallowIncreaseForApplicationCacheQuota):
2490         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2491         (LayoutTestController::reset):
2492         * DumpRenderTree/qt/LayoutTestControllerQt.h:
2493         (LayoutTestController::disallowIncreaseForApplicationCacheQuota):
2494         (LayoutTestController::dumpApplicationCacheDelegateCallbacks):
2495         Expose a function to prevent automatically increasing the per-origin
2496         quota when the quota is reached. This allows us to test what happens
2497         when a user would "disallow" a quota increase.
2498
2499         * DumpRenderTree/mac/UIDelegate.mm:
2500         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:spaceNeeded:]):
2501         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2502         (WebCore::DumpRenderTree::dumpApplicationCacheQuota):
2503         Respect the new disallowIncrease flag and dump the spaceNeeded values
2504         for ports that implement application cache quotas.
2505
2506 2011-07-12  Ojan Vafai  <ojan@chromium.org>
2507
2508         fix flakiness dashboard to work with new crash log filenames
2509         https://bugs.webkit.org/show_bug.cgi?id=64393
2510
2511         Reviewed by Adam Barth.
2512
2513         * TestResultServer/static-dashboards/flakiness_dashboard.html:
2514
2515 2011-07-12  Adam Barth  <abarth@webkit.org>
2516
2517         Add a happy animation to garden-o-matic when there are no failures
2518         https://bugs.webkit.org/show_bug.cgi?id=64382
2519
2520         Reviewed by Ojan Vafai.
2521
2522         As requested by Ojan.
2523
2524         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
2525         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
2526         * Scripts/webkitpy/tool/servers/data/gardeningserver/partytime.gif: Added.
2527         * Scripts/webkitpy/tool/servers/gardeningserver.py:
2528
2529 2011-07-12  Adam Barth  <abarth@webkit.org>
2530
2531         sheriffbot can't perform rollouts
2532         https://bugs.webkit.org/show_bug.cgi?id=64370
2533
2534         Reviewed by Eric Seidel.
2535
2536         Now that we're not updating the working copy before processing every
2537         IRC message, the rollout command needs to update it manually.
2538
2539         * Scripts/webkitpy/tool/bot/irc_command.py:
2540         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
2541
2542 2011-07-12  Adam Barth  <abarth@webkit.org>
2543
2544         Fix the grammar in the FIXME comment.
2545
2546         * Scripts/webkitpy/common/net/resultsjsonparser.py:
2547
2548 2011-07-12  Eric Seidel  <eric@webkit.org>
2549
2550         NRWT should open test results page with Safari trunk, not the system provided one on Mac
2551         https://bugs.webkit.org/show_bug.cgi?id=64346
2552
2553         Reviewed by Adam Barth.
2554
2555         To fix this I implemented Port.show_results_html_file in Mac, Gtk and Qt ports with
2556         implementations (mostly) matching those found in old-run-webkit-tests.
2557         There are still some minor differences for Qt which Qt hackers may wish to tweak.
2558
2559         I had to add a WebKitPort._port_flag_for_scripts method (similar to flag() in common.config.ports.py)
2560         for the Qt/Gtk ports which always require a flag passed to scripts.
2561
2562         While trying to test this, I found that FactoryTest.test_chromium_gpu_linux
2563         was using a real filesystem (due to assert_platform) which due to
2564         global variables in config.py was causing set-webkit-configuration to have
2565         an affect on unit test results!  So I fixed this by making FactoryTest.assert_port
2566         pass a mock file system whenever calling factory.get.
2567
2568         Unfortunately TestPort was depending on always being passed a None filesystem
2569         and asserting filesystem._tests (only true for unit_test_filesystem()).
2570         So I just removed the FactoryTest.test_test and FactoryTest.test_dryrun tests
2571         deciding that they were pretty much useless anyway.  If others feel strongly
2572         I'm happy to fix this in a different way.
2573
2574         * Scripts/webkitpy/common/system/executive.py:
2575          - default arguments in python are screwy.  They use a single shared
2576            instance, so it's better to use argument=None and then argument = argument or Default()
2577            if you have any chance of mutating (or returning) the default argument.
2578         * Scripts/webkitpy/layout_tests/port/config.py:
2579          - This code is wrong.  We don't need to use a global variable here (as far as I can tell).
2580            I'm not fixing it in this patch, but I've marked it with a FIXME and we can convert to
2581            storing the results of the read on the Config object (which should only be created once during normal operation).
2582            Unit tests shouldn't be hitting the disk anyway.  It's possible Config should move off of Port and onto Tool/Host directly.
2583         * Scripts/webkitpy/layout_tests/port/factory.py:
2584         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
2585         * Scripts/webkitpy/layout_tests/port/gtk.py:
2586         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py: Added.
2587         * Scripts/webkitpy/layout_tests/port/mac.py:
2588         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2589         * Scripts/webkitpy/layout_tests/port/qt.py:
2590         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
2591         * Scripts/webkitpy/layout_tests/port/webkit.py:
2592
2593 2011-07-12  Adam Roben  <aroben@apple.com>
2594
2595         Teach TestFailures to recognize when run-webkit-tests gets killed by buildbot
2596
2597         Fixes <http://webkit.org/b/64358> TestFailures page thinks all tests passed in
2598         http://build.webkit.org/builders/Windows%207%20Release%20(Tests)/builds/14672
2599
2600         Reviewed by Daniel Bates.
2601
2602         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
2603         (Builder.prototype.getNumberOfFailingTests): If run-webkit-tests exited with a non-zero
2604         exit status but we didn't find any failure counts, assume that there was some error that
2605         caused run-webkit-tests to die early (like being killed by buildbot due to a timeout).
2606
2607         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
2608         Added a new test that shows that we get a failingTestCount of -1 when run-webkit-tests dies
2609         early.
2610         (runGetNumberOfFailingTestsTest): Moved most code here from the only pre-existing test in this
2611         file.
2612
2613         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
2614         (LayoutTestResultsLoader.prototype.start): Bump the cache number so old cached data that was
2615         tainted by the bug fixed in this patch will be evicted.
2616
2617 2011-07-12  Adam Barth  <abarth@webkit.org>
2618
2619         cr-linux-ews complains about tests that aren't actually failing
2620         https://bugs.webkit.org/show_bug.cgi?id=64367
2621
2622         The underlying problem here is that full_results.json doesn't have
2623         enough information to determine whether a given test result was
2624         expected because whether an actual result is expected depends on
2625         whether full_results.json was generated duing a test run that included
2626         pixel tests.
2627
2628         The right long-term solution is to make full_results.json a complete
2629         description of what happened durning a test run.  In this patch
2630         however, to stop the spam, this patch makes the jsonresultsparser
2631         ignore pixel failures.
2632
2633         (I'm landing this patch unreviewed in order to stop the bots from
2634         spamming.  I'll happily address any review comments in a follow-up
2635         patch.)
2636
2637         * Scripts/webkitpy/common/net/resultsjsonparser.py:
2638         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
2639
2640 2011-07-12  Adam Barth  <abarth@webkit.org>
2641
2642         Re-work garden-o-matic UI to begin to look like the mocks
2643         https://bugs.webkit.org/show_bug.cgi?id=64334
2644
2645         Reviewed by Dimitri Glazkov.
2646
2647         This patch changes how we display failures to be more like the mock the
2648         dglazkov made. We now can display the same information much more
2649         compactly.
2650
2651         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
2652         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
2653         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
2654         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css: Added.
2655         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
2656         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
2657         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
2658         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
2659         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
2660         * Scripts/webkitpy/tool/servers/gardeningserver.py:
2661
2662 2011-07-11  Adam Roben  <aroben@apple.com>
2663
2664         Extract TestFailures's bug-filing code into two new classes
2665
2666         Fixes <http://webkit.org/b/64300> TestFailures page's new-bug-filing code is a mess!
2667
2668         Reviewed by Darin Adler and Adam Barth.
2669
2670         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js:
2671         (Buildbot.prototype.resultsDirectoryURL): Changed to return a URI-encoded URL. Otherwise the
2672         URL isn't valid (and it's harder to mock this function).
2673
2674         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot_unittests.js:
2675         Added. Just tests the above change (for now).
2676
2677         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
2678         Added a license header, enclosed everything in a closure to avoid polluting the global
2679         namespace, changed the test name to actually describe the passing condition, and replaced
2680         uses of equals() with equal(). (The latter is the actual name of the function, and matches
2681         deepEqual, etc.)
2682
2683         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm.js: Added.
2684         (NewBugForm): This class knows how to construct a <form> used for filing new bugs in
2685         Bugzilla based on some parameters.
2686         (NewBugForm.prototype.domElement): Creates and returns the <form> element. Code came from
2687         ViewController.prototype._domForNewAndExistingBugs.
2688
2689         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm_unittests.js:
2690         Added. Tests the above code.
2691
2692         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm.js: Added.
2693         (TestFailureBugForm): This class knows how to construct a <form> element used for filing new
2694         bugs specifically about test failures. Code came from
2695         ViewController.prototype._domForNewAndExistingBugs.
2696         (TestFailureBugForm.prototype.domElement): Slightly customizes the <form> element returned
2697         by NewBugForm.
2698         (TestFailureBugForm.prototype._computeOperatingSystem):
2699         (TestFailureBugForm.prototype._computePlatform):
2700         (TestFailureBugForm.prototype._createBugTitle):
2701         (TestFailureBugForm.prototype._failingResultsHTMLURL):
2702         (TestFailureBugForm.prototype._failingRevision):
2703         (TestFailureBugForm.prototype._passingRevision):
2704         (TestFailureBugForm.prototype._regressionRangeString):
2705         Code came from ViewController.prototype._domForNewAndExistingBugs. I broke it out into
2706         separate functions to break up the rat's nest a bit.
2707
2708         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm_unittests.js:
2709         Added. Tests the above code.
2710
2711         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
2712         (ViewController.prototype._domForNewAndExistingBugs): Moved code from here to
2713         TestFailureBugForm/NewBugForm, and changed this code to use a TestFailureBugForm.
2714
2715         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
2716         NewBugForm/TestFailureBugForm.js.
2717
2718         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
2719         Added new tests and required files.
2720
2721 2011-07-12  Adam Roben  <aroben@apple.com>
2722
2723         Test that no intermediate WTF::Strings are created when concatenating with string literals
2724
2725         Test for <http://webkit.org/b/63330> Concatenating string literals and WTF::Strings using
2726         operator+ is suboptimal
2727
2728         Reviewed by Darin Adler.
2729
2730         * TestWebKitAPI/Tests/WTF/StringOperators.cpp: Added.
2731         (TestWebKitAPI::TEST): Test that a bunch of different string concatenation expressions don't
2732         create any intermediate WTF::Strings while they're being evaluated.
2733
2734         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2735         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
2736         Added new file.
2737
2738 2011-07-12  Eric Seidel  <eric@webkit.org>
2739
2740         [Qt] NRWT should pick up the right httpd config file
2741         https://bugs.webkit.org/show_bug.cgi?id=64086
2742
2743         Reviewed by Adam Barth.
2744
2745         This is more fall-out from the Port class inappropriately encapsulating
2746         both platform and port knowledge.  We need to split out some of this
2747         platform knowledge into a separate class which can be better shared
2748         between ports.
2749
2750         The fix was to move all the _path_to_apache_config_file logic down
2751         into the WebKitPort baseclass so that all of the ports can find the
2752         right apache config file, regardless of what platform they may be running on.
2753
2754         I did not move this down into base.Port (even though I considered it).
2755         Chromium duplicates some of this logic, but since they have separate
2756         subclasses for each port-OS combination (e.g. ChromiumMac, ChromiumWin)
2757         they wouldn't notice the change.  Eventually we'll move this logic
2758         out of Port entirely, and then it will be shared by all ports.
2759
2760         I also cleaned up the http-config logic for ORWT while I was there,
2761         although since we're killing that code, I'm happy to revert that part
2762         of the change if changing it is viewed as needlessly risky.
2763
2764         During this cleanup, I noticed that no ports use the "default" httpd.conf
2765         which is left over from Mac OS X Tiger (and old linux distros) which used
2766         Apache 1.3.  I've removed httpd.conf (and associated support in ORWT) since
2767         we no longer support any ports which use this httpd.conf.
2768
2769         * Scripts/webkitperl/httpd.pm:
2770         (getHTTPDConfigPathForTestDirectory):
2771         * Scripts/webkitpy/layout_tests/port/base.py:
2772         * Scripts/webkitpy/layout_tests/port/gtk.py:
2773         * Scripts/webkitpy/layout_tests/port/mac.py:
2774         * Scripts/webkitpy/layout_tests/port/qt.py:
2775         * Scripts/webkitpy/layout_tests/port/webkit.py:
2776         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2777         * Scripts/webkitpy/layout_tests/port/win.py:
2778         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
2779
2780 2011-07-11  Hironori Bono  <hbono@chromium.org>
2781
2782         Reviewed by Adam Roben.
2783
2784         Implement layoutTestController.setTextDirection for WebKit2 and Windows.
2785         https://bugs.webkit.org/show_bug.cgi?id=61931
2786
2787         This change implements layoutTextController.setTextDirection for WebKit2
2788         and Windows so we can run a layout test added by r87770 on them. Each
2789         implementation adds a wrapper function for Editor::setBaseWritingDirection()
2790         so LayoutTestController can call it.
2791
2792         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
2793         (LayoutTestController::setTextDirection): Implemented the binding function
2794         for layoutTestController.setTextDirection (Windows).
2795         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
2796         Added a binding function for layoutTestController.setTextDirection (WebKit2).
2797         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp: ditto.
2798         (WTR::LayoutTestController::setTextDirection):
2799         * WebKitTestRunner/InjectedBundle/LayoutTestController.h: ditto.
2800
2801 2011-07-11  Raphael Kubo da Costa  <kubo@profusion.mobi>
2802
2803         [EFL] ImageDiff: Make sure gEcoreEvas is destroyed at the right time.
2804         https://bugs.webkit.org/show_bug.cgi?id=64293
2805
2806         Reviewed by Kent Tamura.
2807
2808         gEcoreEvas was being destroyed after shutdownEfl() was called, so the
2809         call to ecore_evas_free() failed.
2810
2811         We now shut it down manually before shutdownEfl() is called.
2812
2813         * DumpRenderTree/efl/ImageDiff.cpp:
2814         (main): Destroy gEcoreEvas before shutting down the EFL.
2815
2816 2011-07-11  Ryosuke Niwa  <rniwa@webkit.org>
2817
2818         Make Chromium Mac and Linux tests bot core builders
2819         https://bugs.webkit.org/show_bug.cgi?id=63196
2820
2821         Reviewed by Adam Barth.
2822
2823         Added Chromium Mac and Linux Release Tests bots core builders.
2824
2825         Also rearranged the order of bots so that test bots show up right next to
2826         respective builders.
2827
2828         * BuildSlaveSupport/build.webkit.org-config/config.json:
2829         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
2830         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2831
2832 2011-07-11  Raphael Kubo da Costa  <kubo@profusion.mobi>
2833
2834         [EFL] DRT: Add code related to font management.
2835         https://bugs.webkit.org/show_bug.cgi?id=63989
2836
2837         Reviewed by Kent Tamura.
2838
2839         The added files are responsible for managing fontconfig and adding the
2840         fonts required by some layout tests to fontconfig's path.
2841
2842         * DumpRenderTree/efl/FontManagement.cpp: Added.
2843         (getFontDirectories):
2844         (getFontFiles):
2845         (addFontDirectories):
2846         (addFontFiles):
2847         (addFontsToEnvironment):
2848         * DumpRenderTree/efl/FontManagement.h: Added.
2849
2850 2011-07-08  Dirk Pranke  <dpranke@chromium.org>
2851
2852         nrwt: linting fixes
2853         https://bugs.webkit.org/show_bug.cgi?id=64225
2854
2855         Reviewed by Eric Siedel.
2856
2857         Miscellaneous linting fixes. The most notable change is that
2858         we add public attributes for user, executive, filesystem, and
2859         options on the Port object, so we don't have to refer to the
2860         "protected" versions all over the place".
2861
2862         * Scripts/webkitpy/layout_tests/controllers/manager.py:
2863         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
2864         * Scripts/webkitpy/layout_tests/controllers/message_broker.py:
2865         * Scripts/webkitpy/layout_tests/controllers/message_broker_unittest.py:
2866         * Scripts/webkitpy/layout_tests/controllers/worker.py:
2867         * Scripts/webkitpy/layout_tests/port/base.py:
2868         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2869
2870 2011-07-11  Ryosuke Niwa  <rniwa@webkit.org>
2871
2872         webkit-patch roll-chromium-deps no longer works
2873         https://bugs.webkit.org/show_bug.cgi?id=64324
2874
2875         Reviewed by James Robinson.
2876
2877         Lower the minimum commit message length to 10.
2878
2879         * Scripts/webkitpy/tool/steps/commit.py:
2880
2881 2011-07-11  Alice Boxhall  <aboxhall@chromium.org>
2882
2883         Convert json_results_generator.py to output version 4 JSON.
2884         https://bugs.webkit.org/show_bug.cgi?id=60869
2885
2886         Reviewed by Ojan Vafai.
2887
2888         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
2889         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2890         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2891
2892 2011-07-11  Martin Robinson  <mrobinson@igalia.com>
2893
2894         [GTK] [NRWT] Pixel tests do not work
2895         https://bugs.webkit.org/show_bug.cgi?id=64091
2896
2897         Reviewed by Eric Seidel.
2898
2899         Fix pixel test runs for GTK+ new-run-webkit-tests by exposing the
2900         path to the ImageDiff binary.
2901
2902         * Scripts/webkitpy/layout_tests/port/gtk.py: Implement _path_to_image_diff.
2903
2904 2011-07-11  Dirk Pranke  <dpranke@chromium.org>
2905
2906         Fix hang on win32 in a multiprocessing test that shouldn't have been running.
2907
2908         Unreviewed, build fix.
2909
2910         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2911
2912 2011-07-11  Adam Roben  <aroben@apple.com>
2913
2914         Don't count leaks as test failures on TestFailures's front page
2915
2916         As a bonus, this patch adds our first unit test.
2917
2918         Fixes <http://webkit.org/b/64303> REGRESSION (r90489): TestFailures page says far more tests
2919         are failing on the Leaks bot than actually are
2920
2921         Reviewed by Adam Barth.
2922
2923         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
2924         (Builder.prototype.getNumberOfFailingTests): Exclude lines containing "leak" when summing
2925         failure counts.
2926
2927         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
2928         Added. Contains a simple test to show that leaks aren't counted as test failures.
2929
2930         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
2931         Added. This is the test harness.
2932
2933 2011-06-14  Mike Stegeman  <mrstegeman@gmail.com>
2934
2935         [GTK] Add API to allow setting local storage database path
2936         https://bugs.webkit.org/show_bug.cgi?id=62091
2937
2938         Reviewed by Martin Robinson and Gustavo Noronha.
2939
2940         Expose the path of the localStorage databases through a setting
2941         to allow HTML5 localStorage to be persistent. New setting is:
2942         html5-local-storage-database-path
2943
2944         * DumpRenderTree/gtk/DumpRenderTree.cpp:
2945         (resetDefaultsToConsistentValues):
2946
2947 2011-07-11  Adam Barth  <abarth@webkit.org>
2948
2949         garden-o-matic should be able to rebaseline tests
2950         https://bugs.webkit.org/show_bug.cgi?id=64186
2951
2952         Reviewed by Eric Seidel.
2953
2954         This patch contains a simple HTTP server binding for the
2955         rebaseline-test command.  The frontend will appear in another patch.
2956
2957         * Scripts/webkitpy/tool/servers/gardeningserver.py:
2958         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
2959         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
2960
2961 2011-07-11  Adam Barth  <abarth@webkit.org>
2962
2963         Add some basic UI for showing regression ranges
2964         https://bugs.webkit.org/show_bug.cgi?id=64243
2965
2966         Reviewed by Eric Seidel.
2967
2968         This UI is all temporary.  It exists mostly just to have some buttons
2969         to click to exercise the code.
2970
2971         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
2972         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
2973         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
2974         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
2975         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
2976
2977 2011-07-11  Adam Barth  <abarth@webkit.org>
2978
2979         garden-o-matic should be able to determine which revisions caused a given failure
2980         https://bugs.webkit.org/show_bug.cgi?id=64189
2981
2982         Reviewed by Adam Roben.
2983
2984         Walking the failure history looking for failures turns out to be
2985         slightly tricky because the network requests are asynchronous.
2986         Currently we do all the fetches serially and our cache is unbounded.
2987         We'll probably optimize both those parameters eventually.
2988
2989         This patch also generalizes some functionality in the unit testing
2990         framework to make testing this sort of code easier.
2991
2992         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
2993         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
2994         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
2995
2996 2011-07-11  Adam Barth  <abarth@webkit.org>
2997
2998         Add a webkit-patch command for rebaselining an individual test
2999         https://bugs.webkit.org/show_bug.cgi?id=64246
3000
3001         Reviewed by Eric Seidel.
3002
3003         This patch introduces a command that's able to rebaseline a single
3004         test.  Currently, the command works only with the build.chromium.org
3005         buildbots, but extending it to work with the build.webkit.org bots
3006         shouldn't be that hard.
3007
3008         A complete rebaseling tool should also include an "optimize baselines"
3009         command (which moves/deletes baselines in order to reduce the number of
3010         expected results files), but that will come in a future patch.
3011
3012         Really BuilderToPort should be merged into builders.py, but I'm going
3013         to save that for a future patch as well.  (We need to stop shaving yaks
3014         at some point.)
3015
3016         * Scripts/webkitpy/tool/commands/rebaseline.py:
3017
3018 2011-07-11  Adam Barth  <abarth@webkit.org>
3019
3020         garden-o-matic should be able to roll out patches
3021         https://bugs.webkit.org/show_bug.cgi?id=64185
3022
3023         Reviewed by Eric Seidel.
3024
3025         This gardening server API is simply a binding to the rollout machinery
3026         already present in webkit-patch.
3027
3028         * Scripts/webkitpy/tool/commands/gardenomatic.py:
3029         * Scripts/webkitpy/tool/servers/gardeningserver.py:
3030         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Added.
3031
3032 2011-07-11  Csaba Osztrogonác  <ossy@webkit.org>
3033
3034         [Qt][Mac] Unreviewed fix after r90746.
3035
3036         * Scripts/run-webkit-tests:
3037         (useNewRunWebKitTests): Really disable NRWT for qt-mac platform.
3038
3039 2011-07-11  Adam Roben  <aroben@apple.com>
3040
3041         Teach TestFailures that ORWT's results.html file might be missing due to all tests passing
3042
3043         There are three reasons why we might fail to fetch ORWT's results.html:
3044           - All tests passed, so no results.html was generated
3045           - Some error during the test run caused results.html not to be generated (e.g., ORWT
3046             timed out)
3047           - Some network error occurred when fetching results.html
3048
3049         We were failing to account for the first possibility in some cases. For test runs before
3050         r89610, we first check build.webkit.org/json to determine how many tests failed and whether
3051         ORWT exited early due to too many failures; if all tests passed then we don't fetch
3052         results.html at all. r89610 changed ORWT to put information in results.html about exiting
3053         early due to too many failures, so we no longer needed to check build.webkit.org/json to get
3054         that information, and in r89619 I changed TestFailures to do just that. But I forgot that we
3055         still needed to check build.webkit.org/json to find out if all tests passed!
3056
3057         Now, for test runs after r89610, we check results.html first, and then check
3058         build.webkit.org/json if we fail to fetch results.html. This lets us distinguish between all
3059         tests passing and the error cases.
3060
3061         Fixes <http://webkit.org/b/64280> TestFailures page can't pinpoint that r90699 caused 13
3062         tests to fail on Windows 7 Release (WebKit2 Tests)
3063
3064         Reviewed by David Kilzer.
3065
3066         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
3067         (LayoutTestResultsLoader.prototype.start): Bumped the cache version so that old, buggy
3068         cached data will get evicted. We were marking builds where all tests passed as errors!
3069         (LayoutTestResultsLoader.prototype._fetchAndParseORWTResults): Added success/error callback
3070         parameters to the fetchAndParseResultsHTML helper function, and added a similar
3071         fetchNumberOfFailingTests function that fetches data from build.webkit.org/json (code came
3072         from later in the function). For test runs before r89610, we first check
3073         build.webkit.org/json then check results.html. For builds after r89610, we first check
3074         results.html and then check build.webkit.org/json if we couldn't fetch results.html.
3075
3076 2011-07-11  Csaba Osztrogonác  <ossy@webkit.org>
3077
3078         new-run-webkit-tests does not support qt-arm or qt-4.8 results
3079         https://bugs.webkit.org/show_bug.cgi?id=64071
3080
3081         [Qt] NRWT should pick up the right httpd config file
3082         https://bugs.webkit.org/show_bug.cgi?id=64086
3083
3084         Reviewed by Zoltán Herczeg.
3085
3086         * Scripts/run-webkit-tests: Disable NRWT for qt-mac, qt-arm and qt-4.8 platforms until fix.
3087         (useNewRunWebKitTests):
3088
3089 2011-07-11  Kenichi Ishibashi  <bashi@chromium.org>
3090
3091         Add TestNetscapePlugIn/Tests/FormValue.cpp to DRT build files
3092         https://bugs.webkit.org/show_bug.cgi?id=64248
3093
3094         Reviewed by Kent Tamura.
3095
3096         * DumpRenderTree/DumpRenderTree.gypi: Added FormValue.cpp.
3097         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Ditto.
3098         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro: Ditto.
3099         * GNUmakefile.am: Ditto.
3100
3101 2011-07-10  Kenichi Ishibashi  <bashi@chromium.org>
3102
3103         Let plugins participate in form submission
3104         https://bugs.webkit.org/show_bug.cgi?id=13061
3105
3106         Adds NPPVformValue support to the testing plugin.
3107
3108         Reviewed by Anders Carlsson.
3109
3110         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added FormValue.cpp
3111         * DumpRenderTree/TestNetScapePlugIn/Tests/FormValue.cpp: Added.
3112         (FormValue::FormValue): Ditto.
3113         (FormValue::NPP_GetValue): Ditto.
3114
3115 2011-07-11  Roland Steiner  <rolandsteiner@chromium.org>
3116
3117         Reviewed by Tony Chang.
3118
3119         Complete functions in filesystem.py
3120         https://bugs.webkit.org/show_bug.cgi?id=63528
3121
3122         - missing text file functions added
3123         - functions sorted
3124         - removed 'append' optional parameters (were unused)
3125         - adapted filesystem_mock in the same way
3126
3127         * Scripts/webkitpy/common/system/filesystem.py:
3128         * Scripts/webkitpy/common/system/filesystem_mock.py:
3129         * Scripts/webkitpy/common/system/filesystem_unittest.py:
3130
3131 2011-07-10  Adam Barth  <abarth@webkit.org>
3132
3133         Clean up style in fallback path calculation
3134         https://bugs.webkit.org/show_bug.cgi?id=64239
3135
3136         Reviewed by Daniel Bates.
3137
3138         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
3139         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
3140         * Scripts/webkitpy/layout_tests/port/mac.py:
3141         * Scripts/webkitpy/layout_tests/port/win.py:
3142
3143 2011-07-10  Balazs Kelemen  <kbalazs@webkit.org>
3144
3145         [Qt][WK2] Implement the rest of WTR::PlatformWebView
3146         https://bugs.webkit.org/show_bug.cgi?id=63630
3147
3148         Reviewed by Kenneth Rohde Christiansen.
3149
3150         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
3151         (WTR::PlatformWebView::windowFrame):
3152         (WTR::PlatformWebView::setWindowFrame):
3153
3154 2011-07-10  Mark Rowe  <mrowe@apple.com>
3155
3156         Teach build-webkit how to find the latest WebKitSystemInterface binary.
3157
3158         * Scripts/build-webkit:
3159
3160 2011-07-09  Adam Roben  <aroben@apple.com>
3161
3162         Teach TestFailures to abbreviate the examples of test flakiness
3163
3164         These lists can get quite long, and it's not really helpful in most cases to have soooooo
3165         many examples of flakiness.
3166
3167         Fixes <http://webkit.org/b/64203> Lists of flaky revisions on TestFailures page can get so
3168         long they're hard to navigate
3169
3170         Reviewed by Dan Bates.
3171
3172         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
3173         (FlakyLayoutTestDetector.prototype.flakinessExamples): If we have more than a certain number
3174         of examples, replace the middle items with a separator.
3175
3176         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
3177         (.flakiness-example-separator): Added styles for the separator.
3178
3179         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
3180         (ViewController.prototype._domForPossiblyFlakyTests): Use a vertical ellipsis to represent
3181         the separator.
3182
3183 2011-07-09  Dirk Pranke  <dpranke@chromium.org>
3184
3185         nrwt: stack traces from worker-side exceptions aren't very useful inside test-webkitpy
3186         https://bugs.webkit.org/show_bug.cgi?id=64218
3187
3188         Reviewed by Eric Seidel.
3189
3190         Exceptions aren't picklable and can't be sent across the
3191         manager/worker message queue without losing information. NRWT
3192         handles this by turning the stack trace into a set of strings,
3193         and logging the strings when we receive an exception from the
3194         worker. However, when you are running tests and something
3195         crashes on the worker side, test-webkitpy prints the
3196         manager-side stack trace, which is just confusing and useless.
3197
3198         This patch changes the logic so that exceptions are passed
3199         through as-is when the worker and manager are in the same
3200         process (the --worker-model=inline option). This increases the
3201         code paths slightly but makes crashes much more useful.
3202
3203         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3204         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
3205         * Scripts/webkitpy/layout_tests/controllers/message_broker.py:
3206         * Scripts/webkitpy/layout_tests/controllers/worker.py:
3207         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3208
3209 2011-07-09  Darin Fisher  <darin@chromium.org>
3210
3211         Eliminate bad dependency on gfx::Point.
3212         https://bugs.webkit.org/show_bug.cgi?id=64228
3213
3214         Reviewed by Kent Tamura.
3215
3216         * DumpRenderTree/chromium/EventSender.cpp:
3217         (initMouseEvent):
3218
3219 2011-07-08  Ojan Vafai  <ojan@chromium.org>
3220
3221         pull static dashboard files into the appengine server from the chromium repository
3222         https://bugs.webkit.org/show_bug.cgi?id=64208
3223
3224         Reviewed by Dirk Pranke.
3225
3226         These files belong in the WebKit repo since they are tied primarily to the webkit tests.
3227         They have some extra bits to support chromium's gtests, but that seems fine.
3228         Mainly, this will allow other WebKit hackers to hack on the dashboards.
3229
3230         As a nice side-effect, we can now push the dashboard files when we do appengine pushes
3231         instead of the weird thing we used to do of pulling them from the Chromium repository
3232         and storing them in the appengine datastore. This allows for cleaning up a lot of code
3233         and will likely make the dashboards load a bit faster.
3234
3235         The new JS files don't fully match WebKit style, but I'd like to clean that up in a
3236         followup patch if possible to maintain my sanity with this patch.
3237
3238         * TestResultServer/app.yaml:
3239         * TestResultServer/handlers/dashboardhandler.py: Removed.
3240         * TestResultServer/handlers/menu.py:
3241         * TestResultServer/main.py:
3242         * TestResultServer/model/dashboardfile.py: Removed.
3243         * TestResultServer/static-dashboards/LICENSE.dygraph.txt: Added.
3244         * TestResultServer/static-dashboards/README.dygraph.txt: Added.
3245         * TestResultServer/static-dashboards/README.webtreemap.txt: Added.
3246         * TestResultServer/static-dashboards/aggregate_results.html: Added.
3247         * TestResultServer/static-dashboards/builders.js: Added.
3248         * TestResultServer/static-dashboards/dashboard_base.js: Added.
3249         * TestResultServer/static-dashboards/dygraph-combined.js: Added.
3250         * TestResultServer/static-dashboards/flakiness_dashboard.html: Added.
3251         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js: Added.
3252         * TestResultServer/static-dashboards/timeline_explorer.html: Added.
3253         * TestResultServer/static-dashboards/treemap.html: Added.
3254         * TestResultServer/static-dashboards/webtreemap.css: Added.
3255         * TestResultServer/static-dashboards/webtreemap.js: Added.
3256         * TestResultServer/stylesheets/dashboardfile.css: Removed.
3257         * TestResultServer/stylesheets/menu.css:
3258         * TestResultServer/templates/dashboardfilelist.html: Removed.
3259         * TestResultServer/templates/menu.html:
3260
3261 2011-07-08  Adam Roben  <aroben@apple.com>
3262
3263         Make TestFailures's list of flaky tests look more like the list of non-flaky tests
3264
3265         Fixes <http://webkit.org/b/64204> TestFailures page's flaky tests list is ugly!
3266
3267         Reviewed by Daniel Bates.
3268
3269         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
3270         (#failure-history, #possibly-flaky-tests): Expanded this rule to apply to the list of flaky
3271         tests.
3272         (#failure-history > li, #possibly-flaky-tests > li): Ditto, but moved the 50px left padding
3273         from here...
3274         (#failure-history > li): ...to here.
3275
3276         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
3277         (ViewController.prototype._domForPossiblyFlakyTests): Give the list an id attribute for
3278         styling purposes.
3279
3280 2011-07-08  Jeffrey Pfau  <jpfau@apple.com>
3281
3282         Unreviewed, add myself as committer.
3283
3284         * Scripts/webkitpy/common/config/committers.py:
3285
3286 2011-07-08  Adam Barth  <abarth@webkit.org>
3287
3288         TestResultsServer should keep old test results
3289         https://bugs.webkit.org/show_bug.cgi?id=64199
3290
3291         Reviewed by Ojan Vafai.
3292
3293         Having historical data will help us do failure archeology.
3294
3295         * TestResultServer/handlers/testfilehandler.py:
3296         * TestResultServer/model/testfile.py:
3297         * TestResultServer/templates/showfilelist.jsonp: Added.
3298
3299 2011-07-08  Adam Roben  <aroben@apple.com>
3300
3301         Remove commit-log-editor's dependency on Module::Load::Conditional
3302
3303         This module isn't available in Perl 5.8.8 (the version used on Leopard).
3304
3305         Fixes <http://webkit.org/b/64198> REGRESSION (r90583):
3306         webkitpy.common.checkout.checkout_unittest failing on Leopard
3307
3308         Reviewed by Daniel Bates.
3309
3310         * Scripts/commit-log-editor: Use the new loadTermReadKey() function instead of
3311         Module::Load::Conditional::can_load.
3312         (loadTermReadKey): Added. Tries to load Term::ReadKey and returns true if it's successful.
3313
3314 2011-07-08  Adam Roben  <aroben@apple.com>
3315
3316         Make checkout_unittest more robust against files moving around
3317
3318         Fixes <http://webkit.org/b/64197> checkout_unittest contains ugly, fragile code to find the
3319         Scripts directory
3320
3321         Reviewed by Adam Barth.
3322
3323         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
3324         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Instantiate a real SCM
3325         object and use it to get the path to the Scripts directory, rather than hard-coding the
3326         relative path from this file to Scripts.
3327
3328 2011-07-08  Adam Roben  <aroben@apple.com>
3329
3330         Make TestFailures remember that run-webkit-tests timed out, even across reloads
3331
3332         Fixes <http://webkit.org/b/64193> TestFailures page incorrectly thinks all tests passed in
3333         http://build.webkit.org/builders/Windows%207%20Release%20(Tests)/builds/14589 after a reload
3334
3335         Reviewed by Daniel Bates.
3336
3337         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
3338         (LayoutTestResultsLoader.prototype.start): Store an "error" attribute in the cached data.
3339         When true, it indicates that there was an error retrieving the results for this build and
3340         that the errorCallback should be called.
3341
3342 2011-07-08  Adam Barth  <abarth@webkit.org>
3343
3344         Teach garden-o-matic how to display test results
3345         https://bugs.webkit.org/show_bug.cgi?id=64141
3346
3347         Reviewed by Ojan Vafai.
3348
3349         This patch includes basic infrastructure for probing build.chromium.org
3350         for test results.  We only handle text and image tests, not anything
3351         complicated like reftests.  Also, we're using the revision/build
3352         independent results store on the server, so we're avoiding that
3353         complication for now.
3354
3355         It's slightly hacky that we need to probe the server to see what kinds
3356         of results exist.  A better solution would be to add CORS support to
3357         the server or to use the local server to help.
3358
3359         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
3360         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
3361         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
3362         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
3363         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
3364         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
3365
3366 2011-07-08  Dirk Pranke  <dpranke@chromium.org>
3367
3368         REGRESSION(90419) NRWT's httpd locking is broken for --child-processes=1
3369         https://bugs.webkit.org/show_bug.cgi?id=64092
3370
3371         Reviewed by Tony Chang.
3372
3373         The code was incorrectly creating empty shards in the
3374         shard_in_two case.
3375
3376         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3377         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3378
3379 2011-07-08  Adam Roben  <aroben@apple.com>
3380
3381         Teach buildbot to figure out how many webkitpy/webkitperl tests failed
3382
3383         Fixes <http://webkit.org/b/64192> It's hard to tell how many test-webkitpy/test-webkitperl
3384         tests failed when looking at build.webkit.org
3385
3386         Reviewed by Eric Seidel.
3387
3388         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
3389         (TestWithFailureCount): New class that represents a test build step which has an associated
3390         failure count. Eventually we should move more of our test classes to deriving from this.
3391         (TestWithFailureCount.countFailures): Method for subclasses to override to say how many
3392         failures occurred.
3393
3394         (TestWithFailureCount.commandComplete):
3395         (TestWithFailureCount.evaluateCommand):
3396         (TestWithFailureCount.getText):
3397         (TestWithFailureCount.getText2):
3398         These were all based on RunGtkAPITests.
3399
3400         (RunPythonTests): Changed to inherit from TestWithFailureCount.
3401         (RunPythonTests.countFailures): Parses the test-webkitpy output looking for the count of
3402         failures.
3403         (RunPerlTests): Changed to inherit from TestWithFailureCount.
3404         (RunPerlTests.countFailures): Parses the test-webkitperl output looking for the count of
3405         failures.
3406
3407 2011-07-08  Adam Roben  <aroben@apple.com>
3408
3409         Ensure $CHANGE_LOG_EMAIL_ADDRESS is set when testing webkitpy's commit-log-editor integration
3410
3411         Fixes <http://webkit.org/b/64180> REGRESSION (r90564): test-webkitpy failing on multiple
3412         bots due to commit-log-editor errors
3413
3414         Reviewed by Adam Barth.
3415
3416         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
3417         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Set
3418         $CHANGE_LOG_EMAIL_ADDRESS to the patch author's email address. This ensures that
3419         commit-log-editor can find a value for the committer's email, and that the committer and
3420         author email addresses match, which will prevent commit-log-editor from inserting a "Patch
3421         by" line in the commit message.
3422
3423 2011-07-08  Andreas Kling  <kling@webkit.org>
3424
3425         [Qt][WK2] Views should expose QActions for basic navigation.
3426         https://bugs.webkit.org/show_bug.cgi?id=64174
3427
3428         Reviewed by Benjamin Poulain.
3429
3430         Bring the toolbar in Qt's MiniBrowser back to life.
3431
3432         * MiniBrowser/qt/BrowserView.cpp:
3433         (BrowserView::navigationAction):
3434         * MiniBrowser/qt/BrowserView.h:
3435         * MiniBrowser/qt/BrowserWindow.cpp:
3436         (BrowserWindow::BrowserWindow):
3437
3438 2011-07-08  Adam Roben  <aroben@apple.com>
3439
3440         Teach TestFailures how to find test names in commit-log-editor-style commit messages
3441
3442         TestFailures was relying on Trac turning the list of modified files in our commit messages
3443         into an HTML list. But Trac only does this when the list of modified files is indented.
3444         commit-log-editor doesn't indent the file list, so the list wasn't being turned into an HTML
3445         list, which was confusing TestFailures.
3446
3447         TestFailures now does much simpler parsing of the commit message (i.e., just a substring
3448         search) without relying at all on its structure.
3449
3450         Fixes <http://webkit.org/b/64173> TestFailures page fails to blame r90608 for breaking
3451         fast/dom/HTMLProgressElement/progress-element-markup.html on Windows
3452
3453         Reviewed by David Kilzer.
3454
3455         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
3456         (Trac.prototype.getCommitDataForRevisionRange): Instead of trying to parse the commit
3457         message, just return its text.
3458
3459         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
3460         (ViewController.prototype._domForRegressionRange): Instead of searching for test names in
3461         each commit's list of modified files, just search for test names anywhere in the commit's
3462         message.
3463
3464 2011-07-08  Adam Barth  <abarth@webkit.org>
3465
3466         sheriffbot is too spammy in IRC
3467         https://bugs.webkit.org/show_bug.cgi?id=64153
3468
3469         Reviewed by Eric Seidel.
3470
3471         Reporting failures in IRC worked well when the tree was greener than it
3472         is today.  Nowadays, this feature mostly just results in spam about
3473         false positives.  If we reach a greener state, we can bring this code
3474         back.
3475
3476         (Another possibility is to restrict this feature to builder bots,
3477         rather than including testers, as we did before this patch.)
3478
3479         * Scripts/webkitpy/tool/commands/sheriffbot.py:
3480         * Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
3481
3482 2011-07-08  Adam Roben  <aroben@apple.com>
3483
3484         Ensure commit-log-editor adds a "Patch by" line when the author and committer are different
3485
3486         Previously we were only adding a "Patch by" line when the ChangeLog contained a "Reviewed
3487         by" line. But some patches (like rollout patches) don't contain that line. Now we always add
3488         "Patch by" to the commit log regardless of the ChangeLog's contents.
3489
3490         Fixes <http://webkit.org/b/64127> Committer for r90588 is commit-queue@webkit.org, but
3491         should have been abarth@webkit.org
3492
3493         Reviewed by Anders Carlsson.
3494
3495         * Scripts/commit-log-editor:
3496         (createCommitMessage): Try to put the "Patch by" line just above the "Reviewed by" line, as
3497         before. If there is no "Reviewed by" line, try to put it just above the first modified file.
3498         If all else fails, put it at the end of the commit message.
3499         (patchAuthorshipString): Added. Code came from createCommitMessage.
3500
3501 2011-07-08  Herczeg Zoltan  <zherczeg@webkit.org>
3502
3503         Reviewed by Andreas Kling.
3504
3505         Adding myself as a reviewer.
3506
3507         * Scripts/webkitpy/common/config/committers.py:
3508
3509 2011-07-07  Kevin Ollivier  <kevino@theolliviers.com>
3510
3511         [wx] Unreviewed build fix, remove old files from the tree and update paths to new ones.
3512
3513         * wx/build: Removed.
3514         * wx/build/build_utils.py: Removed.
3515         * wx/build/settings.py: Removed.
3516         * wx/build/waf_extensions.py: Removed.
3517         * wx/build/wxpresets.py: Removed.
3518         * wx/packaging/build-mac-installer.py:
3519
3520 2011-07-07  Kevin Ollivier  <kevino@theolliviers.com>
3521
3522         [wx] Unreviewed build fix, more fixes for install name issues.
3523
3524         * wx/packaging/build-mac-installer.py:
3525
3526 2011-07-07  Dirk Pranke  <dpranke@chromium.org>
3527
3528         test-webkitpy fails on chromium win
3529         https://bugs.webkit.org/show_bug.cgi?id=64137
3530
3531         Reviewed by Eric Seidel.
3532
3533         test-webkitpy doesn't play nicely with the multiprocessing
3534         module on win32. An earlier change actually reenabled the tests
3535         on win32 by mistake.
3536
3537         This patch also fixes a few cases where path names will trip
3538         things up on win32 (e.g., by testing unix paths on windows).
3539         We do not lose any real coverage here with those fixes.
3540
3541         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
3542         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
3543         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
3544         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3545
3546 2011-07-07  Leandro Pereira  <leandro@profusion.mobi>
3547
3548         [ImageDiff] Calculate/print difference right after reading baseline image.
3549         https://bugs.webkit.org/show_bug.cgi?id=64117
3550
3551         Reviewed by Kent Tamura.
3552
3553         Otherwise, ImageDiff will block on fgets() until the universe ends or you ^C
3554         it.  Whichever happens first.
3555
3556         * DumpRenderTree/efl/ImageDiff.cpp:
3557         (main):
3558
3559 2011-07-07  Raphael Kubo da Costa  <kubo@profusion.mobi>
3560
3561         [EFL] DRT: Leak cairo_t to make the pixel tests stop crashing
3562         https://bugs.webkit.org/show_bug.cgi?id=64107
3563
3564         Reviewed by Kent Tamura.
3565
3566         The pointer is later managed by BitmapContext, but it was being
3567         de-refed and deleted earlier by our RefPtr, causing crashes in all
3568         pixel tests.
3569
3570         We now call leakRef() to make sure it is not removed when our
3571         createBitmapContextFromWebView goes out of context.
3572
3573         * DumpRenderTree/efl/PixelDumpSupportEfl.cpp:
3574         (createBitmapContextFromWebView):
3575
3576 2011-07-07  Adam Roben  <aroben@apple.com>
3577
3578         Update TestFailures's title and header immediately upon navigation
3579
3580         Fixes <http://webkit.org/b/64125> TestFailures page seems slow to react on link clicks
3581
3582         Reviewed by Daniel Bates.
3583
3584         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
3585         (ViewController.prototype.loaded): Create and store an h1 element for displaying the page's
3586         title.
3587         (ViewController.prototype._displayBuilder): Set the page's title and clear out any old
3588         content immediately instead of waiting for the first set of results for the history
3589         analyzer. As we receive new results from the analyzer we'll just clear out the main content
3590         area and replace it with the new info.
3591         (ViewController.prototype._displayTesters): Set the page's title and clear out any old
3592         content immediately instead of waiting for the list of testers from the buildbot. Sprinkled
3593         in a little use of our removeAllChildren helper, too.
3594         (ViewController.prototype._setTitle): Added. Set's the page's title and the header text.
3595
3596 2011-07-07  Adam Roben  <aroben@apple.com>
3597
3598         Make Checkout use SCM's Executive instead of conjuring up its own
3599
3600         This will improve integration with the rest of webkitpy, particularly when invoked via
3601         webkit-patch.
3602
3603         Fixes <http://webkit.org/b/64115> REGRESSION (r90564): webkitpy's Checkout class uses
3604         Executive inappropriately
3605
3606         Reviewed by Adam Barth.
3607
3608         * Scripts/webkitpy/common/checkout/checkout.py:
3609         (Checkout.commit_message_for_this_commit): Use SCM.run instead of creating an Executive for
3610         our own use. SCM might have some super-special Executive that it uses under the covers, and
3611         we want to use it, too!
3612
3613         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
3614         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Mock the SCM.run
3615         method to call through to Executive.run_command.
3616
3617 2011-07-07  Adam Roben  <aroben@apple.com>
3618
3619         Make Term/ReadKey.pm an optional dependency of commit-log-editor
3620
3621         When Term/ReadKey.pm isn't present, the --regenerate-log switch will be non-functional.
3622
3623         Fixes <http://webkit.org/b/64113> REGRESSION (r90564):
3624         webkitpy.common.checkout.checkout_unittest failing on GTK bots due to missing
3625         Term/ReadKey.pm module
3626
3627         Reviewed by Xan Lopez.
3628
3629         * Scripts/commit-log-editor: Use Module::Load::Conditional::can_load to only load
3630         Term/ReadKey.pm if it's present. If it isn't present, ignore the --regenerate-log switch so
3631         that existing commit messages will be preserved. (Users can manually remove existing commit
3632         messages using git-reset.)
3633
3634 2011-07-07  Adam Roben  <aroben@apple.com>
3635
3636         Keep testing that commit messages containing Unicode are handled properly by webkitpy
3637
3638         Fixes <http://webkit.org/b/64109> REGRESSION (r90571): test-webkitpy no longer tests that we
3639         correctly handle Unicode in commit messages
3640
3641         Reviewed by Eric Seidel.
3642
3643         * Scripts/webkitpy/common/checkout/checkout_unittest.py: Put some Unicode characters in
3644         places in the ChangeLog that will end up being included in the commit message.
3645
3646 2011-07-07  Andreas Kling  <kling@webkit.org>
3647
3648         [Qt][WK2] Remove Symbian code.
3649         https://bugs.webkit.org/show_bug.cgi?id=64101
3650
3651         Reviewed by Benjamin Poulain.
3652
3653         * MiniBrowser/qt/BrowserWindow.cpp:
3654         (BrowserWindow::BrowserWindow):
3655         (BrowserWindow::screenshot):
3656         * MiniBrowser/qt/MiniBrowser.pro:
3657         * MiniBrowser/qt/MiniBrowserApplication.h:
3658         (WindowOptions::WindowOptions):
3659
3660 2011-07-07  Andreas Kling  <kling@webkit.org>
3661
3662         [Qt][WK2] Don't support ridiculous matrix of QT_NO_FEATURE combinations.
3663         https://bugs.webkit.org/show_bug.cgi?id=64099
3664
3665         Reviewed by Benjamin Poulain.
3666
3667         * MiniBrowser/qt/BrowserWindow.cpp:
3668         (BrowserWindow::openFile):
3669         (BrowserWindow::screenshot):
3670         (BrowserWindow::loadURLListFromFile):
3671
3672 2011-07-07  Adam Roben  <aroben@apple.com>
3673
3674         Completely ignore too-many-failures builds in TestFailures in most circumstances
3675
3676         Because a semi-arbitrary subset of tests are run in a too-many-failures build, we can't
3677         really use them to perform regression analysis. The only time we want to pay attention to
3678         too-many-failures builds is when we're trying to explain when the current bout of
3679         too-many-failures started.
3680
3681         Fixes <http://webkit.org/b/64106> TestFailures page sometimes claims a test started failing
3682         in a build that didn't even run it (because it exited early due to too many failues)
3683
3684         Reviewed by David Kilzer.
3685
3686         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
3687         (LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): Removed old, broken
3688         too-many-failures handling that would cause us to blame builds that didn't even run a given
3689         test for breaking it. Instead, skip over all too-many-failures builds unless the most recent
3690         build was itself a too-many-failures build.
3691
3692 2011-07-07  Adam Roben  <aroben@apple.com>
3693
3694         Teach webkitpy's Checkout class to use commit-log-editor to create commit messages
3695
3696         Fixes <http://webkit.org/b/26755> webkit-patch's commit messages are less readable than
3697         commit-log-editor's
3698
3699         Reviewed by David Kilzer.
3700
3701         * Scripts/webkitpy/common/checkout/checkout.py:
3702         (Checkout.commit_message_for_this_commit): Run commit-log-editor, passing it the paths of
3703         the modified ChangeLogs, to generate the commit message, rather than trying to generate one
3704         ourselves.
3705
3706         * Scripts/webkitpy/common/checkout/checkout_unittest.py: Updated the expected commit message
3707         to match commit-log-editor's format.
3708         (CommitMessageForThisCommitTest.setUp): Write the ChangeLogs into Tools and LayoutTests
3709         directories so we can see how the various entries get labeled in the commit message.
3710         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Create a mock SCM
3711         instance that knows how to find commit-log-editor and pass it to our Checkout instance.
3712         Don't bother capturing output, since there shouldn't be any.
3713
3714 2011-07-07  Adam Roben  <aroben@apple.com>
3715
3716         Add a --print-log option to commit-log-editor
3717
3718         When specified, commit-log-editor takes a set of ChangeLog paths on the command line,
3719         generates a commit message from those ChangeLogs, prints it to stdout, and exits.
3720
3721         Prep work for fixing <http://webkit.org/b/26755> webkit-patch's commit messages are less
3722         readable than commit-log-editor's
3723
3724         Reviewed by David Kilzer.
3725
3726         * Scripts/commit-log-editor: Changed to use Getopt::Long to parse options. Added --print-log
3727         option, which calls through to createCommitMessage, prints the result, and exits.
3728         (printUsageAndExit): Renamed from "usage" for clarity. Beefed up the usage statement to
3729         explain commit-log-editor's 2.5 (normal, --print-log, and --help) modes of operation.
3730
3731 2011-07-07  Adam Roben  <aroben@apple.com>
3732
3733         Extract commit-log-editor's commit-message-generation code into a separate function
3734
3735         As a bonus, we now skip a bunch of work in the case where we are generating a commit message
3736         for staged git changes.
3737
3738         Prep work for fixing <http://webkit.org/b/26755> webkit-patch's commit messages are less
3739         readable than commit-log-editor's
3740
3741         Reviewed by David Kilzer.
3742
3743         * Scripts/commit-log-editor: Moved the default definition of $endl up toward the top of the
3744         file so that it will be defined even if we aren't operating on an existing log message.
3745         Moved a bunch of top-level code to generate the commit message from here...
3746         (createCommitMessage): ...to here.
3747
3748 2011-07-07  Carlos Garcia Campos  <cgarcia@igalia.com>
3749
3750         [GTK] Open links in a new window when clicking with the middle button in MiniBrowser
3751         https://bugs.webkit.org/show_bug.cgi?id=63988
3752
3753         Reviewed by Martin Robinson.
3754
3755         * MiniBrowser/gtk/BrowserWindow.c:
3756         (browserWindowConstructed):
3757         (decidePolicyForNavigationAction): Ignore the action if a link is
3758         clicked with the middle buttonm and open the link in a new window.
3759         (browserWindowPolicyClientInit):
3760
3761 2011-07-07  Adam Barth  <abarth@webkit.org>
3762
3763         Remove style_references.py
3764         https://bugs.webkit.org/show_bug.cgi?id=64070
3765
3766         Reviewed by Eric Seidel.
3767
3768         We never really adopted the _references design, and this file is a bit
3769         of an orphan at this point.
3770
3771         * Scripts/check-webkit-style:
3772         * Scripts/webkitpy/style/checker.py:
3773         * Scripts/webkitpy/style/checker_unittest.py:
3774         * Scripts/webkitpy/style/checkers/python.py:
3775         * Scripts/webkitpy/style/checkers/test_expectations.py:
3776         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
3777         * Scripts/webkitpy/style/main_unittest.py:
3778         * Scripts/webkitpy/style_references.py: Removed.
3779
3780 2011-07-07  Eric Seidel  <eric@webkit.org>
3781
3782         Fix WebKit2 expected results search paths for Mac and Qt under new-run-webkit-tests
3783         https://bugs.webkit.org/show_bug.cgi?id=64056
3784
3785         Unreviewed.
3786
3787         I had written this unittest prior to landing my change, but
3788         had named it qt_unitest.py, so it wasn't running (nor was
3789         it added to my git repo).
3790
3791         Once I fixe the name of the unittest file, it was very easy
3792         to fix the error in webkit.py (which already had a FIXME).
3793
3794         * Scripts/webkitpy/layout_tests/port/qt_unittest.py: Added.
3795         * Scripts/webkitpy/layout_tests/port/webkit.py:
3796
3797 2011-07-07  Eric Seidel  <eric@webkit.org>
3798
3799         Unreviewed.  Add more debug logging about which test expectations we're using.
3800
3801         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
3802         * Scripts/webkitpy/layout_tests/port/webkit.py:
3803
3804 2011-07-06  Adam Barth  <abarth@webkit.org>
3805
3806         Remove unexpected_results.json
3807         https://bugs.webkit.org/show_bug.cgi?id=64065
3808
3809         Reviewed by Eric Seidel.
3810
3811         No one should be using this file any more.  This patch removes it from
3812         the face of the earth.  This patch also removes two NRWT-specific
3813         command line options that were implemented using
3814         unexpected_results.json.
3815
3816         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
3817         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3818         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3819         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3820         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3821         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
3822         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
3823
3824 2011-07-07  Adam Barth  <abarth@webkit.org>
3825
3826         REGRESSION(90520): chromium GPU tests produce output in the LayoutTests directory
3827         https://bugs.webkit.org/show_bug.cgi?id=64067
3828
3829         Reviewed by Eric Seidel.
3830
3831         The author of r90520 forgot to change the chromium_gpu.py implemenation
3832         of this method as well.
3833
3834         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
3835
3836 2011-07-06  Eric Seidel  <eric@webkit.org>
3837
3838         Fix WebKit2 expected results search paths for Mac and Qt under new-run-webkit-tests
3839         https://bugs.webkit.org/show_bug.cgi?id=64056
3840
3841         Reviewed by Adam Barth.
3842
3843         This fixes the MacPort's version of baseline_search_path to include
3844         wk2 fallback like WebKitPort's does.  This also re-works the QtPort
3845         to set self._name (as other ports do) so that WebKitPort's default
3846         implementations will understand Qt operating system flavors correctly
3847         (including when running wk2 tests).
3848
3849         In trying to test this, I realized that nearly all of our port-tests
3850         use the default FileSystem(), User() and Executive() objects, which
3851         is really bad, because they expose details of your system in the unittest results!
3852
3853         Once I fixed many of the Port() calls to pass MockFileSystem(), then
3854         MacPort.test_tests_for_other_platforms started failing, and I had to
3855         teach MockFileSystem.glob how to handle directories to make it work again.
3856
3857         I removed the useless PortTestCase.test_baseline_search_path and replaced
3858         it with more useful test_baseline_search_path tests in MacPortTest and QtPortTest.
3859
3860         * Scripts/webkitpy/common/config/ports.py:
3861         * Scripts/webkitpy/common/system/filesystem_mock.py:
3862         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
3863         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3864         * Scripts/webkitpy/layout_tests/port/mac.py:
3865         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3866         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
3867         * Scripts/webkitpy/layout_tests/port/qt.py:
3868         * Scripts/webkitpy/layout_tests/port/test_files.py:
3869         * Scripts/webkitpy/layout_tests/port/win.py:
3870         * Scripts/webkitpy/to_be_moved/deduplicate_tests.py:
3871         * Scripts/webkitpy/tool/mocktool.py:
3872
3873 2011-07-06  Adam Barth  <abarth@webkit.org>
3874
3875         Wean rebaseline-server off unexpected_results.json
3876         https://bugs.webkit.org/show_bug.cgi?id=64063
3877
3878         Reviewed by Eric Seidel.
3879
3880         In the process of changing this code to use full_results.json, I
3881         noticed that the code was broken (because it wasn't tested).  This
3882         patch also adds test coverage for the broken code.
3883
3884         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
3885         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
3886         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
3887
3888 2011-07-06  Adam Barth  <abarth@webkit.org>
3889
3890         Wean resultsjsonparser off unexpected_results.json
3891         https://bugs.webkit.org/show_bug.cgi?id=64061
3892
3893         Reviewed by Eric Seidel.
3894
3895         We're about to delete unexpected_results.json, so we need to remove all
3896         the users.
3897
3898         * Scripts/webkitpy/common/config/ports.py:
3899         * Scripts/webkitpy/common/net/resultsjsonparser.py:
3900         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
3901         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
3902
3903 2011-07-06  Adam Barth  <abarth@webkit.org>
3904
3905         Cleanup result_summary.py
3906         https://bugs.webkit.org/show_bug.cgi?id=64057
3907
3908         Reviewed by Eric Seidel.
3909
3910         This class had a bunch of out-of-date docstrings that no longer make
3911         any sense.
3912
3913         * Scripts/webkitpy/layout_tests/models/result_summary.py:
3914
3915 2011-07-06  Adam Barth  <abarth@webkit.org>
3916
3917         Tweak the ChangeLog for DEPS rolls to say "Unreviewed" so that the
3918         patches can be landed by the commit-queue.
3919
3920         * Scripts/webkitpy/tool/steps/preparechangelogfordepsroll.py:
3921
3922 2011-07-06  Adam Barth  <abarth@webkit.org>
3923
3924         Normalize callers of logging.getLogger to use __name__
3925         https://bugs.webkit.org/show_bug.cgi?id=64054
3926
3927         Reviewed by Eric Seidel.
3928
3929         Most callers were already using __name__.  This patch makes everyone
3930         consistent.
3931
3932         * Scripts/webkitpy/common/checkout/diff_parser.py:
3933         * Scripts/webkitpy/common/checkout/scm/svn.py:
3934         * Scripts/webkitpy/common/net/statusserver.py:
3935         * Scripts/webkitpy/common/system/executive.py:
3936         * Scripts/webkitpy/common/system/file_lock.py:
3937         * Scripts/webkitpy/common/system/user.py:
3938         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
3939         * Scripts/webkitpy/layout_tests/models/result_summary.py:
3940         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
3941         * Scripts/webkitpy/layout_tests/port/chromium.py:
3942         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
3943         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
3944         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
3945         * Scripts/webkitpy/layout_tests/port/gtk.py:
3946         * Scripts/webkitpy/layout_tests/port/http_lock.py:
3947         * Scripts/webkitpy/layout_tests/port/mac.py:
3948         * Scripts/webkitpy/layout_tests/port/qt.py:
3949         * Scripts/webkitpy/layout_tests/port/server_process.py:
3950         * Scripts/webkitpy/layout_tests/port/webkit.py:
3951         * Scripts/webkitpy/layout_tests/port/win.py:
3952         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
3953         * Scripts/webkitpy/layout_tests/servers/http_server.py:
3954         * Scripts/webkitpy/layout_tests/servers/http_server_base.py:
3955         * Scripts/webkitpy/layout_tests/servers/websocket_server.py:
3956         * Scripts/webkitpy/layout_tests/views/printing.py:
3957         * Scripts/webkitpy/python24/versioning.py:
3958         * Scripts/webkitpy/style/checker.py:
3959         * Scripts/webkitpy/style/checkers/test_expectations.py:
3960         * Scripts/webkitpy/style/patchreader.py:
3961         * Scripts/webkitpy/to_be_moved/update_webgl_conformance_tests.py:
3962
3963 2011-07-06  Adam Barth  <abarth@webkit.org>
3964
3965         Move model classes out of layout_package and into models
3966         https://bugs.webkit.org/show_bug.cgi?id=64053
3967
3968         Reviewed by Eric Seidel.
3969
3970         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
3971         * Scripts/webkitpy/common/net/layouttestresults.py:
3972         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
3973         * Scripts/webkitpy/common/net/resultsjsonparser.py:
3974         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
3975         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3976         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3977         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
3978         * Scripts/webkitpy/layout_tests/controllers/worker.py:
3979         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
3980         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
3981         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
3982         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py: Removed.
3983         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py: Removed.
3984         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py: Removed.
3985         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py: Removed.
3986         * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py: Removed.
3987         * Scripts/webkitpy/layout_tests/layout_package/test_input.py: Removed.
3988         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
3989         * Scripts/webkitpy/layout_tests/layout_package/test_results.py: Removed.
3990         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py: Removed.
3991         * Scripts/webkitpy/layout_tests/models: Added.
3992         * Scripts/webkitpy/layout_tests/models/__init__.py: Added.
3993         * Scripts/webkitpy/layout_tests/models/result_summary.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/result_summary.py.
3994         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_expectations.py.
3995         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py.
3996         * Scripts/webkitpy/layout_tests/models/test_failures.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_failures.py.
3997         * Scripts/webkitpy/layout_tests/models/test_failures_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py.
3998         * Scripts/webkitpy/layout_tests/models/test_input.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_input.py.
3999         * Scripts/webkitpy/layout_tests/models/test_results.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_results.py.
4000         * Scripts/webkitpy/layout_tests/models/test_results_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py.
4001         * Scripts/webkitpy/layout_tests/port/chromium.py:
4002         * Scripts/webkitpy/layout_tests/views/printing.py:
4003         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
4004         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
4005         * Scripts/webkitpy/style_references.py:
4006         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
4007         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
4008         * Scripts/webkitpy/tool/bot/flakytestreporter_unittest.py:
4009         * Scripts/webkitpy/tool/commands/queues_unittest.py:
4010         * Scripts/webkitpy/tool/commands/rebaseline.py:
4011
4012 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
4013
4014         nrwt: fix typo preventing http startup on windows
4015         https://bugs.webkit.org/show_bug.cgi?id=64050
4016
4017         Reviewed by Eric Siedel.
4018
4019         NRWT checks whether it needs to start the servers by looking
4020         for "/http/" in the test names. After r90520, the leading slash
4021         is not present, so the check isn't quite right.
4022
4023         * Scripts/webkitpy/layout_tests/controllers/manager.py:
4024         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
4025
4026 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
4027
4028         nrwt: remove --use-apache from the command line
4029         https://bugs.webkit.org/show_bug.cgi?id=63358
4030
4031         Reviewed by Adam Barth.
4032
4033         Re-land the fix in bug 63358 - the prior fix was flawed in that
4034         we were trying to start the http server during check_sys_deps()
4035         but not actually setting up the environment properly. This
4036         broken prior to this change, but exposed by it, since the patch
4037         changes the chromium port to call the base class's checks.
4038
4039         * Scripts/webkitpy/common/system/executive.py:
4040         * Scripts/webkitpy/layout_tests/port/base.py:
4041         * Scripts/webkitpy/layout_tests/port/chromium.py:
4042         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
4043         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
4044         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
4045         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
4046         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
4047         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
4048
4049 2011-07-06  Adam Barth  <abarth@webkit.org>
4050
4051         Move view-related code out of layout_package and into views
4052         https://bugs.webkit.org/show_bug.cgi?id=64051
4053
4054         Reviewed by Eric Seidel.
4055
4056         * Scripts/webkitpy/layout_tests/controllers/manager.py:
4057         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
4058         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
4059         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
4060         * Scripts/webkitpy/layout_tests/layout_package/metered_stream.py: Removed.
4061         * Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py: Removed.
4062         * Scripts/webkitpy/layout_tests/layout_package/printing.py: Removed.
4063         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py: Removed.
4064         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
4065         * Scripts/webkitpy/layout_tests/views: Added.
4066         * Scripts/webkitpy/layout_tests/views/__init__.py: Added.
4067         * Scripts/webkitpy/layout_tests/views/metered_stream.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/metered_stream.py.
4068         * Scripts/webkitpy/layout_tests/views/metered_stream_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py.
4069         * Scripts/webkitpy/layout_tests/views/printing.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/printing.py.
4070         * Scripts/webkitpy/layout_tests/views/printing_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py.
4071
4072 2011-07-06  Adam Barth  <abarth@webkit.org>
4073
4074         Add goofy comments to explain why we have these files.
4075
4076         * Scripts/webkitpy/layout_tests/controllers/__init__.py:
4077         * Scripts/webkitpy/layout_tests/layout_package/__init__.py:
4078
4079 2011-07-06  Adam Barth  <abarth@webkit.org>
4080
4081         Move a bunch of classes out of layout_tests.layout_package and into layout_tests.controllers
4082         https://bugs.webkit.org/show_bug.cgi?id=64048
4083
4084         Reviewed by Eric Seidel.
4085
4086         * Scripts/webkitpy/layout_tests/controllers: Added.
4087         * Scripts/webkitpy/layout_tests/controllers/__init__.py: Added.
4088         * Scripts/webkitpy/layout_tests/controllers/manager.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager.py.
4089         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py.
4090         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py.
4091         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py.
4092         * Scripts/webkitpy/layout_tests/controllers/message_broker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker.py.
4093         * Scripts/webkitpy/layout_tests/controllers/message_broker_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py.
4094         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py.
4095         * Scripts/webkitpy/layout_tests/controllers/worker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/worker.py.
4096         * Scripts/webkitpy/layout_tests/layout_package/manager.py: Removed.
4097         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py: Removed.
4098         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py: Removed.
4099         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py: Removed.
4100         * Scripts/webkitpy/layout_tests/layout_package/message_broker.py: Removed.
4101         * Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py: Removed.
4102         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
4103         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: Removed.
4104         * Scripts/webkitpy/layout_tests/layout_package/worker.py: Removed.
4105         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
4106
4107 2011-07-06  Adam Barth  <abarth@webkit.org>
4108
4109         Attempt to fix unittest on Leopard.  Let's enumerate the keys in sorted
4110         order so the test results are predictable.
4111
4112         * Scripts/webkitpy/tool/bot/irc_command.py:
4113         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
4114
4115 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
4116
4117         Reviewed by Eric Siedel.
4118
4119         nrwt: make the code be consistent about using test names instead of filenames or paths
4120
4121         https://bugs.webkit.org/show_bug.cgi?id=63597
4122
4123         It can be hard to figure out where in the code we're assuming
4124         test names are listed as unix-style relative filenames and where
4125         they are either absolute paths or relative paths following the
4126         host filesystem convention.
4127
4128         This patch changes things so that everything outside of the
4129         Port object uses (and must assume) unix-style relative
4130         filenames (with one exception, which is specifying host-local
4131         filenames as a list of test arguments on the command line).
4132
4133         This should make things clearer and more consistent, and also
4134         removes a lot of path conversion calls.
4135
4136         The changes in this patch outside of port/base.py are largely
4137         cosmetic changes of "filename" to "test". We add a few routines
4138         to the base Port implementation to manipulate test names to
4139         split them, figure out if we're referring to a directory of
4140         tests, etc.
4141
4142         * Scripts/webkitpy/common/net/layouttestresults.py:
4143         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
4144         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
4145         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
4146         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
4147         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
4148         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
4149         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
4150         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
4151         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
4152         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
4153         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
4154         * Scripts/webkitpy/layout_tests/layout_package/test_input.py:
4155         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
4156         * Scripts/webkitpy/layout_tests/layout_package/test_results.py:
4157         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py:
4158         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
4159         * Scripts/webkitpy/layout_tests/port/base.py:
4160         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
4161         * Scripts/webkitpy/layout_tests/port/chromium.py:
4162         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
4163         * Scripts/webkitpy/layout_tests/port/dryrun.py:
4164         * Scripts/webkitpy/layout_tests/port/test.py:
4165         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
4166         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
4167         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
4168         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
4169         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
4170         * Scripts/webkitpy/tool/commands/rebaseline.py:
4171         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
4172
4173 2011-07-06  Chris Guillory  <ctguil@chromium.org>
4174
4175         No review necessary.
4176
4177         Adding myself to committers list in committers.py.
4178         https://bugs.webkit.org/show_bug.cgi?id=63478
4179
4180         * Scripts/webkitpy/common/config/committers.py:
4181
4182 2011-07-06  J