608d060241248efcfe60501ba3bae023f1786fa4
[WebKit-https.git] / Tools / ChangeLog
1 2011-06-04  Kevin Ollivier  <kevino@theolliviers.com>
2
3         [wx] Unreviewed build fixes for recent trunk changes.
4
5         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
6         (LayoutTestController::setTextDirection):
7         * waf/build/settings.py:
8
9 2011-06-04  Martin Robinson  <mrobinson@igalia.com>
10
11         Fix the GTK+ build.
12
13         * GNUmakefile.am: Add some missing newline escape.
14
15 2011-06-03  Martin Robinson  <mrobinson@igalia.com>
16
17         Try to fix the GTK+ bulid.
18
19         * GNUmakefile.am: Add JSDOMWrapper.cpp to the libWebCoreInternals
20         source list.
21
22 2011-06-03  Dirk Pranke  <dpranke@chromium.org>
23
24         Reviewed by Adam Barth.
25
26         nrwt: should skip chromium test expectation overrides on upstream bots
27         https://bugs.webkit.org/show_bug.cgi?id=61888
28
29         * Scripts/webkitpy/layout_tests/port/chromium.py:
30         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
31
32 2011-06-03  Dirk Pranke  <dpranke@chromium.org>
33
34         Reviewed by Ojan Vafai.
35
36         new-run-webkit-tests: minor logging, code cleanup
37         https://bugs.webkit.org/show_bug.cgi?id=62055
38
39         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
40         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
41
42 2011-06-03  Martin Robinson  <mrobinson@igalia.com>
43
44         Try to fix the build failure on the GTK+ 32-bit Release bot.
45
46         * GNUmakefile.am: Add JSInternal.{cpp,h} to the BUILT_SOURCES list.
47
48 2011-05-31  Martin Robinson  <mrobinson@igalia.com>
49
50         Reviewed by Ryosuke Niwa.
51
52         [GTK] Support smart replace for the pasteboard
53         https://bugs.webkit.org/show_bug.cgi?id=61734
54
55         Call into the DumpRenderTreeSupportGtk hooks to toggle smart replace.
56         This is required to produce consistent test results.
57
58         * DumpRenderTree/gtk/DumpRenderTree.cpp:
59         (resetDefaultsToConsistentValues): Set the smart replace setting to on by default.
60         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
61         (LayoutTestController::setSmartInsertDeleteEnabled): Added an implementation of this
62         methods which calls into DumpRenderTreeSupportGtk.
63
64 2011-06-03  Martin Robinson  <mrobinson@igalia.com>
65
66         Reviewed by Dimitri Glazkov.
67
68         Teach Gtk build about window.internals
69         https://bugs.webkit.org/show_bug.cgi?id=61071
70
71         * DumpRenderTree/gtk/DumpRenderTree.cpp:
72         (webViewWindowObjectCleared): Inject the new internals object when the
73         window object clears.
74         * GNUmakefile.am: Add new source files and includes necessary for window.internals.
75         Also do a little bit of cleanup.
76
77 2011-06-03  Dirk Pranke  <dpranke@chromium.org>
78
79         Reviewed by Tony Chang.
80
81         Worker may not be stopped after LT.
82         https://bugs.webkit.org/show_bug.cgi?id=61847
83
84         Re-submit a variant of the first fix for this bug. It's possible
85         that actually calling cleanup() in the worker threads was taking
86         too long and we triggered the assertions in the manager thread.
87
88         I've bumped up the timeout in the manager and added better
89         diagnostics. I also cleaned up the cleanup() code a bit because
90         there was some unnecessary if-checks.
91
92         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
93         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
94
95 2011-06-03  Dirk Pranke  <dpranke@chromium.org>
96
97         Reviewed by Tony Chang.
98
99         nrwt: fix duplicate logging regression
100         https://bugs.webkit.org/show_bug.cgi?id=61981
101
102         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
103
104 2011-06-03  Robert Kroeger  <rjkroege@chromium.org>
105
106         Reviewed by Adam Barth.
107
108         Modified addTouchPoint to correctly re-use touchPoint id values starting
109         from the lowest not-un-use touchPoint id as is done by the embedder
110         APIs. (Fixes BUGCR36415)
111
112         https://bugs.webkit.org/show_bug.cgi?id=62016
113
114         * DumpRenderTree/chromium/EventSender.cpp:
115         (EventSender::addTouchPoint):
116
117 2011-06-03  Cary Clark  <caryclark@google.com>
118
119         Reviewed by Eric Seidel.
120
121         Draw DRT scroll bar thumb with CG when WebKit renders with Skia
122         https://bugs.webkit.org/show_bug.cgi?id=61534
123
124         Skia on Mac uses Skia to render WebKit, and CG
125         to render UI elements. The CG elements need a
126         transcribed graphics context that preserves the
127         canvas matrix, and the canvas clip.
128
129         The SkiaBitLocker utility class sets up a CGContext
130         from the SkCanvas, locks the bitmap's bits, and
131         releases the lock when the class goes out of scope.
132
133         The SkiaBitLocker implementation is described by
134         this patch:
135         http://codereview.chromium.org/7031006/
136
137         When WebKit uses Skia for rendering, the WebCanvas
138         is an SkCanvas. Create an equivalent CGContext to
139         draw the scrollbar thumb.
140
141         As the Chrome Mac port does not yet use Skia, this
142         patch has no effect and requires no tests.
143
144         * DumpRenderTree/chromium/WebThemeEngineDRTMac.mm:
145         (WebThemeEngineDRTMac::paintHIThemeScrollbarThumb):
146
147 2011-06-03  Andras Becsi  <abecsi@webkit.org>
148
149         Unreviewed build fix.
150
151         Fix typo.
152
153         * Scripts/old-run-webkit-tests:
154
155 2011-06-03  Andras Becsi  <abecsi@webkit.org>
156
157         Reviewed by Csaba Osztrogonác.
158
159         Print error message when there is no httpd present on the system
160         https://bugs.webkit.org/show_bug.cgi?id=61939
161
162         * Scripts/old-run-webkit-tests: Propose to run with --no-httpd.
163         * Scripts/webkitperl/httpd.pm:
164
165 2011-06-03  Mikhail Naganov  <mnaganov@chromium.org>
166
167         Reviewed by Yury Semikhatsky.
168
169         Web Inspector: [Chromium] cpu-profiler-profiling layout test is flaky on Linux Debug.
170         https://bugs.webkit.org/show_bug.cgi?id=61533
171
172         Skip cpu-profiler-profiling in debug, implement a fast headless alternative.
173
174         * DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
175         (DRTDevToolsAgent::setJavaScriptProfilingEnabled):
176         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
177         * DumpRenderTree/chromium/LayoutTestController.cpp:
178         (LayoutTestController::LayoutTestController):
179         (LayoutTestController::setJavaScriptProfilingEnabled):
180         * DumpRenderTree/chromium/LayoutTestController.h:
181         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
182         (LayoutTestController::setJavaScriptProfilingEnabled):
183         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
184         (WTR::LayoutTestController::setJavaScriptProfilingEnabled):
185         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
186
187 2011-06-02  Dirk Pranke  <dpranke@chromium.org>
188
189         Reviewed by Tony Chang.
190
191         NRWT needs a way to log to a file without including backspaces without needing --verbose
192         https://bugs.webkit.org/show_bug.cgi?id=60328
193
194         This change modifies NRWT so that if it can tell if it is
195         writing to a terminal, it will overwrite any messages printed
196         in the 'one-line-progress' and 'updates' print options, and if
197         not (or if --verbose was specified) the messages will be printed
198         without backspaces.
199
200         If the overwriting updates are disabled, one-line-progress will
201         only print every 10 seconds instead of on every update. This
202         ensures that we still get updates once in a while while the
203         tests are running, even if we're not in --verbose mode. Note
204         that one-line-progress will now still be printed in --verbose,
205         since it's useful as long as you're not flooded with the
206         updates.
207
208         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
209         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
210         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
211
212 2011-06-02  Martin Robinson  <mrobinson@igalia.com>
213
214         Reviewed by Gustavo Noronha Silva.
215
216         [GTK] autogen.sh is run twice for each buld on the bots
217         https://bugs.webkit.org/show_bug.cgi?id=61951
218
219         * Scripts/webkitdirs.pm: Only check the previous autotools arguments
220         for the WebKit project. This prevents build-jsc runs from forcing subsequent
221         build-webkit runs to re-rerun autogen.sh.
222
223 2011-06-02  Sheriff Bot  <webkit.review.bot@gmail.com>
224
225         Unreviewed, rolling out r87946.
226         http://trac.webkit.org/changeset/87946
227         https://bugs.webkit.org/show_bug.cgi?id=61965
228
229         Cause NRWT to break (Requested by abarth|gardener on #webkit).
230
231         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
232
233 2011-06-02  Dimitri Glazkov  <dglazkov@chromium.org>
234
235         Reviewed by Darin Adler.
236
237         Add build logistics and plumbing for window.internals object.
238         https://bugs.webkit.org/show_bug.cgi?id=60313
239
240         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added linking new WebCoreTestSupport library.
241         * DumpRenderTree/chromium/TestShell.cpp:
242         (TestShell::bindJSObjectsToWindow): Added injection code.
243         * DumpRenderTree/mac/FrameLoadDelegate.mm:
244         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]): Ditto.
245
246 2011-06-02  Hao Zheng  <zhenghao@chromium.org>
247
248         Reviewed by Dirk Pranke.
249
250         Call cleanup() explicitly at the end of worker.run().
251         https://bugs.webkit.org/show_bug.cgi?id=61847
252
253         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
254
255 2011-06-02  Dimitri Glazkov  <dglazkov@chromium.org>
256
257         Unreviewed, rolling out r87926.
258         http://trac.webkit.org/changeset/87926
259         https://bugs.webkit.org/show_bug.cgi?id=60313
260
261         Fails to find WebCoreTestSupport.dylib on bots.
262
263         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
264         * DumpRenderTree/chromium/TestShell.cpp:
265         (TestShell::bindJSObjectsToWindow):
266         * DumpRenderTree/mac/FrameLoadDelegate.mm:
267         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]):
268
269 2011-05-21  Dimitri Glazkov  <dglazkov@chromium.org>
270
271         Reviewed by Darin Adler.
272
273         Add build logistics and plumbing for window.internals object.
274         https://bugs.webkit.org/show_bug.cgi?id=60313
275
276         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added linking new WebCoreTestSupport library.
277         * DumpRenderTree/chromium/TestShell.cpp:
278         (TestShell::bindJSObjectsToWindow): Added injection code.
279         * DumpRenderTree/mac/FrameLoadDelegate.mm:
280         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]): Ditto.
281
282 2011-06-02  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
283
284         Reviewed by Eric Seidel.
285
286         Adjust efl coding style to more efl directories
287         https://bugs.webkit.org/show_bug.cgi?id=61903
288
289         * Scripts/webkitpy/style/checker.py: Adjust efl coding style to efl directories.
290
291 2011-06-01  Dirk Pranke  <dpranke@chromium.org>
292
293         Reviewed by Tony Chang.
294
295         old-run-webkit-tests: add support for audio files
296         https://bugs.webkit.org/show_bug.cgi?id=57992
297
298         * Scripts/old-run-webkit-tests:
299
300 2011-06-01  Jochen Eisinger  <jochen@chromium.org>
301
302         Reviewed by Adam Barth.
303
304         Implement layoutTestController.setImagesAllowed for controlling the respective behavior of the WebPermissionClient
305         https://bugs.webkit.org/show_bug.cgi?id=34314
306
307         * DumpRenderTree/DumpRenderTree.gypi:
308         * DumpRenderTree/chromium/LayoutTestController.cpp:
309         (LayoutTestController::LayoutTestController):
310         (LayoutTestController::setImagesAllowed):
311         (LayoutTestController::setStorageAllowed):
312         * DumpRenderTree/chromium/LayoutTestController.h:
313         * DumpRenderTree/chromium/WebPermissions.cpp: Copied from Tools/DumpRenderTree/chromium/WebPermissions.h.
314         (WebPermissions::WebPermissions):
315         (WebPermissions::~WebPermissions):
316         (WebPermissions::allowImages):
317         (WebPermissions::allowStorage):
318         (WebPermissions::setImagesAllowed):
319         (WebPermissions::setStorageAllowed):
320         (WebPermissions::reset):
321         * DumpRenderTree/chromium/WebPermissions.h:
322
323 2011-06-01  Noel Gordon  <noel.gordon@gmail.com>
324
325         Reviewed by Tony Chang.
326
327         Test plugin should support event logging on the windows port.
328         https://bugs.webkit.org/show_bug.cgi?id=61721
329
330         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
331         (handleEventWin): Well, handle them.
332
333 2011-06-01  Daniel Cheng  <dcheng@chromium.org>
334
335         Reviewed by Tony Chang.
336
337         [chromium] Use correct file name for dragging out images.
338         https://bugs.webkit.org/show_bug.cgi?id=24887
339
340         * DumpRenderTree/chromium/EventSender.cpp:
341         (EventSender::EventSender):
342         (EventSender::dumpFilenameBeingDragged):
343         * DumpRenderTree/chromium/EventSender.h:
344
345 2011-06-01  Dirk Pranke  <dpranke@chromium.org>
346
347         Reviewed by Tony Chang.
348
349         nrwt: make webaudio work w/o content-length header
350         https://bugs.webkit.org/show_bug.cgi?id=61819
351
352         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
353         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
354         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
355         * Scripts/webkitpy/layout_tests/port/webkit.py:
356
357 2011-06-01  Viatcheslav Ostapenko  <ostapenko.viatcheslav@nokia.com>
358
359         Reviewed by Andreas Kling.
360
361         [Qt][WK2] Add QGLWidget viewport support to MiniBrowser
362         https://bugs.webkit.org/show_bug.cgi?id=61802
363
364         Add command line parameter and menu item to MiniBrowser application 
365         enabling use of QGLWidget for browser viewport.
366
367         * MiniBrowser/qt/BrowserWindow.cpp:
368         (BrowserWindow::BrowserWindow):
369         (BrowserWindow::toggleGLViewport):
370         * MiniBrowser/qt/BrowserWindow.h:
371         * MiniBrowser/qt/MiniBrowserApplication.cpp:
372         (MiniBrowserApplication::handleUserOptions):
373         * MiniBrowser/qt/MiniBrowserApplication.h:
374         (WindowOptions::WindowOptions):
375
376 2011-06-01  Martin Robinson  <mrobinson@igalia.com>
377
378         Reviewed by Xan Lopez.
379
380         Print out autogen arguments when they differ from the last set, to help
381         with debugging build issues.
382
383         * Scripts/webkitdirs.pm: Print out previous and current build arguments
384         when they differ.
385
386 2011-06-01  Carlos Garcia Campos  <cgarcia@igalia.com>
387
388         Reviewed by Martin Robinson.
389
390         [GTK] run-launcher should run MiniBrowser in WebKit2
391         https://bugs.webkit.org/show_bug.cgi?id=61858
392
393         * Scripts/run-launcher: Check whether -2 command line option has
394         been passed and run MiniBrowser instead of GtkLauncher.
395
396 2011-05-31  Hao Zheng  <zhenghao@chromium.org>
397
398         Reviewed by Dirk Pranke.
399
400         Fix exception in verbose printing mode.
401         https://bugs.webkit.org/show_bug.cgi?id=61723
402
403         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
404
405 2011-05-31  Yong Li  <yoli@rim.com>
406
407         Reviewed by Eric Seidel.
408
409         https://bugs.webkit.org/show_bug.cgi?id=54807
410         We have been assuming plain bitfields (like "int a : 31") are always signed integers.
411         However some compilers can treat them as unsigned. For example, RVCT 4.0 states plain
412         bitfields (declared without either signed or unsigned qualifiers) are treats as unsigned.
413         http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dui0348c/Babjddhe.html
414         Although we can use "--signed-bitfields" flag to make RVCT 4.0 behave as most other compilers,
415         always using "signed"/"unsigned" qualifier to declare integral type bitfields is still a good
416         rule we should have in order to make our code independent from compilers and compiler flags.
417
418         * Scripts/webkitpy/style/checkers/cpp.py:
419
420 2011-05-31  Hironori Bono  <hbono@chromium.org>
421
422         Reviewed by Eric Seidel.
423
424         Add a JavaScript function layoutTestController.setTextDirection().
425         https://bugs.webkit.org/show_bug.cgi?id=50952
426
427         This change adds a new JavaScript function setTextDirection() to the
428         LayoutTestController class of Safari and Chrome to write layout tests
429         that changes the text direction of an editable element.
430
431         * DumpRenderTree/LayoutTestController.cpp:
432         (setTextDirectionCallback):
433         (LayoutTestController::staticFunctions):
434         * DumpRenderTree/LayoutTestController.h:
435         * DumpRenderTree/chromium/LayoutTestController.cpp:
436         (LayoutTestController::LayoutTestController):
437         (LayoutTestController::setTextDirection):
438         * DumpRenderTree/chromium/LayoutTestController.h:
439         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
440         (LayoutTestController::setTextDirection):
441         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
442         (LayoutTestController::setTextDirection):
443         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
444         (LayoutTestController::setTextDirection):
445
446 2011-05-31  Ademar de Souza Reis Jr  <ademar.reis@openbossa.org>
447
448         Reviewed by Simon Hausmann.
449
450         [Qt] User-agent list of Qt test browsers is very outdated
451         https://bugs.webkit.org/show_bug.cgi?id=61531
452
453         The user can edit the user-agent manually, so there's no need to have
454         a super complete and up to date list, but it's nice to have it
455         minimally updated.
456
457         * MiniBrowser/qt/useragentlist.txt:
458         * QtTestBrowser/useragentlist.txt:
459
460 2011-03-30  Martin Robinson  <mrobinson@igalia.com>
461
462         Reviewed by Adam Roben.
463
464         [GTK] [WebKit2] Implement a basic WebKitTestRunner
465         https://bugs.webkit.org/show_bug.cgi?id=57068
466
467         Add an implementation of WebKitTestRunner for GTK+.
468
469         * Scripts/build-webkittestrunner: Added knowledge of GTK+ TestRunner.
470         * Scripts/old-run-webkit-tests: Ditto.
471         * Scripts/run-launcher: Ditto.
472         * Scripts/webkitdirs.pm: Ditto.
473         * WebKitTestRunner/GNUmakefile.am: Added.
474         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
475         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp: Added.
476         * WebKitTestRunner/InjectedBundle/gtk/InjectedBundleGtk.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
477         * WebKitTestRunner/InjectedBundle/gtk/LayoutTestControllerGtk.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
478         * WebKitTestRunner/PlatformWebView.h:
479         * WebKitTestRunner/gtk/PlatformWebViewGtk.cpp: Added.
480         * WebKitTestRunner/gtk/TestControllerGtk.cpp: Added.
481         * WebKitTestRunner/gtk/TestInvocationGtk.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
482         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
483         * WebKitTestRunner/gtk/main.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
484         (main): Added.
485
486 2011-05-31  Dirk Pranke  <dpranke@chromium.org>
487
488         Reviewed by Tony Chang.
489
490         new-run-webkit-tests: doesn't wait for children if it gets a ctrl-c
491         https://bugs.webkit.org/show_bug.cgi?id=60241
492
493         * Scripts/new-run-webkit-tests:
494
495 2011-05-31  Eric Seidel  <eric@webkit.org>
496
497         Reviewed by David Kilzer.
498
499         webkitpy should create zips with zip -9
500         https://bugs.webkit.org/show_bug.cgi?id=61789
501
502         Dave Kilzer suggested we add this to make zips uploaded to bugzilla smaller.
503
504         * Scripts/webkitpy/common/system/workspace.py:
505         * Scripts/webkitpy/common/system/workspace_unittest.py:
506
507 2011-05-31  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
508
509         Reviewed by Csaba Osztrogonác.
510
511         [Qt] Implement find feature for QtTestBrowser
512         https://bugs.webkit.org/show_bug.cgi?id=61425
513
514         Buildfix for --minimal and CONFIG+=qt_minimal build.
515
516         * QtTestBrowser/launcherwindow.cpp:
517         (LauncherWindow::LauncherWindow):
518         (LauncherWindow::createChrome):
519         * QtTestBrowser/launcherwindow.h:
520
521 2011-05-31  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
522
523         Reviewed by Csaba Osztrogonác.
524
525         [Qt] Implement find feature for QtTestBrowser
526         https://bugs.webkit.org/show_bug.cgi?id=61425
527
528         * QtTestBrowser/launcherwindow.cpp:
529         (LauncherWindow::LauncherWindow):
530         (LauncherWindow::createChrome):
531         (LauncherWindow::showFindBar):
532         (LauncherWindow::find):
533         * QtTestBrowser/launcherwindow.h:
534
535 2011-05-30  Carlos Garcia Campos  <cgarcia@igalia.com>
536
537         Reviewed by Martin Robinson.
538
539         [GTK] Implement JavaScript dialogs in MiniBrowser
540         https://bugs.webkit.org/show_bug.cgi?id=61725
541
542         * MiniBrowser/gtk/BrowserWindow.c:
543         (WKStringGetCString): Helper function to get a UTF-8 c string from
544         a WKStringRef.
545         (WKURLGetCString): Helper function to get a UTF-8 c string from a
546         WKURLRef.
547         (browserWindowUpdateURL): Update to use WKURLGetCString.
548         (didReceiveTitleForFrame): Update to use WKStringGetCString.
549         (createMessageDialog): Create a GtkMessageDialog.
550         (runJavaScriptAlert):
551         (runJavaScriptConfirm):
552         (runJavaScriptPrompt):
553         (browserWindowUIClientInit):
554
555 2011-05-29  Yuta Kitamura  <yutak@chromium.org>
556
557         Reviewed by Kent Tamura.
558
559         WebSocket closing handshake
560         https://bugs.webkit.org/show_bug.cgi?id=35721
561
562         * Scripts/webkitpy/thirdparty/__init__.py:
563         Pull in pywebsocket 0.6b1. We need to update pywebsocket
564         to get the right behavior of closing handshake.
565
566 2011-05-29  Daniel Bates  <dbates@rim.com>
567
568         Reviewed by David Kilzer.
569
570         REGRESSION (r86515): svn-apply ignores diffs that omit line count in chunk range
571         https://bugs.webkit.org/show_bug.cgi?id=61162
572
573         Fixes an issue where svn-apply may ignore a diff that contains a chunk range line
574         that omits a line count. In particular, the chunk range regular expression does
575         not match a chunk range line that omits a line count. GNU diff(1) will omit the
576         line count in the chunk range if the line count is exactly 1. For example, appending
577         a new line to the end of an existing file F that contains exactly one line of text will
578         be represented in a diff with a chunk range line that omits the line count for F.
579
580         * Scripts/VCSUtils.pm:
581           (parseChunkRange): Added.
582         * Scripts/webkitperl/VCSUtils_unittest/parseChunkRange.pl: Added.
583         * Scripts/webkitperl/VCSUtils_unittest/parseDiff.pl:
584           - Added unit test "Git: Append new line to the end of an existing file".
585
586 2011-05-28  Adam Barth  <abarth@webkit.org>
587
588         Reviewed by Eric Seidel.
589
590         EWS builds patches that fail to build twice, which seems useless and slows down the bots
591         https://bugs.webkit.org/show_bug.cgi?id=55585
592
593         This patch switches all the early warning system bots over to the new
594         PatchAnalysisTask-based infrastructure.  This patch makes these bots
595         more efficient (in the case where patches fail to build) and paves the
596         way for running tests on these bots!
597
598         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
599         * Scripts/webkitpy/tool/bot/earlywarningsystemtask.py:
600         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
601         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
602         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
603         * Scripts/webkitpy/tool/commands/queues.py:
604
605 2011-05-28  Kenichi Ishibashi  <bashi@chromium.org>
606
607         Reviewed by Kent Tamura.
608
609         Add bashi as a committer.
610
611         https://bugs.webkit.org/show_bug.cgi?id=61685
612
613         * Scripts/webkitpy/common/config/committers.py:
614
615 2011-05-27  Adam Barth  <abarth@webkit.org>
616
617         Reviewed by Eric Seidel.
618
619         When checking whether the tree is red, the EWS posts a link to the wrong log
620         https://bugs.webkit.org/show_bug.cgi?id=61072
621
622         We need to cache the original script error because that contains the
623         failure log we want to upload.  If we don't cache that script error,
624         self._script_error will get overwritten when we sanity check the clean
625         tree (and it also has test failures).
626
627         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
628             - This testing approach is slightly goofy.  We'd like to use
629               assertRaisesRegexp, but that's not available until Python 2.7.
630         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
631
632 2011-05-27  Jochen Eisinger  <jochen@chromium.org>
633
634         Reviewed by Adam Barth.
635
636         Add layoutTestController.setStorageAllowed() to control whether access
637         to the localStorage API is enabled via the WebPermissionClient
638         https://bugs.webkit.org/show_bug.cgi?id=61581
639
640         * DumpRenderTree/chromium/LayoutTestController.cpp:
641         (LayoutTestController::LayoutTestController):
642         (LayoutTestController::setStorageAllowed):
643         * DumpRenderTree/chromium/LayoutTestController.h:
644         * DumpRenderTree/chromium/TestShell.cpp:
645         (TestShell::TestShell):
646         (TestShell::createNewWindow):
647         * DumpRenderTree/chromium/TestShell.h:
648         * DumpRenderTree/chromium/WebPermissions.h: Added.
649         (WebPermissions::WebPermissions):
650         (WebPermissions::allowStorage):
651         (WebPermissions::setStorageAllowed):
652         (WebPermissions::reset):
653
654 2011-05-27  Dirk Pranke  <dpranke@chromium.org>
655
656         Reviewed by Eric Seidel.
657
658         NRWT: clean up metered_stream code in preparation for 'nooverwriting' patch
659         https://bugs.webkit.org/show_bug.cgi?id=60326
660
661         This patch removes a lot of the complexity from the
662         metered_stream implementation that was unnecessary since there
663         was only one caller and the logic could be coordinated better.
664
665         There should be no functional changes in this patch, just code
666         getting deleted and cleaned up.
667
668         * Scripts/webkitpy/layout_tests/layout_package/metered_stream.py:
669         * Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py:
670         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
671         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
672
673 2011-05-27  Dirk Pranke  <dpranke@chromium.org>
674
675         Reviewed by Eric Seidel.
676
677         NRWT: minor cleanup in printing module
678         https://bugs.webkit.org/show_bug.cgi?id=60329
679
680         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
681
682 2011-05-27  Steve Lacey  <sjl@chromium.org>
683
684         Reviewed by David Levin.
685
686         Add sjl@chromium.org as a committer.
687
688         https://bugs.webkit.org/show_bug.cgi?id=61672
689
690         * Scripts/webkitpy/common/config/committers.py:
691
692 2011-05-27  Dirk Pranke  <dpranke@chromium.org>
693
694         Reviewed by Adam Barth.
695
696         NRWT: remove --print detailed-progress
697         https://bugs.webkit.org/show_bug.cgi?id=60324
698
699         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
700         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
701         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
702         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
703
704 2011-05-27  Dirk Pranke  <dpranke@chromium.org>
705
706         Reviewed by Ojan Vafai.
707
708         NRWT: debug messages from the workers are being logged twice
709         https://bugs.webkit.org/show_bug.cgi?id=60428
710
711         It looks like when the workers are run in separate processes
712         we end up getting two copies of every log message they print.
713         This has to do with the multiprocessing module on UNIX cloning
714         the log configuration in a way I wasn't expecting, and so two
715         log handlers end up getting registered.
716
717         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
718         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
719         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
720         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
721
722 2011-05-27  Adam Roben  <aroben@apple.com>
723
724         Always decode tester names from location.hash
725
726         When navigating directly to a tester-specific URL (e.g.,
727         <http://build.webkit.org/TestFailures/#/Windows%207%20Release%20(Tests)>), you could end up
728         with a URL-encoded tester name in the page, and then the encoded name could make its way
729         into Bugzilla bugs. Bad!
730
731         Reviewed by David Kilzer.
732
733         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
734         (ViewController.prototype.parseHash): Pass the tester name we got from location.hash through
735         decodeURIComponent before trying to use it.
736
737 2011-05-27  Adam Roben  <aroben@apple.com>
738
739         Remove testing code that snuck into r87528
740
741         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
742         (ViewController.prototype._domForNewBugLink): Remove a line I was using for some testing.
743
744 2011-05-27  Adam Roben  <aroben@apple.com>
745
746         Add links to TestFailures to easily file bugs about failing tests
747
748         Fixes <http://webkit.org/b/61636> <rdar://problem/9514886> TestFailures page should have
749         links to file bugs for failing tests
750
751         Reviewed by David Kilzer.
752
753         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
754         (Builder.prototype.resultsPageURL): New helper function to return the URL to results.html
755         for a particular build.
756         (Builder.prototype._getFailingTests): Use resultsPageURL.
757
758         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
759         (ViewController.prototype._displayBuilder): Add links to file bugs about each set of
760         failures.
761         (ViewController.prototype._domForBuildName): Use resultsPageURL.
762         (ViewController.prototype._domForNewBugLink): Added. Creates an element that represents the
763         new bug link. The bug has a title, description, and URL already filled in for you!
764
765 2011-05-27  Sheriff Bot  <webkit.review.bot@gmail.com>
766
767         Unreviewed, rolling out r87464.
768         http://trac.webkit.org/changeset/87464
769         https://bugs.webkit.org/show_bug.cgi?id=61643
770
771         client-close/server-close tests are flaky on Chromium Mac and
772         Windows (Requested by aklein on #webkit).
773
774         * Scripts/webkitpy/thirdparty/__init__.py:
775
776 2011-05-27  Alexey Proskuryakov  <ap@apple.com>
777
778         Added Douglas Davidson to contributors list for easier CC'ing.
779
780         * Scripts/webkitpy/common/config/committers.py:
781
782 2011-05-26  Mihai Parparita  <mihaip@chromium.org>
783
784         Reviewed by Adam Barth.
785
786         Fix worldID and destinationDomain argument names
787         https://bugs.webkit.org/show_bug.cgi?id=61571
788
789         As part of working on r87423, I noticed a couple of inconsistencies in
790         argument names:
791         - We would use worldId in .h files but worldID in .cpp files.
792           Standardize on the latter
793         - SecurityOrigin::addOriginAccessWhitelistEntry would take a
794           destinationDomains argument, even though the actual parameter was for
795           a single domain (renamed to be singular).
796
797         * DumpRenderTree/LayoutTestController.h:
798         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
799         (LayoutTestController::evaluateScriptInIsolatedWorld):
800
801 2011-05-26  Yuta Kitamura  <yutak@chromium.org>
802
803         Reviewed by Kent Tamura.
804
805         WebSocket closing handshake
806         https://bugs.webkit.org/show_bug.cgi?id=35721
807
808         * Scripts/webkitpy/thirdparty/__init__.py:
809         Pull in pywebsocket 0.6b1. We need to update pywebsocket
810         to get the right behavior of closing handshake.
811
812 2011-05-26  Qi Zhang  <qi.2.zhang@nokia.com>
813
814         Reviewed by Andreas Kling.
815
816         WebKitTestRunner needs layoutTestController.setAuthorAndUserStylesEnabled
817         https://bugs.webkit.org/show_bug.cgi?id=42676
818
819         Implemented layoutTestController.setAuthorAndUserStylesEnabled for WebKitTestRunner.
820
821         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
822         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
823         (WTR::InjectedBundle::beginTesting):
824         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
825         (WTR::LayoutTestController::setAuthorAndUserStylesEnabled):
826         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
827
828 2011-05-26  Ademar de Souza Reis Jr.  <ademar.reis@openbossa.org>
829
830         Reviewed by Andreas Kling.
831
832         [Qt] QtTestBrowser "Switch User Agent" dialog is broken
833         https://bugs.webkit.org/show_bug.cgi?id=61527
834
835         The setMaximumHeight() call is unecessary and breaks the dialog at
836         least on my setup.
837
838         * QtTestBrowser/launcherwindow.cpp:
839         (LauncherWindow::showUserAgentDialog):
840
841 2011-05-02  Robert Hogan  <robert@webkit.org>
842
843         Reviewed by Adam Roben.
844
845         [Qt] Allow popup windows from plugins when initiated by a user gesture
846
847         https://bugs.webkit.org/show_bug.cgi?id=41292
848
849         Support parameters introduced in plugin-initiate-popup-window.html
850
851         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
852         (webkit_test_plugin_new_instance):
853         (webkit_test_plugin_handle_event):
854
855 2011-05-26  Mihai Parparita  <mihaip@chromium.org>
856
857         Reviewed by Adam Barth.
858
859         Support cross-origin XMLHttpRequest in isolated worlds
860         https://bugs.webkit.org/show_bug.cgi?id=59843
861
862         Expose setIsolatedWorldSecurityOrigin in LayoutTestController.
863
864         * DumpRenderTree/chromium/LayoutTestController.cpp:
865         (LayoutTestController::LayoutTestController):
866         (LayoutTestController::setIsolatedWorldSecurityOrigin):
867         * DumpRenderTree/chromium/LayoutTestController.h:
868
869 2011-05-25  Brian Weinstein  <bweinstein@apple.com>
870
871         Reviewed by Adam Roben.
872
873         WebKit2: Status bar, toolbar, and menu bar checks should be in the injected bundle
874         https://bugs.webkit.org/show_bug.cgi?id=61474
875         <rdar://problem/9468337>
876
877         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
878         (WTR::InjectedBundlePage::InjectedBundlePage): Add empty entries in the WKBundlePageUIClient.
879
880 2011-05-26  Andreas Kling  <kling@webkit.org>
881
882         Unreviewed, correct address to webkit-committers mailing list.
883
884         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
885
886 2011-05-25  Adam Roben  <aroben@apple.com>
887
888         Don't assume that builds finish in the order they were started
889
890         Fixes <http://webkit.org/b/61464> TestFailures page shows failures from a not-most-recent
891         build when one slave is taking a long time to finish a build
892
893         Reviewed by Antti Koivisto.
894
895         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
896         (Builder.prototype.getMostRecentCompletedBuildNumber): Don't assume that all builds that are
897         newer than the oldest in-progress build are also themselves in progress. Just check whether
898         each build is actually in progress (as signified by its presence in the currentBuilds set).
899
900 2011-05-25  Adam Roben  <aroben@apple.com>
901
902         Correctly detect cases where only a single test failed/crashed/timed out
903
904         Fixes <http://webkit.org/b/61463> TestFailures page doesn't show testers with only a single
905         failing test
906
907         Reviewed by Darin Adler.
908
909         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
910         (Builder.prototype.getNumberOfFailingTests): Don't require "test cases" to be plural. It
911         will be singular when only a single test fails/crashes/times out.
912
913         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
914         (ViewController.prototype._displayTesters): Only pluralize "test" when more than one test
915         failed/crashed/timed out.
916
917 2011-05-26  Tom Hudson  <tomhudson@google.com>
918
919         Reviewed by Tony Chang.
920
921         Add flags to Chromium ImageDiff to write image comparison metrics on
922         stdout.
923         https://bugs.webkit.org/show_bug.cgi?id=60569
924
925         * DumpRenderTree/chromium/ImageDiff.cpp:
926         (Image::createFromStdin): Fix spelling.
927         (maxOf3): New function to do 3-way maximum.
928         (getRedComponent): Extract red component from ABGR packed encoding.
929         (getGreenComponent): Extract green component from ABGR packed encoding.
930         (getBlueComponent): Extract blue component from ABGR packed encoding.
931         (weightedPercentageDifferent): Determine % of pixels different in two
932         images multiplied by intensity difference of each pixel.
933         (printHelp): Document new arguments.
934         (compareImages): Parameterize.
935         (untestedCompareImages): Parameterize.
936         (diffImages): Parameterize; write results on stdout if requested.
937         (main): Parse new arguments.
938
939 2011-05-25  Qi Zhang  <qi.2.zhang@nokia.com>
940
941         Reviewed by Simon Hausmann.
942
943         WebKitTestRunner needs layoutTestController.setPrivateBrowsingEnabled
944         https://bugs.webkit.org/show_bug.cgi?id=42697
945
946         Implemented layoutTestController.setPrivateBrowsingEnabled for WebKitTestRunner.
947
948         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
949         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
950         (WTR::InjectedBundle::beginTesting):
951         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
952         (WTR::LayoutTestController::setPrivateBrowsingEnabled):
953         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
954
955 2011-05-25  Jon Honeycutt  <jhoneycutt@apple.com>
956
957         REGRESSION (WebKit2): Crash in Flash on USA Today photo gallery
958         https://bugs.webkit.org/show_bug.cgi?id=61428
959         <rdar://problem/9457006>
960
961         Reviewed by Adam Roben.
962
963         The crash occurs when Flash posts a message to a window that it
964         creates, and in processing the message, it calls NPN_Evaluate to
965         evaluate JavaScript that removes the plug-in from the page. Flash then
966         crashes when we return to Flash code.
967
968         This test emulates that behavior.
969
970         * DumpRenderTree/TestNetscapePlugIn/win/CallJSThatDestroysPlugin.cpp: Added.
971         (CallJSThatDestroysPlugin::CallJSThatDestroysPlugin):
972         Initialize member vars.
973         (CallJSThatDestroysPlugin::~CallJSThatDestroysPlugin):
974         Remove our custom property from the message window, and destroy it.
975         (CallJSThatDestroysPlugin::NPP_Destroy):
976         Set m_isDestroyed, log that the plug-in was destroyed, and notify the
977         layout test controller that we're done.
978         (wndProc):
979         Get the PluginTest object, and call its runTest() function.
980         (CallJSThatDestroysPlugin::NPP_New):
981         Setup the test: register a class for the message-only window, create
982         it, and post a message to it to run the test.
983         (CallJSThatDestroysPlugin::runTest):
984         Execute JS that removes the plug-in from the page, and if we're not
985         destroyed, log a success message.
986
987         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
988         Add new test to project.
989
990 2011-05-25  Tony Chang  <tony@chromium.org>
991
992         Reviewed by Adam Barth.
993
994         webkitpy's NetworkTransaction shouldn't use mechanize
995         https://bugs.webkit.org/show_bug.cgi?id=61450
996
997         mechanize.HTTPError is just an alias for urllib2.HTTPError, so
998         use urllib2 directly.  Re-add NetworkTransaction to
999         test_results_uploader.py (reverting r87124).
1000
1001         * Scripts/webkitpy/common/net/networktransaction.py:
1002         * Scripts/webkitpy/layout_tests/layout_package/test_results_uploader.py:
1003
1004 2011-05-25  Kevin Ollivier  <kevino@theolliviers.com>
1005
1006         [Chromium] Unreviewed build fix. Do not define WTF/JS_EXPORT_PRIVATE to JS_EXPORTDATA
1007         until the port begins using those export macros.
1008
1009         * DumpRenderTree/chromium/config.h:
1010
1011 2011-05-25  Sheriff Bot  <webkit.review.bot@gmail.com>
1012
1013         Unreviewed, rolling out r87257.
1014         http://trac.webkit.org/changeset/87257
1015         https://bugs.webkit.org/show_bug.cgi?id=61457
1016
1017         This broke layout tests, see bug 61431. (Requested by
1018         hwennborg on #webkit).
1019
1020         * DumpRenderTree/chromium/LayoutTestController.cpp:
1021         (LayoutTestController::LayoutTestController):
1022         (LayoutTestController::clearAllDatabases):
1023         * DumpRenderTree/chromium/LayoutTestController.h:
1024
1025 2011-05-25  Qi Zhang  <qi.2.zhang@nokia.com>
1026
1027         Reviewed by Andreas Kling.
1028
1029         [Qt] fast/css/disabled-author-styles.html failed
1030         https://bugs.webkit.org/show_bug.cgi?id=61438
1031
1032         Implemented API setAuthorAndUserStylesEnabled in LayoutTestControllerQt.
1033
1034         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1035         (WebCore::WebPage::resetSettings):
1036         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1037         (LayoutTestController::setAuthorAndUserStylesEnabled):
1038         * DumpRenderTree/qt/LayoutTestControllerQt.h:
1039
1040 2011-05-25  Adam Roben  <aroben@apple.com>
1041
1042         Identify, rather than skip, builds where run-webkit-tests exited early due to too many failures
1043
1044         Fixes <http://webkit.org/b/61441> TestFailures page should show when run-webkit-tests
1045         started exiting early due to too many crashes, timeouts, or failures
1046
1047         Reviewed by David Kilzer.
1048
1049         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
1050         (Builder.prototype.getNumberOfFailingTests): Rather than returning -1 ("unknown") when
1051         run-webkit-tests exits early due to too many failures, pass that information in another
1052         argument to the callback.
1053         (Builder.prototype.startFetchingBuildHistory): Updated the documentation comment to reflect
1054         the new structure of the object passed to the callback.
1055         (Builder.prototype._getFailingTests): Updated to expect a tooManyFailures boolean from
1056         getNumberOfFailingTests and to pass that along to our own callbacks.
1057         (Builder.prototype._incorporateBuildHistory): Updated to expect a tooManyFailures boolean
1058         from _getFailingTests and to store that value in the history object.
1059
1060         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
1061         (.info): Added.
1062
1063         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1064         (ViewController.prototype._displayBuilder): Updated for change in structure to the history
1065         object and to add a note when run-webkit-tests exited early due to too many failures.
1066         (ViewController.prototype._displayTesters): Renamed testersAndFailureCounts to
1067         latestBuildInfos since each entry now contains more than just the tester and failure count.
1068         Now displays a message for testers where the latest build exited early due to too many
1069         failures. Updated to expect a tooManyFailures boolean from getNumberOfFailingTests and to
1070         store that value in latestBuildInfos.
1071
1072 2011-05-24  Keishi Hattori  <keishi@webkit.org>
1073
1074         Reviewed by Kent Tamura.
1075
1076         Disable textfield implementation of <input type=color>. Add INPUT_COLOR feature flag. Add input color sanitizer.
1077         https://bugs.webkit.org/show_bug.cgi?id=61273
1078
1079         * Scripts/build-webkit: Added INPUT_COLOR feature flag.
1080
1081 2011-05-24  Greg Simon  <gregsimon@chromium.org>
1082
1083         Reviewed by Dimitri Glazkov.
1084
1085         Control Indexeddb backends from LayoutTestController
1086         https://bugs.webkit.org/show_bug.cgi?id=61000
1087
1088         * DumpRenderTree/chromium/LayoutTestController.cpp:
1089         (LayoutTestController::LayoutTestController):
1090         (LayoutTestController::setOverrideIndexedDBBackingStore):
1091         (LayoutTestController::clearAllDatabases):
1092         * DumpRenderTree/chromium/LayoutTestController.h:
1093
1094 2011-05-24  Robert Hogan  <robert@webkit.org>
1095
1096         Reviewed by Ryosuke Niwa.
1097
1098         [Qt] Fix fast/events/selectstart-by-single-click-with-shift.html
1099         https://bugs.webkit.org/show_bug.cgi?id=61393
1100
1101         * DumpRenderTree/qt/EventSenderQt.h:
1102         * DumpRenderTree/qt/EventSenderQt.cpp:
1103         (getModifiers): static function for interpreting modifiers
1104         (EventSender::mouseDown): Respect keyboard modifiers
1105         (EventSender::keyDown): use new static function
1106         (EventSender::sendOrQueueEvent): whitespace fix
1107
1108 2011-05-24  Jay Civelli  <jcivelli@chromium.org>
1109
1110         Reviewed by Adam Barth.
1111
1112         Adding MHTML reading support.
1113         https://bugs.webkit.org/show_bug.cgi?id=7168
1114
1115         * Scripts/build-webkit:
1116         * Scripts/old-run-webkit-tests:
1117         * Scripts/webkitperl/features.pm:
1118         * Scripts/webkitpy/layout_tests/port/test_files.py:
1119         * Scripts/webkitpy/layout_tests/port/webkit.py:
1120
1121 2011-05-24  Stephen White  <senorblanco@chromium.org>
1122
1123         Reviewed by Kenneth Russell.
1124
1125         Remove most style checks on Source/ThirdParty/glu, since it doesn't
1126         follow WebKit style.
1127         https://bugs.webkit.org/show_bug.cgi?id=61365
1128
1129         * Scripts/webkitpy/style/checker.py:
1130
1131 2011-05-24  Adam Roben  <aroben@apple.com>
1132
1133         Don't allow forcing builds through the build.webkit.org web interface
1134
1135         Someone has been forcing bogus builds.
1136
1137         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
1138
1139 2011-05-24  Qi Zhang  <qi.2.zhang@nokia.com>
1140
1141         Reviewed by Adam Roben.
1142
1143         WebKitTestRunner needs layoutTestController.setJavaScriptCanAccessClipboard
1144         https://bugs.webkit.org/show_bug.cgi?id=42669
1145
1146         Fixes the regression issue caused by patch 94265. The patch didn't intialize the setting before the 
1147         layouttest starting, it will make the test cases after editing/execCommand/clipboard-access.html failed.
1148         Because that case changed the setting.
1149
1150         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
1151         (WTR::InjectedBundle::beginTesting):
1152
1153 2011-05-24  Csaba Osztrogonác  <ossy@webkit.org>
1154
1155         Reviewed by Kenneth Rohde Christiansen.
1156
1157         [ORWT]Make --skipped=only option work with parameters with LayoutTests/ prefix
1158         https://bugs.webkit.org/show_bug.cgi?id=61052
1159
1160         * Scripts/old-run-webkit-tests:
1161
1162 2011-05-24  Adam Roben  <aroben@apple.com>
1163
1164         Make TestFailures show how many tests are failing on each tester, and omit testers with no failures
1165
1166         Fixes <http://webkit.org/b/61063> <rdar://problem/9460533> TestFailures page shows testers
1167         that don't have any failing tests, which isn't useful
1168
1169         Reviewed by David Kilzer.
1170
1171         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js:
1172         (Buildbot.prototype.getTesters): Renamed from getTesterNames. Now returns Builder objects
1173         instead of name strings.
1174         (Buildbot.prototype._buildersForNames): Added. Helper function to convert an array of
1175         builder names into an array of builders.
1176
1177         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
1178         (Builder.prototype.getMostRecentCompletedBuildNumber): Added. Returns the build number of
1179         the most recently completed build, or -1 if there is no such build.
1180         (Builder.prototype.getNumberOfFailingTests): Added. Returns the number of tests that failed
1181         in the given build, or -1 if the number could not be determined. Some of this code came from
1182         _getFailingTests.
1183         (Builder.prototype._getBuildJSON): Added. Code came from _getFailingTests.
1184         (Builder.prototype._getFailingTests): Changed to use new _getBuildJSON and
1185         getNumberOfFailingTests functions.
1186
1187         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
1188         (ViewController.prototype._displayTesters): Get the current number of test failures for each
1189         tester and show it in the list. Omit testers that have no failures at all. We keep the
1190         testers and failure counts in an array and sort it before displaying the current data, as
1191         the order in which data will be fetched is unpredictable.
1192
1193 2011-05-23  Tony Chang  <tony@chromium.org>
1194
1195         Reviewed by Ojan Vafai.
1196
1197         [nrwt] don't use NetworkTranslation when uploading results
1198         https://bugs.webkit.org/show_bug.cgi?id=61315
1199
1200         This avoids autoinstalling mechanize.  It didn't matter anyway
1201         because test_results_uploader.py wasn't using mechanize to make the
1202         request so NetworkTransaction::run wasn't doing anything extra
1203         (it only matters if it catches mechanize.HTTPError).
1204
1205         * Scripts/webkitpy/layout_tests/layout_package/test_results_uploader.py:
1206
1207 2011-05-23  Tony Chang  <tony@chromium.org>
1208
1209         Reviewed by Dirk Pranke.
1210
1211         [chromium] fix the rebaseline tool to work with Linux GPU failures
1212         https://bugs.webkit.org/show_bug.cgi?id=61312
1213
1214         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
1215
1216 2011-05-23  Darin Adler  <darin@apple.com>
1217
1218         Ignore generated .pyc files in a directory full of python code.
1219
1220         * Scripts/webkitpy/common/checkout/scm: Added property svn:ignore.
1221
1222 2011-05-23  Adam Barth  <abarth@webkit.org>
1223
1224         Reviewed by Eric Seidel.
1225
1226         sheriffbot should give more details about the failures in IRC
1227         https://bugs.webkit.org/show_bug.cgi?id=61233
1228
1229         With this patch, sheriffbot will annonce the set of failing tests,
1230         which might help folks triage the problem.
1231
1232         * Scripts/webkitpy/tool/commands/sheriffbot.py:
1233         * Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
1234
1235 2011-05-23  Sheriff Bot  <webkit.review.bot@gmail.com>
1236
1237         Unreviewed, rolling out r87078.
1238         http://trac.webkit.org/changeset/87078
1239         https://bugs.webkit.org/show_bug.cgi?id=61294
1240
1241         Patch contains errors and seems likely to break a bunch of
1242         bots (Requested by abarth on #webkit).
1243
1244         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
1245         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
1246         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
1247
1248 2011-05-23  Alice Boxhall  <aboxhall@chromium.org>
1249
1250         Reviewed by Ojan Vafai.
1251
1252         Convert json_results_generator.py to output version 4 JSON.
1253         https://bugs.webkit.org/show_bug.cgi?id=60869
1254
1255         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
1256         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
1257         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
1258
1259 2011-05-23  Patrick Gansterer  <paroga@webkit.org>
1260
1261         Reviewed by Adam Barth.
1262
1263         Move generation of shell_command from port to executive
1264         https://bugs.webkit.org/show_bug.cgi?id=61251
1265
1266         After move it's possible to use the code in other places too.
1267
1268         * Scripts/webkitpy/common/config/ports.py:
1269         * Scripts/webkitpy/common/system/executive.py:
1270
1271 2011-05-23  Qi Zhang  <qi.2.zhang@nokia.com>
1272
1273         Reviewed by Darin Adler.
1274
1275         WebKitTestRunner needs layoutTestController.setJavaScriptCanAccessClipboard
1276         https://bugs.webkit.org/show_bug.cgi?id=42669
1277
1278         Implement setJavaScriptCanAccessClipboard in WebKitTestRunner.
1279
1280         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
1281         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1282         (WTR::LayoutTestController::setJavaScriptCanAccessClipboard):
1283         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1284
1285 2011-05-22  Robert Hogan  <robert@webkit.org>
1286
1287         Reviewed by Kenneth Rohde Christiansen.
1288
1289         Fix policyDelegate in Qt DRT
1290         https://bugs.webkit.org/show_bug.cgi?id=61247
1291
1292         Use the 'policy delegate' implemented in FrameLoaderClient
1293         for layout tests. The partial implementation in DumpRenderTreeQt
1294         is redundant, so remove it.
1295
1296         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1297         (WebCore::WebPage::acceptNavigationRequest):
1298         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1299         (LayoutTestController::waitForPolicyDelegate):
1300
1301 2011-05-22  Patrick Gansterer  <paroga@webkit.org>
1302
1303         Reviewed by Darin Adler.
1304
1305         Use double quotes for filename arguments
1306         https://bugs.webkit.org/show_bug.cgi?id=61250
1307
1308         Windows does not support single quotes on the command line,
1309         so replace them with double quotes.
1310
1311         * Scripts/prepare-ChangeLog:
1312
1313 2011-05-20  Simon Fraser  <simon.fraser@apple.com>
1314
1315         Reviewed by Sam Weinig.
1316
1317         WebKitTestRunner needs layoutTestController.pauseTransitionAtTimeOnElementWithId
1318         https://bugs.webkit.org/show_bug.cgi?id=42550
1319         
1320         Implement pauseTransitionAtTimeOnElementWithId in WebKitTestRunner.
1321
1322         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
1323         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1324         Copy the code for pausing animations, FIXME comments and all.
1325         (WTR::LayoutTestController::pauseTransitionAtTimeOnElementWithId):
1326         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1327
1328 2011-05-20  Alok Priyadarshi  <alokp@chromium.org>
1329
1330         Reviewed by Stephen White.
1331
1332         Adding myself to the committers list.
1333
1334         * Scripts/webkitpy/common/config/committers.py:
1335
1336 2011-05-19  Adam Roben  <aroben@apple.com>
1337
1338         Test that the WebKit2 UI process doesn't crash when starting a download
1339
1340         Test for <http://webkit.org/b/61142> <rdar://problem/9471680> REGRESSION (r86812): Crash
1341         (preceded by assertion) in fastMalloc when downloading a file
1342
1343         Reviewed by Darin Adler.
1344
1345         * TestWebKitAPI/Tests/WebKit2/18-characters.html: Added.
1346
1347         * TestWebKitAPI/Tests/WebKit2/DownloadDecideDestinationCrash.cpp: Added.
1348         (TestWebKitAPI::decidePolicyForNavigationAction): Start a download.
1349         (TestWebKitAPI::decideDestinationWithSuggestedFilename): Record that the download was
1350         started, cancel the download, and return a bogus string.
1351
1352         (TestWebKitAPI::setContextDownloadClient):
1353         (TestWebKitAPI::setPagePolicyClient):
1354         Simple helper functions.
1355
1356         (TestWebKitAPI::TEST): Load 18-characters.html, which should trigger a download thanks to
1357         our policy client, and run until we know that the download was started. If we haven't
1358         crashed, we win!
1359
1360         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1361         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
1362         * TestWebKitAPI/win/copy-resources.cmd:
1363         Added new files.
1364
1365 2011-05-20  Kent Tamura  <tkent@chromium.org>
1366
1367         Reviewed by Ryosuke Niwa.
1368
1369         Fix style errors in DumpRenderTree/chromium/.
1370         https://bugs.webkit.org/show_bug.cgi?id=61172
1371
1372         * DumpRenderTree/chromium/CppBoundClass.cpp:
1373         (CppBoundClass::getAsCppVariant):
1374         (CppBoundClass::bindToJavascript):
1375         * DumpRenderTree/chromium/CppBoundClass.h:
1376         (CppBoundClass::GetterCallback::~GetterCallback):
1377         (CppBoundClass::CppBoundClass):
1378         (CppBoundClass::Callback::~Callback):
1379         (CppBoundClass::MemberCallback::MemberCallback):
1380         (CppBoundClass::MemberCallback::~MemberCallback):
1381         (CppBoundClass::MemberGetterCallback::MemberGetterCallback):
1382         (CppBoundClass::MemberGetterCallback::~MemberGetterCallback):
1383         * DumpRenderTree/chromium/CppVariant.h:
1384         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
1385         (DRTDevToolsAgent::~DRTDevToolsAgent):
1386         * DumpRenderTree/chromium/DRTDevToolsClient.cpp:
1387         (DRTDevToolsClient::sendFrontendLoaded):
1388         * DumpRenderTree/chromium/DRTDevToolsClient.h:
1389         * DumpRenderTree/chromium/DumpRenderTree.cpp:
1390         (main):
1391         * DumpRenderTree/chromium/EventSender.cpp:
1392         (SavedEvent::SavedEvent):
1393         (EventSender::EventSender):
1394         (EventSender::keyDown):
1395         * DumpRenderTree/chromium/EventSender.h:
1396         * DumpRenderTree/chromium/ImageDiff.cpp:
1397         (Image::Image):
1398         * DumpRenderTree/chromium/LayoutTestController.cpp:
1399         (LayoutTestController::LayoutTestController):
1400         (LayoutTestController::WorkQueue::reset):
1401         (WorkItemBackForward::WorkItemBackForward):
1402         (WorkItemLoadingScript::WorkItemLoadingScript):
1403         (WorkItemNonLoadingScript::WorkItemNonLoadingScript):
1404         (WorkItemLoad::WorkItemLoad):
1405         (WorkItemLoadHTMLString::WorkItemLoadHTMLString):
1406         (LayoutTestController::pathToLocalResource):
1407         * DumpRenderTree/chromium/LayoutTestController.h:
1408         (LayoutTestController::WorkItem::~WorkItem):
1409         (LayoutTestController::WorkQueue::WorkQueue):
1410         * DumpRenderTree/chromium/MockSpellCheck.cpp:
1411         (MockSpellCheck::MockSpellCheck):
1412         (MockSpellCheck::~MockSpellCheck):
1413         * DumpRenderTree/chromium/MockSpellCheck.h:
1414         * DumpRenderTree/chromium/NotificationPresenter.h:
1415         (NotificationPresenter::NotificationPresenter):
1416         * DumpRenderTree/chromium/Task.cpp:
1417         (WebTask::WebTask):
1418         * DumpRenderTree/chromium/Task.h:
1419         (TaskList::TaskList):
1420         * DumpRenderTree/chromium/TestEventPrinter.cpp:
1421         (DRTPrinter::DRTPrinter):
1422         (TestShellPrinter::TestShellPrinter):
1423         * DumpRenderTree/chromium/TestNavigationController.cpp:
1424         (TestNavigationEntry::TestNavigationEntry):
1425         (TestNavigationEntry::~TestNavigationEntry):
1426         (TestNavigationController::TestNavigationController):
1427         (TestNavigationController::reload):
1428         (TestNavigationController::loadEntry):
1429         (TestNavigationController::didNavigateToEntry):
1430         * DumpRenderTree/chromium/TestNavigationController.h:
1431         (TestShellExtraData::TestShellExtraData):
1432         * DumpRenderTree/chromium/TestShell.cpp:
1433         (dumpHistoryItem):
1434         * DumpRenderTree/chromium/TestShell.h:
1435         (TestParams::TestParams):
1436         * DumpRenderTree/chromium/TestShellWin.cpp:
1437         (TestShell::waitTestFinished):
1438         * DumpRenderTree/chromium/TestWebWorker.h:
1439         (TestWebWorker::startWorkerContext):
1440         (TestWebWorker::terminateWorkerContext):
1441         (TestWebWorker::postMessageToWorkerContext):
1442         (TestWebWorker::clientDestroyed):
1443         (TestWebWorker::postMessageToWorkerObject):
1444         (TestWebWorker::postExceptionToWorkerObject):
1445         (TestWebWorker::postConsoleMessageToWorkerObject):
1446         (TestWebWorker::confirmMessageFromWorkerObject):
1447         (TestWebWorker::reportPendingActivity):
1448         (TestWebWorker::workerContextClosed):
1449         (TestWebWorker::~TestWebWorker):
1450         * DumpRenderTree/chromium/TextInputController.cpp:
1451         * DumpRenderTree/chromium/WebThemeControlDRTWin.h:
1452         * DumpRenderTree/chromium/WebThemeEngineDRTMac.h:
1453         * DumpRenderTree/chromium/WebThemeEngineDRTWin.cpp:
1454         (WebThemeEngineDRTWin::paintButton):
1455         (WebThemeEngineDRTWin::paintMenuList):
1456         (WebThemeEngineDRTWin::paintTrackbar):
1457         * DumpRenderTree/chromium/WebThemeEngineDRTWin.h:
1458         (WebThemeEngineDRTWin::WebThemeEngineDRTWin):
1459         * DumpRenderTree/chromium/WebViewHost.cpp:
1460         (WebViewHost::navigate):
1461         (WebViewHost::updateSessionHistory):
1462         (WebViewHost::paintInvalidatedRegion):
1463         * DumpRenderTree/chromium/WebViewHost.h:
1464
1465 2011-05-19  Dmitry Lomov  <dslomov@google.com>
1466
1467         Reviewed by Adam Roben.
1468
1469         Detect hangs in run-api-tests
1470         https://bugs.webkit.org/show_bug.cgi?id=48043
1471
1472         * Scripts/run-api-tests: Added test timeouts
1473
1474 2011-05-19  Dmitry Lomov  <dslomov@google.com>
1475
1476         Reviewed by Adam Roben.
1477
1478         run-api-tests should run one test per process
1479         https://bugs.webkit.org/show_bug.cgi?id=61088
1480
1481         * Scripts/run-api-tests: Resurrecting the previous revison of this file, with fixes to system call under Windows, 
1482         return code, and parsing GTest output format.
1483
1484 2011-05-15  Robert Hogan  <robert@webkit.org>
1485
1486         Reviewed by Antonio Gomes.
1487
1488         [Qt][GTK] plugins/get-url-with-javascript-url.html fails
1489         https://bugs.webkit.org/show_bug.cgi?id=60834
1490         Fix unix test plugin for plugins/get-url-with-javascript-url.html
1491
1492         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
1493         (webkit_test_plugin_new_stream):
1494         (webkit_test_plugin_write_ready):
1495         (webkit_test_plugin_write):
1496
1497 2011-05-19  Sheriff Bot  <webkit.review.bot@gmail.com>
1498
1499         Unreviewed, rolling out r86869, r86873, r86875, and r86877.
1500         http://trac.webkit.org/changeset/86869
1501         http://trac.webkit.org/changeset/86873
1502         http://trac.webkit.org/changeset/86875
1503         http://trac.webkit.org/changeset/86877
1504         https://bugs.webkit.org/show_bug.cgi?id=61139
1505
1506         broke builds and debug DRT (Requested by rniwa on #webkit).
1507
1508         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
1509         * DumpRenderTree/chromium/TestShell.cpp:
1510         (TestShell::bindJSObjectsToWindow):
1511         * DumpRenderTree/mac/FrameLoadDelegate.mm:
1512         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]):
1513
1514 2011-05-18  Dimitri Glazkov  <dglazkov@chromium.org>
1515
1516         Reviewed by Darin Adler.
1517
1518         Add build logistics and plumbing for window.internals object.
1519         https://bugs.webkit.org/show_bug.cgi?id=60313
1520
1521         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added linking new WebCoreTestSupport library.
1522         * DumpRenderTree/chromium/TestShell.cpp:
1523         (TestShell::bindJSObjectsToWindow): Added injection code.
1524         * DumpRenderTree/mac/FrameLoadDelegate.mm:
1525         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]): Ditto.
1526
1527 2011-05-19  Zsolt Fehér  <h490267@stud.u-szeged.hu>
1528
1529         Reviewed by Csaba Osztrogonác.
1530
1531         [Qt] Implement eventSender.scalePageBy
1532         https://bugs.webkit.org/show_bug.cgi?id=60015
1533
1534         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1535         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
1536         * DumpRenderTree/qt/EventSenderQt.cpp:
1537         (EventSender::scalePageBy):
1538         * DumpRenderTree/qt/EventSenderQt.h:
1539
1540 2011-05-19  Kenichi Ishibashi  <bashi@chromium.org>
1541
1542         Reviewed by Kent Tamura.
1543
1544         [Chromium] IME candidate window appears wrong position in an iframe
1545         https://bugs.webkit.org/show_bug.cgi?id=61023
1546
1547         Call the focused frame's firstRectForCharacterRange() instead of the
1548         main frame so that DRT behaves as the same as Chromium.
1549
1550         * DumpRenderTree/chromium/TextInputController.cpp:
1551         (TextInputController::firstRectForCharacterRange):
1552
1553 2011-05-18  Adam Roben  <aroben@apple.com>
1554
1555         Add a new page to build.webkit.org to help find when tests started failing
1556
1557         The page is accessible at <http://build.webkit.org/TestFailures/>. It is pretty minimalist
1558         right now, but already shows some useful information. It's somewhat similar to webkit-patch
1559         failure-reason and sheriffbot, and perhaps can be combined with them eventually. It's a
1560         little more convenient than either of them, though, because it's all done in the browser
1561         (and thus it's easy to go directly to the relevant test results).
1562
1563         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js: Added.
1564         (Buildbot): This class represents a Buildbot server.
1565         (Buildbot.prototype.buildURL): Returns the URL for the summary page for a particular build.
1566         (Buildbot.prototype.builderNamed): Returns a Builder with the given name.
1567         (Buildbot.prototype.getTesterNames): Fetches the names of all testers and passes them to the
1568         callback.
1569         (Buildbot.prototype.parseBuildName): Breaks up a build name into its constituent parts. Must
1570         be implemented by a derived class that understands this server's build naming scheme.
1571         (Buildbot.prototype.resultsDirectoryURL): Returns the URL for the results directory for a
1572         particular build.
1573
1574         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js: Added.
1575         (Builder): This class represents one builder on the buildbot.
1576         (Builder.prototype.buildURL): Returns the URL for the summary page for a particular build.
1577         (Builder.prototype.failureDiagnosisTextAndURL): Returns data that provides a little more
1578         information about a particular test failure.
1579         (Builder.prototype.startFetchingBuildHistory): Periodically calls the callback with
1580         information about when tests started failing.
1581         (Builder.prototype.resultsDirectoryURL): Returns the URL for the results directory for a
1582         particular build.
1583         (Builder.prototype._getBuildNames): Fetches the names of all builds and passes them to the
1584         callback.
1585         (Builder.prototype._getFailingTests): Fetches the results.html page for the given build and
1586         extracts all the failing tests listed in it, passing them to the callback.
1587         (Builder.prototype._incorporateBuildHistory): Gets the failing tests for the specified
1588         build, merges them into the build history, and calls the callback telling it whether the
1589         next build should be fetched to provide more information.
1590
1591         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
1592         Added. Just some simple styles.
1593
1594         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js: Added.
1595         (createDefinitionList): Takes an array of pairs and turns them into a DL element.
1596         (getResource): Wrapper around XMLHttpRequest.
1597         (Array.prototype.findFirst): Finds the first element matching the given predicate and
1598         returns it.
1599         (Array.prototype.last): Returns the last element of the array.
1600
1601         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js: Added.
1602         (ViewController): This class contains the main logic for displaying the page.
1603         (ViewController.loaded): Just calls through to parseHash.
1604         (ViewController.parseHash): Either starts analyzing failures on a particular builder, or
1605         shows the list of all testers so one can be chosen. This function is called when the page
1606         loads and whenever we get a hashchange event.
1607         (ViewController._displayBuilder): Asks the builder to fetch build history, and displays it
1608         as it is fetched. The display ends up grouping tests by when they started failing.
1609         (ViewController._displayTesters): Gets the list of testers and displays it.
1610
1611         (ViewController._domForBuildName):
1612         (ViewController._domForFailedTest):
1613         Helper functions to create descriptions and links for a particular build or failed test.
1614
1615         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBuildbot.js: Added.
1616         (WebKitBuildbot): Calls up to the base class constructor with the correct base URL.
1617         (WebKitBuildbot.prototype.parseBuildName): Parses a build.webkit.org-style build name.
1618
1619         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Added. Just
1620         loads all the files and sets up a ViewController, which does the rest.
1621
1622         * BuildSlaveSupport/build.webkit.org-config/templates/root.html: Added a link to the new
1623         page.
1624
1625 2011-05-18  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
1626
1627         Reviewed by Csaba Osztrogonác.
1628
1629         [Qt] Implement layoutTestController.setValueForUser()
1630         https://bugs.webkit.org/show_bug.cgi?id=60956
1631
1632         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1633         (LayoutTestController::setValueForUser):
1634         * DumpRenderTree/qt/LayoutTestControllerQt.h:
1635
1636 2011-05-16  Robert Hogan  <robert@webkit.org>
1637
1638         Reviewed by Kenneth Rohde Christiansen.
1639
1640         plugins/invalidate_rect.html fails on linux ports
1641
1642         - Make the unix test netscape plugin recognize the onPaintEvent
1643           and windowedPlugin parameters.
1644
1645         https://bugs.webkit.org/show_bug.cgi?id=54051
1646
1647         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
1648         (handleEventX11):
1649         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1650         (LayoutTestController::displayInvalidatedRegion):
1651         * DumpRenderTree/qt/LayoutTestControllerQt.h:
1652         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
1653         (webkit_test_plugin_new_instance):
1654         (webkit_test_plugin_handle_event):
1655
1656 2011-05-17  Tony Chang  <tony@chromium.org>
1657
1658         Reviewed by Ojan Vafai.
1659
1660         [chromium] move Lucid 64 bit results into LayoutTests/platform/chromium-linux
1661         https://bugs.webkit.org/show_bug.cgi?id=60895
1662
1663         Update the tools to handle the move.  Lucid 32 results now go in
1664         chromium-linux-x86 and the default platform on Linux is now x86_64.
1665
1666         * Scripts/webkitpy/layout_tests/deduplicate_tests.py: Default to x86_64
1667         * Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py:
1668         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py: Default to Lucid
1669         * Scripts/webkitpy/layout_tests/port/base.py: ditto
1670         * Scripts/webkitpy/layout_tests/port/chromium_linux.py: Update directory fallback and default to Lucid 64
1671         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
1672         * Scripts/webkitpy/layout_tests/port/test.py: Update tests to default to x86_64
1673         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py: Update bot names.
1674         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py: Whitespace cleanup.
1675
1676 2011-05-17  Adam Roben  <aroben@apple.com>
1677
1678         Make run-api-tests work on Windows when there are spaces in the path
1679
1680         Fixes <http://webkit.org/b/60954> REGRESSION (r86511): run-api-tests fails if there are
1681         spaces in the path to TestWebKitAPI.exe
1682
1683         Reviewed by David Levin.
1684
1685         * Scripts/run-api-tests:
1686         (runTestTool): Use the "direct object" form of system() to avoid having the path to
1687         TestWebKitAPI.exe be split by the shell.
1688
1689 2011-05-16  Ian Henderson  <ianh@apple.com>
1690
1691         Reviewed by Joseph Pecoraro.
1692
1693         Page::goToItem doesn't work while loading is deferred
1694         https://bugs.webkit.org/show_bug.cgi?id=60412
1695
1696         Add setDefersLoading and goBack methods to LayoutTestController.  We
1697         need to use goBack() instead of history.back() because the latter goes
1698         through NavigationScheduler, hence doesn't exhibit the bug.
1699
1700         * DumpRenderTree/LayoutTestController.cpp:
1701         (goBackCallback):
1702         (setDefersLoadingCallback):
1703         (LayoutTestController::staticFunctions):
1704         * DumpRenderTree/LayoutTestController.h:
1705         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1706         (LayoutTestController::goBack):
1707         (LayoutTestController::setDefersLoading):
1708         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
1709         (LayoutTestController::goBack):
1710         (LayoutTestController::setDefersLoading):
1711         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1712         (LayoutTestController::goBack):
1713         (LayoutTestController::setDefersLoading):
1714         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
1715         (LayoutTestController::goBack):
1716         (LayoutTestController::setDefersLoading):
1717
1718 2011-05-16  Sam Weinig  <sam@webkit.org>
1719
1720         Reviewed by Anders Carlsson.
1721
1722         TestWebKitAPI should build with clang if it can
1723         https://bugs.webkit.org/show_bug.cgi?id=60918
1724
1725         * TestWebKitAPI/Configurations/CompilerVersion.xcconfig:
1726         Update CompilerVersion.xcconfig to match others.
1727
1728         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1729         Add CompilerVersion.xcconfig to the project.
1730
1731         * TestWebKitAPI/Tests/WebKit2/FailedLoad.cpp:
1732         (TestWebKitAPI::didFailProvisionalLoadWithErrorForFrame):
1733         * TestWebKitAPI/Tests/WebKit2/PageLoadDidChangeLocationWithinPageForFrame.cpp:
1734         (TestWebKitAPI::didSameDocumentNavigationForFrame):
1735         Add some casts to quiet warnings from clang.
1736
1737 2011-05-16  Dirk Pranke  <dpranke@chromium.org>
1738
1739         Reviewed by David Levin.
1740
1741         add dpranke as a reviewer
1742         https://bugs.webkit.org/show_bug.cgi?id=60919
1743
1744         * Scripts/webkitpy/common/config/committers.py:
1745
1746 2011-05-16  Robert Hogan  <robert@webkit.org>
1747
1748         Reviewed by Anders Carlsson.
1749
1750         [Gtk] plugins/get-url-notify-with-url-that-fails-to-load.html crashes on buildbot
1751
1752         https://bugs.webkit.org/show_bug.cgi?id=60838
1753
1754         The unix test plugin needs to call the test's NPP_URLNotify
1755         rather than the browsers.
1756
1757         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
1758         (webkit_test_plugin_url_notify):
1759
1760 2011-05-16  Brent Fulgham  <bfulgham@webkit.org>
1761
1762         Rubber-stamped by Adam Roben.
1763
1764         Adding myself as a reviewer.
1765         * Scripts/webkitpy/common/config/committers.py:
1766
1767 2011-05-16  Sam Weinig  <sam@webkit.org>
1768
1769         Fix typo pointed out by Dave Levin.
1770
1771         * TestWebKitAPI/PlatformUtilities.h:
1772         (TestWebKitAPI::Util::assertWKStringEqual):
1773         Strig -> String.
1774
1775 2011-05-16  Sam Weinig  <sam@webkit.org>
1776
1777         Reviewed by David Levin.
1778
1779         Convert api tester over to using gtest expectations directly
1780         https://bugs.webkit.org/show_bug.cgi?id=60862
1781
1782         * TestWebKitAPI/PlatformUtilities.cpp:
1783         (TestWebKitAPI::Util::toSTD):
1784         * TestWebKitAPI/PlatformUtilities.h:
1785         (TestWebKitAPI::Util::assertWKStrigEqual):
1786         Add convenience macro to compare WK2 strings. Add some overloads of toSTD to make
1787         the implementation of the macro simpler.
1788
1789         * TestWebKitAPI/Test.h:
1790         Remove TEST_ASSERT forwarder.
1791
1792         [Test changes elided]
1793
1794 2011-05-16  David Kilzer  <ddkilzer@apple.com>
1795
1796         <http://webkit.org/b/60913> C++ exceptions should not be enabled when building with llvm-gcc-4.2
1797         <rdar://problem/9446430>
1798
1799         Reviewed by Mark Rowe.
1800
1801         * MiniBrowser/Configurations/Base.xcconfig: Fixed typo.
1802         * TestWebKitAPI/Configurations/Base.xcconfig: Ditto.
1803         * WebKitTestRunner/Configurations/Base.xcconfig: Ditto.
1804
1805 2011-05-16  Jon Lee  <jonlee@apple.com>
1806
1807         Reviewed by Simon Fraser.
1808
1809         Can't horizontally scroll iframes and overflow because wheel events are always accepted
1810         https://bugs.webkit.org/show_bug.cgi?id=60779
1811
1812         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1813         (WTR::InjectedBundlePage::InjectedBundlePage): Set new default method to nil.
1814
1815 2011-05-16  Ademar de Souza Reis Jr.  <ademar.reis@openbossa.org>
1816
1817         Reviewed by Csaba Osztrogonác.
1818
1819         [Qt] Tools.pro misses include(common.pri)
1820         https://bugs.webkit.org/show_bug.cgi?id=60883
1821
1822         It includes features.pri, which depends on common.pri (the build is
1823         not broken today with the default options, but this is the right
1824         thing to do).
1825
1826         * Tools.pro:
1827
1828 2011-05-15  Dominic Cooney  <dominicc@chromium.org>
1829
1830         Unreviewed: adding myself to committers.py.
1831
1832         * Scripts/webkitpy/common/config/committers.py:
1833
1834 2011-05-15  Daniel Bates  <dbates@webkit.org>
1835
1836         Reviewed by Chris Jerdonek.
1837
1838         svn-apply can't handle Git diff that only changes executable bit
1839         https://bugs.webkit.org/show_bug.cgi?id=60848
1840
1841         Fixes an issue where svn-{apply, unapply} fail to handle a Git diff for a file
1842         that only has an executable bit change.
1843
1844         * Scripts/VCSUtils.pm:
1845           - Modified regex $chunkRangeRegEx to match chunk range with ending text, say: @@ -2,6 +2,18 @@ foo().
1846         (parseDiff): Modified to count the number of text chunks in the diff. This should also help towards
1847         fixing <https://bugs.webkit.org/show_bug.cgi?id=29684>.
1848         * Scripts/svn-apply:
1849         (patch): Only use standard patch tool if the patch is non-empty and for a text file (i.e. not a binary patch).
1850         * Scripts/svn-unapply:
1851         (patch): Ditto.
1852         * Scripts/webkitperl/VCSUtils_unittest/parseDiff.pl:
1853           - Updated expected results for test cases to account for numTextChunks.
1854           - Fixed malformed chunk range in test cases: "Git: simple", and "Git: Git diff followed by SVN diff".
1855           - Added test case "Git: file that only has an executable bit change".
1856         * Scripts/webkitperl/VCSUtils_unittest/parseDiffWithMockFiles.pl:
1857           - Updated expected results for test cases to account for numTextChunks.
1858
1859 2011-05-15  Sam Weinig  <sam@webkit.org>
1860
1861         Fix windows build.
1862
1863         * TestWebKitAPI/win/main.cpp:
1864         (main):
1865
1866 2011-05-15  Sam Weinig  <sam@webkit.org>
1867
1868         Reviewed by David Levin.
1869
1870         Make API tester more gtest friendly
1871         https://bugs.webkit.org/show_bug.cgi?id=60859
1872
1873         * Scripts/run-api-tests:
1874         Numerous changes:
1875         - Remove logic to run tests one at a time.
1876         - Remove logic to build up a list of tests, instead just pass --gtest_list_tests to 
1877           the test runner and allow gtest to print it for us.
1878         - Forward stdout from the test tool to stdout, even in non-verbose runs.
1879         - Run test tool from the current working directory instead of the root webkit
1880           directory.
1881
1882         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1883         Move plist to be with other bundle sources.
1884
1885         * TestWebKitAPI/TestsController.cpp:
1886         (TestWebKitAPI::TestsController::TestsController):
1887         (TestWebKitAPI::TestsController::run):
1888         * TestWebKitAPI/TestsController.h:
1889         * TestWebKitAPI/mac/main.mm:
1890         Simplify TestController down to a single run function. We should consider removing the class
1891         entirely at some point.
1892
1893 2011-05-15  Sheriff Bot  <webkit.review.bot@gmail.com>
1894
1895         Unreviewed, rolling out r86504.
1896         http://trac.webkit.org/changeset/86504
1897         https://bugs.webkit.org/show_bug.cgi?id=60853
1898
1899         "Broke Qt EventSender in editing/selection tests" (Requested
1900         by mwenge on #webkit).
1901
1902         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1903         (WebCore::WebPage::WebPage):
1904         (WebCore::DumpRenderTree::DumpRenderTree):
1905         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
1906         (WebCore::DumpRenderTree::initJSObjects):
1907         * DumpRenderTree/qt/DumpRenderTreeQt.h:
1908         (WebCore::DumpRenderTree::eventSender):
1909         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
1910         (webkit_test_plugin_new_instance):
1911         (webkit_test_plugin_set_window):
1912         (webkit_test_plugin_handle_event):
1913
1914 2011-05-12  Robert Hogan  <robert@webkit.org>
1915
1916         Reviewed by Benjamin Poulain.
1917
1918         [Qt] fix http/tests/plugins/plugin-document-has-focus.html
1919
1920         The support for this test added to the unix test plugin here
1921         may allow other platforms to pass it (nearly everyone
1922         skips it). On Qt it required a bit of trickiness with
1923         the page's EventSender object to get it working fully though,
1924         so only unskipping Qt here.
1925
1926         https://bugs.webkit.org/show_bug.cgi?id=60722
1927
1928         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1929         (WebCore::WebPage::WebPage):
1930         (WebCore::WebPage::eventSender):
1931         (WebCore::DumpRenderTree::DumpRenderTree):
1932         (WebCore::DumpRenderTree::initJSObjects):
1933         (WebCore::DumpRenderTree::createWindow):
1934         * DumpRenderTree/qt/DumpRenderTreeQt.h:
1935         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
1936         (webkit_test_plugin_new_instance):
1937         (webkit_test_plugin_set_window):
1938         (webkit_test_plugin_handle_event):
1939
1940 2011-05-14  Arno Renevier  <arno@renevier.net>
1941
1942         Reviewed by Martin Robinson.
1943
1944         make run-gtk-tests executable
1945         https://bugs.webkit.org/show_bug.cgi?id=59278
1946
1947         * Scripts/run-gtk-tests:
1948
1949 2011-05-14  Anders Carlsson  <andersca@apple.com>
1950
1951         Reviewed by Kevin Decker.
1952
1953         NPP_URLNotify is not called if a URL passed to NPN_GetURLNotify fails to load
1954         https://bugs.webkit.org/show_bug.cgi?id=60823
1955         <rdar://problem/9430386>
1956
1957         Add test that calls NPN_GetURLNotify on a URL that can't be loaded, and checks if NPP_URLNotify is called.
1958
1959         * DumpRenderTree/DumpRenderTree.gypi:
1960         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
1961         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLNotifyWithURLThatFailsToLoad.cpp: Added.
1962         (GetURLNotifyWithURLThatFailsToLoad::GetURLNotifyWithURLThatFailsToLoad):
1963         (GetURLNotifyWithURLThatFailsToLoad::NPP_New):
1964         (GetURLNotifyWithURLThatFailsToLoad::NPP_URLNotify):
1965         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
1966         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
1967         * GNUmakefile.am:
1968
1969 2011-05-14  Sheriff Bot  <webkit.review.bot@gmail.com>
1970
1971         Unreviewed, rolling out r86476.
1972         http://trac.webkit.org/changeset/86476
1973         https://bugs.webkit.org/show_bug.cgi?id=60832
1974
1975         It made plugins/get-url-notify-with-url-that-fails-to-
1976         load.html crash on GTK (Requested by Ossy on #webkit).
1977
1978         * DumpRenderTree/DumpRenderTree.gypi:
1979         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
1980         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLNotifyWithURLThatFailsToLoad.cpp: Removed.
1981         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
1982         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
1983         * GNUmakefile.am:
1984
1985 2011-05-13  Anders Carlsson  <andersca@apple.com>
1986
1987         Reviewed by Sam Weinig.
1988
1989         NPP_URLNotify is not called if a URL passed to NPN_GetURLNotify fails to load
1990         https://bugs.webkit.org/show_bug.cgi?id=60823
1991         <rdar://problem/9430386>
1992
1993         Add test that calls NPN_GetURLNotify on a URL that can't be loaded, and checks if NPP_URLNotify is called.
1994
1995         * DumpRenderTree/DumpRenderTree.gypi:
1996         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
1997         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLNotifyWithURLThatFailsToLoad.cpp: Added.
1998         (GetURLNotifyWithURLThatFailsToLoad::GetURLNotifyWithURLThatFailsToLoad):
1999         (GetURLNotifyWithURLThatFailsToLoad::NPP_New):
2000         (GetURLNotifyWithURLThatFailsToLoad::NPP_URLNotify):
2001         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
2002         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
2003         * GNUmakefile.am:
2004
2005 2011-05-13  Anders Carlsson  <andersca@apple.com>
2006
2007         Build fix.
2008
2009         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLWithJavaScriptURL.cpp:
2010
2011 2011-05-13  Anders Carlsson  <andersca@apple.com>
2012
2013         Reviewed by Sam Weinig.
2014
2015         NPN_GetURL with a javascript: URL doesn't send back the result in the stream
2016         https://bugs.webkit.org/show_bug.cgi?id=60810
2017
2018         Add a test that calls NPN_GetURL on a javascript: URL and checks that it gets back the expected result.
2019
2020         * DumpRenderTree/DumpRenderTree.gypi:
2021         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
2022         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
2023         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
2024         * GNUmakefile.am:
2025         Add files.
2026
2027         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
2028         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
2029         Add new NPN and NPP wrappers and sort them according to the order they appear in npfunctions.h
2030
2031         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLWithJavaScriptURL.cpp: Added.
2032         (GetURLWithJavaScriptURL::GetURLWithJavaScriptURL):
2033         (GetURLWithJavaScriptURL::NPP_New):
2034         (GetURLWithJavaScriptURL::NPP_NewStream):
2035         (GetURLWithJavaScriptURL::NPP_DestroyStream):
2036         (GetURLWithJavaScriptURL::NPP_WriteReady):
2037         (GetURLWithJavaScriptURL::NPP_Write):
2038         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
2039         (NPP_NewStream):
2040         (NPP_WriteReady):
2041         (NPP_Write):
2042         (NPP_URLNotify):
2043         call the PluginTest NPP wrappers.
2044
2045 2011-05-13  Sheriff Bot  <webkit.review.bot@gmail.com>
2046
2047         Unreviewed, rolling out r86447.
2048         http://trac.webkit.org/changeset/86447
2049         https://bugs.webkit.org/show_bug.cgi?id=60809
2050
2051         "Broke some uses of EventSender object on Qt" (Requested by
2052         mwenge on #webkit).
2053
2054         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2055         (WebCore::WebPage::WebPage):
2056         (WebCore::DumpRenderTree::DumpRenderTree):
2057         (WebCore::DumpRenderTree::initJSObjects):
2058         (WebCore::DumpRenderTree::createWindow):
2059         * DumpRenderTree/qt/DumpRenderTreeQt.h:
2060         (WebCore::DumpRenderTree::eventSender):
2061         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
2062         (webkit_test_plugin_new_instance):
2063         (webkit_test_plugin_set_window):
2064         (webkit_test_plugin_handle_event):
2065
2066 2011-05-13  Robert Hogan  <robert@webkit.org>
2067
2068         Reviewed by Benjamin Poulain.
2069
2070         [Qt] fix http/tests/plugins/plugin-document-has-focus.html
2071
2072         The support for this test added to the unix test plugin here
2073         may allow other platforms to pass it (nearly everyone
2074         skips it). On Qt it required a bit of trickiness with
2075         the page's EventSender object to get it working fully though,
2076         so only unskipping Qt here.
2077
2078         https://bugs.webkit.org/show_bug.cgi?id=60722
2079
2080         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2081         (WebCore::WebPage::WebPage):
2082         (WebCore::WebPage::eventSender):
2083         (WebCore::DumpRenderTree::DumpRenderTree):
2084         (WebCore::DumpRenderTree::initJSObjects):
2085         (WebCore::DumpRenderTree::createWindow):
2086         * DumpRenderTree/qt/DumpRenderTreeQt.h:
2087         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
2088         (webkit_test_plugin_new_instance):
2089         (webkit_test_plugin_set_window):
2090         (webkit_test_plugin_handle_event):
2091
2092 2011-05-13  Adam Roben  <aroben@apple.com>
2093
2094         Exclude the Leopard NRWT bot from the core builder set
2095
2096         Fixes <http://webkit.org/b/60392> Leopard Intel Release (NRWT) shouldn't be a core builder
2097
2098         Reviewed by Darin Adler.
2099
2100         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
2101         (BuildBot.__init__): Tighten up the Leopard regexp to only allow the non-NRWT Leopard bots
2102         in the core builder set.
2103
2104         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
2105         (BuildBotTest.test_builder_name_regexps): Updated to match the current set of builders and
2106         regexps.
2107
2108 2011-05-13  Adam Roben  <aroben@apple.com>
2109
2110         Group all the NRWT bots together
2111
2112         This also gets the new NRWT bots out of the middle of Apple's bots. Since these new NRWT
2113         bots are experimental, it seems best not to clutter up the non-experimental bots with
2114         experimental ones.
2115
2116         Fixes <http://webkit.org/b/60769> New "(NRWT)" bots are right in the middle of Apple's bots
2117         on build.webkit.org
2118
2119         Reviewed by Dan Bernstein.
2120
2121         * BuildSlaveSupport/build.webkit.org-config/config.json: Moved the "(NRWT)" bots next to the
2122         Chromium testers, which also use new-run-webkit-tests.
2123
2124 2011-05-13  Adam Roben  <aroben@apple.com>
2125
2126         Add a plugin test that dumps the plugin window's rect
2127
2128         Part of the test for <http://webkit.org/b/60194> <rdar://problem/9383760> REGRESSION
2129         (r79167): Windowed plugins in Google Reader don't move when the article list is scrolled
2130
2131         Reviewed by Dan Bernstein.
2132
2133         * DumpRenderTree/TestNetscapePlugIn/Tests/win/DumpWindowRect.cpp: Added.
2134         (DumpWindowRect::DumpWindowRect): Just call up to the base class.
2135         (DumpWindowRect::performWindowGeometryTest): Find our window rect relative to the test
2136         harness window and log it.
2137
2138         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Added DumpWindowRect.
2139
2140 2011-05-13  Adam Roben  <aroben@apple.com>
2141
2142         Move code required for most tests of plugin window geometry into a base class
2143
2144         WebKit2's asynchronous plugin window positioning makes testing window geometry tricky. By
2145         having all the trickiness in a shared base class, writing new window geometry tests will be
2146         much easier.
2147
2148         Fixes <http://webkit.org/b/60776> It's hard to write tests that check plugin window geometry
2149
2150         Reviewed by Anders Carlsson.
2151
2152         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowRegionIsSetToClipRect.cpp:
2153         (WindowRegionIsSetToClipRect::WindowRegionIsSetToClipRect): Changed to inherit from the new
2154         WindowGeometryTest base class. Most of the code from this class was moved to the new base
2155         class.
2156         (WindowRegionIsSetToClipRect::performWindowGeometryTest): Renamed from checkWindowRegion.
2157         This is called by the base class when it is safe to check the window's geometry. Note that
2158         the m_didCheckWindowRegion logic was removed completely; it has never been needed and was
2159         just vestigial code from an earlier version of this test.
2160
2161         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Added new
2162         WindowGeometryTest class.
2163
2164         * DumpRenderTree/TestNetscapePlugIn/win/WindowGeometryTest.cpp: Added. Code came from
2165         WindowRegionIsSetToClipRect.
2166         (WindowGeometryTest::finishTest): This is the only change from WindowRegionIsSetToClipRect.
2167         We now call the new virtual performWindowGeometryTest function instead of checkWindowRegion.
2168         Subclasses can implement that function to perform their own tests.
2169
2170         * DumpRenderTree/TestNetscapePlugIn/win/WindowGeometryTest.h: Added.
2171
2172 2011-05-13  Zoltan Herczeg  <zherczeg@webkit.org>
2173
2174         Reviewed by Csaba Osztrogonác.
2175
2176         Missing 2nd #EOF if WebProcess is killed.
2177         https://bugs.webkit.org/show_bug.cgi?id=60761
2178
2179         If the WebProcess is killed, only one #EOF is yielded
2180         by WebKitTestRunner. However, run-webkit-test wait for
2181         a second #EOF, which signals the finishing of pixel dump.
2182         Since this never happens, run-webkit-test waits forever.
2183
2184         * WebKitTestRunner/TestInvocation.cpp:
2185         (WTR::TestInvocation::dump):
2186         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
2187         * WebKitTestRunner/TestInvocation.h:
2188
2189 2011-05-13  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
2190
2191         Reviewed by Csaba Osztrogonác.
2192
2193         [NRWT] chromium-linux tests shouldn't run on other platforms
2194         https://bugs.webkit.org/show_bug.cgi?id=60162
2195
2196         Remove _tests_for_other_platforms defs because r82753
2197         implemented it properly in the baseclass in webkit.py.
2198
2199         * Scripts/webkitpy/layout_tests/port/gtk.py:
2200         * Scripts/webkitpy/layout_tests/port/qt.py:
2201         * Scripts/webkitpy/layout_tests/port/win.py:
2202
2203 2011-05-12  Ryuan Choi  <ryuan.choi@samsung.com>
2204
2205         Reviewed by Antonio Gomes.
2206
2207         [EFL] Remove unused option of EWeblauncher
2208         https://bugs.webkit.org/show_bug.cgi?id=60747
2209
2210         Remove sudo-workaround option.
2211         This option is not used and make other option wrong.
2212
2213         * EWebLauncher/main.c:
2214
2215 2011-05-12  Lucas Forschler  <lforschler@apple.com>
2216
2217         Reviewed by Stephanie Lewis.
2218
2219         Fix config.json to address missing WK2 builder/tester config.
2220         Fix naming of machines (xserve not macpro)
2221
2222         * BuildSlaveSupport/build.webkit.org-config/config.json:
2223
2224 2011-05-12  Lucas Forschler  <lforschler@apple.com>
2225
2226         Reviewed by Stephanie Lewis.
2227
2228         Add SnowLeopard Debug builder and Test bot.
2229         apple-xserve-8 = SL Debug Tester.
2230         apple-xserve-9 = SL Debug Builder.
2231         
2232         * BuildSlaveSupport/build.webkit.org-config/config.json:
2233
2234 2011-05-12  Adam Barth  <abarth@webkit.org>
2235
2236         This patch disables sheriffbot's bug postings.  He's just been way too
2237         spammy recently due to test flakiness.
2238
2239         * Scripts/webkitpy/tool/commands/sheriffbot.py:
2240         * Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
2241
2242 2011-05-11  Brian Weinstein  <bweinstein@apple.com>
2243
2244         Reviewed by Anders Carlsson.
2245
2246         After relaunching an unresponsive web process, web page stops responding to mouse movement
2247         https://bugs.webkit.org/show_bug.cgi?id=60669
2248         <rdar://problem/9409934>
2249         
2250         This test follows the steps needed to reproduce this bug:
2251         1) Hang the web process
2252         2) While the web process is hung, move the mouse (so WebPageProxy thinks it is processing a mouse move event)
2253         3) Kill and restart the WebProcess
2254         4) Move the mouse
2255         
2256         We are testing that we receive mouse move events after the WebProcess has restarted.
2257
2258         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash.cpp: Added.
2259         (TestWebKitAPI::didFinishLoadForFrame): Set a flag that we finished loading.
2260         (TestWebKitAPI::processDidBecomeUnresponsive): Kill the web process once we become unresponsive.
2261         (TestWebKitAPI::processDidCrash): Reload the page when the process crashes, which restarts the web process.
2262         (TestWebKitAPI::setPageLoaderClient):
2263         (TestWebKitAPI::TEST):
2264         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash_Bundle.cpp: Added.
2265         (TestWebKitAPI::MouseMoveAfterCrashTest::MouseMoveAfterCrashTest):
2266         (TestWebKitAPI::MouseMoveAfterCrashTest::didReceiveMessage):
2267         * TestWebKitAPI/Tests/WebKit2/mouse-move-listener.html: Added.
2268
2269         * TestWebKitAPI/PlatformWebView.h:
2270         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
2271         (TestWebKitAPI::PlatformWebView::simulateMouseMove): Sends the messages that simulate a mouse move.
2272         * TestWebKitAPI/win/PlatformWebViewWin.cpp:
2273         (TestWebKitAPI::PlatformWebView::simulateMouseMove): Ditto.
2274
2275         Add new files.
2276         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2277         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
2278         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
2279         * TestWebKitAPI/win/copy-resources.cmd:
2280
2281 2011-05-12  Tomasz Morawski  <t.morawski@samsung.com>
2282
2283         Reviewed by Antonio Gomes.
2284
2285         [EFL] Added initialization/shutdown for ecore_file library
2286         https://bugs.webkit.org/show_bug.cgi?id=60457
2287
2288         Added initialization and shuthown for ecore_file library and replaced
2289         realpath function to ecore_file_realpath.
2290
2291         * EWebLauncher/main.c:
2292         (quit):
2293         (findThemePath):
2294         (main):
2295
2296 2011-05-12  Young Han Lee  <joybro@company100.net>
2297
2298         Reviewed by Kenneth Rohde Christiansen.
2299
2300         [Qt] Implement layoutTestController.layerTreeAsText()
2301         https://bugs.webkit.org/show_bug.cgi?id=60367
2302
2303         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2304         (LayoutTestController::layerTreeAsText):
2305         * DumpRenderTree/qt/LayoutTestControllerQt.h:
2306
2307 2011-05-11  Eric Seidel  <eric@webkit.org>
2308
2309         Reviewed by David Levin.
2310
2311         Teach validate-committer-lists how to validate committers bugzilla_email()
2312         https://bugs.webkit.org/show_bug.cgi?id=60651
2313
2314         It turns out we currently have 9 committers with invalid values for their
2315         bugzilla_email field (meaning we don't know what email is their bugzilla account).
2316         "Alexander Kellett" <lypanov@mac.com>
2317         "Andre Boule" <aboule@apple.com>
2318         "Cameron McCormack" <cam@webkit.org>
2319         "Chris Petersen" <cpetersen@apple.com>
2320         "Eli Fidler" <eli@staikos.net>
2321         "Matt Lilek" <webkit@mattlilek.com>
2322         "Pierre d'Herbemont" <pdherbemont@free.fr>
2323         "Scott Violet" <sky@chromium.org>
2324         "Adam Treat" <treat@kde.org>
2325
2326         Most of those haven't committed in a very long time, so it's not suprising
2327         that we haven't noticed that their bugzilla emails are bogus.
2328
2329         * Scripts/validate-committer-lists:
2330         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
2331         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
2332
2333 2011-05-11  Dmitry Lomov  <dslomov@google.com>
2334
2335         Reviewed by David Levin.
2336
2337         https://bugs.webkit.org/show_bug.cgi?id=59561
2338         Switch TestWebKitAPI to GTest
2339
2340         Fixes Makefiles and WinCairo build as well
2341
2342         * Makefile:
2343         * Scripts/build-api-tests:
2344         * Scripts/run-api-tests:
2345         * TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
2346         * TestWebKitAPI/Test.h:
2347         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2348         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
2349         (TestWebKitAPI::TEST):
2350         * TestWebKitAPI/TestsController.cpp:
2351         (TestWebKitAPI::TestsController::TestsController):
2352         (TestWebKitAPI::TestsController::dumpTestNames):
2353         (TestWebKitAPI::TestsController::runTestNamed):
2354         (TestWebKitAPI::TestsController::runAllTests):
2355         * TestWebKitAPI/TestsController.h:
2356         * TestWebKitAPI/mac/main.mm:
2357         (main):
2358
2359 2011-05-11  Eric Seidel  <eric@webkit.org>
2360
2361         Reviewed by Alexey Proskuryakov.
2362
2363         Add Luke Macpherson to committers.py as a Contributor
2364         https://bugs.webkit.org/show_bug.cgi?id=60639
2365
2366         * Scripts/webkitpy/common/config/committers.py:
2367
2368 2011-05-11  Noam Rosenthal  <noam.rosenthal@nokia.com>
2369
2370         Reviewed by Kenneth Rohde Christiansen.
2371
2372         [Texmap][Qt] Upstream texture-mapper changes from Qt's WebKit2 branch
2373         https://bugs.webkit.org/show_bug.cgi?id=60439
2374
2375         Patch 4/12: Update the build script to enable 3D rendering by default for Qt.
2376
2377         * Scripts/build-webkit:
2378
2379 2011-05-11  Brian Weinstein  <bweinstein@apple.com>
2380
2381         Reviewed by Adam Roben.
2382
2383         REGRESSION (r86161): WKPreferences API test failing on Windows
2384         https://bugs.webkit.org/show_bug.cgi?id=60658
2385
2386         * TestWebKitAPI/Tests/WebKit2/WKPreferences.cpp:
2387         (TestWebKitAPI::TEST):
2388
2389 2011-05-11  Alice Boxhall  <aboxhall@chromium.org>
2390
2391         Reviewed by Ojan Vafai.
2392
2393         Modify jsonresults_unittest.py to use a dict format for its test data, and modify jsonresults.py to flatten hierarchical directory structures in input JSON.
2394         https://bugs.webkit.org/show_bug.cgi?id=60521
2395
2396         First steps towards making the results JSON all hierarchical.
2397
2398         * TestResultServer/model/jsonresults.py:
2399         * TestResultServer/model/jsonresults_unittest.py:
2400
2401 2011-05-11  Kevin Ollivier  <kevino@theolliviers.com>
2402
2403         [wx] Build fix, adding new LayoutTestController method stubs.
2404
2405         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
2406         (LayoutTestController::localStorageDiskUsageForOrigin):
2407         (LayoutTestController::applicationCacheDiskUsageForOrigin):
2408
2409 2011-05-11  Jessie Berlin  <jberlin@apple.com>
2410
2411         Reviewed by Darin Adler.
2412
2413         [WebKit2] exceededDatabaseQuota function on WKPageUIClient needs to provide information
2414         about the usage for both the origin overall and the database in question.
2415         https://bugs.webkit.org/show_bug.cgi?id=60573
2416
2417         * WebKitTestRunner/TestController.cpp:
2418         (WTR::exceededDatabaseQuota):
2419
2420 2011-05-10  Stephen White  <senorblanco@chromium.org>
2421
2422         Reviewed by David Levin.
2423
2424         Adding myself as a reviewer.
2425         https://bugs.webkit.org/show_bug.cgi?id=60587
2426
2427         * Scripts/webkitpy/common/config/committers.py:
2428
2429 2011-05-11  Eric Seidel  <eric@webkit.org>
2430
2431         Reviewed by Csaba Osztrogonác.
2432
2433         queues.webkit.org should show 1000+ instead of 1000 when we hit the query limit
2434         https://bugs.webkit.org/show_bug.cgi?id=60614
2435
2436         AppEngine has a result limit of 1000.  We hit that when trying
2437         to find out how many patches the style-queue (and other queues)
2438         have passed in the last 30 days.  So lets display 1000+ instead
2439         of 1000 to avoid confusion. (Ossy was asking about this in #webkit tonight.)
2440
2441         I also made the status pages show "never" instead of just "ago" when we
2442         haven't seen a pass or boot message in the last 1000 messages.
2443
2444         * QueueStatusServer/handlers/queuestatus.py:
2445         * QueueStatusServer/templates/queuestatus.html:
2446
2447 2011-05-11  Adam Barth  <abarth@webkit.org>
2448
2449         Reviewed by Eric Seidel.
2450
2451         sheriffbot shouldn't spam bugs when it can't figure out who to blame
2452         https://bugs.webkit.org/show_bug.cgi?id=60613
2453
2454         This patch limits the number of bugs sheriffbot will spam to three.
2455         Ideally, we'd come up with some smarter huerstics, but hopefully this
2456         will cut down on the bugmail spam for slow bots.
2457
2458         * Scripts/webkitpy/tool/commands/sheriffbot.py:
2459
2460 2011-05-10  Eric Seidel  <eric@webkit.org>
2461
2462         Reviewed by Adam Barth.
2463
2464         EWS bots should be robust against test-webkitpy hangs
2465         https://bugs.webkit.org/show_bug.cgi?id=60579
2466
2467         Added a killAfterTimeout subroutine, which takes a child_pid
2468         and timeout, and will setup a watchdog process to kill that child
2469         after the timeout.  killAfterTimeout will waitpid on the child.
2470
2471         * EWSTools/start-queue.sh:
2472
2473 2011-05-10  MORITA Hajime  <morrita@google.com>
2474
2475         Unreviewed, moved myself to the reviewers list.
2476
2477         * Scripts/webkitpy/common/config/committers.py:
2478
2479 2011-05-10  Anton D'Auria  <adauria@apple.com>
2480
2481         Reviewed by David Levin.
2482
2483         StorageTracker should report actual local storage usage on disk
2484         https://bugs.webkit.org/show_bug.cgi?id=60558
2485
2486         * DumpRenderTree/LayoutTestController.cpp:
2487         (localStorageDiskUsageForOriginCallback):
2488         (LayoutTestController::staticFunctions):
2489         * DumpRenderTree/LayoutTestController.h:
2490         * DumpRenderTree/chromium/LayoutTestController.cpp:
2491         (LayoutTestController::LayoutTestController):
2492         (LayoutTestController::localStorageDiskUsageForOrigin):
2493         * DumpRenderTree/chromium/LayoutTestController.h:
2494         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2495         (LayoutTestController::localStorageDiskUsageForOrigin):
2496         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
2497         (LayoutTestController::localStorageDiskUsageForOrigin):
2498         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2499         (LayoutTestController::localStorageDiskUsageForOrigin):
2500         * DumpRenderTree/qt/LayoutTestControllerQt.h:
2501         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
2502         (LayoutTestController::localStorageDiskUsageForOrigin):
2503
2504 2011-05-10  Anton D'Auria  <adauria@apple.com>
2505
2506         Reviewed by Alexey Proskuryakov.
2507
2508         Add ApplicationCache API to get disk usage per origin
2509         https://bugs.webkit.org/show_bug.cgi?id=57127
2510
2511         Added test that loads two frames with ~15Kb image, checks usage,
2512         deletes app cache, verifies usage is 0.
2513
2514         * DumpRenderTree/LayoutTestController.cpp:
2515         (applicationCacheDiskUsageForOriginCallback):
2516         (LayoutTestController::staticFunctions):
2517         * DumpRenderTree/LayoutTestController.h:
2518         * DumpRenderTree/chromium/LayoutTestController.cpp: Added stub.
2519         (LayoutTestController::LayoutTestController):
2520         (LayoutTestController::applicationCacheDiskUsageForOrigin):
2521         * DumpRenderTree/chromium/LayoutTestController.h:
2522         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: Added stub.
2523         (LayoutTestController::applicationCacheDiskUsageForOrigin):
2524         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
2525         (LayoutTestController::applicationCacheDiskUsageForOrigin):
2526         * DumpRenderTree/qt/LayoutTestControllerQt.cpp: Added stub.
2527         (LayoutTestController::applicationCacheDiskUsageForOrigin):
2528         * DumpRenderTree/qt/LayoutTestControllerQt.h:
2529         * DumpRenderTree/win/LayoutTestControllerWin.cpp: Added stub.
2530         (LayoutTestController::applicationCacheDiskUsageForOrigin):
2531
2532 2011-05-10  Lucas De Marchi  <lucas.demarchi@profusion.mobi>
2533
2534         Unreviewed, adding Rafael Antognolli as a committer.
2535
2536         * Scripts/webkitpy/common/config/committers.py:
2537
2538 2011-05-10  Chris Evans  <cevans@chromium.org>
2539
2540         Reviewed by Adam Barth.
2541
2542         Add WebCore::Setting to block displaying and/or running insecure content on secure pages
2543         https://bugs.webkit.org/show_bug.cgi?id=58378
2544
2545         * DumpRenderTree/chromium/LayoutTestController.cpp:
2546         (LayoutTestController::LayoutTestController):
2547         (LayoutTestController::setAllowDisplayOfInsecureContent):
2548         (LayoutTestController::setAllowRunningOfInsecureContent):
2549         * DumpRenderTree/chromium/LayoutTestController.h:
2550         * DumpRenderTree/chromium/WebPreferences.cpp:
2551         (WebPreferences::reset):
2552         (WebPreferences::applyTo):
2553         * DumpRenderTree/chromium/WebPreferences.h:
2554         Add plumbing for new settings to Chromium DRT.
2555
2556 2011-05-10  Enrica Casucci  <enrica@apple.com>
2557
2558         Reviewed by Beth Dakin.
2559
2560         Added my name to the reviewers list and removed my name from
2561         to committer-only list.
2562
2563         * Scripts/webkitpy/common/config/committers.py:
2564
2565 2011-05-10  David Levin  <levin@chromium.org>
2566
2567         Reviewed by Shinichiro Hamaji.
2568
2569         check-webkit-style shouldn't filter the errors using the modified lines for the xml and test expectations checkers.
2570         https://bugs.webkit.org/show_bug.cgi?id=60466
2571
2572         * Scripts/webkitpy/style/checker.py: (unrelated) style fix.
2573         * Scripts/webkitpy/style/checker_unittest.py: Remove the file_path check since
2574           that property was removed.
2575         * Scripts/webkitpy/style/checkers/test_expectations.py: Turn off the line
2576           filtering for errors.
2577         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Add testing to
2578           verify that the line filtering is turned off.
2579         * Scripts/webkitpy/style/checkers/xml.py: Turn off the line filtering for errors
2580           and remove the unused file_path parameter.
2581         * Scripts/webkitpy/style/checkers/xml_unittest.py: Add testing to verify that the
2582           line filtering is turned off (and fix test_no_error).
2583         * Scripts/webkitpy/style/error_handlers.py: Added a way to turn off the filtering
2584           of errors based on the lines changed in the patch.
2585         * Scripts/webkitpy/style/error_handlers_unittest.py: Added a test for turning off
2586           the filtering.
2587
2588 2011-05-10  Adam Roben  <aroben@apple.com>
2589
2590         Update some plugin tests to work with WebKit2's asynchronous windowed plugin positioning on Windows
2591
2592         Fixes <http://webkit.org/b/60442> <rdar://problem/9403200> REGRESSION (r85795):
2593         platform/win/plugins/window-geometry-initialized-before-set-window.html and
2594         platform/win/plugins/window-region-is-set-to-clip-rect.html failing on Windows 7 Release
2595         (WebKit2 Tests)
2596
2597         Reviewed by Anders Carlsson.
2598
2599         * DumpRenderTree/TestNetscapePlugIn/Tests/win/NPNInvalidateRectInvalidatesWindow.cpp:
2600         (NPNInvalidateRectInvalidatesWindow::NPP_SetWindow): Moved some code from here to the new
2601         WindowedPluginTest::testHarnessWindow function.
2602
2603         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowGeometryInitializedBeforeSetWindow.cpp:
2604         (WindowGeometryInitializedBeforeSetWindow::NPP_SetWindow): Only check the window's size, not
2605         its position. In WebKit2, the position won't have been set yet, so we can't test it. But it
2606         turns out only the size mattered for <http://webkit.org/b/46716>, which this test was
2607         created for.
2608
2609         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowRegionIsSetToClipRect.cpp:
2610         (WindowRegionIsSetToClipRect::WindowRegionIsSetToClipRect): Converted to be a
2611         WindowedPluginTest, and initialized new members.
2612         (WindowRegionIsSetToClipRect::startTest): Make our plugin window visible, and set things up
2613         to ensure we get a WM_WINDOWPOSCHANGED message about it becoming visible.
2614         (WindowRegionIsSetToClipRect::finishTest): Check that the window region is what we expect,
2615         then clean up and tell the test harness we're done.
2616         (WindowRegionIsSetToClipRect::checkWindowRegion): Moved code to check the window region here
2617         from NPP_SetWindow. Changed to use the fancy new log() function instead of pluginLog().
2618         (WindowRegionIsSetToClipRect::showTestHarnessWindowIfNeeded): Show the test harness's window
2619         if it is hidden. This is needed to ensure we get a WM_WINDOWPOSCHANGED message when our own
2620         window becomes visible.
2621         (WindowRegionIsSetToClipRect::hideTestHarnessWindowIfNeeded): Hide the test harness's window
2622         if it was hidden at the start of the test.
2623         (WindowRegionIsSetToClipRect::wndProc): When our timer fires, tell the UI process to paint.
2624         This should cause our window to become visible and our window region to get set. When we get
2625         a WM_WINDOWPOSCHANGED message, we can finally finish the test.
2626         (WindowRegionIsSetToClipRect::NPP_GetValue): Return our script object.
2627         (WindowRegionIsSetToClipRect::ScriptObject::hasMethod): We only have one method:
2628         "startTest".
2629         (WindowRegionIsSetToClipRect::ScriptObject::invoke): Call through to the PluginTest object.
2630
2631         * DumpRenderTree/TestNetscapePlugIn/win/WindowedPluginTest.cpp:
2632         (WindowedPluginTest::testHarnessWindow):
2633         * DumpRenderTree/TestNetscapePlugIn/win/WindowedPluginTest.h:
2634         Added this helper function. Code came from NPNInvalidateRectInvalidatesWindow.
2635
2636 2011-05-10  Csaba Osztrogonác  <ossy@webkit.org>
2637
2638         Reviewed by Kenneth Rohde Christiansen.
2639
2640         [Qt] Add new qt-4.8 platform for layout testing
2641         https://bugs.webkit.org/show_bug.cgi?id=60544
2642
2643         * Scripts/old-run-webkit-tests:
2644         * Scripts/webkitdirs.pm:
2645
2646 2011-05-09  Sheriff Bot  <webkit.review.bot@gmail.com>
2647
2648         Unreviewed, rolling out r86108.
2649         http://trac.webkit.org/changeset/86108
2650         https://bugs.webkit.org/show_bug.cgi?id=60533
2651
2652         Broke makefile based builds and Cairo (Requested by weinig on
2653         #webkit).
2654
2655         * Scripts/build-api-tests:
2656         * Scripts/run-api-tests:
2657         * TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
2658         * TestWebKitAPI/Test.h:
2659         (TestWebKitAPI::Test::~Test):
2660         (TestWebKitAPI::Test::name):
2661         (TestWebKitAPI::Test::Register::Register):
2662         (TestWebKitAPI::Test::Register::create):
2663         (TestWebKitAPI::Test::Test):
2664         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2665         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
2666         (TestWebKitAPI::TEST):
2667         * TestWebKitAPI/TestsController.cpp:
2668         (TestWebKitAPI::TestsController::TestsController):
2669         (TestWebKitAPI::TestsController::dumpTestNames):
2670         (TestWebKitAPI::TestsController::runTestNamed):
2671         (TestWebKitAPI::TestsController::testFailed):
2672         (TestWebKitAPI::TestsController::registerCreateTestFunction):
2673         * TestWebKitAPI/TestsController.h:
2674         * TestWebKitAPI/mac/main.mm:
2675         (main):
2676
2677 2011-05-09  Hayato Ito  <hayato@chromium.org>
2678
2679         Reviewed by Ojan Vafai.
2680
2681         Skip reftests if pixel tests are disabled.
2682
2683         https://bugs.webkit.org/show_bug.cgi?id=59078
2684
2685         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
2686         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
2687         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
2688         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2689
2690 2011-05-09  Chang Shu  <cshu@webkit.org>
2691
2692         Reviewed by Kenneth Rohde Christiansen.
2693
2694         [Qt] DumpRenderTreeQt needs an implementation of unmarkText
2695         https://bugs.webkit.org/show_bug.cgi?id=60499
2696
2697         The implementation is similar to gtk.
2698
2699         * DumpRenderTree/qt/TextInputControllerQt.cpp:
2700         (TextInputController::unmarkText):
2701         * DumpRenderTree/qt/TextInputControllerQt.h:
2702
2703 2011-05-09  James Robinson  <jamesr@chromium.org>
2704
2705         Reviewed by Eric Seidel.
2706
2707         [chromium] Teach new-run-webkit-tests to follow symlinks when trying to determine the arch of DumpRenderTree on linux
2708         https://bugs.webkit.org/show_bug.cgi?id=60520
2709
2710         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2711
2712 2011-05-09  Dmitry Lomov  <dslomov@google.com>
2713
2714         Reviewed by Sam Weinig.
2715
2716         Switch TestWebKitAPI to GTest
2717         https://bugs.webkit.org/show_bug.cgi?id=59561
2718
2719         * Scripts/build-api-tests:
2720         * Scripts/run-api-tests:
2721         * TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
2722         * TestWebKitAPI/Test.h:
2723         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2724         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
2725         (TestWebKitAPI::TEST):
2726         * TestWebKitAPI/TestsController.cpp:
2727         (TestWebKitAPI::TestsController::TestsController):
2728         (TestWebKitAPI::TestsController::dumpTestNames):
2729         (TestWebKitAPI::TestsController::runTestNamed):
2730         (TestWebKitAPI::TestsController::runAllTests):
2731         * TestWebKitAPI/TestsController.h:
2732         * TestWebKitAPI/mac/main.mm:
2733         (main):
2734
2735 2011-05-09  Igor Oliveira  <igor.oliveira@openbossa.org>
2736
2737         Reviewed by Andreas Kling.
2738
2739         [Qt] DRT eventSender.addTouchPoint method is not setting the start position
2740         https://bugs.webkit.org/show_bug.cgi?id=60487
2741
2742         When adding a new touch point, the start position of this point needs to be set.
2743
2744         * DumpRenderTree/qt/EventSenderQt.cpp:
2745         (EventSender::addTouchPoint):
2746
2747 2011-05-09  Robert Hogan  <robert@webkit.org>
2748
2749         Reviewed by Adam Roben.
2750
2751         https://bugs.webkit.org/show_bug.cgi?id=57137
2752
2753         Unskip plugins/embed-prefers-plugins-for-images.html
2754
2755         Add image/png mime-type to test netscape plugin on Unix.
2756
2757         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
2758         (NP_GetMIMEDescription):
2759
2760 2011-05-09  Alice Boxhall  <aboxhall@chromium.org>
2761
2762         Reviewed by Ojan Vafai.
2763
2764         Pull jsonresults_unittest test cases out into separate test methods.
2765         https://bugs.webkit.org/show_bug.cgi?id=60418
2766
2767         * TestResultServer/model/jsonresults_unittest.py:
2768
2769 2011-05-09  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
2770
2771         Reviewed by Andreas Kling.
2772
2773         [Qt] fix fast/dom/Geolocation/window-close-crash.html
2774         https://bugs.webkit.org/show_bug.cgi?id=60458
2775
2776         Update all pages with Mock position/error/permission.
2777
2778         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2779         (LayoutTestController::setGeolocationPermission):
2780         (LayoutTestController::setMockGeolocationError):
2781         (LayoutTestController::setMockGeolocationPosition):
2782
2783 2011-05-09  Shishir Agrawal  <shishir@chromium.org>
2784
2785         Reviewed by Tony Gentilcore.
2786
2787         Implement Page Visibility API.
2788         https://bugs.webkit.org/show_bug.cgi?id=54181
2789
2790         * DumpRenderTree/LayoutTestController.cpp:
2791         (setPageVisibilityCallback):
2792         (resetPageVisibilityCallback):
2793         (LayoutTestController::staticFunctions):
2794         * DumpRenderTree/LayoutTestController.h:
2795         (LayoutTestController::setPageVisibility):
2796         (LayoutTestController::resetPageVisibility):
2797         * DumpRenderTree/chromium/LayoutTestController.cpp:
2798         (LayoutTestController::LayoutTestController):
2799         (LayoutTestController::resetPageVisibility):
2800         (LayoutTestController::setPageVisibility):
2801         * DumpRenderTree/chromium/LayoutTestController.h:
2802
2803 2011-05-07  Adam Barth  <abarth@webkit.org>
2804
2805         Reviewed by Eric Seidel.
2806
2807         Testing EWS spins on patches with a large number of failures
2808         https://bugs.webkit.org/show_bug.cgi?id=60441
2809
2810         In cases where the tree has a small number of persistent failures but a
2811         patch has a large number of failures, we weren't believing any results.
2812         That lead to us retrying the patch forever while we waited for
2813         trustworthy results that would never come.
2814
2815         This patch loosens the semantics of unexpected_failures to return a
2816         subset (rather than the exact set) of unexpected failures.  In the case
2817         where the tree has a bounded number of failures and the patch has an
2818         unbounded number of failures, we can accurately compute such a subset
2819         and reject the patch.
2820
2821         * Scripts/webkitpy/tool/bot/expectedfailures.py:
2822         * Scripts/webkitpy/tool/bot/expectedfailures_unittest.py:
2823         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
2824         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
2825         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
2826
2827 2011-05-07  Joe Mason  <jmason@rim.com>
2828
2829         Reviewed by Daniel Bates.
2830
2831         FrameLoader::isProcessingUserGesture is wrong in dispatchWillPerformClientRedirect
2832         https://bugs.webkit.org/show_bug.cgi?id=52211
2833
2834         Dump isUserProcessingGesture in willPerformClientRedirect so that the layout tests can be extended.
2835         (For the Mac and Chromium ports - other ports don't support dumping user gestures in DRT.)
2836
2837         * DumpRenderTree/chromium/WebViewHost.cpp:
2838         (WebViewHost::willPerformClientRedirect):
2839         (WebViewHost::didStartProvisionalLoad):
2840         * DumpRenderTree/mac/FrameLoadDelegate.mm:
2841         (-[FrameLoadDelegate webView:willPerformClientRedirectToURL:delay:fireDate:forFrame:]):
2842
2843 2011-05-07  Daniel Bates  <dbates@webkit.org>
2844
2845         Attempt to fix the build; remove some more references to isTiger(), which was removed in changeset 86007
2846         <http://trac.webkit.org/changeset/86007> (bug #60438).
2847
2848         * Scripts/old-run-webkit-tests:
2849           - Removed reference to isTiger() from openDumpTool(), captureSavedCrashLog(), and
2850             file local variable. 
2851
2852 2011-05-07  Pratik Solanki  <psolanki@apple.com>
2853
2854         Reviewed by Darin Adler.
2855
2856         Remove references to Tiger in scripts
2857         https://bugs.webkit.org/show_bug.cgi?id=60438
2858
2859         Remove isTiger() and all of the code that depends on it.
2860
2861         * Scripts/build-webkit:
2862         * Scripts/ensure-valid-python:
2863         * Scripts/gdb-safari:
2864         * Scripts/webkitdirs.pm:
2865
2866 2011-05-06  David Grogan  <dgrogan@chromium.org>
2867
2868         Reviewed by Eric Seidel.
2869
2870         update link to message about mac 64bit git binaries
2871         https://bugs.webkit.org/show_bug.cgi?id=60411
2872
2873         The original message implied (correctly, at-the-time) that
2874         http://code.google.com/p/git-osx-installer/ doesn't provide 64-bit
2875         binaries. The new message notes that they do. This might prevent someone
2876         else from spending a few minutes confused.
2877
2878         * Scripts/webkitpy/common/checkout/scm/git.py:
2879
2880 2011-05-06  Dirk Pranke  <dpranke@chromium.org>
2881
2882         Reviewed by Ojan Vafai.
2883
2884         new-run-webkit-tests: crashes in race with crashing DRT
2885         https://bugs.webkit.org/show_bug.cgi?id=60238
2886
2887         This fixes a race where if we wrote the URL to DRT and
2888         DRT crashed immediately, we could get a broken pipe
2889         error back during the write() and not handle it correctly,
2890         causing NRWT to also crash.
2891
2892         * Scripts/webkitpy/layout_tests/port/server_process.py:
2893         * Scripts/webkitpy/layout_tests/port/webkit.py:
2894
2895 2011-05-06  Adam Barth  <abarth@webkit.org>
2896
2897         Reviewed by Eric Seidel.
2898
2899         Add a --skip-integrationtests mode to test-webkitpy
2900         https://bugs.webkit.org/show_bug.cgi?id=60354
2901
2902         These two test suites account for 2/3 of test-webkitpy's running time
2903         because they're not actually unit tests.  They're integration tests.
2904         This patch adds the ability to skip integration tests by passing a
2905         command line argument to test-webkitpy.  The buildbots will continue to
2906         run these integration tests to make sure they don't regress.  This
2907         option is for use by developers when working on parts of the code that
2908         are well-covered by unit tests.
2909
2910         * Scripts/webkitpy/common/system/file_lock_integrationtest.py: Copied from Tools/Scripts/webkitpy/common/system/file_lock_unittest.py.
2911         * Scripts/webkitpy/common/system/file_lock_unittest.py: Removed.
2912         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py: Copied from Tools/Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py.
2913         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py: Removed.
2914         * Scripts/webkitpy/test/main.py:
2915
2916 2011-05-06  Adam Barth  <abarth@webkit.org>
2917
2918         Reviewed by Eric Seidel.
2919
2920         The results.zip generated by the Chromium EWS are too big
2921         https://bugs.webkit.org/show_bug.cgi?id=60352
2922
2923         Currently results.zip created by the Chromium EWS bot contain roughly
2924         1000 tests that we expect to fail.  These test results are not helpful
2925         to contributors.
2926
2927         There are several approaches to solving this problem:
2928
2929         1) We could be more selective about which files to include in the zip.
2930         2) We could ask new-run-webkit-tests not to generate results for
2931            expected failures.
2932         3) We could ask new-run-webkit-tests not to run tests that we expect to
2933            fail.
2934
2935         This patch takes approach (3).  Currently, we're ignoring the outcome
2936         of failing tests anyway, so there's not much point in actually running
2937         them.  As a side benefit, this will speed up the test run by about 6%.
2938         In the future, we might revisit this decision and choose another
2939         approach.
2940
2941         * Scripts/webkitpy/common/config/ports.py:
2942         * Scripts/webkitpy/common/config/ports_unittest.py:
2943
2944 2011-05-06  Gabor Rapcsanyi  <rgabor@webkit.org>
2945
2946         Reviewed by Csaba Osztrogonác.
2947
2948         [Qt] Use qt-arm platform instead of qt-linux on ARM platform.
2949
2950         * Scripts/old-run-webkit-tests:
2951
2952 2011-05-06  Adam Barth  <abarth@webkit.org>
2953
2954         Reviewed by Ojan Vafai.
2955
2956         Add an option to NRWT to skip failing tests
2957         https://bugs.webkit.org/show_bug.cgi?id=60295
2958
2959         This option is similar to the way old-run-webkit-tests works.  We're
2960         planning to use this option on the testing EWS bots because they don't
2961         use information from failing tests yet.
2962
2963         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
2964         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2965         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2966
2967 2011-05-04  Philippe Normand  <pnormand@igalia.com>
2968
2969         Reviewed by Dimitri Glazkov.
2970
2971         Implement LayoutTestController::pseudoShadowId()
2972         https://bugs.webkit.org/show_bug.cgi?id=60034
2973
2974         shadowPseudoId support in the various LayoutTestController
2975         implementations.
2976
2977         * DumpRenderTree/LayoutTestController.cpp:
2978         (shadowPseudoIdCallback):
2979         (LayoutTestController::staticFunctions):
2980         * DumpRenderTree/LayoutTestController.h:
2981         * DumpRenderTree/chromium/LayoutTestController.cpp:
2982         (LayoutTestController::LayoutTestController):
2983         (LayoutTestController::shadowPseudoId):
2984         * DumpRenderTree/chromium/LayoutTestController.h:
2985         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
2986         (LayoutTestController::shadowPseudoId):
2987         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
2988         (LayoutTestController::shadowPseudoId):
2989         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
2990         (LayoutTestController::shadowPseudoId):
2991         * DumpRenderTree/qt/LayoutTestControllerQt.h:
2992         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
2993         (LayoutTestController::shadowPseudoId):
2994         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
2995         (LayoutTestController::shadowPseudoId):
2996         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
2997         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
2998         (WTR::LayoutTestController::shadowPseudoId):
2999         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
3000
3001 2011-05-06  Adam Barth  <abarth@webkit.org>
3002
3003         EWS can't reject patches
3004         https://bugs.webkit.org/show_bug.cgi?id=60345
3005
3006         We're supposed to use patch.id() for the attachment ID, silly.
3007
3008         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
3009         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
3010
3011 2011-05-05  Steve Lacey  <sjl@chromium.org>
3012
3013         Reviewed by Ojan Vafai.
3014
3015         Fix generation of results html file by passing relative path to scm
3016         methods.
3017         Generate diff images for results html file.
3018
3019         https://bugs.webkit.org/show_bug.cgi?id=60086
3020
3021         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
3022         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
3023
3024 2011-05-05  Andy Estes  <aestes@apple.com>
3025
3026         Reviewed by Daniel Bates.
3027
3028         update-webkit should handle Internal using a different VCS than OpenSource.
3029         https://bugs.webkit.org/show_bug.cgi?id=60310
3030
3031         isGit() and isSVN() cache their results, so Internal will always be
3032         updated using the VCS commands used for OpenSource. If Internal uses a
3033         different VCS, this will fail. Fix this by calling isGitDirectory() and
3034         isSVNDirectory() instead, which do not cache their results.
3035
3036         * Scripts/update-webkit:
3037
3038 2011-05-05  Dirk Pranke  <dpranke@chromium.org>
3039
3040         Reviewed by Ojan Vafai.
3041
3042         nrwt: revamp integration tests incl. tests for servers
3043         https://bugs.webkit.org/show_bug.cgi?id=59985
3044
3045         This change adds the ability to run "integration tests" that
3046         depend on (and affect) the local environment on your machine.
3047         These tests can be run manually by invoking python on the
3048         test file in question and passing the '-i' flag, but are
3049         not run automatically as part of test-webkitpy.
3050
3051         Integration tests are identified by being methods named as
3052         "integration_test_*" instead of "test_*", and we use a
3053         custom TestLoader instead of the default one provided by
3054         unittest.
3055
3056         This change also adds a protected "_default_to_apache" method
3057         to the port object so we can do the right thing when starting
3058         and stopping web servers with less mechanism.
3059
3060         * Scripts/webkitpy/layout_tests/port/base.py:
3061         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
3062         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
3063         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
3064         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3065         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
3066
3067 2011-05-05  Adam Roben  <aroben@apple.com>
3068
3069         Robustify DRT's pixel-dumping code against semi-random alpha channel values on Windows
3070
3071         DRT generates a pixel dump on Windows using the WM_PRINT message. This message (just like
3072         most of GDI) doesn't write anything meaningful into the alpha channel. In some cases, the
3073         alpha channel will even get filled with zeros (like when ::GradientFill is used). To
3074         safeguard against this, we fill the alpha channel with constant full opacity after the
3075         WM_PRINT message is handled to ensure we end up with an opaque image. (Note that we can't
3076         just ignore the alpha channel when creating the PNG; Mac's pixel results have an alpha
3077         channel, and we want to be able to compare Mac and Windows pixel results, so Windows pixel
3078         results need an alpha channel, too.)
3079
3080         Fixes <http://webkit.org/b/60300> draws-gradient.html's pixel results can't be correctly
3081         generated due to alpha channel weirdness
3082
3083         Reviewed by Sam Weinig.
3084
3085         * DumpRenderTree/win/PixelDumpSupportWin.cpp:
3086         (makeAlphaChannelOpaque): Added this helper function to fill a bitmap's alpha channel with
3087         constant full opacity.
3088         (createBitmapContextFromWebView): Make the bitmap's alpha channel opaque before wrapping it
3089         in a CGContext or cairo_surface.
3090
3091 2011-05-05  Adam Roben  <aroben@apple.com>
3092
3093         Add support for layoutTestController.dumpAsText(true) to WebKitTestRunner
3094
3095         Passing true tells the test harness to dump pixels even though we're not dumping a render
3096         tree.
3097
3098         Fixes <http://webkit.org/b/60298> Calling layoutTestController.dumpAsText(true) in
3099         WebKitTestRunner doesn't result in pixel results being generated
3100
3101         Reviewed by Sam Weinig.
3102
3103         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl: Added dumpPixels
3104         argument to dumpAsText.
3105
3106         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
3107         (WTR::InjectedBundlePage::dump): Updated for dumpAsText change.
3108
3109         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
3110         (WTR::LayoutTestController::dumpAsText): Store the dumpPixels argument.
3111
3112         * WebKitTestRunner/InjectedBundle/LayoutTestController.h: Updated for IDL change.
3113
3114         * WebKitTestRunner/win/InjectedBundle.vcproj: Added .idl files as a convenience for Windows
3115         developers.
3116
3117 2011-05-05  Dirk Pranke  <dpranke@chromium.org>
3118
3119         Reviewed by Ojan Vafai.
3120
3121         new-run-webkit-tests: merge os, version fields in test config
3122         https://bugs.webkit.org/show_bug.cgi?id=60004
3123
3124         This change removes the 'o/s' field from the TestConfiguration
3125         object, and changes the 'MAC', 'WIN', and 'LINUX' modifiers in
3126         the expectations file to macros that expand to the version names
3127         for each platform.
3128
3129         This means that you can once again put any subset of the
3130         operating system versions on a single line, e.g.:
3131           BUGX LEOPARD XP LINUX : fast/html/foo.html = FAIL
3132
3133         This change will break any existing test_expectations.txt
3134         file that uses the "OS-VERSION" macros, which are no longer
3135         needed. It turns out they were never needed, as just VERSION
3136         by itself was sufficient. The test_expectations.txt file should
3137         be updated accordingly prior to landing this change.
3138
3139         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
3140         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
3141         * Scripts/webkitpy/layout_tests/port/base.py:
3142
3143 2011-05-05  Tony Chang  <tony@chromium.org>
3144
3145         Reviewed by Adam Barth.
3146
3147         [chromium] switch from Lucida to free fonts for Hindi and Thai
3148         https://bugs.webkit.org/show_bug.cgi?id=60287
3149
3150         The fonts are provided by the ttf-thai-tlwg and ttf-indic-fonts
3151         packages on Lucid. There are actually multiple Thai fonts in the
3152         package, so I just picked one.
3153
3154         * DumpRenderTree/chromium/TestShellGtk.cpp:
3155         (setupFontconfig):
3156
3157 2011-05-05  Tony Chang  <tony@chromium.org>
3158
3159         Reviewed by Ojan Vafai.
3160
3161         [chromium] google-chrome-64 builds should now fallback to chromium-linux-x86_64 results
3162         https://bugs.webkit.org/show_bug.cgi?id=60290
3163
3164         It used to be that google-chrome-32 was hardy32, google-chrome-64 was
3165         hardy64, chromium-linux was hardy32 and chromium-linux-x86_64 was
3166         lucid64.  Because of that, google-chrome-64 used to skip the
3167         chromium-linux-x86_64 results since they were more incorrect.
3168
3169         Now that we're switching to lucid for google-chrome, we want
3170         google-chrome-64 to fall back to chromium-linux-x86_64.
3171
3172         * Scripts/webkitpy/layout_tests/port/google_chrome.py:
3173
3174 2011-05-05  Dirk Pranke  <dpranke@chromium.org>
3175
3176         Reviewed by Ojan Vafai.
3177
3178         new-run-webkit-tests: merge TestExpectations, TestExpectationsFile classes
3179         https://bugs.webkit.org/show_bug.cgi?id=60002
3180
3181         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3182         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
3183         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
3184         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
3185         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
3186         * Scripts/webkitpy/style/checkers/test_expectations.py:
3187
3188 2011-05-05  Eric Seidel  <eric@webkit.org>
3189
3190         Reviewed by Adam Barth.
3191
3192         commit-queue should reject patches which fail ewses
3193         https://bugs.webkit.org/show_bug.cgi?id=47534
3194
3195         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
3196         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
3197         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
3198
3199 2011-05-05  Chang Shu  <cshu@webkit.org>
3200
3201         Reviewed by Andreas Kling.
3202
3203         [Qt][WK2] Support PageGroup in QGraphicsWKView/QWKPage constructors
3204         https://bugs.webkit.org/show_bug.cgi?id=59638
3205
3206         1. Call QGraphicsWKView constructor with pageGroupRef.
3207         2. Disable setFontFamily calls as it breaks Qt test fonts.
3208
3209         * WebKitTestRunner/TestController.cpp:
3210         (WTR::TestController::resetStateToConsistentValues):
3211         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
3212         (WTR::WebView::WebView):
3213         (WTR::PlatformWebView::PlatformWebView):
3214
3215 2011-05-05  Joseph Pecoraro  <joepeck@webkit.org>
3216
3217         Reviewed by David Kilzer.
3218
3219         Add --only-webkit option to build-webkit
3220         https://bugs.webkit.org/show_bug.cgi?id=60245
3221
3222         * Scripts/build-webkit: If --only-webkit was specified
3223         just build the WebKit project, and not its dependencies.
3224
3225 2011-05-05  Jeff Miller  <jeffm@apple.com>
3226
3227         Reviewed by Darin Adler.
3228
3229         Add didNotHandleWheelEvent callback to WKPageUIClient
3230         https://bugs.webkit.org/show_bug.cgi?id=60203
3231
3232         * MiniBrowser/gtk/BrowserWindow.c:
3233         (browserWindowUIClientInit): Stub out didNotHandleWheelEvent callback.
3234         * MiniBrowser/mac/BrowserWindowController.m:
3235         (-[BrowserWindowController awakeFromNib]): Stub out didNotHandleWheelEvent callback.
3236         * MiniBrowser/win/BrowserView.cpp:
3237         (BrowserView::create): Stub out didNotHandleWheelEvent callback.
3238         * WebKitTestRunner/TestController.cpp:
3239         (WTR::TestController::createOtherPage): Stub out didNotHandleWheelEvent callback.
3240         (WTR::TestController::initialize): Stub out didNotHandleWheelEvent callback.
3241
3242 2011-05-04  Eric Seidel  <eric@webkit.org>
3243
3244         Reviewed by Adam Barth.
3245
3246         Add more missing imports from scm refactor
3247         https://bugs.webkit.org/show_bug.cgi?id=60239
3248
3249         * Scripts/webkitpy/common/checkout/scm/git.py:
3250         * Scripts/webkitpy/common/checkout/scm/svn.py:
3251
3252 2011-05-04  Eric Seidel  <eric@webkit.org>
3253
3254         Reviewed by Adam Barth.
3255
3256         sheriff-bot is having more exception trouble
3257         https://bugs.webkit.org/show_bug.cgi?id=60233
3258
3259         A webpage-forced build was causing SB to throw an exception.
3260         We've long had this bug, but we just didn't notice it until now.
3261
3262         In debugging this I found we fetch an absurd amount of data now
3263         that we use buildbot's json (instead of xmlrpc), so I added
3264         the filter=1 parameter to reduce it a little.  That required
3265         me to handle the case where filter=1 would strip
3266         result=0 in some cases (it's unclear why it does that).
3267
3268         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
3269         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
3270
3271 2011-05-04  Cris Neckar  <cdn@chromium.org>
3272
3273         Unreviewed, Adding Cris Neckar as a committer
3274
3275         * Scripts/webkitpy/common/config/committers.py:
3276
3277 2011-05-04  Siddharth Mathur  <siddharth.mathur@nokia.com>
3278
3279         Reviewed by Laszlo Gombos.
3280
3281         [Qt][Wk2][Symbian] Capabilities for Qt MiniBrowser and QtWebProcess on Symbian
3282         https://bugs.webkit.org/show_bug.cgi?id=59462
3283
3284         * MiniBrowser/qt/MiniBrowser.pro: Add Symbian OS capabilities to executable. Also enable data paging. 
3285
3286 2011-05-04  Eric Seidel  <eric@webkit.org>
3287
3288         Reviewed by Adam Barth.
3289
3290         The testing EWS bots should upload layout-test-results.zip to bugs after failures
3291         https://bugs.webkit.org/show_bug.cgi?id=60223
3292
3293         This required sharing a bit of code between the commit-queue
3294         (which already knew how to do this) and the new EWS testing bots.
3295
3296         In the process I also cleaned up EWS testing a little and
3297         removed some dead code from the commit-queue.
3298
3299         * Scripts/webkitpy/common/config/committervalidator.py:
3300         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
3301         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
3302         * Scripts/webkitpy/tool/commands/queues.py:
3303         * Scripts/webkitpy/tool/commands/queues_unittest.py:
3304
3305 2011-05-04  James Kozianski  <koz@chromium.org>
3306
3307         Reviewed by Eric Seidel.
3308
3309         Add method to make a Port able to retrieve TestOutputSets from its builders.
3310         https://bugs.webkit.org/show_bug.cgi?id=58211
3311
3312         This will be used in the new rebaseline script - TestOutputs are
3313         potential new baselines.
3314
3315         * Scripts/webkitpy/common/config/urls.py:
3316         * Scripts/webkitpy/layout_tests/port/base.py:
3317         * Scripts/webkitpy/layout_tests/port/chromium.py:
3318         * Scripts/webkitpy/layout_tests/port/factory.py:
3319         * Scripts/webkitpy/layout_tests/port/webkit.py:
3320
3321 2011-05-04  Eric Seidel  <eric@webkit.org>
3322
3323         Reviewed by Ojan Vafai.
3324
3325         sheriff-bot can't parse results.html
3326         https://bugs.webkit.org/show_bug.cgi?id=60196
3327
3328         The problem was that the NRWT bots are currently core
3329         and so sheriff-bot was trying to diagnose failures on
3330         them.  It was trying to parse their results.html file
3331         and BeautifulSoup was throwing an exception.
3332
3333         It seems our version of BeautifulSoup was quite old
3334         and unable to handle <script> tags correctly.
3335         This moves us to 3.2 which doesn't seem to have this bug.
3336
3337         * Scripts/webkitpy/thirdparty/BeautifulSoup.py:
3338
3339 2011-05-04  Adam Barth  <abarth@webkit.org>
3340
3341         Reviewed by Eric Seidel.
3342
3343         Make the EWS test logging less verbose
3344         https://bugs.webkit.org/show_bug.cgi?id=60201
3345
3346         This should help us generate less than a hundred MB of logs per hour.
3347
3348         * Scripts/webkitpy/common/config/ports.py:
3349         * Scripts/webkitpy/common/config/ports_unittest.py:
3350
3351 2011-05-04  Tao Bai  <michaelbai@chromium.org>
3352
3353         Reviewed by David Kilzer.
3354
3355         Populate touch-icon url to FrameLoaderClient
3356         https://bugs.webkit.org/show_bug.cgi?id=59143
3357
3358         Disable touch icon by default. 
3359
3360         * Scripts/build-webkit:
3361
3362 2011-05-04  Justin Novosad  <junov@chromium.org>
3363
3364         Reviewed by Kenneth Russell.
3365
3366         [Chromium] Add gpu acceleration flags to DumpRenderTree
3367         https://bugs.webkit.org/show_bug.cgi?id=60168
3368         Adding support for command line flags 
3369         enable-accelerated-2d-canvas an
3370         enable-accelerated-drawing in DumpRenderTree.
3371
3372         * DumpRenderTree/chromium/DumpRenderTree.cpp:
3373         (main):
3374         * DumpRenderTree/chromium/TestShell.cpp:
3375         (TestShell::TestShell):
3376         (TestShell::resetWebSettings):
3377         * DumpRenderTree/chromium/TestShell.h:
3378         (TestShell::setLegacyAccelerated2dCanvasEnabled):
3379         (TestShell::setAcceleratedDrawingEnabled):
3380         * DumpRenderTree/chromium/WebPreferences.cpp:
3381         (WebPreferences::reset):
3382         (WebPreferences::applyTo):
3383         * DumpRenderTree/chromium/WebPreferences.h:
3384
3385 2011-05-04  Ademar de Souza Reis Jr.  <ademar.reis@openbossa.org>
3386
3387         Reviewed by Csaba Osztrogonác.
3388
3389         [Qt] run-qtwebkit-tests misses exec bit
3390         https://bugs.webkit.org/show_bug.cgi?id=60188
3391
3392         Also removing an unecessary import (mostly to let this bugfix goes
3393         through the system, as this is just a chmod).
3394
3395         * Scripts/run-qtwebkit-tests:
3396
3397 2011-05-04  Adam Barth  <abarth@webkit.org>
3398
3399         Fix import of ircproxy module to actually import the module.  I don't
3400         really understand how this code could have worked before.
3401
3402         * Scripts/webkitpy/common/host.py:
3403
3404 2011-05-04  Siddharth Mathur  <siddharth.mathur@nokia.com>
3405
3406         Reviewed by Andreas Kling.
3407
3408         [Qt][Wk2][Symbian] Minor UI facelift for MiniBrowser
3409         https://bugs.webkit.org/show_bug.cgi?id=60172
3410
3411         * MiniBrowser/qt/BrowserWindow.cpp: Bring in UI layout changes from QtTestBrowser
3412         (BrowserWindow::BrowserWindow):
3413
3414 2011-05-04  Mihai Parparita  <mihaip@chromium.org>
3415
3416         Reviewed by Ojan Vafai.
3417
3418         [NRWT] --new-baseline should put baselines in the directory specified by --additional-platform-directory
3419         https://bugs.webkit.org/show_bug.cgi?id=60088
3420         
3421         Makes --new-baseline put new baselines into the first entry in the
3422         baseline search path, including any additional directories passed in via
3423         --additional-platform-directory.
3424         
3425         Simplifies baseline_path() by making it return the first entry in the
3426         baseline search path instead of requiring each port to have an
3427         implementation (should not result in any behavior change).
3428
3429         * Scripts/webkitpy/layout_tests/port/base.py:
3430         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3431         * Scripts/webkitpy/layout_tests/port/chromium.py:
3432         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
3433         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
3434         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
3435         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
3436         * Scripts/webkitpy/layout_tests/port/mac.py:
3437         * Scripts/webkitpy/layout_tests/port/test.py:
3438         * Scripts/webkitpy/layout_tests/port/webkit.py:
3439
3440 2011-05-04  Adam Barth  <abarth@webkit.org>
3441
3442         Reviewed by Eric Seidel.
3443
3444         Fix circular dependency in webkitpy
3445         https://bugs.webkit.org/show_bug.cgi?id=60075
3446
3447         These functions don't belong in bugzilla.py.  They only exist there
3448         because they are old.  Really, these functions shouldn't be free
3449         functions at all, but that's a patch for another day.
3450
3451         * Scripts/webkitpy/common/checkout/changelog.py:
3452         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
3453         * Scripts/webkitpy/common/checkout/checkout.py:
3454         * Scripts/webkitpy/common/config/urls.py:
3455         * Scripts/webkitpy/common/net/bugzilla/__init__.py:
3456         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
3457         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
3458         * Scripts/webkitpy/style/checkers/changelog.py:
3459         * Scripts/webkitpy/tool/bot/irc_command.py:
3460         * Scripts/webkitpy/tool/bot/sheriff.py:
3461         * Scripts/webkitpy/tool/commands/upload.py:
3462
3463 2011-05-03  Pratik Solanki  <psolanki@apple.com>
3464
3465         Reviewed by Dan Bernstein.
3466
3467         https://bugs.webkit.org/show_bug.cgi?id=60141
3468         Remove reference to libWebKitSystemInterfaceTiger.a from build-webkit
3469
3470         * Scripts/build-webkit:
3471
3472 2011-05-03  Mihai Parparita  <mihaip@chromium.org>
3473
3474         Reviewed by Ojan Vafai.
3475
3476         Update rebaseline queue server to handle hierarchical test results
3477         https://bugs.webkit.org/show_bug.cgi?id=60063
3478         
3479         Update test result parsing done by the rebaseline queue server to handle
3480         the hierarchical test result output that is generated as of r85254.
3481
3482         * RebaselineQueueServer/templates/builder-queue-edit.html:
3483
3484 2011-05-03  Eric Seidel  <eric@webkit.org>
3485
3486         Reviewed by Adam Roben.
3487
3488         webkit-patch post-commits HEAD^.. throws an exception
3489         https://bugs.webkit.org/show_bug.cgi?id=60027
3490
3491         * Scripts/webkitpy/common/checkout/scm/git.py:
3492
3493 2011-05-03  Luiz Agostini  <luiz.agostini@openbossa.org>
3494
3495         Reviewed by Kenneth Rohde Christiansen.
3496
3497         [Qt] Layout tests may timeout if loadFinished receives false
3498         https://bugs.webkit.org/show_bug.cgi?id=60064
3499
3500         When using layoutTestController.queueLoad(), if QWebPage::loadFinished(bool) receives false
3501         then the layout test may timout. It happens that QWebPage::loadFinished() receives false when
3502         a response is ignored, like it should be for 204 responses. It is a problem for creating
3503         tests for 204 responses for example.
3504
3505         This change has been tested and it causes no regressions in layout tests.
3506
3507         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
3508         (LayoutTestController::maybeDump):
3509
3510 2011-05-03  Dirk Pranke  <dpranke@chromium.org>
3511
3512         Unreviewed, rolling out r85659.
3513         http://trac.webkit.org/changeset/85659
3514         https://bugs.webkit.org/show_bug.cgi?id=59977
3515
3516         restarting isn't working correctly
3517
3518         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
3519         * Scripts/webkitpy/layout_tests/port/http_server.py:
3520         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
3521         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
3522
3523 2011-05-03  Dirk Pranke  <dpranke@chromium.org>
3524
3525         Reviewed by Ojan Vafai.
3526
3527         new-run-webkit-tests: fix http server startup/shutdown
3528
3529         Previous versions of the code had three problems that made startup
3530         and shutdown flaky. The first is that it would throw exceptions
3531         if it couldn't delete stale log files, which was overly
3532         paranoid. The second is that some of the exceptions weren't
3533         defined properly. The third, and most important, is that it was
3534         using urllib to check if ports were available, which was leaving
3535         sockets in a half-closed state, and keeping ports from being
3536         reused. By switching to raw sockets, we are able to now reliably
3537         restart.
3538
3539         This change also switches the code to using Executives to stop
3540         processes, which will let us delete a bunch of code in the
3541         port/* implementations and fix a weird layering problem in a
3542         subsequent patch.
3543
3544         https://bugs.webkit.org/show_bug.cgi?id=59977
3545
3546         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
3547         * Scripts/webkitpy/layout_tests/port/http_server.py:
3548         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
3549         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
3550
3551 2011-05-03  Dirk Pranke  <dpranke@chromium.org>
3552
3553         Reviewed by Adam Barth.
3554
3555         fix failing printing_unittest on win32
3556         https://bugs.webkit.org/show_bug.cgi?id=59963
3557
3558         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
3559
3560 2011-05-03  Adam Roben  <aroben@apple.com>
3561
3562         Fix most strict PassOwnPtr violations on Windows
3563
3564         Fixes <http://webkit.org/b/60025> Windows should (almost) build with strict PassOwnPtr
3565         enabled
3566
3567         Reviewed by Anders Carlsson.
3568
3569         * WebKitAPITest/Test.h: Use adoptPtr.
3570
3571 2011-05-03  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
3572
3573         Reviewed by Andreas Kling.
3574
3575         Add myself to the committers list.
3576
3577         * Scripts/webkitpy/common/config/committers.py:
3578
3579 2011-05-03  James Robinson  <jamesr@chromium.org>
3580
3581         Reviewed by Kenneth Russell.
3582
3583         [chromium] Layout test animations/3d/replace-filling-transform.html failing
3584         https://bugs.webkit.org/show_bug.cgi?id=59839
3585
3586         Add animations/3d to the list of tests run in the chromium-gpu configuration since these tests exercise the compositor.
3587
3588         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
3589
3590 2011-05-03  Lucas De Marchi  <lucas.demarchi@profusion.mobi>
3591
3592         Unreviewed, rolling out r85471.
3593         http://trac.webkit.org/changeset/85471
3594         https://bugs.webkit.org/show_bug.cgi?id=59922
3595
3596         The rule of thumb in EFL is that you init in all the places
3597         you use it. So, if in your application you call ecore_evas_*
3598         functions, you have to call ecore_evas_init() first.
3599
3600         * EWebLauncher/main.c: always call *_init() functions when you use
3601         that namespace.
3602         (quit):
3603         (main):
3604
3605 2011-05-02  Ojan Vafai  <ojan@chromium.org>
3606
3607         Reviewed by Tony Chang.
3608
3609         remove no longer used method
3610         https://bugs.webkit.org/show_bug.cgi?id=59981
3611
3612         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
3613         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
3614
3615 2011-05-02  Ojan Vafai  <ojan@chromium.org>
3616
3617         Reviewed by Tony Chang.
3618
3619         add image toggling behavior from old-run-webkit-tests
3620         https://bugs.webkit.org/show_bug.cgi?id=59978
3621
3622         Output -diffs.html files. This matches old-run-webkit-tests.
3623         In theory, we could use data urls in the results.html file, but
3624         data urls don't currently have access to local files.
3625
3626         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
3627
3628 2011-05-02  Tony Chang  <tony@chromium.org>
3629
3630         Reviewed by Ojan Vafai.
3631
3632         [nrwt] remove some obsolete references to .checksum files
3633         https://bugs.webkit.org/show_bug.cgi?id=59971
3634
3635         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
3636         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
3637         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
3638         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
3639         * Scripts/webkitpy/layout_tests/port/test.py:
3640         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
3641
3642 2011-05-02  Adam Roben  <aroben@apple.com>
3643
3644         Take advantage of implicit nullptr_t -> PassOwnPtr conversion
3645
3646         Fixes <http://webkit.org/b/59964> Implicit conversion from std::nullptr_t to PassOwnPtr
3647         doesn't work, but sho