Tools/ChangeLog

   1 2011-05-18  Adam Roben  <aroben@apple.com>
   2
   3         Add a new page to build.webkit.org to help find when tests started failing
   4
   5         The page is accessible at <http://build.webkit.org/TestFailures/>. It is pretty minimalist
   6         right now, but already shows some useful information. It's somewhat similar to webkit-patch
   7         failure-reason and sheriffbot, and perhaps can be combined with them eventually. It's a
   8         little more convenient than either of them, though, because it's all done in the browser
   9         (and thus it's easy to go directly to the relevant test results).
  10
  11         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js: Added.
  12         (Buildbot): This class represents a Buildbot server.
  13         (Buildbot.prototype.buildURL): Returns the URL for the summary page for a particular build.
  14         (Buildbot.prototype.builderNamed): Returns a Builder with the given name.
  15         (Buildbot.prototype.getTesterNames): Fetches the names of all testers and passes them to the
  16         callback.
  17         (Buildbot.prototype.parseBuildName): Breaks up a build name into its constituent parts. Must
  18         be implemented by a derived class that understands this server's build naming scheme.
  19         (Buildbot.prototype.resultsDirectoryURL): Returns the URL for the results directory for a
  20         particular build.
  21
  22         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js: Added.
  23         (Builder): This class represents one builder on the buildbot.
  24         (Builder.prototype.buildURL): Returns the URL for the summary page for a particular build.
  25         (Builder.prototype.failureDiagnosisTextAndURL): Returns data that provides a little more
  26         information about a particular test failure.
  27         (Builder.prototype.startFetchingBuildHistory): Periodically calls the callback with
  28         information about when tests started failing.
  29         (Builder.prototype.resultsDirectoryURL): Returns the URL for the results directory for a
  30         particular build.
  31         (Builder.prototype._getBuildNames): Fetches the names of all builds and passes them to the
  32         callback.
  33         (Builder.prototype._getFailingTests): Fetches the results.html page for the given build and
  34         extracts all the failing tests listed in it, passing them to the callback.
  35         (Builder.prototype._incorporateBuildHistory): Gets the failing tests for the specified
  36         build, merges them into the build history, and calls the callback telling it whether the
  37         next build should be fetched to provide more information.
  38
  39         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
  40         Added. Just some simple styles.
  41
  42         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js: Added.
  43         (createDefinitionList): Takes an array of pairs and turns them into a DL element.
  44         (getResource): Wrapper around XMLHttpRequest.
  45         (Array.prototype.findFirst): Finds the first element matching the given predicate and
  46         returns it.
  47         (Array.prototype.last): Returns the last element of the array.
  48
  49         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js: Added.
  50         (ViewController): This class contains the main logic for displaying the page.
  51         (ViewController.loaded): Just calls through to parseHash.
  52         (ViewController.parseHash): Either starts analyzing failures on a particular builder, or
  53         shows the list of all testers so one can be chosen. This function is called when the page
  54         loads and whenever we get a hashchange event.
  55         (ViewController._displayBuilder): Asks the builder to fetch build history, and displays it
  56         as it is fetched. The display ends up grouping tests by when they started failing.
  57         (ViewController._displayTesters): Gets the list of testers and displays it.
  58
  59         (ViewController._domForBuildName):
  60         (ViewController._domForFailedTest):
  61         Helper functions to create descriptions and links for a particular build or failed test.
  62
  63         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBuildbot.js: Added.
  64         (WebKitBuildbot): Calls up to the base class constructor with the correct base URL.
  65         (WebKitBuildbot.prototype.parseBuildName): Parses a build.webkit.org-style build name.
  66
  67         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Added. Just
  68         loads all the files and sets up a ViewController, which does the rest.
  69
  70         * BuildSlaveSupport/build.webkit.org-config/templates/root.html: Added a link to the new
  71         page.
  72
  73 2011-05-18  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
  74
  75         Reviewed by Csaba Osztrogonác.
  76
  77         [Qt] Implement layoutTestController.setValueForUser()
  78         https://bugs.webkit.org/show_bug.cgi?id=60956
  79
  80         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
  81         (LayoutTestController::setValueForUser):
  82         * DumpRenderTree/qt/LayoutTestControllerQt.h:
  83
  84 2011-05-16  Robert Hogan  <robert@webkit.org>
  85
  86         Reviewed by Kenneth Rohde Christiansen.
  87
  88         plugins/invalidate_rect.html fails on linux ports
  89
  90         - Make the unix test netscape plugin recognize the onPaintEvent
  91           and windowedPlugin parameters.
  92
  93         https://bugs.webkit.org/show_bug.cgi?id=54051
  94
  95         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
  96         (handleEventX11):
  97         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
  98         (LayoutTestController::displayInvalidatedRegion):
  99         * DumpRenderTree/qt/LayoutTestControllerQt.h:
 100         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
 101         (webkit_test_plugin_new_instance):
 102         (webkit_test_plugin_handle_event):
 103
 104 2011-05-17  Tony Chang  <tony@chromium.org>
 105
 106         Reviewed by Ojan Vafai.
 107
 108         [chromium] move Lucid 64 bit results into LayoutTests/platform/chromium-linux
 109         https://bugs.webkit.org/show_bug.cgi?id=60895
 110
 111         Update the tools to handle the move.  Lucid 32 results now go in
 112         chromium-linux-x86 and the default platform on Linux is now x86_64.
 113
 114         * Scripts/webkitpy/layout_tests/deduplicate_tests.py: Default to x86_64
 115         * Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py:
 116         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py: Default to Lucid
 117         * Scripts/webkitpy/layout_tests/port/base.py: ditto
 118         * Scripts/webkitpy/layout_tests/port/chromium_linux.py: Update directory fallback and default to Lucid 64
 119         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
 120         * Scripts/webkitpy/layout_tests/port/test.py: Update tests to default to x86_64
 121         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py: Update bot names.
 122         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py: Whitespace cleanup.
 123
 124 2011-05-17  Adam Roben  <aroben@apple.com>
 125
 126         Make run-api-tests work on Windows when there are spaces in the path
 127
 128         Fixes <http://webkit.org/b/60954> REGRESSION (r86511): run-api-tests fails if there are
 129         spaces in the path to TestWebKitAPI.exe
 130
 131         Reviewed by David Levin.
 132
 133         * Scripts/run-api-tests:
 134         (runTestTool): Use the "direct object" form of system() to avoid having the path to
 135         TestWebKitAPI.exe be split by the shell.
 136
 137 2011-05-16  Ian Henderson  <ianh@apple.com>
 138
 139         Reviewed by Joseph Pecoraro.
 140
 141         Page::goToItem doesn't work while loading is deferred
 142         https://bugs.webkit.org/show_bug.cgi?id=60412
 143
 144         Add setDefersLoading and goBack methods to LayoutTestController.  We
 145         need to use goBack() instead of history.back() because the latter goes
 146         through NavigationScheduler, hence doesn't exhibit the bug.
 147
 148         * DumpRenderTree/LayoutTestController.cpp:
 149         (goBackCallback):
 150         (setDefersLoadingCallback):
 151         (LayoutTestController::staticFunctions):
 152         * DumpRenderTree/LayoutTestController.h:
 153         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
 154         (LayoutTestController::goBack):
 155         (LayoutTestController::setDefersLoading):
 156         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
 157         (LayoutTestController::goBack):
 158         (LayoutTestController::setDefersLoading):
 159         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
 160         (LayoutTestController::goBack):
 161         (LayoutTestController::setDefersLoading):
 162         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
 163         (LayoutTestController::goBack):
 164         (LayoutTestController::setDefersLoading):
 165
 166 2011-05-16  Sam Weinig  <sam@webkit.org>
 167
 168         Reviewed by Anders Carlsson.
 169
 170         TestWebKitAPI should build with clang if it can
 171         https://bugs.webkit.org/show_bug.cgi?id=60918
 172
 173         * TestWebKitAPI/Configurations/CompilerVersion.xcconfig:
 174         Update CompilerVersion.xcconfig to match others.
 175
 176         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
 177         Add CompilerVersion.xcconfig to the project.
 178
 179         * TestWebKitAPI/Tests/WebKit2/FailedLoad.cpp:
 180         (TestWebKitAPI::didFailProvisionalLoadWithErrorForFrame):
 181         * TestWebKitAPI/Tests/WebKit2/PageLoadDidChangeLocationWithinPageForFrame.cpp:
 182         (TestWebKitAPI::didSameDocumentNavigationForFrame):
 183         Add some casts to quiet warnings from clang.
 184
 185 2011-05-16  Dirk Pranke  <dpranke@chromium.org>
 186
 187         Reviewed by David Levin.
 188
 189         add dpranke as a reviewer
 190         https://bugs.webkit.org/show_bug.cgi?id=60919
 191
 192         * Scripts/webkitpy/common/config/committers.py:
 193
 194 2011-05-16  Robert Hogan  <robert@webkit.org>
 195
 196         Reviewed by Anders Carlsson.
 197
 198         [Gtk] plugins/get-url-notify-with-url-that-fails-to-load.html crashes on buildbot
 199
 200         https://bugs.webkit.org/show_bug.cgi?id=60838
 201
 202         The unix test plugin needs to call the test's NPP_URLNotify
 203         rather than the browsers.
 204
 205         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
 206         (webkit_test_plugin_url_notify):
 207
 208 2011-05-16  Brent Fulgham  <bfulgham@webkit.org>
 209
 210         Rubber-stamped by Adam Roben.
 211
 212         Adding myself as a reviewer.
 213         * Scripts/webkitpy/common/config/committers.py:
 214
 215 2011-05-16  Sam Weinig  <sam@webkit.org>
 216
 217         Fix typo pointed out by Dave Levin.
 218
 219         * TestWebKitAPI/PlatformUtilities.h:
 220         (TestWebKitAPI::Util::assertWKStringEqual):
 221         Strig -> String.
 222
 223 2011-05-16  Sam Weinig  <sam@webkit.org>
 224
 225         Reviewed by David Levin.
 226
 227         Convert api tester over to using gtest expectations directly
 228         https://bugs.webkit.org/show_bug.cgi?id=60862
 229
 230         * TestWebKitAPI/PlatformUtilities.cpp:
 231         (TestWebKitAPI::Util::toSTD):
 232         * TestWebKitAPI/PlatformUtilities.h:
 233         (TestWebKitAPI::Util::assertWKStrigEqual):
 234         Add convenience macro to compare WK2 strings. Add some overloads of toSTD to make
 235         the implementation of the macro simpler.
 236
 237         * TestWebKitAPI/Test.h:
 238         Remove TEST_ASSERT forwarder.
 239
 240         [Test changes elided]
 241
 242 2011-05-16  David Kilzer  <ddkilzer@apple.com>
 243
 244         <http://webkit.org/b/60913> C++ exceptions should not be enabled when building with llvm-gcc-4.2
 245         <rdar://problem/9446430>
 246
 247         Reviewed by Mark Rowe.
 248
 249         * MiniBrowser/Configurations/Base.xcconfig: Fixed typo.
 250         * TestWebKitAPI/Configurations/Base.xcconfig: Ditto.
 251         * WebKitTestRunner/Configurations/Base.xcconfig: Ditto.
 252
 253 2011-05-16  Jon Lee  <jonlee@apple.com>
 254
 255         Reviewed by Simon Fraser.
 256
 257         Can't horizontally scroll iframes and overflow because wheel events are always accepted
 258         https://bugs.webkit.org/show_bug.cgi?id=60779
 259
 260         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
 261         (WTR::InjectedBundlePage::InjectedBundlePage): Set new default method to nil.
 262
 263 2011-05-16  Ademar de Souza Reis Jr.  <ademar.reis@openbossa.org>
 264
 265         Reviewed by Csaba Osztrogonác.
 266
 267         [Qt] Tools.pro misses include(common.pri)
 268         https://bugs.webkit.org/show_bug.cgi?id=60883
 269
 270         It includes features.pri, which depends on common.pri (the build is
 271         not broken today with the default options, but this is the right
 272         thing to do).
 273
 274         * Tools.pro:
 275
 276 2011-05-15  Dominic Cooney  <dominicc@chromium.org>
 277
 278         Unreviewed: adding myself to committers.py.
 279
 280         * Scripts/webkitpy/common/config/committers.py:
 281
 282 2011-05-15  Daniel Bates  <dbates@webkit.org>
 283
 284         Reviewed by Chris Jerdonek.
 285
 286         svn-apply can't handle Git diff that only changes executable bit
 287         https://bugs.webkit.org/show_bug.cgi?id=60848
 288
 289         Fixes an issue where svn-{apply, unapply} fail to handle a Git diff for a file
 290         that only has an executable bit change.
 291
 292         * Scripts/VCSUtils.pm:
 293           - Modified regex $chunkRangeRegEx to match chunk range with ending text, say: @@ -2,6 +2,18 @@ foo().
 294         (parseDiff): Modified to count the number of text chunks in the diff. This should also help towards
 295         fixing <https://bugs.webkit.org/show_bug.cgi?id=29684>.
 296         * Scripts/svn-apply:
 297         (patch): Only use standard patch tool if the patch is non-empty and for a text file (i.e. not a binary patch).
 298         * Scripts/svn-unapply:
 299         (patch): Ditto.
 300         * Scripts/webkitperl/VCSUtils_unittest/parseDiff.pl:
 301           - Updated expected results for test cases to account for numTextChunks.
 302           - Fixed malformed chunk range in test cases: "Git: simple", and "Git: Git diff followed by SVN diff".
 303           - Added test case "Git: file that only has an executable bit change".
 304         * Scripts/webkitperl/VCSUtils_unittest/parseDiffWithMockFiles.pl:
 305           - Updated expected results for test cases to account for numTextChunks.
 306
 307 2011-05-15  Sam Weinig  <sam@webkit.org>
 308
 309         Fix windows build.
 310
 311         * TestWebKitAPI/win/main.cpp:
 312         (main):
 313
 314 2011-05-15  Sam Weinig  <sam@webkit.org>
 315
 316         Reviewed by David Levin.
 317
 318         Make API tester more gtest friendly
 319         https://bugs.webkit.org/show_bug.cgi?id=60859
 320
 321         * Scripts/run-api-tests:
 322         Numerous changes:
 323         - Remove logic to run tests one at a time.
 324         - Remove logic to build up a list of tests, instead just pass --gtest_list_tests to
 325           the test runner and allow gtest to print it for us.
 326         - Forward stdout from the test tool to stdout, even in non-verbose runs.
 327         - Run test tool from the current working directory instead of the root webkit
 328           directory.
 329
 330         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
 331         Move plist to be with other bundle sources.
 332
 333         * TestWebKitAPI/TestsController.cpp:
 334         (TestWebKitAPI::TestsController::TestsController):
 335         (TestWebKitAPI::TestsController::run):
 336         * TestWebKitAPI/TestsController.h:
 337         * TestWebKitAPI/mac/main.mm:
 338         Simplify TestController down to a single run function. We should consider removing the class
 339         entirely at some point.
 340
 341 2011-05-15  Sheriff Bot  <webkit.review.bot@gmail.com>
 342
 343         Unreviewed, rolling out r86504.
 344         http://trac.webkit.org/changeset/86504
 345         https://bugs.webkit.org/show_bug.cgi?id=60853
 346
 347         "Broke Qt EventSender in editing/selection tests" (Requested
 348         by mwenge on #webkit).
 349
 350         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
 351         (WebCore::WebPage::WebPage):
 352         (WebCore::DumpRenderTree::DumpRenderTree):
 353         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
 354         (WebCore::DumpRenderTree::initJSObjects):
 355         * DumpRenderTree/qt/DumpRenderTreeQt.h:
 356         (WebCore::DumpRenderTree::eventSender):
 357         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
 358         (webkit_test_plugin_new_instance):
 359         (webkit_test_plugin_set_window):
 360         (webkit_test_plugin_handle_event):
 361
 362 2011-05-12  Robert Hogan  <robert@webkit.org>
 363
 364         Reviewed by Benjamin Poulain.
 365
 366         [Qt] fix http/tests/plugins/plugin-document-has-focus.html
 367
 368         The support for this test added to the unix test plugin here
 369         may allow other platforms to pass it (nearly everyone
 370         skips it). On Qt it required a bit of trickiness with
 371         the page's EventSender object to get it working fully though,
 372         so only unskipping Qt here.
 373
 374         https://bugs.webkit.org/show_bug.cgi?id=60722
 375
 376         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
 377         (WebCore::WebPage::WebPage):
 378         (WebCore::WebPage::eventSender):
 379         (WebCore::DumpRenderTree::DumpRenderTree):
 380         (WebCore::DumpRenderTree::initJSObjects):
 381         (WebCore::DumpRenderTree::createWindow):
 382         * DumpRenderTree/qt/DumpRenderTreeQt.h:
 383         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
 384         (webkit_test_plugin_new_instance):
 385         (webkit_test_plugin_set_window):
 386         (webkit_test_plugin_handle_event):
 387
 388 2011-05-14  Arno Renevier  <arno@renevier.net>
 389
 390         Reviewed by Martin Robinson.
 391
 392         make run-gtk-tests executable
 393         https://bugs.webkit.org/show_bug.cgi?id=59278
 394
 395         * Scripts/run-gtk-tests:
 396
 397 2011-05-14  Anders Carlsson  <andersca@apple.com>
 398
 399         Reviewed by Kevin Decker.
 400
 401         NPP_URLNotify is not called if a URL passed to NPN_GetURLNotify fails to load
 402         https://bugs.webkit.org/show_bug.cgi?id=60823
 403         <rdar://problem/9430386>
 404
 405         Add test that calls NPN_GetURLNotify on a URL that can't be loaded, and checks if NPP_URLNotify is called.
 406
 407         * DumpRenderTree/DumpRenderTree.gypi:
 408         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
 409         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLNotifyWithURLThatFailsToLoad.cpp: Added.
 410         (GetURLNotifyWithURLThatFailsToLoad::GetURLNotifyWithURLThatFailsToLoad):
 411         (GetURLNotifyWithURLThatFailsToLoad::NPP_New):
 412         (GetURLNotifyWithURLThatFailsToLoad::NPP_URLNotify):
 413         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
 414         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
 415         * GNUmakefile.am:
 416
 417 2011-05-14  Sheriff Bot  <webkit.review.bot@gmail.com>
 418
 419         Unreviewed, rolling out r86476.
 420         http://trac.webkit.org/changeset/86476
 421         https://bugs.webkit.org/show_bug.cgi?id=60832
 422
 423         It made plugins/get-url-notify-with-url-that-fails-to-
 424         load.html crash on GTK (Requested by Ossy on #webkit).
 425
 426         * DumpRenderTree/DumpRenderTree.gypi:
 427         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
 428         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLNotifyWithURLThatFailsToLoad.cpp: Removed.
 429         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
 430         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
 431         * GNUmakefile.am:
 432
 433 2011-05-13  Anders Carlsson  <andersca@apple.com>
 434
 435         Reviewed by Sam Weinig.
 436
 437         NPP_URLNotify is not called if a URL passed to NPN_GetURLNotify fails to load
 438         https://bugs.webkit.org/show_bug.cgi?id=60823
 439         <rdar://problem/9430386>
 440
 441         Add test that calls NPN_GetURLNotify on a URL that can't be loaded, and checks if NPP_URLNotify is called.
 442
 443         * DumpRenderTree/DumpRenderTree.gypi:
 444         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
 445         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLNotifyWithURLThatFailsToLoad.cpp: Added.
 446         (GetURLNotifyWithURLThatFailsToLoad::GetURLNotifyWithURLThatFailsToLoad):
 447         (GetURLNotifyWithURLThatFailsToLoad::NPP_New):
 448         (GetURLNotifyWithURLThatFailsToLoad::NPP_URLNotify):
 449         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
 450         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
 451         * GNUmakefile.am:
 452
 453 2011-05-13  Anders Carlsson  <andersca@apple.com>
 454
 455         Build fix.
 456
 457         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLWithJavaScriptURL.cpp:
 458
 459 2011-05-13  Anders Carlsson  <andersca@apple.com>
 460
 461         Reviewed by Sam Weinig.
 462
 463         NPN_GetURL with a javascript: URL doesn't send back the result in the stream
 464         https://bugs.webkit.org/show_bug.cgi?id=60810
 465
 466         Add a test that calls NPN_GetURL on a javascript: URL and checks that it gets back the expected result.
 467
 468         * DumpRenderTree/DumpRenderTree.gypi:
 469         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
 470         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
 471         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
 472         * GNUmakefile.am:
 473         Add files.
 474
 475         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
 476         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
 477         Add new NPN and NPP wrappers and sort them according to the order they appear in npfunctions.h
 478
 479         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLWithJavaScriptURL.cpp: Added.
 480         (GetURLWithJavaScriptURL::GetURLWithJavaScriptURL):
 481         (GetURLWithJavaScriptURL::NPP_New):
 482         (GetURLWithJavaScriptURL::NPP_NewStream):
 483         (GetURLWithJavaScriptURL::NPP_DestroyStream):
 484         (GetURLWithJavaScriptURL::NPP_WriteReady):
 485         (GetURLWithJavaScriptURL::NPP_Write):
 486         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
 487         (NPP_NewStream):
 488         (NPP_WriteReady):
 489         (NPP_Write):
 490         (NPP_URLNotify):
 491         call the PluginTest NPP wrappers.
 492
 493 2011-05-13  Sheriff Bot  <webkit.review.bot@gmail.com>
 494
 495         Unreviewed, rolling out r86447.
 496         http://trac.webkit.org/changeset/86447
 497         https://bugs.webkit.org/show_bug.cgi?id=60809
 498
 499         "Broke some uses of EventSender object on Qt" (Requested by
 500         mwenge on #webkit).
 501
 502         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
 503         (WebCore::WebPage::WebPage):
 504         (WebCore::DumpRenderTree::DumpRenderTree):
 505         (WebCore::DumpRenderTree::initJSObjects):
 506         (WebCore::DumpRenderTree::createWindow):
 507         * DumpRenderTree/qt/DumpRenderTreeQt.h:
 508         (WebCore::DumpRenderTree::eventSender):
 509         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
 510         (webkit_test_plugin_new_instance):
 511         (webkit_test_plugin_set_window):
 512         (webkit_test_plugin_handle_event):
 513
 514 2011-05-13  Robert Hogan  <robert@webkit.org>
 515
 516         Reviewed by Benjamin Poulain.
 517
 518         [Qt] fix http/tests/plugins/plugin-document-has-focus.html
 519
 520         The support for this test added to the unix test plugin here
 521         may allow other platforms to pass it (nearly everyone
 522         skips it). On Qt it required a bit of trickiness with
 523         the page's EventSender object to get it working fully though,
 524         so only unskipping Qt here.
 525
 526         https://bugs.webkit.org/show_bug.cgi?id=60722
 527
 528         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
 529         (WebCore::WebPage::WebPage):
 530         (WebCore::WebPage::eventSender):
 531         (WebCore::DumpRenderTree::DumpRenderTree):
 532         (WebCore::DumpRenderTree::initJSObjects):
 533         (WebCore::DumpRenderTree::createWindow):
 534         * DumpRenderTree/qt/DumpRenderTreeQt.h:
 535         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
 536         (webkit_test_plugin_new_instance):
 537         (webkit_test_plugin_set_window):
 538         (webkit_test_plugin_handle_event):
 539
 540 2011-05-13  Adam Roben  <aroben@apple.com>
 541
 542         Exclude the Leopard NRWT bot from the core builder set
 543
 544         Fixes <http://webkit.org/b/60392> Leopard Intel Release (NRWT) shouldn't be a core builder
 545
 546         Reviewed by Darin Adler.
 547
 548         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
 549         (BuildBot.__init__): Tighten up the Leopard regexp to only allow the non-NRWT Leopard bots
 550         in the core builder set.
 551
 552         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
 553         (BuildBotTest.test_builder_name_regexps): Updated to match the current set of builders and
 554         regexps.
 555
 556 2011-05-13  Adam Roben  <aroben@apple.com>
 557
 558         Group all the NRWT bots together
 559
 560         This also gets the new NRWT bots out of the middle of Apple's bots. Since these new NRWT
 561         bots are experimental, it seems best not to clutter up the non-experimental bots with
 562         experimental ones.
 563
 564         Fixes <http://webkit.org/b/60769> New "(NRWT)" bots are right in the middle of Apple's bots
 565         on build.webkit.org
 566
 567         Reviewed by Dan Bernstein.
 568
 569         * BuildSlaveSupport/build.webkit.org-config/config.json: Moved the "(NRWT)" bots next to the
 570         Chromium testers, which also use new-run-webkit-tests.
 571
 572 2011-05-13  Adam Roben  <aroben@apple.com>
 573
 574         Add a plugin test that dumps the plugin window's rect
 575
 576         Part of the test for <http://webkit.org/b/60194> <rdar://problem/9383760> REGRESSION
 577         (r79167): Windowed plugins in Google Reader don't move when the article list is scrolled
 578
 579         Reviewed by Dan Bernstein.
 580
 581         * DumpRenderTree/TestNetscapePlugIn/Tests/win/DumpWindowRect.cpp: Added.
 582         (DumpWindowRect::DumpWindowRect): Just call up to the base class.
 583         (DumpWindowRect::performWindowGeometryTest): Find our window rect relative to the test
 584         harness window and log it.
 585
 586         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Added DumpWindowRect.
 587
 588 2011-05-13  Adam Roben  <aroben@apple.com>
 589
 590         Move code required for most tests of plugin window geometry into a base class
 591
 592         WebKit2's asynchronous plugin window positioning makes testing window geometry tricky. By
 593         having all the trickiness in a shared base class, writing new window geometry tests will be
 594         much easier.
 595
 596         Fixes <http://webkit.org/b/60776> It's hard to write tests that check plugin window geometry
 597
 598         Reviewed by Anders Carlsson.
 599
 600         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowRegionIsSetToClipRect.cpp:
 601         (WindowRegionIsSetToClipRect::WindowRegionIsSetToClipRect): Changed to inherit from the new
 602         WindowGeometryTest base class. Most of the code from this class was moved to the new base
 603         class.
 604         (WindowRegionIsSetToClipRect::performWindowGeometryTest): Renamed from checkWindowRegion.
 605         This is called by the base class when it is safe to check the window's geometry. Note that
 606         the m_didCheckWindowRegion logic was removed completely; it has never been needed and was
 607         just vestigial code from an earlier version of this test.
 608
 609         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Added new
 610         WindowGeometryTest class.
 611
 612         * DumpRenderTree/TestNetscapePlugIn/win/WindowGeometryTest.cpp: Added. Code came from
 613         WindowRegionIsSetToClipRect.
 614         (WindowGeometryTest::finishTest): This is the only change from WindowRegionIsSetToClipRect.
 615         We now call the new virtual performWindowGeometryTest function instead of checkWindowRegion.
 616         Subclasses can implement that function to perform their own tests.
 617
 618         * DumpRenderTree/TestNetscapePlugIn/win/WindowGeometryTest.h: Added.
 619
 620 2011-05-13  Zoltan Herczeg  <zherczeg@webkit.org>
 621
 622         Reviewed by Csaba Osztrogonác.
 623
 624         Missing 2nd #EOF if WebProcess is killed.
 625         https://bugs.webkit.org/show_bug.cgi?id=60761
 626
 627         If the WebProcess is killed, only one #EOF is yielded
 628         by WebKitTestRunner. However, run-webkit-test wait for
 629         a second #EOF, which signals the finishing of pixel dump.
 630         Since this never happens, run-webkit-test waits forever.
 631
 632         * WebKitTestRunner/TestInvocation.cpp:
 633         (WTR::TestInvocation::dump):
 634         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
 635         * WebKitTestRunner/TestInvocation.h:
 636
 637 2011-05-13  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
 638
 639         Reviewed by Csaba Osztrogonác.
 640
 641         [NRWT] chromium-linux tests shouldn't run on other platforms
 642         https://bugs.webkit.org/show_bug.cgi?id=60162
 643
 644         Remove _tests_for_other_platforms defs because r82753
 645         implemented it properly in the baseclass in webkit.py.
 646
 647         * Scripts/webkitpy/layout_tests/port/gtk.py:
 648         * Scripts/webkitpy/layout_tests/port/qt.py:
 649         * Scripts/webkitpy/layout_tests/port/win.py:
 650
 651 2011-05-12  Ryuan Choi  <ryuan.choi@samsung.com>
 652
 653         Reviewed by Antonio Gomes.
 654
 655         [EFL] Remove unused option of EWeblauncher
 656         https://bugs.webkit.org/show_bug.cgi?id=60747
 657
 658         Remove sudo-workaround option.
 659         This option is not used and make other option wrong.
 660
 661         * EWebLauncher/main.c:
 662
 663 2011-05-12  Lucas Forschler  <lforschler@apple.com>
 664
 665         Reviewed by Stephanie Lewis.
 666
 667         Fix config.json to address missing WK2 builder/tester config.
 668         Fix naming of machines (xserve not macpro)
 669
 670         * BuildSlaveSupport/build.webkit.org-config/config.json:
 671
 672 2011-05-12  Lucas Forschler  <lforschler@apple.com>
 673
 674         Reviewed by Stephanie Lewis.
 675
 676         Add SnowLeopard Debug builder and Test bot.
 677         apple-xserve-8 = SL Debug Tester.
 678         apple-xserve-9 = SL Debug Builder.
 679
 680         * BuildSlaveSupport/build.webkit.org-config/config.json:
 681
 682 2011-05-12  Adam Barth  <abarth@webkit.org>
 683
 684         This patch disables sheriffbot's bug postings.  He's just been way too
 685         spammy recently due to test flakiness.
 686
 687         * Scripts/webkitpy/tool/commands/sheriffbot.py:
 688         * Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
 689
 690 2011-05-11  Brian Weinstein  <bweinstein@apple.com>
 691
 692         Reviewed by Anders Carlsson.
 693
 694         After relaunching an unresponsive web process, web page stops responding to mouse movement
 695         https://bugs.webkit.org/show_bug.cgi?id=60669
 696         <rdar://problem/9409934>
 697
 698         This test follows the steps needed to reproduce this bug:
 699         1) Hang the web process
 700         2) While the web process is hung, move the mouse (so WebPageProxy thinks it is processing a mouse move event)
 701         3) Kill and restart the WebProcess
 702         4) Move the mouse
 703
 704         We are testing that we receive mouse move events after the WebProcess has restarted.
 705
 706         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash.cpp: Added.
 707         (TestWebKitAPI::didFinishLoadForFrame): Set a flag that we finished loading.
 708         (TestWebKitAPI::processDidBecomeUnresponsive): Kill the web process once we become unresponsive.
 709         (TestWebKitAPI::processDidCrash): Reload the page when the process crashes, which restarts the web process.
 710         (TestWebKitAPI::setPageLoaderClient):
 711         (TestWebKitAPI::TEST):
 712         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash_Bundle.cpp: Added.
 713         (TestWebKitAPI::MouseMoveAfterCrashTest::MouseMoveAfterCrashTest):
 714         (TestWebKitAPI::MouseMoveAfterCrashTest::didReceiveMessage):
 715         * TestWebKitAPI/Tests/WebKit2/mouse-move-listener.html: Added.
 716
 717         * TestWebKitAPI/PlatformWebView.h:
 718         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
 719         (TestWebKitAPI::PlatformWebView::simulateMouseMove): Sends the messages that simulate a mouse move.
 720         * TestWebKitAPI/win/PlatformWebViewWin.cpp:
 721         (TestWebKitAPI::PlatformWebView::simulateMouseMove): Ditto.
 722
 723         Add new files.
 724         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
 725         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
 726         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
 727         * TestWebKitAPI/win/copy-resources.cmd:
 728
 729 2011-05-12  Tomasz Morawski  <t.morawski@samsung.com>
 730
 731         Reviewed by Antonio Gomes.
 732
 733         [EFL] Added initialization/shutdown for ecore_file library
 734         https://bugs.webkit.org/show_bug.cgi?id=60457
 735
 736         Added initialization and shuthown for ecore_file library and replaced
 737         realpath function to ecore_file_realpath.
 738
 739         * EWebLauncher/main.c:
 740         (quit):
 741         (findThemePath):
 742         (main):
 743
 744 2011-05-12  Young Han Lee  <joybro@company100.net>
 745
 746         Reviewed by Kenneth Rohde Christiansen.
 747
 748         [Qt] Implement layoutTestController.layerTreeAsText()
 749         https://bugs.webkit.org/show_bug.cgi?id=60367
 750
 751         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
 752         (LayoutTestController::layerTreeAsText):
 753         * DumpRenderTree/qt/LayoutTestControllerQt.h:
 754
 755 2011-05-11  Eric Seidel  <eric@webkit.org>
 756
 757         Reviewed by David Levin.
 758
 759         Teach validate-committer-lists how to validate committers bugzilla_email()
 760         https://bugs.webkit.org/show_bug.cgi?id=60651
 761
 762         It turns out we currently have 9 committers with invalid values for their
 763         bugzilla_email field (meaning we don't know what email is their bugzilla account).
 764         "Alexander Kellett" <lypanov@mac.com>
 765         "Andre Boule" <aboule@apple.com>
 766         "Cameron McCormack" <cam@webkit.org>
 767         "Chris Petersen" <cpetersen@apple.com>
 768         "Eli Fidler" <eli@staikos.net>
 769         "Matt Lilek" <webkit@mattlilek.com>
 770         "Pierre d'Herbemont" <pdherbemont@free.fr>
 771         "Scott Violet" <sky@chromium.org>
 772         "Adam Treat" <treat@kde.org>
 773
 774         Most of those haven't committed in a very long time, so it's not suprising
 775         that we haven't noticed that their bugzilla emails are bogus.
 776
 777         * Scripts/validate-committer-lists:
 778         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
 779         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
 780
 781 2011-05-11  Dmitry Lomov  <dslomov@google.com>
 782
 783         Reviewed by David Levin.
 784
 785         https://bugs.webkit.org/show_bug.cgi?id=59561
 786         Switch TestWebKitAPI to GTest
 787
 788         Fixes Makefiles and WinCairo build as well
 789
 790         * Makefile:
 791         * Scripts/build-api-tests:
 792         * Scripts/run-api-tests:
 793         * TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
 794         * TestWebKitAPI/Test.h:
 795         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
 796         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
 797         (TestWebKitAPI::TEST):
 798         * TestWebKitAPI/TestsController.cpp:
 799         (TestWebKitAPI::TestsController::TestsController):
 800         (TestWebKitAPI::TestsController::dumpTestNames):
 801         (TestWebKitAPI::TestsController::runTestNamed):
 802         (TestWebKitAPI::TestsController::runAllTests):
 803         * TestWebKitAPI/TestsController.h:
 804         * TestWebKitAPI/mac/main.mm:
 805         (main):
 806
 807 2011-05-11  Eric Seidel  <eric@webkit.org>
 808
 809         Reviewed by Alexey Proskuryakov.
 810
 811         Add Luke Macpherson to committers.py as a Contributor
 812         https://bugs.webkit.org/show_bug.cgi?id=60639
 813
 814         * Scripts/webkitpy/common/config/committers.py:
 815
 816 2011-05-11  Noam Rosenthal  <noam.rosenthal@nokia.com>
 817
 818         Reviewed by Kenneth Rohde Christiansen.
 819
 820         [Texmap][Qt] Upstream texture-mapper changes from Qt's WebKit2 branch
 821         https://bugs.webkit.org/show_bug.cgi?id=60439
 822
 823         Patch 4/12: Update the build script to enable 3D rendering by default for Qt.
 824
 825         * Scripts/build-webkit:
 826
 827 2011-05-11  Brian Weinstein  <bweinstein@apple.com>
 828
 829         Reviewed by Adam Roben.
 830
 831         REGRESSION (r86161): WKPreferences API test failing on Windows
 832         https://bugs.webkit.org/show_bug.cgi?id=60658
 833
 834         * TestWebKitAPI/Tests/WebKit2/WKPreferences.cpp:
 835         (TestWebKitAPI::TEST):
 836
 837 2011-05-11  Alice Boxhall  <aboxhall@chromium.org>
 838
 839         Reviewed by Ojan Vafai.
 840
 841         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.
 842         https://bugs.webkit.org/show_bug.cgi?id=60521
 843
 844         First steps towards making the results JSON all hierarchical.
 845
 846         * TestResultServer/model/jsonresults.py:
 847         * TestResultServer/model/jsonresults_unittest.py:
 848
 849 2011-05-11  Kevin Ollivier  <kevino@theolliviers.com>
 850
 851         [wx] Build fix, adding new LayoutTestController method stubs.
 852
 853         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
 854         (LayoutTestController::localStorageDiskUsageForOrigin):
 855         (LayoutTestController::applicationCacheDiskUsageForOrigin):
 856
 857 2011-05-11  Jessie Berlin  <jberlin@apple.com>
 858
 859         Reviewed by Darin Adler.
 860
 861         [WebKit2] exceededDatabaseQuota function on WKPageUIClient needs to provide information
 862         about the usage for both the origin overall and the database in question.
 863         https://bugs.webkit.org/show_bug.cgi?id=60573
 864
 865         * WebKitTestRunner/TestController.cpp:
 866         (WTR::exceededDatabaseQuota):
 867
 868 2011-05-10  Stephen White  <senorblanco@chromium.org>
 869
 870         Reviewed by David Levin.
 871
 872         Adding myself as a reviewer.
 873         https://bugs.webkit.org/show_bug.cgi?id=60587
 874
 875         * Scripts/webkitpy/common/config/committers.py:
 876
 877 2011-05-11  Eric Seidel  <eric@webkit.org>
 878
 879         Reviewed by Csaba Osztrogonác.
 880
 881         queues.webkit.org should show 1000+ instead of 1000 when we hit the query limit
 882         https://bugs.webkit.org/show_bug.cgi?id=60614
 883
 884         AppEngine has a result limit of 1000.  We hit that when trying
 885         to find out how many patches the style-queue (and other queues)
 886         have passed in the last 30 days.  So lets display 1000+ instead
 887         of 1000 to avoid confusion. (Ossy was asking about this in #webkit tonight.)
 888
 889         I also made the status pages show "never" instead of just "ago" when we
 890         haven't seen a pass or boot message in the last 1000 messages.
 891
 892         * QueueStatusServer/handlers/queuestatus.py:
 893         * QueueStatusServer/templates/queuestatus.html:
 894
 895 2011-05-11  Adam Barth  <abarth@webkit.org>
 896
 897         Reviewed by Eric Seidel.
 898
 899         sheriffbot shouldn't spam bugs when it can't figure out who to blame
 900         https://bugs.webkit.org/show_bug.cgi?id=60613
 901
 902         This patch limits the number of bugs sheriffbot will spam to three.
 903         Ideally, we'd come up with some smarter huerstics, but hopefully this
 904         will cut down on the bugmail spam for slow bots.
 905
 906         * Scripts/webkitpy/tool/commands/sheriffbot.py:
 907
 908 2011-05-10  Eric Seidel  <eric@webkit.org>
 909
 910         Reviewed by Adam Barth.
 911
 912         EWS bots should be robust against test-webkitpy hangs
 913         https://bugs.webkit.org/show_bug.cgi?id=60579
 914
 915         Added a killAfterTimeout subroutine, which takes a child_pid
 916         and timeout, and will setup a watchdog process to kill that child
 917         after the timeout.  killAfterTimeout will waitpid on the child.
 918
 919         * EWSTools/start-queue.sh:
 920
 921 2011-05-10  MORITA Hajime  <morrita@google.com>
 922
 923         Unreviewed, moved myself to the reviewers list.
 924
 925         * Scripts/webkitpy/common/config/committers.py:
 926
 927 2011-05-10  Anton D'Auria  <adauria@apple.com>
 928
 929         Reviewed by David Levin.
 930
 931         StorageTracker should report actual local storage usage on disk
 932         https://bugs.webkit.org/show_bug.cgi?id=60558
 933
 934         * DumpRenderTree/LayoutTestController.cpp:
 935         (localStorageDiskUsageForOriginCallback):
 936         (LayoutTestController::staticFunctions):
 937         * DumpRenderTree/LayoutTestController.h:
 938         * DumpRenderTree/chromium/LayoutTestController.cpp:
 939         (LayoutTestController::LayoutTestController):
 940         (LayoutTestController::localStorageDiskUsageForOrigin):
 941         * DumpRenderTree/chromium/LayoutTestController.h:
 942         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
 943         (LayoutTestController::localStorageDiskUsageForOrigin):
 944         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
 945         (LayoutTestController::localStorageDiskUsageForOrigin):
 946         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
 947         (LayoutTestController::localStorageDiskUsageForOrigin):
 948         * DumpRenderTree/qt/LayoutTestControllerQt.h:
 949         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
 950         (LayoutTestController::localStorageDiskUsageForOrigin):
 951
 952 2011-05-10  Anton D'Auria  <adauria@apple.com>
 953
 954         Reviewed by Alexey Proskuryakov.
 955
 956         Add ApplicationCache API to get disk usage per origin
 957         https://bugs.webkit.org/show_bug.cgi?id=57127
 958
 959         Added test that loads two frames with ~15Kb image, checks usage,
 960         deletes app cache, verifies usage is 0.
 961
 962         * DumpRenderTree/LayoutTestController.cpp:
 963         (applicationCacheDiskUsageForOriginCallback):
 964         (LayoutTestController::staticFunctions):
 965         * DumpRenderTree/LayoutTestController.h:
 966         * DumpRenderTree/chromium/LayoutTestController.cpp: Added stub.
 967         (LayoutTestController::LayoutTestController):
 968         (LayoutTestController::applicationCacheDiskUsageForOrigin):
 969         * DumpRenderTree/chromium/LayoutTestController.h:
 970         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: Added stub.
 971         (LayoutTestController::applicationCacheDiskUsageForOrigin):
 972         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
 973         (LayoutTestController::applicationCacheDiskUsageForOrigin):
 974         * DumpRenderTree/qt/LayoutTestControllerQt.cpp: Added stub.
 975         (LayoutTestController::applicationCacheDiskUsageForOrigin):
 976         * DumpRenderTree/qt/LayoutTestControllerQt.h:
 977         * DumpRenderTree/win/LayoutTestControllerWin.cpp: Added stub.
 978         (LayoutTestController::applicationCacheDiskUsageForOrigin):
 979
 980 2011-05-10  Lucas De Marchi  <lucas.demarchi@profusion.mobi>
 981
 982         Unreviewed, adding Rafael Antognolli as a committer.
 983
 984         * Scripts/webkitpy/common/config/committers.py:
 985
 986 2011-05-10  Chris Evans  <cevans@chromium.org>
 987
 988         Reviewed by Adam Barth.
 989
 990         Add WebCore::Setting to block displaying and/or running insecure content on secure pages
 991         https://bugs.webkit.org/show_bug.cgi?id=58378
 992
 993         * DumpRenderTree/chromium/LayoutTestController.cpp:
 994         (LayoutTestController::LayoutTestController):
 995         (LayoutTestController::setAllowDisplayOfInsecureContent):
 996         (LayoutTestController::setAllowRunningOfInsecureContent):
 997         * DumpRenderTree/chromium/LayoutTestController.h:
 998         * DumpRenderTree/chromium/WebPreferences.cpp:
 999         (WebPreferences::reset):
1000         (WebPreferences::applyTo):
1001         * DumpRenderTree/chromium/WebPreferences.h:
1002         Add plumbing for new settings to Chromium DRT.
1003
1004 2011-05-10  Enrica Casucci  <enrica@apple.com>
1005
1006         Reviewed by Beth Dakin.
1007
1008         Added my name to the reviewers list and removed my name from
1009         to committer-only list.
1010
1011         * Scripts/webkitpy/common/config/committers.py:
1012
1013 2011-05-10  David Levin  <levin@chromium.org>
1014
1015         Reviewed by Shinichiro Hamaji.
1016
1017         check-webkit-style shouldn't filter the errors using the modified lines for the xml and test expectations checkers.
1018         https://bugs.webkit.org/show_bug.cgi?id=60466
1019
1020         * Scripts/webkitpy/style/checker.py: (unrelated) style fix.
1021         * Scripts/webkitpy/style/checker_unittest.py: Remove the file_path check since
1022           that property was removed.
1023         * Scripts/webkitpy/style/checkers/test_expectations.py: Turn off the line
1024           filtering for errors.
1025         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Add testing to
1026           verify that the line filtering is turned off.
1027         * Scripts/webkitpy/style/checkers/xml.py: Turn off the line filtering for errors
1028           and remove the unused file_path parameter.
1029         * Scripts/webkitpy/style/checkers/xml_unittest.py: Add testing to verify that the
1030           line filtering is turned off (and fix test_no_error).
1031         * Scripts/webkitpy/style/error_handlers.py: Added a way to turn off the filtering
1032           of errors based on the lines changed in the patch.
1033         * Scripts/webkitpy/style/error_handlers_unittest.py: Added a test for turning off
1034           the filtering.
1035
1036 2011-05-10  Adam Roben  <aroben@apple.com>
1037
1038         Update some plugin tests to work with WebKit2's asynchronous windowed plugin positioning on Windows
1039
1040         Fixes <http://webkit.org/b/60442> <rdar://problem/9403200> REGRESSION (r85795):
1041         platform/win/plugins/window-geometry-initialized-before-set-window.html and
1042         platform/win/plugins/window-region-is-set-to-clip-rect.html failing on Windows 7 Release
1043         (WebKit2 Tests)
1044
1045         Reviewed by Anders Carlsson.
1046
1047         * DumpRenderTree/TestNetscapePlugIn/Tests/win/NPNInvalidateRectInvalidatesWindow.cpp:
1048         (NPNInvalidateRectInvalidatesWindow::NPP_SetWindow): Moved some code from here to the new
1049         WindowedPluginTest::testHarnessWindow function.
1050
1051         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowGeometryInitializedBeforeSetWindow.cpp:
1052         (WindowGeometryInitializedBeforeSetWindow::NPP_SetWindow): Only check the window's size, not
1053         its position. In WebKit2, the position won't have been set yet, so we can't test it. But it
1054         turns out only the size mattered for <http://webkit.org/b/46716>, which this test was
1055         created for.
1056
1057         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowRegionIsSetToClipRect.cpp:
1058         (WindowRegionIsSetToClipRect::WindowRegionIsSetToClipRect): Converted to be a
1059         WindowedPluginTest, and initialized new members.
1060         (WindowRegionIsSetToClipRect::startTest): Make our plugin window visible, and set things up
1061         to ensure we get a WM_WINDOWPOSCHANGED message about it becoming visible.
1062         (WindowRegionIsSetToClipRect::finishTest): Check that the window region is what we expect,
1063         then clean up and tell the test harness we're done.
1064         (WindowRegionIsSetToClipRect::checkWindowRegion): Moved code to check the window region here
1065         from NPP_SetWindow. Changed to use the fancy new log() function instead of pluginLog().
1066         (WindowRegionIsSetToClipRect::showTestHarnessWindowIfNeeded): Show the test harness's window
1067         if it is hidden. This is needed to ensure we get a WM_WINDOWPOSCHANGED message when our own
1068         window becomes visible.
1069         (WindowRegionIsSetToClipRect::hideTestHarnessWindowIfNeeded): Hide the test harness's window
1070         if it was hidden at the start of the test.
1071         (WindowRegionIsSetToClipRect::wndProc): When our timer fires, tell the UI process to paint.
1072         This should cause our window to become visible and our window region to get set. When we get
1073         a WM_WINDOWPOSCHANGED message, we can finally finish the test.
1074         (WindowRegionIsSetToClipRect::NPP_GetValue): Return our script object.
1075         (WindowRegionIsSetToClipRect::ScriptObject::hasMethod): We only have one method:
1076         "startTest".
1077         (WindowRegionIsSetToClipRect::ScriptObject::invoke): Call through to the PluginTest object.
1078
1079         * DumpRenderTree/TestNetscapePlugIn/win/WindowedPluginTest.cpp:
1080         (WindowedPluginTest::testHarnessWindow):
1081         * DumpRenderTree/TestNetscapePlugIn/win/WindowedPluginTest.h:
1082         Added this helper function. Code came from NPNInvalidateRectInvalidatesWindow.
1083
1084 2011-05-10  Csaba Osztrogonác  <ossy@webkit.org>
1085
1086         Reviewed by Kenneth Rohde Christiansen.
1087
1088         [Qt] Add new qt-4.8 platform for layout testing
1089         https://bugs.webkit.org/show_bug.cgi?id=60544
1090
1091         * Scripts/old-run-webkit-tests:
1092         * Scripts/webkitdirs.pm:
1093
1094 2011-05-09  Sheriff Bot  <webkit.review.bot@gmail.com>
1095
1096         Unreviewed, rolling out r86108.
1097         http://trac.webkit.org/changeset/86108
1098         https://bugs.webkit.org/show_bug.cgi?id=60533
1099
1100         Broke makefile based builds and Cairo (Requested by weinig on
1101         #webkit).
1102
1103         * Scripts/build-api-tests:
1104         * Scripts/run-api-tests:
1105         * TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
1106         * TestWebKitAPI/Test.h:
1107         (TestWebKitAPI::Test::~Test):
1108         (TestWebKitAPI::Test::name):
1109         (TestWebKitAPI::Test::Register::Register):
1110         (TestWebKitAPI::Test::Register::create):
1111         (TestWebKitAPI::Test::Test):
1112         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1113         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
1114         (TestWebKitAPI::TEST):
1115         * TestWebKitAPI/TestsController.cpp:
1116         (TestWebKitAPI::TestsController::TestsController):
1117         (TestWebKitAPI::TestsController::dumpTestNames):
1118         (TestWebKitAPI::TestsController::runTestNamed):
1119         (TestWebKitAPI::TestsController::testFailed):
1120         (TestWebKitAPI::TestsController::registerCreateTestFunction):
1121         * TestWebKitAPI/TestsController.h:
1122         * TestWebKitAPI/mac/main.mm:
1123         (main):
1124
1125 2011-05-09  Hayato Ito  <hayato@chromium.org>
1126
1127         Reviewed by Ojan Vafai.
1128
1129         Skip reftests if pixel tests are disabled.
1130
1131         https://bugs.webkit.org/show_bug.cgi?id=59078
1132
1133         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
1134         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
1135         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
1136         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1137
1138 2011-05-09  Chang Shu  <cshu@webkit.org>
1139
1140         Reviewed by Kenneth Rohde Christiansen.
1141
1142         [Qt] DumpRenderTreeQt needs an implementation of unmarkText
1143         https://bugs.webkit.org/show_bug.cgi?id=60499
1144
1145         The implementation is similar to gtk.
1146
1147         * DumpRenderTree/qt/TextInputControllerQt.cpp:
1148         (TextInputController::unmarkText):
1149         * DumpRenderTree/qt/TextInputControllerQt.h:
1150
1151 2011-05-09  James Robinson  <jamesr@chromium.org>
1152
1153         Reviewed by Eric Seidel.
1154
1155         [chromium] Teach new-run-webkit-tests to follow symlinks when trying to determine the arch of DumpRenderTree on linux
1156         https://bugs.webkit.org/show_bug.cgi?id=60520
1157
1158         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1159
1160 2011-05-09  Dmitry Lomov  <dslomov@google.com>
1161
1162         Reviewed by Sam Weinig.
1163
1164         Switch TestWebKitAPI to GTest
1165         https://bugs.webkit.org/show_bug.cgi?id=59561
1166
1167         * Scripts/build-api-tests:
1168         * Scripts/run-api-tests:
1169         * TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
1170         * TestWebKitAPI/Test.h:
1171         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
1172         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
1173         (TestWebKitAPI::TEST):
1174         * TestWebKitAPI/TestsController.cpp:
1175         (TestWebKitAPI::TestsController::TestsController):
1176         (TestWebKitAPI::TestsController::dumpTestNames):
1177         (TestWebKitAPI::TestsController::runTestNamed):
1178         (TestWebKitAPI::TestsController::runAllTests):
1179         * TestWebKitAPI/TestsController.h:
1180         * TestWebKitAPI/mac/main.mm:
1181         (main):
1182
1183 2011-05-09  Igor Oliveira  <igor.oliveira@openbossa.org>
1184
1185         Reviewed by Andreas Kling.
1186
1187         [Qt] DRT eventSender.addTouchPoint method is not setting the start position
1188         https://bugs.webkit.org/show_bug.cgi?id=60487
1189
1190         When adding a new touch point, the start position of this point needs to be set.
1191
1192         * DumpRenderTree/qt/EventSenderQt.cpp:
1193         (EventSender::addTouchPoint):
1194
1195 2011-05-09  Robert Hogan  <robert@webkit.org>
1196
1197         Reviewed by Adam Roben.
1198
1199         https://bugs.webkit.org/show_bug.cgi?id=57137
1200
1201         Unskip plugins/embed-prefers-plugins-for-images.html
1202
1203         Add image/png mime-type to test netscape plugin on Unix.
1204
1205         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
1206         (NP_GetMIMEDescription):
1207
1208 2011-05-09  Alice Boxhall  <aboxhall@chromium.org>
1209
1210         Reviewed by Ojan Vafai.
1211
1212         Pull jsonresults_unittest test cases out into separate test methods.
1213         https://bugs.webkit.org/show_bug.cgi?id=60418
1214
1215         * TestResultServer/model/jsonresults_unittest.py:
1216
1217 2011-05-09  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
1218
1219         Reviewed by Andreas Kling.
1220
1221         [Qt] fix fast/dom/Geolocation/window-close-crash.html
1222         https://bugs.webkit.org/show_bug.cgi?id=60458
1223
1224         Update all pages with Mock position/error/permission.
1225
1226         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1227         (LayoutTestController::setGeolocationPermission):
1228         (LayoutTestController::setMockGeolocationError):
1229         (LayoutTestController::setMockGeolocationPosition):
1230
1231 2011-05-09  Shishir Agrawal  <shishir@chromium.org>
1232
1233         Reviewed by Tony Gentilcore.
1234
1235         Implement Page Visibility API.
1236         https://bugs.webkit.org/show_bug.cgi?id=54181
1237
1238         * DumpRenderTree/LayoutTestController.cpp:
1239         (setPageVisibilityCallback):
1240         (resetPageVisibilityCallback):
1241         (LayoutTestController::staticFunctions):
1242         * DumpRenderTree/LayoutTestController.h:
1243         (LayoutTestController::setPageVisibility):
1244         (LayoutTestController::resetPageVisibility):
1245         * DumpRenderTree/chromium/LayoutTestController.cpp:
1246         (LayoutTestController::LayoutTestController):
1247         (LayoutTestController::resetPageVisibility):
1248         (LayoutTestController::setPageVisibility):
1249         * DumpRenderTree/chromium/LayoutTestController.h:
1250
1251 2011-05-07  Adam Barth  <abarth@webkit.org>
1252
1253         Reviewed by Eric Seidel.
1254
1255         Testing EWS spins on patches with a large number of failures
1256         https://bugs.webkit.org/show_bug.cgi?id=60441
1257
1258         In cases where the tree has a small number of persistent failures but a
1259         patch has a large number of failures, we weren't believing any results.
1260         That lead to us retrying the patch forever while we waited for
1261         trustworthy results that would never come.
1262
1263         This patch loosens the semantics of unexpected_failures to return a
1264         subset (rather than the exact set) of unexpected failures.  In the case
1265         where the tree has a bounded number of failures and the patch has an
1266         unbounded number of failures, we can accurately compute such a subset
1267         and reject the patch.
1268
1269         * Scripts/webkitpy/tool/bot/expectedfailures.py:
1270         * Scripts/webkitpy/tool/bot/expectedfailures_unittest.py:
1271         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
1272         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
1273         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
1274
1275 2011-05-07  Joe Mason  <jmason@rim.com>
1276
1277         Reviewed by Daniel Bates.
1278
1279         FrameLoader::isProcessingUserGesture is wrong in dispatchWillPerformClientRedirect
1280         https://bugs.webkit.org/show_bug.cgi?id=52211
1281
1282         Dump isUserProcessingGesture in willPerformClientRedirect so that the layout tests can be extended.
1283         (For the Mac and Chromium ports - other ports don't support dumping user gestures in DRT.)
1284
1285         * DumpRenderTree/chromium/WebViewHost.cpp:
1286         (WebViewHost::willPerformClientRedirect):
1287         (WebViewHost::didStartProvisionalLoad):
1288         * DumpRenderTree/mac/FrameLoadDelegate.mm:
1289         (-[FrameLoadDelegate webView:willPerformClientRedirectToURL:delay:fireDate:forFrame:]):
1290
1291 2011-05-07  Daniel Bates  <dbates@webkit.org>
1292
1293         Attempt to fix the build; remove some more references to isTiger(), which was removed in changeset 86007
1294         <http://trac.webkit.org/changeset/86007> (bug #60438).
1295
1296         * Scripts/old-run-webkit-tests:
1297           - Removed reference to isTiger() from openDumpTool(), captureSavedCrashLog(), and
1298             file local variable.
1299
1300 2011-05-07  Pratik Solanki  <psolanki@apple.com>
1301
1302         Reviewed by Darin Adler.
1303
1304         Remove references to Tiger in scripts
1305         https://bugs.webkit.org/show_bug.cgi?id=60438
1306
1307         Remove isTiger() and all of the code that depends on it.
1308
1309         * Scripts/build-webkit:
1310         * Scripts/ensure-valid-python:
1311         * Scripts/gdb-safari:
1312         * Scripts/webkitdirs.pm:
1313
1314 2011-05-06  David Grogan  <dgrogan@chromium.org>
1315
1316         Reviewed by Eric Seidel.
1317
1318         update link to message about mac 64bit git binaries
1319         https://bugs.webkit.org/show_bug.cgi?id=60411
1320
1321         The original message implied (correctly, at-the-time) that
1322         http://code.google.com/p/git-osx-installer/ doesn't provide 64-bit
1323         binaries. The new message notes that they do. This might prevent someone
1324         else from spending a few minutes confused.
1325
1326         * Scripts/webkitpy/common/checkout/scm/git.py:
1327
1328 2011-05-06  Dirk Pranke  <dpranke@chromium.org>
1329
1330         Reviewed by Ojan Vafai.
1331
1332         new-run-webkit-tests: crashes in race with crashing DRT
1333         https://bugs.webkit.org/show_bug.cgi?id=60238
1334
1335         This fixes a race where if we wrote the URL to DRT and
1336         DRT crashed immediately, we could get a broken pipe
1337         error back during the write() and not handle it correctly,
1338         causing NRWT to also crash.
1339
1340         * Scripts/webkitpy/layout_tests/port/server_process.py:
1341         * Scripts/webkitpy/layout_tests/port/webkit.py:
1342
1343 2011-05-06  Adam Barth  <abarth@webkit.org>
1344
1345         Reviewed by Eric Seidel.
1346
1347         Add a --skip-integrationtests mode to test-webkitpy
1348         https://bugs.webkit.org/show_bug.cgi?id=60354
1349
1350         These two test suites account for 2/3 of test-webkitpy's running time
1351         because they're not actually unit tests.  They're integration tests.
1352         This patch adds the ability to skip integration tests by passing a
1353         command line argument to test-webkitpy.  The buildbots will continue to
1354         run these integration tests to make sure they don't regress.  This
1355         option is for use by developers when working on parts of the code that
1356         are well-covered by unit tests.
1357
1358         * Scripts/webkitpy/common/system/file_lock_integrationtest.py: Copied from Tools/Scripts/webkitpy/common/system/file_lock_unittest.py.
1359         * Scripts/webkitpy/common/system/file_lock_unittest.py: Removed.
1360         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py: Copied from Tools/Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py.
1361         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py: Removed.
1362         * Scripts/webkitpy/test/main.py:
1363
1364 2011-05-06  Adam Barth  <abarth@webkit.org>
1365
1366         Reviewed by Eric Seidel.
1367
1368         The results.zip generated by the Chromium EWS are too big
1369         https://bugs.webkit.org/show_bug.cgi?id=60352
1370
1371         Currently results.zip created by the Chromium EWS bot contain roughly
1372         1000 tests that we expect to fail.  These test results are not helpful
1373         to contributors.
1374
1375         There are several approaches to solving this problem:
1376
1377         1) We could be more selective about which files to include in the zip.
1378         2) We could ask new-run-webkit-tests not to generate results for
1379            expected failures.
1380         3) We could ask new-run-webkit-tests not to run tests that we expect to
1381            fail.
1382
1383         This patch takes approach (3).  Currently, we're ignoring the outcome
1384         of failing tests anyway, so there's not much point in actually running
1385         them.  As a side benefit, this will speed up the test run by about 6%.
1386         In the future, we might revisit this decision and choose another
1387         approach.
1388
1389         * Scripts/webkitpy/common/config/ports.py:
1390         * Scripts/webkitpy/common/config/ports_unittest.py:
1391
1392 2011-05-06  Gabor Rapcsanyi  <rgabor@webkit.org>
1393
1394         Reviewed by Csaba Osztrogonác.
1395
1396         [Qt] Use qt-arm platform instead of qt-linux on ARM platform.
1397
1398         * Scripts/old-run-webkit-tests:
1399
1400 2011-05-06  Adam Barth  <abarth@webkit.org>
1401
1402         Reviewed by Ojan Vafai.
1403
1404         Add an option to NRWT to skip failing tests
1405         https://bugs.webkit.org/show_bug.cgi?id=60295
1406
1407         This option is similar to the way old-run-webkit-tests works.  We're
1408         planning to use this option on the testing EWS bots because they don't
1409         use information from failing tests yet.
1410
1411         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
1412         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1413         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
1414
1415 2011-05-04  Philippe Normand  <pnormand@igalia.com>
1416
1417         Reviewed by Dimitri Glazkov.
1418
1419         Implement LayoutTestController::pseudoShadowId()
1420         https://bugs.webkit.org/show_bug.cgi?id=60034
1421
1422         shadowPseudoId support in the various LayoutTestController
1423         implementations.
1424
1425         * DumpRenderTree/LayoutTestController.cpp:
1426         (shadowPseudoIdCallback):
1427         (LayoutTestController::staticFunctions):
1428         * DumpRenderTree/LayoutTestController.h:
1429         * DumpRenderTree/chromium/LayoutTestController.cpp:
1430         (LayoutTestController::LayoutTestController):
1431         (LayoutTestController::shadowPseudoId):
1432         * DumpRenderTree/chromium/LayoutTestController.h:
1433         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1434         (LayoutTestController::shadowPseudoId):
1435         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
1436         (LayoutTestController::shadowPseudoId):
1437         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1438         (LayoutTestController::shadowPseudoId):
1439         * DumpRenderTree/qt/LayoutTestControllerQt.h:
1440         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
1441         (LayoutTestController::shadowPseudoId):
1442         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
1443         (LayoutTestController::shadowPseudoId):
1444         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
1445         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1446         (WTR::LayoutTestController::shadowPseudoId):
1447         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
1448
1449 2011-05-06  Adam Barth  <abarth@webkit.org>
1450
1451         EWS can't reject patches
1452         https://bugs.webkit.org/show_bug.cgi?id=60345
1453
1454         We're supposed to use patch.id() for the attachment ID, silly.
1455
1456         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
1457         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
1458
1459 2011-05-05  Steve Lacey  <sjl@chromium.org>
1460
1461         Reviewed by Ojan Vafai.
1462
1463         Fix generation of results html file by passing relative path to scm
1464         methods.
1465         Generate diff images for results html file.
1466
1467         https://bugs.webkit.org/show_bug.cgi?id=60086
1468
1469         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
1470         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
1471
1472 2011-05-05  Andy Estes  <aestes@apple.com>
1473
1474         Reviewed by Daniel Bates.
1475
1476         update-webkit should handle Internal using a different VCS than OpenSource.
1477         https://bugs.webkit.org/show_bug.cgi?id=60310
1478
1479         isGit() and isSVN() cache their results, so Internal will always be
1480         updated using the VCS commands used for OpenSource. If Internal uses a
1481         different VCS, this will fail. Fix this by calling isGitDirectory() and
1482         isSVNDirectory() instead, which do not cache their results.
1483
1484         * Scripts/update-webkit:
1485
1486 2011-05-05  Dirk Pranke  <dpranke@chromium.org>
1487
1488         Reviewed by Ojan Vafai.
1489
1490         nrwt: revamp integration tests incl. tests for servers
1491         https://bugs.webkit.org/show_bug.cgi?id=59985
1492
1493         This change adds the ability to run "integration tests" that
1494         depend on (and affect) the local environment on your machine.
1495         These tests can be run manually by invoking python on the
1496         test file in question and passing the '-i' flag, but are
1497         not run automatically as part of test-webkitpy.
1498
1499         Integration tests are identified by being methods named as
1500         "integration_test_*" instead of "test_*", and we use a
1501         custom TestLoader instead of the default one provided by
1502         unittest.
1503
1504         This change also adds a protected "_default_to_apache" method
1505         to the port object so we can do the right thing when starting
1506         and stopping web servers with less mechanism.
1507
1508         * Scripts/webkitpy/layout_tests/port/base.py:
1509         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
1510         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
1511         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1512         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1513         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
1514
1515 2011-05-05  Adam Roben  <aroben@apple.com>
1516
1517         Robustify DRT's pixel-dumping code against semi-random alpha channel values on Windows
1518
1519         DRT generates a pixel dump on Windows using the WM_PRINT message. This message (just like
1520         most of GDI) doesn't write anything meaningful into the alpha channel. In some cases, the
1521         alpha channel will even get filled with zeros (like when ::GradientFill is used). To
1522         safeguard against this, we fill the alpha channel with constant full opacity after the
1523         WM_PRINT message is handled to ensure we end up with an opaque image. (Note that we can't
1524         just ignore the alpha channel when creating the PNG; Mac's pixel results have an alpha
1525         channel, and we want to be able to compare Mac and Windows pixel results, so Windows pixel
1526         results need an alpha channel, too.)
1527
1528         Fixes <http://webkit.org/b/60300> draws-gradient.html's pixel results can't be correctly
1529         generated due to alpha channel weirdness
1530
1531         Reviewed by Sam Weinig.
1532
1533         * DumpRenderTree/win/PixelDumpSupportWin.cpp:
1534         (makeAlphaChannelOpaque): Added this helper function to fill a bitmap's alpha channel with
1535         constant full opacity.
1536         (createBitmapContextFromWebView): Make the bitmap's alpha channel opaque before wrapping it
1537         in a CGContext or cairo_surface.
1538
1539 2011-05-05  Adam Roben  <aroben@apple.com>
1540
1541         Add support for layoutTestController.dumpAsText(true) to WebKitTestRunner
1542
1543         Passing true tells the test harness to dump pixels even though we're not dumping a render
1544         tree.
1545
1546         Fixes <http://webkit.org/b/60298> Calling layoutTestController.dumpAsText(true) in
1547         WebKitTestRunner doesn't result in pixel results being generated
1548
1549         Reviewed by Sam Weinig.
1550
1551         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl: Added dumpPixels
1552         argument to dumpAsText.
1553
1554         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
1555         (WTR::InjectedBundlePage::dump): Updated for dumpAsText change.
1556
1557         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
1558         (WTR::LayoutTestController::dumpAsText): Store the dumpPixels argument.
1559
1560         * WebKitTestRunner/InjectedBundle/LayoutTestController.h: Updated for IDL change.
1561
1562         * WebKitTestRunner/win/InjectedBundle.vcproj: Added .idl files as a convenience for Windows
1563         developers.
1564
1565 2011-05-05  Dirk Pranke  <dpranke@chromium.org>
1566
1567         Reviewed by Ojan Vafai.
1568
1569         new-run-webkit-tests: merge os, version fields in test config
1570         https://bugs.webkit.org/show_bug.cgi?id=60004
1571
1572         This change removes the 'o/s' field from the TestConfiguration
1573         object, and changes the 'MAC', 'WIN', and 'LINUX' modifiers in
1574         the expectations file to macros that expand to the version names
1575         for each platform.
1576
1577         This means that you can once again put any subset of the
1578         operating system versions on a single line, e.g.:
1579           BUGX LEOPARD XP LINUX : fast/html/foo.html = FAIL
1580
1581         This change will break any existing test_expectations.txt
1582         file that uses the "OS-VERSION" macros, which are no longer
1583         needed. It turns out they were never needed, as just VERSION
1584         by itself was sufficient. The test_expectations.txt file should
1585         be updated accordingly prior to landing this change.
1586
1587         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
1588         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
1589         * Scripts/webkitpy/layout_tests/port/base.py:
1590
1591 2011-05-05  Tony Chang  <tony@chromium.org>
1592
1593         Reviewed by Adam Barth.
1594
1595         [chromium] switch from Lucida to free fonts for Hindi and Thai
1596         https://bugs.webkit.org/show_bug.cgi?id=60287
1597
1598         The fonts are provided by the ttf-thai-tlwg and ttf-indic-fonts
1599         packages on Lucid. There are actually multiple Thai fonts in the
1600         package, so I just picked one.
1601
1602         * DumpRenderTree/chromium/TestShellGtk.cpp:
1603         (setupFontconfig):
1604
1605 2011-05-05  Tony Chang  <tony@chromium.org>
1606
1607         Reviewed by Ojan Vafai.
1608
1609         [chromium] google-chrome-64 builds should now fallback to chromium-linux-x86_64 results
1610         https://bugs.webkit.org/show_bug.cgi?id=60290
1611
1612         It used to be that google-chrome-32 was hardy32, google-chrome-64 was
1613         hardy64, chromium-linux was hardy32 and chromium-linux-x86_64 was
1614         lucid64.  Because of that, google-chrome-64 used to skip the
1615         chromium-linux-x86_64 results since they were more incorrect.
1616
1617         Now that we're switching to lucid for google-chrome, we want
1618         google-chrome-64 to fall back to chromium-linux-x86_64.
1619
1620         * Scripts/webkitpy/layout_tests/port/google_chrome.py:
1621
1622 2011-05-05  Dirk Pranke  <dpranke@chromium.org>
1623
1624         Reviewed by Ojan Vafai.
1625
1626         new-run-webkit-tests: merge TestExpectations, TestExpectationsFile classes
1627         https://bugs.webkit.org/show_bug.cgi?id=60002
1628
1629         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
1630         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
1631         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
1632         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
1633         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
1634         * Scripts/webkitpy/style/checkers/test_expectations.py:
1635
1636 2011-05-05  Eric Seidel  <eric@webkit.org>
1637
1638         Reviewed by Adam Barth.
1639
1640         commit-queue should reject patches which fail ewses
1641         https://bugs.webkit.org/show_bug.cgi?id=47534
1642
1643         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
1644         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
1645         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
1646
1647 2011-05-05  Chang Shu  <cshu@webkit.org>
1648
1649         Reviewed by Andreas Kling.
1650
1651         [Qt][WK2] Support PageGroup in QGraphicsWKView/QWKPage constructors
1652         https://bugs.webkit.org/show_bug.cgi?id=59638
1653
1654         1. Call QGraphicsWKView constructor with pageGroupRef.
1655         2. Disable setFontFamily calls as it breaks Qt test fonts.
1656
1657         * WebKitTestRunner/TestController.cpp:
1658         (WTR::TestController::resetStateToConsistentValues):
1659         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
1660         (WTR::WebView::WebView):
1661         (WTR::PlatformWebView::PlatformWebView):
1662
1663 2011-05-05  Joseph Pecoraro  <joepeck@webkit.org>
1664
1665         Reviewed by David Kilzer.
1666
1667         Add --only-webkit option to build-webkit
1668         https://bugs.webkit.org/show_bug.cgi?id=60245
1669
1670         * Scripts/build-webkit: If --only-webkit was specified
1671         just build the WebKit project, and not its dependencies.
1672
1673 2011-05-05  Jeff Miller  <jeffm@apple.com>
1674
1675         Reviewed by Darin Adler.
1676
1677         Add didNotHandleWheelEvent callback to WKPageUIClient
1678         https://bugs.webkit.org/show_bug.cgi?id=60203
1679
1680         * MiniBrowser/gtk/BrowserWindow.c:
1681         (browserWindowUIClientInit): Stub out didNotHandleWheelEvent callback.
1682         * MiniBrowser/mac/BrowserWindowController.m:
1683         (-[BrowserWindowController awakeFromNib]): Stub out didNotHandleWheelEvent callback.
1684         * MiniBrowser/win/BrowserView.cpp:
1685         (BrowserView::create): Stub out didNotHandleWheelEvent callback.
1686         * WebKitTestRunner/TestController.cpp:
1687         (WTR::TestController::createOtherPage): Stub out didNotHandleWheelEvent callback.
1688         (WTR::TestController::initialize): Stub out didNotHandleWheelEvent callback.
1689
1690 2011-05-04  Eric Seidel  <eric@webkit.org>
1691
1692         Reviewed by Adam Barth.
1693
1694         Add more missing imports from scm refactor
1695         https://bugs.webkit.org/show_bug.cgi?id=60239
1696
1697         * Scripts/webkitpy/common/checkout/scm/git.py:
1698         * Scripts/webkitpy/common/checkout/scm/svn.py:
1699
1700 2011-05-04  Eric Seidel  <eric@webkit.org>
1701
1702         Reviewed by Adam Barth.
1703
1704         sheriff-bot is having more exception trouble
1705         https://bugs.webkit.org/show_bug.cgi?id=60233
1706
1707         A webpage-forced build was causing SB to throw an exception.
1708         We've long had this bug, but we just didn't notice it until now.
1709
1710         In debugging this I found we fetch an absurd amount of data now
1711         that we use buildbot's json (instead of xmlrpc), so I added
1712         the filter=1 parameter to reduce it a little.  That required
1713         me to handle the case where filter=1 would strip
1714         result=0 in some cases (it's unclear why it does that).
1715
1716         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
1717         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
1718
1719 2011-05-04  Cris Neckar  <cdn@chromium.org>
1720
1721         Unreviewed, Adding Cris Neckar as a committer
1722
1723         * Scripts/webkitpy/common/config/committers.py:
1724
1725 2011-05-04  Siddharth Mathur  <siddharth.mathur@nokia.com>
1726
1727         Reviewed by Laszlo Gombos.
1728
1729         [Qt][Wk2][Symbian] Capabilities for Qt MiniBrowser and QtWebProcess on Symbian
1730         https://bugs.webkit.org/show_bug.cgi?id=59462
1731
1732         * MiniBrowser/qt/MiniBrowser.pro: Add Symbian OS capabilities to executable. Also enable data paging.
1733
1734 2011-05-04  Eric Seidel  <eric@webkit.org>
1735
1736         Reviewed by Adam Barth.
1737
1738         The testing EWS bots should upload layout-test-results.zip to bugs after failures
1739         https://bugs.webkit.org/show_bug.cgi?id=60223
1740
1741         This required sharing a bit of code between the commit-queue
1742         (which already knew how to do this) and the new EWS testing bots.
1743
1744         In the process I also cleaned up EWS testing a little and
1745         removed some dead code from the commit-queue.
1746
1747         * Scripts/webkitpy/common/config/committervalidator.py:
1748         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
1749         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
1750         * Scripts/webkitpy/tool/commands/queues.py:
1751         * Scripts/webkitpy/tool/commands/queues_unittest.py:
1752
1753 2011-05-04  James Kozianski  <koz@chromium.org>
1754
1755         Reviewed by Eric Seidel.
1756
1757         Add method to make a Port able to retrieve TestOutputSets from its builders.
1758         https://bugs.webkit.org/show_bug.cgi?id=58211
1759
1760         This will be used in the new rebaseline script - TestOutputs are
1761         potential new baselines.
1762
1763         * Scripts/webkitpy/common/config/urls.py:
1764         * Scripts/webkitpy/layout_tests/port/base.py:
1765         * Scripts/webkitpy/layout_tests/port/chromium.py:
1766         * Scripts/webkitpy/layout_tests/port/factory.py:
1767         * Scripts/webkitpy/layout_tests/port/webkit.py:
1768
1769 2011-05-04  Eric Seidel  <eric@webkit.org>
1770
1771         Reviewed by Ojan Vafai.
1772
1773         sheriff-bot can't parse results.html
1774         https://bugs.webkit.org/show_bug.cgi?id=60196
1775
1776         The problem was that the NRWT bots are currently core
1777         and so sheriff-bot was trying to diagnose failures on
1778         them.  It was trying to parse their results.html file
1779         and BeautifulSoup was throwing an exception.
1780
1781         It seems our version of BeautifulSoup was quite old
1782         and unable to handle <script> tags correctly.
1783         This moves us to 3.2 which doesn't seem to have this bug.
1784
1785         * Scripts/webkitpy/thirdparty/BeautifulSoup.py:
1786
1787 2011-05-04  Adam Barth  <abarth@webkit.org>
1788
1789         Reviewed by Eric Seidel.
1790
1791         Make the EWS test logging less verbose
1792         https://bugs.webkit.org/show_bug.cgi?id=60201
1793
1794         This should help us generate less than a hundred MB of logs per hour.
1795
1796         * Scripts/webkitpy/common/config/ports.py:
1797         * Scripts/webkitpy/common/config/ports_unittest.py:
1798
1799 2011-05-04  Tao Bai  <michaelbai@chromium.org>
1800
1801         Reviewed by David Kilzer.
1802
1803         Populate touch-icon url to FrameLoaderClient
1804         https://bugs.webkit.org/show_bug.cgi?id=59143
1805
1806         Disable touch icon by default.
1807
1808         * Scripts/build-webkit:
1809
1810 2011-05-04  Justin Novosad  <junov@chromium.org>
1811
1812         Reviewed by Kenneth Russell.
1813
1814         [Chromium] Add gpu acceleration flags to DumpRenderTree
1815         https://bugs.webkit.org/show_bug.cgi?id=60168
1816         Adding support for command line flags
1817         enable-accelerated-2d-canvas an
1818         enable-accelerated-drawing in DumpRenderTree.
1819
1820         * DumpRenderTree/chromium/DumpRenderTree.cpp:
1821         (main):
1822         * DumpRenderTree/chromium/TestShell.cpp:
1823         (TestShell::TestShell):
1824         (TestShell::resetWebSettings):
1825         * DumpRenderTree/chromium/TestShell.h:
1826         (TestShell::setLegacyAccelerated2dCanvasEnabled):
1827         (TestShell::setAcceleratedDrawingEnabled):
1828         * DumpRenderTree/chromium/WebPreferences.cpp:
1829         (WebPreferences::reset):
1830         (WebPreferences::applyTo):
1831         * DumpRenderTree/chromium/WebPreferences.h:
1832
1833 2011-05-04  Ademar de Souza Reis Jr.  <ademar.reis@openbossa.org>
1834
1835         Reviewed by Csaba Osztrogonác.
1836
1837         [Qt] run-qtwebkit-tests misses exec bit
1838         https://bugs.webkit.org/show_bug.cgi?id=60188
1839
1840         Also removing an unecessary import (mostly to let this bugfix goes
1841         through the system, as this is just a chmod).
1842
1843         * Scripts/run-qtwebkit-tests:
1844
1845 2011-05-04  Adam Barth  <abarth@webkit.org>
1846
1847         Fix import of ircproxy module to actually import the module.  I don't
1848         really understand how this code could have worked before.
1849
1850         * Scripts/webkitpy/common/host.py:
1851
1852 2011-05-04  Siddharth Mathur  <siddharth.mathur@nokia.com>
1853
1854         Reviewed by Andreas Kling.
1855
1856         [Qt][Wk2][Symbian] Minor UI facelift for MiniBrowser
1857         https://bugs.webkit.org/show_bug.cgi?id=60172
1858
1859         * MiniBrowser/qt/BrowserWindow.cpp: Bring in UI layout changes from QtTestBrowser
1860         (BrowserWindow::BrowserWindow):
1861
1862 2011-05-04  Mihai Parparita  <mihaip@chromium.org>
1863
1864         Reviewed by Ojan Vafai.
1865
1866         [NRWT] --new-baseline should put baselines in the directory specified by --additional-platform-directory
1867         https://bugs.webkit.org/show_bug.cgi?id=60088
1868
1869         Makes --new-baseline put new baselines into the first entry in the
1870         baseline search path, including any additional directories passed in via
1871         --additional-platform-directory.
1872
1873         Simplifies baseline_path() by making it return the first entry in the
1874         baseline search path instead of requiring each port to have an
1875         implementation (should not result in any behavior change).
1876
1877         * Scripts/webkitpy/layout_tests/port/base.py:
1878         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
1879         * Scripts/webkitpy/layout_tests/port/chromium.py:
1880         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
1881         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
1882         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
1883         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1884         * Scripts/webkitpy/layout_tests/port/mac.py:
1885         * Scripts/webkitpy/layout_tests/port/test.py:
1886         * Scripts/webkitpy/layout_tests/port/webkit.py:
1887
1888 2011-05-04  Adam Barth  <abarth@webkit.org>
1889
1890         Reviewed by Eric Seidel.
1891
1892         Fix circular dependency in webkitpy
1893         https://bugs.webkit.org/show_bug.cgi?id=60075
1894
1895         These functions don't belong in bugzilla.py.  They only exist there
1896         because they are old.  Really, these functions shouldn't be free
1897         functions at all, but that's a patch for another day.
1898
1899         * Scripts/webkitpy/common/checkout/changelog.py:
1900         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
1901         * Scripts/webkitpy/common/checkout/checkout.py:
1902         * Scripts/webkitpy/common/config/urls.py:
1903         * Scripts/webkitpy/common/net/bugzilla/__init__.py:
1904         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
1905         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
1906         * Scripts/webkitpy/style/checkers/changelog.py:
1907         * Scripts/webkitpy/tool/bot/irc_command.py:
1908         * Scripts/webkitpy/tool/bot/sheriff.py:
1909         * Scripts/webkitpy/tool/commands/upload.py:
1910
1911 2011-05-03  Pratik Solanki  <psolanki@apple.com>
1912
1913         Reviewed by Dan Bernstein.
1914
1915         https://bugs.webkit.org/show_bug.cgi?id=60141
1916         Remove reference to libWebKitSystemInterfaceTiger.a from build-webkit
1917
1918         * Scripts/build-webkit:
1919
1920 2011-05-03  Mihai Parparita  <mihaip@chromium.org>
1921
1922         Reviewed by Ojan Vafai.
1923
1924         Update rebaseline queue server to handle hierarchical test results
1925         https://bugs.webkit.org/show_bug.cgi?id=60063
1926
1927         Update test result parsing done by the rebaseline queue server to handle
1928         the hierarchical test result output that is generated as of r85254.
1929
1930         * RebaselineQueueServer/templates/builder-queue-edit.html:
1931
1932 2011-05-03  Eric Seidel  <eric@webkit.org>
1933
1934         Reviewed by Adam Roben.
1935
1936         webkit-patch post-commits HEAD^.. throws an exception
1937         https://bugs.webkit.org/show_bug.cgi?id=60027
1938
1939         * Scripts/webkitpy/common/checkout/scm/git.py:
1940
1941 2011-05-03  Luiz Agostini  <luiz.agostini@openbossa.org>
1942
1943         Reviewed by Kenneth Rohde Christiansen.
1944
1945         [Qt] Layout tests may timeout if loadFinished receives false
1946         https://bugs.webkit.org/show_bug.cgi?id=60064
1947
1948         When using layoutTestController.queueLoad(), if QWebPage::loadFinished(bool) receives false
1949         then the layout test may timout. It happens that QWebPage::loadFinished() receives false when
1950         a response is ignored, like it should be for 204 responses. It is a problem for creating
1951         tests for 204 responses for example.
1952
1953         This change has been tested and it causes no regressions in layout tests.
1954
1955         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1956         (LayoutTestController::maybeDump):
1957
1958 2011-05-03  Dirk Pranke  <dpranke@chromium.org>
1959
1960         Unreviewed, rolling out r85659.
1961         http://trac.webkit.org/changeset/85659
1962         https://bugs.webkit.org/show_bug.cgi?id=59977
1963
1964         restarting isn't working correctly
1965
1966         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
1967         * Scripts/webkitpy/layout_tests/port/http_server.py:
1968         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
1969         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
1970
1971 2011-05-03  Dirk Pranke  <dpranke@chromium.org>
1972
1973         Reviewed by Ojan Vafai.
1974
1975         new-run-webkit-tests: fix http server startup/shutdown
1976
1977         Previous versions of the code had three problems that made startup
1978         and shutdown flaky. The first is that it would throw exceptions
1979         if it couldn't delete stale log files, which was overly
1980         paranoid. The second is that some of the exceptions weren't
1981         defined properly. The third, and most important, is that it was
1982         using urllib to check if ports were available, which was leaving
1983         sockets in a half-closed state, and keeping ports from being
1984         reused. By switching to raw sockets, we are able to now reliably
1985         restart.
1986
1987         This change also switches the code to using Executives to stop
1988         processes, which will let us delete a bunch of code in the
1989         port/* implementations and fix a weird layering problem in a
1990         subsequent patch.
1991
1992         https://bugs.webkit.org/show_bug.cgi?id=59977
1993
1994         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
1995         * Scripts/webkitpy/layout_tests/port/http_server.py:
1996         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
1997         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
1998
1999 2011-05-03  Dirk Pranke  <dpranke@chromium.org>
2000
2001         Reviewed by Adam Barth.
2002
2003         fix failing printing_unittest on win32
2004         https://bugs.webkit.org/show_bug.cgi?id=59963
2005
2006         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
2007
2008 2011-05-03  Adam Roben  <aroben@apple.com>
2009
2010         Fix most strict PassOwnPtr violations on Windows
2011
2012         Fixes <http://webkit.org/b/60025> Windows should (almost) build with strict PassOwnPtr
2013         enabled
2014
2015         Reviewed by Anders Carlsson.
2016
2017         * WebKitAPITest/Test.h: Use adoptPtr.
2018
2019 2011-05-03  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
2020
2021         Reviewed by Andreas Kling.
2022
2023         Add myself to the committers list.
2024
2025         * Scripts/webkitpy/common/config/committers.py:
2026
2027 2011-05-03  James Robinson  <jamesr@chromium.org>
2028
2029         Reviewed by Kenneth Russell.
2030
2031         [chromium] Layout test animations/3d/replace-filling-transform.html failing
2032         https://bugs.webkit.org/show_bug.cgi?id=59839
2033
2034         Add animations/3d to the list of tests run in the chromium-gpu configuration since these tests exercise the compositor.
2035
2036         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2037
2038 2011-05-03  Lucas De Marchi  <lucas.demarchi@profusion.mobi>
2039
2040         Unreviewed, rolling out r85471.
2041         http://trac.webkit.org/changeset/85471
2042         https://bugs.webkit.org/show_bug.cgi?id=59922
2043
2044         The rule of thumb in EFL is that you init in all the places
2045         you use it. So, if in your application you call ecore_evas_*
2046         functions, you have to call ecore_evas_init() first.
2047
2048         * EWebLauncher/main.c: always call *_init() functions when you use
2049         that namespace.
2050         (quit):
2051         (main):
2052
2053 2011-05-02  Ojan Vafai  <ojan@chromium.org>
2054
2055         Reviewed by Tony Chang.
2056
2057         remove no longer used method
2058         https://bugs.webkit.org/show_bug.cgi?id=59981
2059
2060         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2061         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
2062
2063 2011-05-02  Ojan Vafai  <ojan@chromium.org>
2064
2065         Reviewed by Tony Chang.
2066
2067         add image toggling behavior from old-run-webkit-tests
2068         https://bugs.webkit.org/show_bug.cgi?id=59978
2069
2070         Output -diffs.html files. This matches old-run-webkit-tests.
2071         In theory, we could use data urls in the results.html file, but
2072         data urls don't currently have access to local files.
2073
2074         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
2075
2076 2011-05-02  Tony Chang  <tony@chromium.org>
2077
2078         Reviewed by Ojan Vafai.
2079
2080         [nrwt] remove some obsolete references to .checksum files
2081         https://bugs.webkit.org/show_bug.cgi?id=59971
2082
2083         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
2084         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
2085         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
2086         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
2087         * Scripts/webkitpy/layout_tests/port/test.py:
2088         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
2089
2090 2011-05-02  Adam Roben  <aroben@apple.com>
2091
2092         Take advantage of implicit nullptr_t -> PassOwnPtr conversion
2093
2094         Fixes <http://webkit.org/b/59964> Implicit conversion from std::nullptr_t to PassOwnPtr
2095         doesn't work, but should
2096
2097         Reviewed by Adam Barth.
2098
2099         * DumpRenderTree/chromium/CppBoundClass.h:
2100         * DumpRenderTree/chromium/WebViewHost.cpp:
2101
2102 2011-05-02  Martin Robinson  <mrobinson@igalia.com>
2103
2104         Reviewed by Xan Lopez.
2105
2106         [GTK] Compiler warning in Tools/MiniBrowser/main.c
2107         https://bugs.webkit.org/show_bug.cgi?id=59986
2108
2109         Remove a compiler warning.
2110
2111         * MiniBrowser/gtk/main.c: Use full array initializers to prevent some compilers
2112         from complaining.
2113
2114 2011-05-02  Adam Barth  <abarth@webkit.org>
2115
2116         Reviewed by Eric Seidel.
2117
2118         Add runtests_unittest.py
2119         https://bugs.webkit.org/show_bug.cgi?id=59955
2120
2121         I forgot to add this file in a previous patch.
2122
2123         * Scripts/webkitpy/tool/steps/runtests_unittest.py: Added.
2124
2125 2011-05-02  Adam Barth  <abarth@webkit.org>
2126
2127         Change the testing EWS to log failures explicitly to avoid an infinite loop.
2128
2129         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
2130
2131 2011-05-02  Adrienne Walker  <enne@google.com>
2132
2133         Unreviewed, fix misspelled contributor name.
2134
2135         * Scripts/webkitpy/common/config/committers.py:
2136
2137 2011-05-02  Adam Barth  <abarth@webkit.org>
2138
2139         Reviewed by Eric Seidel.
2140
2141         cr-linux-ews should run tests
2142         https://bugs.webkit.org/show_bug.cgi?id=59272
2143
2144         This patch turns on testing for the cr-linux-ews.  Yay!
2145
2146         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
2147         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
2148         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
2149         * Scripts/webkitpy/tool/steps/runtests.py:
2150         * Scripts/webkitpy/tool/steps/steps_unittest.py:
2151
2152 2011-05-02  Martin Robinson  <mrobinson@igalia.com>
2153
2154         Incorporate a review comment that I mistakenly omitted from my last commit.
2155
2156         * DumpRenderTree/gtk/EventSender.cpp: Use unsigned here instead of unsigned int.
2157
2158 2011-05-02  Adam Barth  <abarth@webkit.org>
2159
2160         Reviewed by Eric Seidel.
2161
2162         EWS should report the list of unexpected failures
2163         https://bugs.webkit.org/show_bug.cgi?id=59923
2164
2165         We cache the results of the first layout test results run on the task
2166         so the EWS can grab them later.  Another option is to include them in
2167         the exception that's thrown, but this approach seemed slightly more
2168         aesthetic.
2169
2170         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
2171         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
2172
2173 2011-05-02  Adam Barth  <abarth@webkit.org>
2174
2175         Reviewed by Eric Seidel.
2176
2177         Add base case for a test-running EWS
2178         https://bugs.webkit.org/show_bug.cgi?id=59920
2179
2180         Later I will subclass this to run the Chromium Linux EWS.
2181
2182         * Scripts/webkitpy/tool/bot/earlywarningsystemtask.py: Added.
2183         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
2184         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
2185
2186 2011-05-02  Anders Carlsson  <andersca@apple.com>
2187
2188         Build fix.
2189
2190         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2191         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
2192
2193 2011-05-02  Nico Weber  <thakis@chromium.org>
2194
2195         Reviewed by Mihai Parparita.
2196
2197         Add --no-webkit2 to the output of "build-webkit -h"
2198         https://bugs.webkit.org/show_bug.cgi?id=59942
2199
2200         * Scripts/build-webkit:
2201
2202 2011-05-02  Martin Robinson  <mrobinson@igalia.com>
2203
2204         Reviewed by Xan Lopez.
2205
2206         [GTK] old-run-webkit-tests should disable the Ubuntu overlay scrollbar
2207         https://bugs.webkit.org/show_bug.cgi?id=59943
2208
2209         Disable the Ubuntu overlay scrollbar while testing, by injecting the
2210         LIBOVERLAY_SCROLLBAR=0 environment variable.
2211
2212         * Scripts/old-run-webkit-tests:
2213
2214 2011-05-02  Martin Robinson  <mrobinson@igalia.com>
2215
2216         Reviewed by Eric Seidel.
2217
2218         [GTK] editing/selection/select-out-of-editable.html failing after r83967
2219         https://bugs.webkit.org/show_bug.cgi?id=58667
2220
2221         Do not trigger mouse down events when the EventSender knows that a mouse button is
2222         already down. These type of situations can confuse Xvfb and cause dragging failures.
2223
2224         * DumpRenderTree/gtk/EventSender.cpp:
2225         (mouseDownCallback): Do not trigger another mouse button down event if the mouse button
2226         is already down. This can interfere with drags in Xvfb.
2227
2228 2011-05-02  Zsolt Fehér  <h490267@stud.u-szeged.hu>
2229
2230         Reviewed by Csaba Osztrogonác.
2231
2232         DRT sideeffect revealed by r71268.
2233         https://bugs.webkit.org/show_bug.cgi?id=48958
2234
2235         Reset QWebSettings::ZoomTextOnly attribute before all tests.
2236
2237         Call setTextSizeMultiplier(1.0) to reset TextZoomFactor and
2238         PageZoomFactor too. It should be done before resetsettings()
2239         to guarantee resetting QWebSettings::ZoomTextOnly correctly.
2240
2241         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
2242         (WebCore::WebPage::resetSettings):
2243         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
2244
2245 2011-05-02  Sheriff Bot  <webkit.review.bot@gmail.com>
2246
2247         Unreviewed, rolling out r85469.
2248         http://trac.webkit.org/changeset/85469
2249         https://bugs.webkit.org/show_bug.cgi?id=59928
2250
2251         Broke  webkitpy-tests on Windows 7, GTK and Qt. (Requested by
2252         bbandix on #webkit).
2253
2254         * Scripts/webkitpy/tool/bot/earlywarningsystemtask.py: Removed.
2255         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
2256         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
2257
2258 2011-05-02  Tomasz Morawski  <t.morawski@samsung.com>
2259
2260         Reviewed by Kenneth Rohde Christiansen.
2261
2262         Cleanup EFL initialization/shutdown in the EWebLauncher
2263         https://bugs.webkit.org/show_bug.cgi?id=59922
2264
2265         Removed ecore_evas_init and edje_init form main function due to they
2266         are also called in the ewk_init. Moved ewk_shutdown to quit function.
2267
2268         * EWebLauncher/main.c:
2269         (quit):
2270         (main):
2271
2272 2011-05-02  Adam Barth  <abarth@webkit.org>
2273
2274         Reviewed by Eric Seidel.
2275
2276         Add base case for a test-running EWS
2277         https://bugs.webkit.org/show_bug.cgi?id=59920
2278
2279         Later I will subclass this to run the Chromium Linux EWS.
2280
2281         * Scripts/webkitpy/tool/bot/earlywarningsystemtask.py: Added.
2282         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
2283         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
2284
2285 2011-05-02  Adam Barth  <abarth@webkit.org>
2286
2287         Reviewed by Eric Seidel.
2288
2289         Add the ability to skip python and perl unit tests on EC2
2290         https://bugs.webkit.org/show_bug.cgi?id=59918
2291
2292         * Scripts/webkitpy/common/config/ports.py:
2293             - This is slightly ugly, but it's useful...
2294         * Scripts/webkitpy/common/config/ports_unittest.py:
2295         * Scripts/webkitpy/tool/mocktool.py:
2296         * Scripts/webkitpy/tool/steps/runtests.py:
2297
2298 2011-05-02  Adam Barth  <abarth@webkit.org>
2299
2300         Reviewed by Eric Seidel.
2301
2302         webkit-patch --non-interactive is incompatible with new-run-webkit-tests
2303         https://bugs.webkit.org/show_bug.cgi?id=59919
2304
2305         new-run-webkit-tests doesn't support --wait-for-httpd (because it
2306         always waits for HTTPD, as I understand things).  It turns out we're
2307         not really using this flag for anything, so we should just remove it.
2308
2309         * Scripts/webkitpy/tool/steps/runtests.py:
2310         * Scripts/webkitpy/tool/steps/steps_unittest.py:
2311
2312 2011-05-01  Eric Seidel  <eric@webkit.org>
2313
2314         Reviewed by Adam Barth.
2315
2316         The bots should learn from expected failures without having to retry
2317         https://bugs.webkit.org/show_bug.cgi?id=59279
2318
2319         Rather than re-trying the whole patch, we can just check whether we now
2320         expect the failures we see with the patch applied, saving a lot of time.
2321
2322         This patch also teaches the CommitQueue how to reject patches even
2323         when the tree is red using the new unexpected_failures() method on
2324         ExpectedFailures.
2325
2326         Finally in testing we found a bug in ExpectedFailures where it would
2327         "trust" results which had many many failures, due to using an
2328         equal sign instead of a less-than sign.
2329
2330         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py
2331         * Scripts/webkitpy/tool/bot/expectedfailures.py
2332         * Scripts/webkitpy/tool/bot/expectedfailures_unittest.py
2333         * Scripts/webkitpy/tool/bot/patchanalysistask.py
2334
2335
2336 2011-05-02  Adam Barth  <abarth@webkit.org>
2337
2338         Reviewed by Eric Seidel.
2339
2340         workspace.py fails to import ScriptError
2341         https://bugs.webkit.org/show_bug.cgi?id=59915
2342
2343         If the workspace fails to create a zip, webkitpy crashes instead of
2344         handling the error as expected.
2345
2346         * Scripts/webkitpy/common/system/workspace.py:
2347         * Scripts/webkitpy/common/system/workspace_unittest.py:
2348         * Scripts/webkitpy/tool/mocktool.py:
2349
2350 2011-05-01  Adam Barth  <abarth@webkit.org>
2351
2352         Remove unadvertantly added file.
2353
2354         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2355
2356 2011-05-01  Adam Barth  <abarth@webkit.org>
2357
2358         Reviewed by Eric Seidel.
2359
2360         expectedfailures.py should report the unexpected failures when asked
2361         https://bugs.webkit.org/show_bug.cgi?id=59916
2362
2363         This information is useful when the EWS runs tests.
2364
2365         * Scripts/webkitpy/tool/bot/expectedfailures.py:
2366         * Scripts/webkitpy/tool/bot/expectedfailures_unittest.py:
2367
2368 2011-05-01  Adam Barth  <abarth@webkit.org>
2369
2370         Reviewed by Eric Seidel.
2371
2372         JSONTestResult needs to handle multiple results
2373         https://bugs.webkit.org/show_bug.cgi?id=59269
2374
2375         This case occurs when a test is flaky because NRWT will run the test
2376         twice and report both results as "actual."
2377
2378         * Scripts/webkitpy/common/net/resultsjsonparser.py:
2379         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
2380
2381 2011-05-01  Adam Barth  <abarth@webkit.org>
2382
2383         Reviewed by Eric Seidel.
2384
2385         Move checkout/api.py to checkout/checkout.py
2386         https://bugs.webkit.org/show_bug.cgi?id=59906
2387
2388         This is more consistent with Python style.
2389
2390         * Scripts/webkitpy/common/checkout/__init__.py:
2391         * Scripts/webkitpy/common/checkout/scm_unittest.py:
2392         * Scripts/webkitpy/common/host.py:
2393         * Scripts/webkitpy/tool/main.py:
2394
2395 2011-05-01  Eric Seidel  <eric@webkit.org>
2396
2397         Reviewed by Adam Barth.
2398
2399         scm.py should be split into many pieces
2400         https://bugs.webkit.org/show_bug.cgi?id=59908
2401
2402         Forgot an AuthenticationError import.
2403
2404         * Scripts/webkitpy/common/checkout/scm/git.py:
2405         * Scripts/webkitpy/common/checkout/scm/svn.py:
2406
2407 2011-05-01  Arno Renevier  <arno@renevier.net>
2408
2409         Reviewed by Eric Seidel.
2410
2411         wrong exception raised in Lighttpd start method
2412         https://bugs.webkit.org/show_bug.cgi?id=59283
2413
2414         raise standard Exception instead of non existent
2415         google.httpd_utils.HttpdNotStarted
2416
2417         * Scripts/webkitpy/layout_tests/port/http_server.py:
2418
2419 2011-05-01  Eric Seidel  <eric@webkit.org>
2420
2421         Reviewed by Adam Barth.
2422
2423         scm.py should be split into many pieces
2424         https://bugs.webkit.org/show_bug.cgi?id=59908
2425
2426         * Scripts/webkitpy/common/checkout/scm/__init__.py:
2427         * Scripts/webkitpy/common/checkout/scm/commitmessage.py: Added.
2428         * Scripts/webkitpy/common/checkout/scm/git.py: Added.
2429         * Scripts/webkitpy/common/checkout/scm/scm.py:
2430         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
2431         * Scripts/webkitpy/common/checkout/scm/svn.py: Added.
2432
2433 2011-05-01  Eric Seidel  <eric@webkit.org>
2434
2435         Reviewed by Adam Barth.
2436
2437         Make most scm.py tests pass in preparation for splitting them up
2438         https://bugs.webkit.org/show_bug.cgi?id=59904
2439
2440         There is one remaining failure:
2441         webkitpy.common.checkout.scm.scm_unittest.GitTestWithMock.test_create_patch
2442         which identifies a difference between the SVN and Git versions of the
2443         SCM api.  I'm not going to bother fixing that until we split out all
2444         the tests.
2445
2446         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
2447
2448 2011-05-01  Patrick Gansterer  <paroga@webkit.org>
2449
2450         [WIN] Unreviewed buildfix after r85440.
2451
2452         * DumpRenderTree/win/UIDelegate.cpp:
2453         (UIDelegate::UIDelegate): Removed accidental added adoptPtr.
2454
2455 2011-05-01  Patrick Gansterer  <paroga@webkit.org>
2456
2457         [WIN] Unreviewed buildfix after r85434.
2458
2459         * DumpRenderTree/win/FrameLoadDelegate.cpp: Added missing header.
2460         * DumpRenderTree/win/UIDelegate.cpp: Ditto.
2461
2462 2011-05-01  Patrick Gansterer  <paroga@webkit.org>
2463
2464         [WIN] Unreviewed buildfix after r85434.
2465
2466         * DumpRenderTree/win/FrameLoadDelegate.cpp:
2467         (FrameLoadDelegate::FrameLoadDelegate):
2468         * DumpRenderTree/win/UIDelegate.cpp:
2469         (DRTUndoManager::DRTUndoManager):
2470         (UIDelegate::UIDelegate):
2471         (UIDelegate::resetUndoManager):
2472
2473 2011-05-01  Eric Seidel  <eric@webkit.org>
2474
2475         Reviewed by Adam Barth.
2476
2477         Move scm.py into its own module in preparation for splitting it up
2478         https://bugs.webkit.org/show_bug.cgi?id=59884
2479
2480         * Scripts/webkitpy/common/checkout/scm/__init__.py: Added.
2481         * Scripts/webkitpy/common/checkout/scm/scm.py: Moved.
2482         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py: Moved.
2483         * Scripts/webkitpy/test/main.py:
2484
2485 2011-05-01  Adam Barth  <abarth@webkit.org>
2486
2487         Reviewed by Eric Seidel.
2488
2489         webkit-patch fails when not run from root
2490         https://bugs.webkit.org/show_bug.cgi?id=59882
2491
2492         We need to set the CWD when running svn so that svn interprets file
2493         paths relative to the root of the working copy.
2494
2495         * Scripts/webkitpy/common/checkout/scm.py:
2496
2497 2011-04-29  Eric Seidel  <eric@webkit.org>
2498
2499         Reviewed by Adam Barth.
2500
2501         land-safely patches should show show their place in the commit-queue
2502         https://bugs.webkit.org/show_bug.cgi?id=59853
2503
2504         This complicated the "should we show the bubbles or the submit-to-ews button"
2505         logic substantially. :(  But now we'll show both the commit-queue position
2506         as well as the submit-to-ews button for land-safely (cq+ only) patches.
2507         Previously we would only show the submit-to-ews button.
2508
2509         * QueueStatusServer/handlers/statusbubble.py:
2510         * QueueStatusServer/templates/statusbubble.html:
2511
2512 2011-04-29  Naoki Takano  <takano.naoki@gmail.com>
2513
2514         Reviewed by Adam Barth.
2515
2516         Add my name to contributers.
2517         https://bugs.webkit.org/show_bug.cgi?id=59854
2518
2519         * Scripts/webkitpy/common/config/committers.py:
2520
2521 2011-04-29  Keith Kyzivat  <keith.kyzivat@nokia.com>
2522
2523         Reviewed by Csaba Osztrogonác.
2524
2525         [Qt] Don't build TestNetscapePlugin if ENABLE_NETSCAPE_PLUGIN_API=0
2526         https://bugs.webkit.org/show_bug.cgi?id=59772
2527
2528         * Tools.pro:
2529
2530 2011-04-29  Stephanie Lewis  <slewis@apple.com>
2531
2532         Rubber-stamped by Adam Roben.
2533
2534         Remove the production configuration guard since installs only happen in production anyway.
2535         Works around a bug in Xcode
2536
2537         * DumpRenderTree/mac/PerlSupport/Makefile:
2538
2539 2011-04-29  Adam Roben  <aroben@apple.com>
2540
2541         Remove a PHP setting that is now provided by httpd.conf
2542
2543         r85155 added this setting, but putting it in httpd.conf makes more sense (since the setting
2544         isn't dependent on paths that could be different on each user's machine).
2545
2546         Reviewed by Sam Weinig.
2547
2548         * Scripts/webkitperl/httpd.pm:
2549         (getDefaultConfigForTestDirectory): Removed log_errors setting.
2550
2551 2011-04-29  Jon Lee  <jonlee@apple.com>
2552
2553         Unreviewed, Adding Jon Lee as a committer
2554
2555         * Scripts/webkitpy/common/config/committers.py:
2556
2557 2011-04-28  Ojan Vafai  <ojan@chromium.org>
2558
2559         Reviewed by Tony Chang.
2560
2561         fix pretty_patch_available and wdiff_available
2562         https://bugs.webkit.org/show_bug.cgi?id=59761
2563
2564         These methods rely on the check_* method being called first.
2565         This would only happen when we actually tried to compute a diff,
2566         but computing the diff happens on a different process.
2567
2568         Also fix setting _wdiff_avialable to True by default.
2569
2570         * Scripts/webkitpy/layout_tests/port/base.py:
2571         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
2572         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
2573
2574 2011-04-29  Ojan Vafai  <ojan@chromium.org>
2575
2576         Reviewed by Eric Seidel.
2577
2578         exclude skipped tests from the summarized results
2579         https://bugs.webkit.org/show_bug.cgi?id=59813
2580
2581         If we find a need fo skipped tests later, we should either create
2582         a skipped.json file or expose this some other way. As it is, the data
2583         is unused and is hugely bloating the size of full_results.json.
2584
2585         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
2586
2587 2011-04-29  Adam Roben  <aroben@apple.com>
2588
2589         Make DumpRenderTreeSupport get installed when the default configuration is built
2590
2591         The default configuration is Production, but in that case the CONFIGURATION environment
2592         variable is set to the empty string.
2593
2594         Reviewed by Sam Weinig.
2595
2596         * DumpRenderTree/mac/PerlSupport/Makefile: Set CONFIGURATION to Production if it isn't
2597         already set so that later logic that depends on CONFIGURATION having that value will work.
2598
2599 2011-04-29  Dean Jackson  <dino@apple.com>
2600
2601         Reviewed by Simon Fraser.
2602
2603         Add ENABLE macro for WebKitAnimation
2604         https://bugs.webkit.org/show_bug.cgi?id=59729
2605
2606         Add new feature to toggle WebKit Animation API.
2607
2608         * Scripts/build-webkit:
2609
2610 2011-04-29  Sam Weinig  <sam@webkit.org>
2611
2612         Reviewed by Adam Roben.
2613
2614         Make production the default build type for WebKitTestRunner, matching
2615         all other projects.
2616
2617         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
2618
2619 2011-04-29  Kevin Ollivier  <kevino@theolliviers.com>
2620
2621         [wx] Unreviewed build fix, add new directory to the build.
2622
2623         * waf/build/settings.py:
2624
2625 2011-04-29  Carlos Garcia Campos  <cgarcia@igalia.com>
2626
2627         Reviewed by Martin Robinson.
2628
2629         [GTK] Implement UI client in minibrowser
2630         https://bugs.webkit.org/show_bug.cgi?id=58416
2631
2632         It also allows to open more than one minibrowser window from the
2633         command line.
2634
2635         * MiniBrowser/gtk/BrowserWindow.c:
2636         (browserWindowFinalize):
2637         (browser_window_init):
2638         (browserWindowConstructed):
2639         (createNewPage):
2640         (showPage):
2641         (closePage):
2642         (runJavaScriptAlert):
2643         (browserWindowUIClientInit):
2644         * MiniBrowser/gtk/main.c:
2645         (loadURI):
2646         (main):
2647
2648 2011-04-29  Carlos Garcia Campos  <cgarcia@igalia.com>
2649
2650         Reviewed by Martin Robinson.
2651
2652         [GTK] Implement page loader client in MiniBrowser
2653         https://bugs.webkit.org/show_bug.cgi?id=58024
2654
2655         A new class BrowserWindow has been added to implement the page
2656         loader client. It makes supporting multiple windows easier. Window
2657         title and url entry are now updated during the loading process.
2658
2659         * MiniBrowser/gtk/BrowserWindow.c: Added.
2660         (activateUriEntryCallback):
2661         (goBackCallback):
2662         (goForwardCallback):
2663         (browserWindowFinalize):
2664         (browserWindowGetProperty):
2665         (browserWindowSetProperty):
2666         (browser_window_init):
2667         (browserWindowConstructed):
2668         (browser_window_class_init):
2669         (browserWindowUpdateTitle):
2670         (browserWindowSetTitle):
2671         (browserWindowSetLoadProgress):
2672         (browserWindowUpdateURL):
2673         (didStartProvisionalLoadForFrame):
2674         (didReceiveServerRedirectForProvisionalLoadForFrame):
2675         (didFailProvisionalLoadWithErrorForFrame):
2676         (didCommitLoadForFrame):
2677         (didFinishDocumentLoadForFrame):
2678         (didFinishLoadForFrame):
2679         (didFailLoadWithErrorForFrame):
2680         (didReceiveTitleForFrame):
2681         (didFirstLayoutForFrame):
2682         (didFirstVisuallyNonEmptyLayoutForFrame):
2683         (didRemoveFrameFromHierarchy):
2684         (didStartProgress):
2685         (didChangeProgress):
2686         (didFinishProgress):
2687         (didBecomeUnresponsive):
2688         (didBecomeResponsive):
2689         (browserWindowLoaderClientInit):
2690         (browser_window_new):
2691         (browser_window_get_view):
2692         * MiniBrowser/gtk/BrowserWindow.h: Added.
2693         * MiniBrowser/gtk/GNUmakefile.am:
2694         * MiniBrowser/gtk/main.c:
2695         (createWindow):
2696
2697 2011-04-28  Stephanie Lewis  <slewis@apple.com>
2698
2699         Reviewed by Mark Rowe.
2700
2701         http://bugs.webkit.org/show_bug.cgi?id=59743
2702         <rdar://problem/9040413> Add an install target for production builds.
2703
2704         * DumpRenderTree/mac/PerlSupport/Makefile:
2705
2706 2011-04-28  Adam Roben  <aroben@apple.com>
2707
2708         Stop and restart ATSServer before each test run on Leopard
2709
2710         This is to work around an ATSServer leak that eventually causes tests
2711         to fail.
2712
2713         Fixes <http://webkit.org/b/51807> <rdar://problem/9273894>
2714         fast/blockflow/broken-ideograph-[small-caps|font].html sometimes fails on Leopard buildbots
2715         (due to ATSServer leak)
2716
2717         Reviewed by Dan Bernstein.
2718
2719         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
2720         (StopATSServer):
2721         (StartATSServer):
2722         Added these new commands to stop and start ATSServer.
2723
2724         (TestFactory.__init__): Stop and start ATSServer on Leopard before running tests.
2725
2726 2011-04-28  Dirk Pranke  <dpranke@chromium.org>
2727
2728         Reviewed by Adam Roben.
2729
2730         add Leopard NRWT bot to the scheduler
2731         https://bugs.webkit.org/show_bug.cgi?id=59766
2732
2733         * BuildSlaveSupport/build.webkit.org-config/config.json:
2734
2735 2011-04-28  Sam Weinig  <sam@webkit.org>
2736
2737         Fix WebKitTestRunner build.
2738
2739         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
2740         The Derived Sources target needs a product name.
2741
2742 2011-04-28  Sam Weinig  <sam@webkit.org>
2743
2744         Reviewed by Mark Rowe.
2745
2746         Add install support for WebKitTestRunner.
2747         <rdar://problem/9349055>
2748
2749         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
2750         Add missing CompilerVersion.xcconfig.
2751
2752         * WebKitTestRunner/Configurations/InjectedBundle.xcconfig:
2753         * WebKitTestRunner/Configurations/WebKitTestRunner.xcconfig:
2754         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
2755         Add install path and tell it not to skip installation.
2756
2757         Add missing CompilerVersion.xcconfig.
2758
2759 2011-04-28  David Levin  <levin@chromium.org>
2760
2761         Reviewed by Adam Barth.
2762
2763         Remove IMAGE_RESIZER related code.
2764         https://bugs.webkit.org/show_bug.cgi?id=59735
2765
2766         * Scripts/build-webkit:
2767
2768 2011-04-28  Sam Weinig  <sam@webkit.org>
2769
2770         Reviewed by Adam Roben.
2771
2772         Make reading skipped files from additional platform
2773         directories work in WebKit2 mode.
2774
2775         * Scripts/old-run-webkit-tests:
2776         (buildPlatformTestHierarchy): Simplify logic to always use the first
2777         and last platforms in the build hierarchy as the test hierarchy, and to
2778         insert the WebKit2 platform in the middle if it exists. Yum.
2779
2780 2011-04-28  Adam Barth  <abarth@webkit.org>
2781
2782         Reviewed by Eric Seidel.
2783
2784         Remove WML
2785         https://bugs.webkit.org/show_bug.cgi?id=59678
2786
2787         Remove build system integration and test harness integration for WML.
2788
2789         * Scripts/build-webkit:
2790         * Scripts/old-run-webkit-tests:
2791         * Scripts/webkitperl/features.pm:
2792         * Scripts/webkitpy/layout_tests/port/webkit.py:
2793         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2794
2795 2011-04-28  Ojan Vafai  <ojan@chromium.org>
2796
2797         Reviewed by Mihai Parparita.
2798
2799         use a hierarchical data structure to store results json
2800         https://bugs.webkit.org/show_bug.cgi?id=59736
2801
2802         Instead of a map from test name to results dict, we now store a hierarchical
2803         data structure.
2804
2805         Old: { foo/bar/baz.html: results_dict, foo/bar/baz2.html: results_dict }
2806         New:
2807         { foo: {
2808             bar: {
2809               baz.html: results_dict,
2810               baz2.html: results_dict } } }
2811
2812         This cut ~52% of the size of times_ms.json. I expect it will cut
2813         30-40% of full_results.json and unexpected_results.json.
2814
2815         * Scripts/webkitpy/common/net/resultsjsonparser.py:
2816         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
2817         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
2818         Using os.sep was just wrong since relative_test_filename canonicalizes separators.
2819         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
2820         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
2821         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
2822         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
2823         Instead of plumbing through the change to all of the rebaselineserver
2824         python and JS, just convert to the old format after reading in the json.
2825
2826 2011-04-28  Eric Seidel  <eric@webkit.org>
2827
2828         Reviewed by Adam Barth.
2829
2830         EWS Bubbles should show commit-queue position when patch is in the commit-queue
2831         https://bugs.webkit.org/show_bug.cgi?id=59744
2832
2833         Add target="_top" to the bubble links.  This was
2834         forgotten when I moved from using an onclick to a normal link.
2835
2836         * QueueStatusServer/templates/statusbubble.html:
2837
2838 2011-04-28  Eric Seidel  <eric@webkit.org>
2839
2840         Reviewed by Adam Barth.
2841
2842         EWS bubbles should show "submit to ews" button if they've not been processed
2843         https://bugs.webkit.org/show_bug.cgi?id=59751
2844
2845         * QueueStatusServer/handlers/statusbubble.py:
2846         * QueueStatusServer/handlers/submittoews.py:
2847         * QueueStatusServer/templates/statusbubble.html:
2848         * QueueStatusServer/templates/submittoews.html:
2849
2850 2011-04-28  Eric Seidel  <eric@webkit.org>
2851
2852         Reviewed by Adam Barth.
2853
2854         EWS Bubbles should show commit-queue position when patch is in the commit-queue
2855         https://bugs.webkit.org/show_bug.cgi?id=59744
2856
2857         This was actually just a 2 line change to statusbubble.py.
2858         I also cleaned up statusbubble.html a little while I was there
2859         (the links are now actual <a> tags and have hover).
2860
2861         * QueueStatusServer/handlers/statusbubble.py:
2862         * QueueStatusServer/templates/statusbubble.html:
2863
2864 2011-04-28  Dirk Pranke  <dpranke@chromium.org>
2865
2866         Reviewed by Eric Seidel.
2867
2868         add a mac leopard release NRWT bot
2869         https://bugs.webkit.org/show_bug.cgi?id=59733
2870
2871         * BuildSlaveSupport/build.webkit.org-config/config.json:
2872
2873 2011-04-28  Ojan Vafai  <ojan@chromium.org>
2874
2875         Avoid crashing when running new-run-webkit-tests outside of
2876         version control.
2877
2878         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
2879
2880 2011-04-28  Xan Lopez  <xlopez@igalia.com>
2881
2882         Unreviewed build fix.
2883
2884         * GNUmakefile.am:
2885
2886 2011-04-28  Ojan Vafai  <ojan@chromium.org>
2887
2888         Reviewed by Eric Seidel.
2889
2890         fix python printing test
2891         https://bugs.webkit.org/show_bug.cgi?id=59719
2892
2893         Now that we run scm commands in summarize_results, --verbose prints
2894         debug logging to stderr.
2895
2896         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
2897
2898 2011-04-28  Ojan Vafai  <ojan@chromium.org>
2899
2900         Reviewed by Tony Chang.
2901
2902         link to tests at a specific revision in the results.html file
2903         https://bugs.webkit.org/show_bug.cgi?id=59711
2904
2905         * Scripts/webkitpy/common/checkout/scm.py:
2906         Add a method to get the head svn revision.
2907         * Scripts/webkitpy/common/checkout/scm_unittest.py:
2908         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
2909         Include the svn revision in the results json.
2910
2911 2011-04-28  Carlos Garcia Campos  <cgarcia@igalia.com>
2912
2913         Reviewed by Sam Weinig.
2914
2915         [GTK] Remove WKURLCreateWithURL()
2916         https://bugs.webkit.org/show_bug.cgi?id=59691
2917
2918         Use WKURLCreateWithUTF8CString() instead of WKURLCreateWithURL().
2919
2920         * MiniBrowser/gtk/main.c:
2921         (activateUriEntryCallback):
2922         (main):
2923
2924 2011-04-28  Carlos Garcia Campos  <cgarcia@igalia.com>
2925
2926         Reviewed by Martin Robinson.
2927
2928         [GTK] WKView should be a GtkWidget
2929         https://bugs.webkit.org/show_bug.cgi?id=59603
2930
2931         Don't use WKViewGetWindow(), use WKView as a GtkWidget instead.
2932
2933         * MiniBrowser/gtk/main.c:
2934         (createWindow):
2935         (main):
2936
2937 2011-04-28  Gabor Rapcsanyi  <rgabor@webkit.org>
2938
2939         Reviewed by Csaba Osztrogonác.
2940
2941         [Qt] Add ARM specific Skipped list
2942         https://bugs.webkit.org/show_bug.cgi?id=59680
2943
2944         * Scripts/old-run-webkit-tests:
2945         * Scripts/webkitdirs.pm:
2946
2947 2011-04-27  Eric Seidel  <eric@webkit.org>
2948
2949         Reviewed by Adam Barth.
2950
2951         error_log.txt should include PHP errors
2952         https://bugs.webkit.org/show_bug.cgi?id=59643
2953
2954         Turns out this was really easy after I finally learned
2955         enough about PHP to do it...
2956
2957         I'm adding this in hopes of better tracking down
2958         our random http failures.
2959
2960         * Scripts/webkitperl/httpd.pm:
2961
2962 2011-04-27  David Levin  <levin@chromium.org>
2963
2964         Reviewed by Adam Barth.
2965
2966         check-webkit-style should flags spacing errors for braces on the same line.
2967         https://bugs.webkit.org/show_bug.cgi?id=59665
2968
2969         * Scripts/webkitpy/style/checkers/cpp.py: Added the check.
2970         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added some tests for the check
2971           and fixed some test cases to not trigger this error since it isn't what they are verifying.
2972
2973 2011-04-27  David Levin  <levin@chromium.org>
2974
2975         Reviewed by Adam Barth.
2976
2977         check-webkit-style error for spaces after periods in a comment has a bad error
2978         message and warns about end of line spaces.
2979         https://bugs.webkit.org/show_bug.cgi?id=59663
2980
2981         * Scripts/webkitpy/style/checkers/cpp.py: Fix both issues.
2982         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added tests for spaces at the end of line
2983         and changed the error messages.
2984
2985 2011-04-27  Ojan Vafai  <ojan@chromium.org>
2986
2987         remove assertion now that platform/mac-tiger no longer exists
2988         https://bugs.webkit.org/show_bug.cgi?id=59660
2989
2990         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
2991
2992 2011-04-27  Ojan Vafai  <ojan@chromium.org>
2993
2994         Reviewed by Eric Seidel.
2995
2996         remove test that didn't make sense
2997         https://bugs.webkit.org/show_bug.cgi?id=59658
2998
2999         It was artifically saying the test wasn't expected to pass,
3000         but the expectations string was empty. The case of unexpected
3001         passes with a real expectations string is tested later in the
3002         same test.
3003
3004         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
3005
3006 2011-04-27  Sam Weinig  <sam@webkit.org>
3007
3008         Reviewed by Adam Roben.
3009
3010         WebKit2 needs to notify the client when a plugin fails to initialize
3011         https://bugs.webkit.org/show_bug.cgi?id=59657
3012
3013         * MiniBrowser/mac/BrowserWindowController.m:
3014         (-[BrowserWindowController awakeFromNib]):
3015         * WebKitTestRunner/TestController.cpp:
3016         (WTR::TestController::initialize):
3017         Add stubs.
3018
3019 2011-04-26  Sam Weinig  <sam@webkit.org>
3020
3021         Reviewed by David Levin.
3022
3023         Add a few more missing options to new-run-webkit-tests
3024         https://bugs.webkit.org/show_bug.cgi?id=59567
3025
3026         Add support for:
3027             --threaded
3028             --gc-between-tests
3029             --complex-text
3030
3031         * Scripts/webkitpy/layout_tests/port/webkit.py:
3032         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3033         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
3034
3035 2011-04-27  Mark Rowe  <mrowe@apple.com>
3036
3037         Build fix.
3038
3039         Fix an OwnPtr-related issue that was missed before the strictness was turned on.
3040
3041         * WebKitTestRunner/TestController.cpp:
3042         (WTR::TestController::runTest):
3043
3044 2011-04-27  Ojan Vafai  <ojan@chromium.org>
3045
3046         Forgot to upload the new times_ms.json file after splitting
3047         it out from full_results.json.
3048
3049         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3050
3051 2011-04-27  Ojan Vafai  <ojan@chromium.org>
3052
3053         Fix snafu in generating the JSON for the chromium tests.
3054
3055         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
3056
3057 2011-04-27  Ojan Vafai  <ojan@chromium.org>
3058
3059         Fix bad merge. This broke the chromium tests generating JSON files.
3060         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
3061
3062 2011-04-27  Mark Rowe  <mrowe@apple.com>
3063
3064         Rubber-stamped by Geoff Garen.
3065
3066         Don't look for results in a location that no longer exists.
3067
3068         * Scripts/old-run-webkit-tests:
3069
3070 2011-04-27  Adrienne Walker  <enne@google.com>
3071
3072         Unreviewed, add Chrome GPU folks to contributors that I'm tired of manually CCing
3073
3074         * Scripts/webkitpy/common/config/committers.py:
3075
3076 2011-04-22  Ojan Vafai  <ojan@chromium.org>
3077
3078         Reviewed by Tony Chang.
3079
3080         only include failures in full_results.json
3081         https://bugs.webkit.org/show_bug.cgi?id=59256
3082
3083         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3084
3085 2011-04-22  Ojan Vafai  <ojan@chromium.org>
3086
3087         Reviewed by Adam Roben.
3088
3089         move times into their own json file
3090         https://bugs.webkit.org/show_bug.cgi?id=59239
3091
3092         The test runtimes are the only case where we need an entry for
3093         every test that was run, including passing tests. Move them into
3094         their own json file so we can prune the other json files to only
3095         include failing tests.
3096
3097         Use a trie-like data structure to hold the tests to avoid including
3098         full test paths. This cuts 52% of the filesize. The end result is ~700k
3099         pre-gzip.
3100
3101         * Scripts/webkitpy/common/net/resultsjsonparser.py:
3102         Exclude times from the parsed TestResults.
3103         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
3104         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
3105         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
3106         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
3107
3108 2011-04-27  Anders Carlsson  <andersca@apple.com>
3109
3110         More casts and more clang++ build fixes.
3111
3112         * DumpRenderTree/mac/DumpRenderTree.mm:
3113         (createWebViewAndOffscreenWindow):
3114         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
3115         (TestWebKitAPI::PlatformWebView::PlatformWebView):
3116         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
3117         (WTR::PlatformWebView::PlatformWebView):
3118
3119 2011-04-27  Tony Chang  <tony@chromium.org>
3120
3121         Reviewed by Ojan Vafai.
3122
3123         lazily autoinstall thirdparty python libraries
3124         https://bugs.webkit.org/show_bug.cgi?id=55723
3125
3126         This helps by only having pywebsocket installed when the bots run.  It
3127         doesn't help much for devs since webkit-patch pulls in almost everything.
3128
3129         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
3130         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
3131         * Scripts/webkitpy/common/net/networktransaction.py:
3132         * Scripts/webkitpy/common/net/networktransaction_unittest.py:
3133         * Scripts/webkitpy/common/net/statusserver.py:
3134         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
3135         * Scripts/webkitpy/thirdparty/__init__.py: Add an import hook so we can install based on
3136             the import command rather than just installing everything.
3137         * Scripts/webkitpy/thirdparty/__init___unittest.py: Added.
3138
3139 2011-04-27  Girish Ramakrishnan  <girish@forwardbias.in>
3140
3141         Add my IRC nick.
3142
3143         * Scripts/webkitpy/common/config/committers.py:
3144
3145 2011-04-27  Eric Seidel  <eric@webkit.org>
3146
3147         Reviewed by Mihai Parparita.
3148
3149         sherrifbot create-bug shouldn't assign bugs to webkit.review.bot
3150         https://bugs.webkit.org/show_bug.cgi?id=59545
3151
3152         To do this, I needed a way to look up contributors by irc-name
3153         (since anyone in #webkit who might use this command may not be a committer).
3154         To lookup contributors, I had to make Contributor a real object.
3155         Which led me to redesign parts of committers.py...
3156         and finally fix one spot in changelog.py where we wanted to be
3157         looking up contributors and not committers.
3158
3159         Overall a pretty simple fix, once you wade through the yak-hair.
3160
3161         This may not prevent *all* possible ways that bugs would get assigned
3162         to webkit.review.bot.  If we don't recognize the requester we will
3163         go through the previous code path (which shouldn't change the assignee
3164         on the bug from the default as far as I can tell).
3165
3166         * Scripts/webkitpy/common/checkout/changelog.py:
3167         * Scripts/webkitpy/common/config/committers.py:
3168         * Scripts/webkitpy/common/config/committers_unittest.py:
3169         * Scripts/webkitpy/tool/bot/irc_command.py:
3170         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
3171
3172 2011-04-27  Yi Shen  <yi.4.shen@nokia.com>
3173
3174         Reviewed by Kenneth Rohde Christiansen.
3175
3176         [Qt] Upstream Symbian platform plugin
3177         https://bugs.webkit.org/show_bug.cgi?id=58435
3178
3179         Exempting directory WebKit/qt/symbian/platformplugin from style guide.
3180
3181         * Scripts/webkitpy/style/checker.py:
3182         * Scripts/webkitpy/style/checker_unittest.py:
3183
3184 2011-04-27  Jeff Miller  <jeffm@apple.com>
3185
3186         Correct my IRC nickname.
3187
3188         * Scripts/webkitpy/common/config/committers.py:
3189
3190 2011-04-27  Lucas De Marchi  <lucas.demarchi@profusion.mobi>
3191
3192         Reviewed by Darin Adler.
3193
3194         Change the email used on bugzilla
3195         https://bugs.webkit.org/show_bug.cgi?id=59582
3196
3197         * Scripts/webkitpy/common/config/committers.py:
3198
3199 2011-04-27  James Kozianski  <koz@chromium.org>
3200
3201         Reviewed by Ojan Vafai.
3202
3203         Extract map from port to builder name.
3204         https://bugs.webkit.org/show_bug.cgi?id=58301
3205
3206         * Scripts/webkitpy/layout_tests/port/__init__.py:
3207         * Scripts/webkitpy/layout_tests/port/builders.py: Added.
3208         * Scripts/webkitpy/layout_tests/port/builders_unittest.py: Added.
3209         * Scripts/webkitpy/layout_tests/port/factory.py:
3210
3211 2011-04-27  Ryuan Choi  <ryuan.choi@samsung.com>
3212
3213         Reviewed by Martin Robinson.
3214
3215         [GTK] Add proxy support to GtkLauncher
3216         https://bugs.webkit.org/show_bug.cgi?id=58852
3217
3218         Add SOUP_TYPE_PROXY_RESOLVER_DEFAULT feature or set SOUP_SESSION_PROXY_URI for developers behind proxy.
3219
3220         * GNUmakefile.am:
3221         * GtkLauncher/main.c:
3222         (main):
3223
3224 2011-04-27  Annie Sullivan  <sullivan@chromium.org>
3225
3226         Reviewed by David Levin.
3227
3228         Adding myself as a Chromium contributor to committers.py.
3229
3230         https://bugs.webkit.org/show_bug.cgi?id=59576
3231
3232         * Scripts/webkitpy/common/config/committers.py:
3233
3234 2011-04-26  David Grogan  <dgrogan@chromium.org>
3235
3236         Reviewed by David Levin.
3237
3238         Adding myself to Chromium contributors list in committers.py
3239         https://bugs.webkit.org/show_bug.cgi?id=59565
3240
3241         * Scripts/webkitpy/common/config/committers.py:
3242
3243 2011-04-26  Chang Shu  <cshu@webkit.org>
3244
3245         Reviewed by Laszlo Gombos.
3246
3247         WebKitTestRunner needs layoutTestController.addOriginAccessWhitelistEntry
3248         https://bugs.webkit.org/show_bug.cgi?id=42544
3249
3250         Also implemented removeOriginAccessWhitelistEntry and resetOriginAccessWhitelists.
3251
3252         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
3253         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
3254         (WTR::InjectedBundle::beginTesting):
3255         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
3256         (WTR::LayoutTestController::addOriginAccessWhitelistEntry):
3257         (WTR::LayoutTestController::removeOriginAccessWhitelistEntry):
3258         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
3259
3260 2011-04-26  Chris Fleizach  <cfleizach@apple.com>
3261
3262         Reviewed by Adele Peterson.
3263
3264         AX: Add more text marker routines to DRT
3265         https://bugs.webkit.org/show_bug.cgi?id=58949
3266
3267         Add a few more methods to help test the text marker system, including nextMarker, previousMarker and
3268         stringForMarkerRange.
3269
3270         * DumpRenderTree/AccessibilityUIElement.cpp:
3271         (nextTextMarkerCallback):
3272         (previousTextMarkerCallback):
3273         (stringForTextMarkerRangeCallback):
3274         (AccessibilityUIElement::previousTextMarker):
3275         (AccessibilityUIElement::nextTextMarker):
3276         (AccessibilityUIElement::stringForTextMarkerRange):
3277         (AccessibilityUIElement::getJSClass):
3278         * DumpRenderTree/AccessibilityUIElement.h:
3279         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
3280         (AccessibilityUIElement::previousTextMarker):
3281         (AccessibilityUIElement::nextTextMarker):
3282         (AccessibilityUIElement::stringForTextMarkerRange):
3283
3284 2011-04-26  Dan Bernstein  <mitz@apple.com>
3285
3286         Build fix.
3287
3288         * DumpRenderTree/mac/Configurations/CompilerVersion.xcconfig:
3289         * MiniBrowser/Configurations/CompilerVersion.xcconfig:
3290         * WebKitTestRunner/Configurations/CompilerVersion.xcconfig:
3291
3292 2011-04-26  Kevin Ollivier  <kevino@theolliviers.com>
3293
3294         Rubberstamped by Eric Seidel.
3295
3296         Enable waf to be used to build other ports
3297         https://bugs.webkit.org/show_bug.cgi?id=58213
3298
3299         * Scripts/build-webkit:
3300         * Scripts/webkitdirs.pm:
3301         * waf: Added.
3302         * waf/build: Added.
3303         * waf/build/build_utils.py: Added.
3304         * waf/build/settings.py: Added.
3305         * waf/build/waf_extensions.py: Added.
3306         * waf/build/wxpresets.py: Added.
3307
3308 2011-04-26  Sam Weinig  <sam@webkit.org>
3309
3310         Reviewed by David Hyatt.
3311
3312         Remove Datagrid from the tree
3313         https://bugs.webkit.org/show_bug.cgi?id=59543
3314
3315         * Scripts/build-webkit:
3316         Remove feature.
3317
3318         * iExploder/iexploder-1.3.2/htdocs/htmltags.in:
3319         * iExploder/iexploder-1.7.2/src/html-tags/webkit:
3320         Remove from fuzzer.
3321
3322 2011-04-26  Brent Fulgham  <bfulgham@webkit.org>
3323
3324         Reviewed by Adam Roben.
3325
3326         Update WinLauncher with command-line arguments to allow the
3327         transparency to be tested.  Use --transparent to create a
3328         transparent view.  Use --desktop to have the view fill the
3329         desktop area of the screen.
3330         https://bugs.webkit.org/show_bug.cgi?id=58300
3331
3332         * WinLauncher/WinLauncher.cpp:
3333         (usesLayeredWebView): New method to indicate that the
3334         program is running with layered windows (alpha-blended
3335         transparent windows).
3336         (shouldUseFullDesktop): New method to indicate that the
3337         program is running across the full desktop (less any
3338         task bars.)
3339         (resizeSubViews): Layered window mode doesn't have any
3340         winapi child windows, so this routine is exited early
3341         for that case.
3342         (subclassForLayeredWindow): Subclass the window so we can
3343         override its defaut message loop.
3344         (computeFullDesktopFrame): Convenience function to get the
3345         desktop region and update the frame dimensions appropriately.
3346         (_tWinMain): Updated to accept the new '--transparent' and
3347         '--desktop' arguments.
3348         (InitInstance): Modified to exit early for the transparent
3349         background case.  WebKit will construct our primary window (with
3350         appropriate settings) so we don't need to create one ourselves.
3351         (WndProc):
3352         (TransparentWndProc): New message loop for transparent windows,
3353         since they need special handling to support dragging.
3354
3355 2011-04-26  Alexandre MAzari  <amazari@igalia.com>
3356
3357         Reviewed by Xan Lopez.
3358
3359         [Gtk] Implement LayoutTestController::setPluginsEnabled
3360         https://bugs.webkit.org/show_bug.cgi?id=59184
3361
3362         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
3363         (LayoutTestController::setPluginsEnabled):
3364
3365 2011-04-26  Chang Shu  <cshu@webkit.org>
3366
3367         Reviewed by Laszlo Gombos.
3368
3369         WebKitTestRunner needs layoutTestController.setAllowFileAccessFromFileURLs
3370         https://bugs.webkit.org/show_bug.cgi?id=57572
3371
3372         Remove the unnecessary code that updates preference.
3373
3374         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
3375         (WTR::InjectedBundle::beginTesting):
3376         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
3377         (WTR::LayoutTestController::setAllowUniversalAccessFromFileURLs):
3378
3379 2011-04-26  Mario Sanchez Prada  <msanchez@igalia.com>
3380
3381         Reviewed by Eric Seidel.
3382
3383         [GTK] accessibility/aria-combobox-expected.txt
3384         https://bugs.webkit.org/show_bug.cgi?id=59399
3385
3386         Implement isExpanded(), as it's used in the layout test.
3387
3388         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3389         (AccessibilityUIElement::isExpanded): Implemented.
3390
3391 2011-04-26  Adrienne Walker  <enne@google.com>
3392
3393         Reviewed by Kent Tamura.
3394
3395         [chromium] Fix memory leak in DRT due to OwnPtr<T*>
3396         https://bugs.webkit.org/show_bug.cgi?id=59489
3397
3398         * DumpRenderTree/chromium/WebViewHost.h:
3399
3400 2011-04-26  Mario Sanchez Prada  <msanchez@igalia.com>
3401
3402         Reviewed by Eric Seidel.
3403
3404         [GTK] Implement isChecked in DRT
3405         https://bugs.webkit.org/show_bug.cgi?id=59385
3406
3407         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
3408         (AccessibilityUIElement::isChecked): Implemented.
3409
3410 2011-04-26  Alejandro G. Castro  <alex@igalia.com>
3411
3412         Unreviewed, rolling out r84949.
3413         http://trac.webkit.org/changeset/84949
3414         https://bugs.webkit.org/show_bug.cgi?id=58852
3415
3416         Broke GTK+ compilation
3417
3418         * GtkLauncher/main.c:
3419         (main):
3420
3421 2011-04-26  Eric Seidel  <eric@webkit.org>
3422
3423         Reviewed by Mihai Parparita.
3424
3425         sheriff-bot should have a create-bug command
3426         https://bugs.webkit.org/show_bug.cgi?id=59491
3427
3428         This patch has some stderr unittest spew, but I will
3429         fix it after my live demo.
3430
3431         * Scripts/webkitpy/tool/bot/irc_command.py:
3432         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
3433         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
3434
3435 2011-04-26  Ryuan Choi  <ryuan.choi@samsung.com>
3436
3437         Reviewed by Martin Robinson.
3438
3439         [GTK] Add proxy support to GtkLauncher
3440         https://bugs.webkit.org/show_bug.cgi?id=58852
3441
3442         Add SOUP_TYPE_PROXY_RESOLVER_DEFAULT feature or set SOUP_SESSION_PROXY_URI for developers behind proxy.
3443
3444         * GtkLauncher/main.c:
3445         (main):
3446
3447 2011-04-26  Dan Bernstein  <mitz@apple.com>
3448
3449         Reviewed by Mark Rowe.
3450
3451         Choose the compiler based on the Xcode version for Snow Leopard debug builds.
3452
3453         * DumpRenderTree/mac/Configurations/Base.xcconfig:
3454         * DumpRenderTree/mac/Configurations/CompilerVersion.xcconfig: Added.
3455         * MiniBrowser/Configurations/Base.xcconfig:
3456         * MiniBrowser/Configurations/CompilerVersion.xcconfig: Added.
3457         * TestWebKitAPI/Configurations/Base.xcconfig:
3458         * TestWebKitAPI/Configurations/CompilerVersion.xcconfig: Added.
3459         * WebKitTestRunner/Configurations/Base.xcconfig:
3460         * WebKitTestRunner/Configurations/CompilerVersion.xcconfig: Added.
3461
3462 2011-04-26  Finnur Thorarinsson  <finnur.webkit@gmail.com>
3463
3464         Reviewed by David Levin.
3465
3466         Adding myself as a Chromium contributors to committers.py.
3467
3468         https://bugs.webkit.org/show_bug.cgi?id=59444
3469
3470         * Scripts/webkitpy/common/config/committers.py:
3471
3472 2011-04-26  Adam Barth  <abarth@webkit.org>
3473
3474         Reviewed by Eric Seidel.
3475
3476         DumpRenderTree/chromium should play nice with strict OwnPtrs
3477         https://bugs.webkit.org/show_bug.cgi?id=59458
3478
3479         All these changes are pretty straight forward.
3480
3481         * DumpRenderTree/chromium/CppBoundClass.cpp:
3482         (GetterPropertyCallback::GetterPropertyCallback):
3483         (CppBoundClass::bindGetterCallback):
3484         * DumpRenderTree/chromium/CppBoundClass.h:
3485         (CppBoundClass::bindProperty):
3486         (CppBoundClass::bindFallbackCallback):
3487         (CppBoundClass::bindFallbackMethod):
3488         * DumpRenderTree/chromium/DRTDevToolsClient.cpp:
3489         (DRTDevToolsClient::DRTDevToolsClient):
3490         * DumpRenderTree/chromium/TestEventPrinter.cpp:
3491         (TestEventPrinter::createDRTPrinter):
3492         (TestEventPrinter::createTestShellPrinter):
3493         * DumpRenderTree/chromium/TestEventPrinter.h:
3494         * DumpRenderTree/chromium/TestShell.cpp:
3495         (TestShell::TestShell):
3496         (TestShell::createMainWindow):
3497         (TestShell::createDRTDevToolsClient):
3498         * DumpRenderTree/chromium/WebViewHost.cpp:
3499         (WebViewHost::geolocationClientMock):
3500         (WebViewHost::speechInputController):
3501         (WebViewHost::deviceOrientationClientMock):
3502         (WebViewHost::reset):
3503         (WebViewHost::navigate):
3504         (WebViewHost::setPendingExtraData):
3505         (WebViewHost::canvas):
3506         * DumpRenderTree/chromium/WebViewHost.h:
3507
3508 2011-04-26  Adam Roben  <aroben@apple.com>
3509
3510         Make DRT/win log loading errors the same way DRT/mac does
3511
3512         Fixes <http://webkit.org/b/59303> [Windows WebKit1 Tests]
3513         http/tests/misc/will-send-request-returns-null-on-redirect.html failing since r84742
3514
3515         Reviewed by Alexey Proskuryakov.
3516
3517         * DumpRenderTree/win/ResourceLoadDelegate.cpp:
3518         (ResourceLoadDelegate::descriptionSuitableForTestResult): Removed the workaround for
3519         <rdar://problem/5064234>, which has long since been fixed, and then made this function match
3520         -[NSError(DRTExtras) _drt_descriptionSuitableForTestResult] more closely.
3521
3522 2011-04-26  Dan Bernstein  <mitz@apple.com>
3523
3524         REGRESSION (84876): build-webkit produces an error message
3525         https://bugs.webkit.org/show_bug.cgi?id=59460
3526
3527         * Scripts/webkitdirs.pm:
3528
3529 2011-04-26  Mihai Parparita  <mihaip@chromium.org>
3530
3531         Reviewed by Eric Seidel.
3532
3533         Fix timestamp comparison in check-for-weak-vtables-and-externals
3534         https://bugs.webkit.org/show_bug.cgi?id=59416
3535
3536         We should only re-run the check-for-weak-vtables-and-externals script
3537         if the executable is more recent than the last run.
3538
3539         * Scripts/check-for-weak-vtables-and-externals:
3540
3541 2011-04-26  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
3542
3543         Reviewed by Csaba Osztrogonác.
3544
3545         [Qt] Implement LayoutTestController::setAutofilled
3546         https://bugs.webkit.org/show_bug.cgi?id=59439
3547
3548         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
3549         (LayoutTestController::setAutofilled):
3550         * DumpRenderTree/qt/LayoutTestControllerQt.h:
3551
3552 2011-04-25  Dan Bernstein  <mitz@apple.com>
3553
3554         Reviewed by Mark Rowe.
3555
3556         Teach build-webkit where newer Xcode versions store the application-wide build settings user default.
3557
3558         * Scripts/webkitdirs.pm:
3559
3560 2011-04-25  Sam Weinig  <sam@webkit.org>
3561
3562         Reviewed by Adam Roben.
3563
3564         Leaks Viewer throws "Cannot post cyclic structures" in Safari 5
3565         https://bugs.webkit.org/show_bug.cgi?id=56090
3566
3567         Make LeaksViewer work with versions of browsers that don't support message
3568         passing of cyclic structures.
3569
3570         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParser.js:
3571         (LeaksParser.this._worker.onmessage):
3572         (LeaksParser):
3573         (LeaksParser.prototype.addLeaksFile):
3574         Use LeaksParserImpl directly if we don't support passing cyclic structures.
3575
3576         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserImpl.js: Copied from BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js.
3577         (LeaksParserImpl):
3578         (LeaksParserImpl.prototype.addLeaksFile):
3579         (LeaksParserImpl.prototype._incorporateLeaks):
3580         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js:
3581         (onmessage):
3582         Factor out LeaksParser implementation into LeaksParserImpl.js.
3583
3584         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Utilities.js:
3585         (workersSupportCyclicStructures):
3586         Add function to detect if workers support passing cyclic structures.
3587
3588         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html:
3589         Add LeaksParserImpl.js include.
3590
3591 2011-04-25  Jeff Miller  <jeffm@apple.com>
3592
3593         Reviewed by Timothy Hatcher.
3594
3595         Treat strings defined with UI_STRING_LOCALIZE_LATER_KEY() as debugging strings in extract-localizable-strings
3596         https://bugs.webkit.org/show_bug.cgi?id=59373
3597
3598         * Scripts/extract-localizable-strings: Add UI_STRING_LOCALIZE_LATER_KEY() to isDebugMacro var.
3599
3600 2011-04-25  Andrew Scherkus  <scherkus@chromium.org>
3601
3602         Reviewed by Dimitri Glazkov.
3603
3604         Adding a few Chromium contributors to committers.py.
3605
3606         https://bugs.webkit.org/show_bug.cgi?id=59327
3607
3608         * Scripts/webkitpy/common/config/committers.py:
3609
3610 2011-04-23  Dominic Cooney  <dominicc@chromium.org>
3611
3612         Reviewed by Dimitri Glazkov.
3613
3614         [V8] Nodes in shadow DOM should not be GCed while their hosts are alive
3615         https://bugs.webkit.org/show_bug.cgi?id=59284
3616
3617         Chromium DRT over-refcounted the NPObject wrappers it used to
3618         marshal shadowRoot nodes back to tests (WebBindings::makeNode
3619         births NPObjects with referenceCount == 1). This masked the GC bug
3620         59284 in tests by making shadowRoot nodes live forever.
3621
3622         * DumpRenderTree/chromium/LayoutTestController.cpp:
3623         (LayoutTestController::shadowRoot):
3624         (LayoutTestController::ensureShadowRoot):
3625
3626 2011-04-23  Kevin Ollivier  <kevino@theolliviers.com>
3627
3628         [wx] Unreviewed build fixes. Add new methods to wx LayoutTestController,
3629         and update the URL for downloading waf.
3630
3631         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
3632         (LayoutTestController::ensureShadowRoot):
3633         (LayoutTestController::removeShadowRoot):
3634         * Scripts/webkitdirs.pm:
3635
3636 2011-04-22  Sheriff Bot  <webkit.review.bot@gmail.com>
3637
3638         Unreviewed, rolling out r84627.
3639         http://trac.webkit.org/changeset/84627
3640         https://bugs.webkit.org/show_bug.cgi?id=59271
3641
3642         It broke Symbian build (Requested by Ossy on #webkit).
3643
3644         * Scripts/webkitpy/style/checker.py:
3645         * Scripts/webkitpy/style/checker_unittest.py:
3646
3647 2011-04-22  Dirk Pranke  <dpranke@chromium.org>
3648
3649         Reviewed by Adam Barth.
3650
3651         rebaseline-chromium-webkit-tests: clean up output
3652         https://bugs.webkit.org/show_bug.cgi?id=59240
3653
3654         This change dramatically changes the output the tool provides
3655         by default. The previous output (more or less) is still there
3656         in -v / verbose / debug mode, but the default output is
3657         way terser and more readable. It mostly says which files are
3658         updated, which are duplicates, and which are not found in an
3659         archive. Feedback desired :).
3660
3661         This change also adds a scm.exists() method to the SCM object
3662         so that we can cleanly figure out what we're doing while
3663         rebaselining.
3664
3665         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
3666         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
3667         * Scripts/webkitpy/common/system/filesystem_mock.py:
3668         * Scripts/webkitpy/common/checkout/scm.py:
3669         * Scripts/webkitpy/common/checkout/scm_unittest.py:
3670
3671 2011-04-22  Ilya Sherman  <isherman@chromium.org>
3672
3673         Reviewed by Alexey Proskuryakov.
3674
3675         Add Ilya Sherman (isherman@chromium.org) to the list of contributors
3676         https://bugs.webkit.org/show_bug.cgi?id=59231
3677
3678         * Scripts/webkitpy/common/config/committers.py:
3679
3680 2011-04-22  Yuta Kitamura  <yutak@chromium.org>
3681
3682         Reviewed by Ojan Vafai.
3683
3684         new-run-webkit-tests: Exception raised when stderr contains binary data
3685         https://bugs.webkit.org/show_bug.cgi?id=59083
3686
3687         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
3688         Use write_binary_f