a619744 | Geoffrey Sneddon | 09 October 2018, 17:34:30 UTC | Only run Windows CI for tools/ changes | 10 October 2018, 14:53:57 UTC |
a04efc1 | Geoffrey Sneddon | 09 October 2018, 15:24:17 UTC | Mark these Windows failures as Py2-only | 10 October 2018, 14:53:57 UTC |
eba61b4 | Geoffrey Sneddon | 09 October 2018, 15:20:25 UTC | Fix OpenSSL on Python 3 on Windows | 10 October 2018, 14:53:57 UTC |
2d09f3c | Geoffrey Sneddon | 08 October 2018, 16:38:24 UTC | Update Windows expectations These were previously marked as XFAIL on Windows, but now XPASS. Likely fixed by virtue of being fixed for Python 3. | 10 October 2018, 14:53:55 UTC |
8c268a6 | Geoffrey Sneddon | 24 September 2018, 14:19:16 UTC | Add tests for tools.wpt.testfiles, fix empty not_heads case | 10 October 2018, 14:51:33 UTC |
921d6e4 | Geoffrey Sneddon | 13 September 2018, 14:46:13 UTC | Start running Windows CI again | 10 October 2018, 14:51:32 UTC |
881c8c1 | Yutaka Hirano | 10 October 2018, 08:41:35 UTC | Simplify "WouldTaintOrigin" concept in media/blink Currently WebMediaPlayer has three predicates: - DidGetOpaqueResponseFromServiceWorker - HasSingleSecurityOrigin - DidPassCORSAccessCheck . These are used to determine whether the response body is available for scripts. They are known to be confusing, and actually MediaElementAudioSourceHandler::WouldTaintOrigin misuses them. This CL merges the three predicates to one, WouldTaintOrigin, to remove the confusion. Now the "response type" concept is available and we don't need a custom CORS check, so this CL removes BaseAudioContext::WouldTaintOrigin. This CL also renames URLData::has_opaque_data_ and its (direct and indirect) data accessors to match the spec. Bug: 849942, 875153 Change-Id: I6acf50169d7445c4ff614e80ac606f79ee577d2a Reviewed-on: https://chromium-review.googlesource.com/c/1238098 Reviewed-by: Fredrik Hubinette <hubbe@chromium.org> Reviewed-by: Kinuko Yasuda <kinuko@chromium.org> Reviewed-by: Raymond Toy <rtoy@chromium.org> Commit-Queue: Yutaka Hirano <yhirano@chromium.org> Cr-Commit-Position: refs/heads/master@{#598258} | 10 October 2018, 08:51:38 UTC |
280f780 | Anders Hartvoll Ruud | 10 October 2018, 08:00:02 UTC | [css-properties-values-api] Proper types from StylePropertyMap iterator. When iterating a StylePropertyMap, any registered custom properties in that map must have the appropriate type according to the registration. Currently all custom properties are always CSSUnparsedValue, which is only correct for unregistered properties and properties registered as "*". Fixed by parsing the value on its way out, similar to how get/getAll implements the same behavior. R=chrishtr@chromium.org Bug: 641877 Change-Id: I5e0b2d064cd0ce4827eabf464a5cf0a0e138c5db Reviewed-on: https://chromium-review.googlesource.com/c/1264540 Commit-Queue: Anders Ruud <andruud@chromium.org> Reviewed-by: Chris Harrelson <chrishtr@chromium.org> Cr-Commit-Position: refs/heads/master@{#598247} | 10 October 2018, 08:16:01 UTC |
813f11b | Anders Hartvoll Ruud | 10 October 2018, 07:42:02 UTC | [css-properties-values-api] Include initial values in computedStyleMap. Registered properties with initial values are currently missing from the computedStyleMap, because such properties are not explicitly stored on individual computed styles. To fix this, do an extra pass over the PropertyRegistrations to "fill in blanks" with initial values. R=chrishtr@chromium.org Bug: 641877 Change-Id: If36cb3ef3d419ff547e8b4073cdd9ec298c539a0 Reviewed-on: https://chromium-review.googlesource.com/c/1264575 Reviewed-by: Chris Harrelson <chrishtr@chromium.org> Commit-Queue: Anders Ruud <andruud@chromium.org> Cr-Commit-Position: refs/heads/master@{#598243} | 10 October 2018, 07:51:08 UTC |
935717d | Manuel Rego Casasnovas | 09 October 2018, 23:48:39 UTC | [css-contain] Make 'contain:layout' (not 'contain:size') suppress baseline measurements The CSSWG has recently resolved that layout containment suppress baseline alignment, while size containment does not: https://github.com/w3c/csswg-drafts/issues/2995 Spec text (https://drafts.csswg.org/css-contain/#containment-layout): "7. For the purpose of the vertical-align property, or any other property whose effects need to relate the position of the containing element's baseline to something other than its descendants, the containing element is treated as having no baseline." And a note in (https://drafts.csswg.org/css-contain/#containment-size): "Note: size containment does not suppress baseline alignment. See layout containment for that." This patch does this change just switching IsContainSize() by IsLayoutSize() in several places related to baseline alignment in the source code. With the patch several WPT tests start to pass. Apart from that, some of the tests under vendor-imports are updated to follow the new behavior. bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=1491235 gecko-commit: 4e07d3243403123217539ce6d328641daa0513da gecko-integration-branch: mozilla-inbound gecko-reviewers: dholbert | 10 October 2018, 07:07:54 UTC |
e15b5eb | Florian Rivoal | 10 October 2018, 00:06:20 UTC | [css-text] white-space collapse at start/end of line (#13387) Related to https://github.com/w3c/csswg-drafts/issues/1997 | 10 October 2018, 00:06:20 UTC |
2792963 | Emilio Cobos Álvarez | 09 October 2018, 19:53:00 UTC | Add a reftest that depends on ancestor intrinsics being cleared on bsize changes. Differential Revision: https://phabricator.services.mozilla.com/D7921 bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=1496833 gecko-commit: f3480613106993d658f9b5a46f41e59ca9934a5e gecko-integration-branch: autoland gecko-reviewers: dholbert | 09 October 2018, 22:16:23 UTC |
3971449 | Philip Jägenstedt | 09 October 2018, 22:11:58 UTC | Add edenchuang as payment-request/ reviewer (#13440) Repeats https://github.com/web-platform-tests/wpt/pull/12352 against the master branch. | 09 October 2018, 22:11:58 UTC |
8b09ece | Robert Ma | 27 September 2018, 20:53:09 UTC | Rework encode/decode of headers in Request This change changes the encoding/decoding of headers in Request and username/password in Authentication: now all keys and values have binary type (because of an implementation detail in the Python 3 standard library, we actually need to re-encode the headers back to bytes in Python 3). Documentation and comments are also improved to clarify the encoding situation. Also add test cases for non-ASCII characters in auth headers. | 09 October 2018, 21:25:09 UTC |
d79ae19 | Robert Ma | 27 September 2018, 20:45:30 UTC | Add a Unicode smoke test to wptserve The test sends a request to wptserve with non-ASCII characters in a header and sets up a simple handler to return the value of that header. The server shouldn't crash in either Python 2 or 3, and the response should not be garbled. The server crashes in Python 2 (#13204). | 09 October 2018, 21:25:09 UTC |
f0686bd | Ms2ger | 24 September 2018, 11:42:16 UTC | Reland "Various test fixes for python3 support. (#11769)" | 09 October 2018, 21:25:09 UTC |
600cd35 | Morten Stenshorne | 09 October 2018, 20:06:55 UTC | [LayoutNG] Auto top/bottom margins may resolve to negative values. The spec doesn't say that auto margin-top or margin-bottom need to resolve to something non-negative, so just allow negative values as well. They don't cause over-constrainedness. Also, when distributing the space to margin-top and margin-bottom, make sure that everything gets distributed. Set margin-top to the half, and set margin-bottom to total space MINUS margin-top. Just dividing both by two might cause us to lose a tiny fraction of the total, due to rounding. Bug: 888270 Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_layout_ng Change-Id: Ia10577851e60707a0e2d67857a1c43a12e28b9d2 Reviewed-on: https://chromium-review.googlesource.com/c/1269855 Commit-Queue: Morten Stenshorne <mstensho@chromium.org> Reviewed-by: Emil A Eklund <eae@chromium.org> Reviewed-by: Christian Biesinger <cbiesinger@chromium.org> Reviewed-by: Aleks Totic <atotic@chromium.org> Cr-Commit-Position: refs/heads/master@{#598052} | 09 October 2018, 20:17:53 UTC |
caff176 | Ehsan Karamad | 09 October 2018, 19:28:32 UTC | No vertical scroll if kVerticalScroll disabled 'vertical-scroll' policy is a policy-controlled feature which allows developers to control scrollability of their web pages and disallow embedded content from blocking vertical scrolling. In line with the implementation of the feature, this CL ensures that any content inside a document which has the feature disabled will not be able to consume vertical scroll deltas. Note that the current implementation of the policy only deals with methods of blocking input from turning into scroll gesture. This CL complements that controlling the consumption of such gestures. Bug: 853485 Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_slimming_paint_v2;master.tryserver.blink:linux_trusty_blink_rel Change-Id: I7db68e5897998ba29f2ad75cef44660563d6b95b Reviewed-on: https://chromium-review.googlesource.com/c/1252530 Reviewed-by: David Bokan <bokan@chromium.org> Reviewed-by: Ehsan Karamad <ekaramad@chromium.org> Commit-Queue: Ehsan Karamad <ekaramad@chromium.org> Cr-Commit-Position: refs/heads/master@{#598035} | 09 October 2018, 19:34:05 UTC |
0ba93eb | James Graham | 09 October 2018, 13:40:49 UTC | Fixup a missing return statement in wpt logging setup bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=1275926 gecko-commit: 8b5926fac04e55e492e998b87920931cd32f3a73 gecko-integration-branch: mozilla-inbound gecko-reviewers: me | 09 October 2018, 18:42:38 UTC |
546a3e6 | pyup-bot | 09 October 2018, 16:43:42 UTC | Update urllib3 from 1.22 to 1.23 | 09 October 2018, 18:32:39 UTC |
25a1c15 | Andy Paicu | 09 October 2018, 12:29:52 UTC | Inherit CSP when self-navigating to local-scheme URL As the linked bug example shows, we should inherit CSP when we navigate to a local-scheme URL (even if we are in a main browsing context). Bug: 799747 Change-Id: I8413aa8e8049461ebcf0ffbf7b04c41d1340af02 Reviewed-on: https://chromium-review.googlesource.com/c/1234337 Reviewed-by: Mike West <mkwst@chromium.org> Commit-Queue: Andy Paicu <andypaicu@chromium.org> Cr-Commit-Position: refs/heads/master@{#597889} | 09 October 2018, 12:40:18 UTC |
63ef84e | Olga Sharonova | 09 October 2018, 12:08:37 UTC | Revert "Resubmit cr/1261477" This reverts commit e2d1946acc8a4c2e0651958d663ef04c3eefe202. Reason for revert: FindIt identified it as a source of flakiness in crbug/893511 Bug: 893511 Original change's description: > Resubmit cr/1261477 > > I reproduced the failure locally, and then re-ran with the changes 100 > time successfully > > TBR=peter@chromium.org > > Change-Id: Ieb60211fcd9c6beb1aec2b152f298335c4d9178b > Reviewed-on: https://chromium-review.googlesource.com/c/1264640 > Reviewed-by: Mugdha Lakhani <nator@chromium.org> > Commit-Queue: Rayan Kanso <rayankans@chromium.org> > Cr-Commit-Position: refs/heads/master@{#597106} TBR=nator@chromium.org,rayankans@chromium.org # Not skipping CQ checks because original CL landed > 1 day ago. Change-Id: I8e0d0d5339f3f2ea8106dc76bf5bf7d4b8f0d243 Reviewed-on: https://chromium-review.googlesource.com/c/1270736 Reviewed-by: Olga Sharonova <olka@chromium.org> Commit-Queue: Olga Sharonova <olka@chromium.org> Cr-Commit-Position: refs/heads/master@{#597883} | 09 October 2018, 12:18:58 UTC |
ba7e07f | Yoshifumi Inoue | 09 October 2018, 11:31:27 UTC | Make Element#innerText not to emit a space after inline-block and before block This patch makes Element#innerText not to emit a space after inline-block and before block for improving interop. This patch fixes regression caused by the patch[1]. [1] http://crrev.com/c/1250825 Make Element#innerText to not collapse white space around inline-block Bug: 892500 Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_layout_ng Change-Id: I4343c0d74f3445263a209697aaf3c24aef82a900 Reviewed-on: https://chromium-review.googlesource.com/c/1264339 Reviewed-by: Xiaocheng Hu <xiaochengh@chromium.org> Commit-Queue: Yoshifumi Inoue <yosin@chromium.org> Cr-Commit-Position: refs/heads/master@{#597879} | 09 October 2018, 11:37:23 UTC |
1f32249 | Philip Jägenstedt | 09 October 2018, 11:09:22 UTC | Remove last remnants of chrome_webdriver (#13432) It got left here because of two concurrent PRs: https://github.com/web-platform-tests/wpt/pull/12726 https://github.com/web-platform-tests/wpt/pull/13277 | 09 October 2018, 11:09:22 UTC |
21913d5 | Andy Paicu | 09 October 2018, 10:58:52 UTC | Directive names should be case-insensitive Spec: https://w3c.github.io/webappsec-csp/#parse-serialized-policy Fixed Chromium's behaviour Added unit tests Added WPT tests Bug: 893119 Change-Id: I0ee1114fd90a933071ff03c9a64b95431a17be95 Reviewed-on: https://chromium-review.googlesource.com/c/1268316 Commit-Queue: Andy Paicu <andypaicu@chromium.org> Reviewed-by: Mike West <mkwst@chromium.org> Cr-Commit-Position: refs/heads/master@{#597873} | 09 October 2018, 11:04:21 UTC |
df87b8c | Andy Paicu | 09 October 2018, 08:22:16 UTC | Added tests to ensure that script blocks are converted to utf-8 before hashing https://github.com/w3c/webappsec-csp/issues/109 Bug: 892570 Change-Id: I52d0ff8ab6abd58de9503992f44ea42df50cb6b7 Reviewed-on: https://chromium-review.googlesource.com/c/1264536 Reviewed-by: Mike West <mkwst@chromium.org> Commit-Queue: Andy Paicu <andypaicu@chromium.org> Cr-Commit-Position: refs/heads/master@{#597836} | 09 October 2018, 08:32:36 UTC |
434ca47 | Florian Rivoal | 08 October 2018, 23:55:30 UTC | [css-text] Control characters should be visible (#13389) Tests for https://github.com/w3c/csswg-drafts/issues/1990 and https://github.com/w3c/csswg-drafts/issues/855 | 08 October 2018, 23:55:30 UTC |
e124e16 | Florian Rivoal | 08 October 2018, 23:23:38 UTC | [css-text-3] Tests for percentage text indent Relates to the spec changes decided in https://github.com/w3c/csswg-drafts/issues/2394 | 08 October 2018, 23:23:38 UTC |
907455e | Philip Jägenstedt | 08 October 2018, 20:13:44 UTC | Fix lint by syncing all support/grid.css | 08 October 2018, 20:39:37 UTC |
d66f261 | Manuel Rego Casasnovas | 08 October 2018, 13:24:37 UTC | [css-grid] Fix percentages in relative offsets for grid items The method LayoutBoxModelObject::RelativePositionOffset() was not considering the case of grid items, where the containing block is the grid area. The patch modifies the method so the new code uses OverrideContainingBlockContentLogicalWidth|Height when required. Two new tests are added, the first one does not use percentages and is already passing. The second one has the very same output but using percentages which was not working before this patch. BUG=835607 TEST=external/wpt/css/css-grid/grid-items/grid-items-relative-offsets-001.html TEST=external/wpt/css/css-grid/grid-items/grid-items-relative-offsets-002.html Change-Id: Icb76f4a521566ad36f87924835b21ae450f2cb24 Reviewed-on: https://chromium-review.googlesource.com/c/1238726 Reviewed-by: Javier Fernandez <jfernandez@igalia.com> Commit-Queue: Manuel Rego <rego@igalia.com> Cr-Commit-Position: refs/heads/master@{#597543} | 08 October 2018, 20:39:37 UTC |
6f0a96e | Philip Jägenstedt | 08 October 2018, 15:02:13 UTC | Drop the use of the `hosts` addon in Travis The hosts list is out of date and this isn't needed, as tools/ci/lib.sh does the hosts fixup when needed. Drive-by: remove a redundant `sudo: required`. | 08 October 2018, 17:11:41 UTC |
a239c7d | Geoffrey Sneddon | 08 October 2018, 16:10:16 UTC | Avoid serve/test_functional.py#test_subprocess_exit relying on hosts Without bind_address: False we use the server address for the stash, and this obviously doesn't work without the hosts being set up. | 08 October 2018, 17:11:41 UTC |
64842c0 | Philip Jägenstedt | 08 October 2018, 14:21:36 UTC | Add one missing conditional for "stability (Firefox Nightly)" (#13423) Missed in https://github.com/web-platform-tests/wpt/pull/13421. | 08 October 2018, 14:21:36 UTC |
058b882 | Philip Jägenstedt | 08 October 2018, 13:59:20 UTC | Use conditionals to avoid running no-op jobs on Travis See documentation: https://docs.travis-ci.com/user/conditional-builds-stages-jobs/ https://docs.travis-ci.com/user/conditions-v1 | 08 October 2018, 14:06:53 UTC |
5810d07 | Rune Lillesveen | 08 October 2018, 13:50:07 UTC | Make :host account for its own pseudo-class specificity. Otherwise `:host div` is equally specific to `div`, which seems undesirable and doesn't match any spec. This is only done for Shadow DOM v1 since there are backward compat issues with Polymer and WebUI. Intent to ship: https://groups.google.com/a/chromium.org/forum/#!topic/blink-dev/gYrR2nR6quI Bug: 857415 Change-Id: Ieaaf9b835ae7136806ced8f6e72fb30f9adf02af Reviewed-on: https://chromium-review.googlesource.com/c/1250963 Commit-Queue: Rune Lillesveen <futhark@chromium.org> Reviewed-by: Emilio Cobos Álvarez <emilio@chromium.org> Cr-Commit-Position: refs/heads/master@{#597549} | 08 October 2018, 14:00:54 UTC |
ddf2b44 | Philip Jägenstedt | 08 October 2018, 13:04:21 UTC | Remove tools/wptserve/.travis.yml It's no longer used after wptserve was merged into this repo. | 08 October 2018, 13:24:23 UTC |
698d355 | Morten Stenshorne | 08 October 2018, 12:59:19 UTC | [LayoutNG] Correct LayoutText::LinesBoundingBox() for vertical-rl. NG needs to produce a rectangle with the block-axis offset relatively to block-start of the container, just like legacy does. Using purely physical coordinates is wrong. This fixes issues with scroll anchoring, and possibly other things too. One test had to be rebaselined, due to a rounding change. LayoutText::DebugRect() calls LinesBoundingBox(), which now flips the block-axis coordinate for vertical-rl. This affects rounding of the size of the rectangle. Bug: 889449 Change-Id: I7053ed7fe05ce443f53a128660d60f27fea7f8b3 Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_layout_ng Reviewed-on: https://chromium-review.googlesource.com/c/1264596 Commit-Queue: Morten Stenshorne <mstensho@chromium.org> Reviewed-by: Koji Ishii <kojii@chromium.org> Cr-Commit-Position: refs/heads/master@{#597541} | 08 October 2018, 13:09:25 UTC |
300f8e2 | Philip Jägenstedt | 08 October 2018, 11:39:02 UTC | Ignore "runner_teardown" messages in `TestRunnerManager.cleanup()` This is an alternative to https://github.com/web-platform-tests/wpt/pull/13409, which adds more handling around this rather than removing it. Fixes https://github.com/web-platform-tests/wpt/issues/13407. | 08 October 2018, 12:01:07 UTC |
a9653ad | Philip Jägenstedt | 07 October 2018, 20:47:35 UTC | Move a comment in TestRunnerManager.wait_finished to make more sense It's certaintly not the logging that does post-stop processing, but `after_test_end` might, so the comment is accurate if moved. The comment was added in commit 7f422ef4b1b1ecf09ff7ffc94b60c6b7bf7a41b4. | 08 October 2018, 10:20:43 UTC |
af795eb | Philip Jägenstedt | 07 October 2018, 21:36:19 UTC | Drop the Thread-/Process- prefixes for thread/process names | 08 October 2018, 09:57:53 UTC |
d30010a | Philip Jägenstedt | 07 October 2018, 19:09:00 UTC | Fix Thread-TestRunner misnomer | 08 October 2018, 09:57:53 UTC |
5e46692 | Philip Jägenstedt | 07 October 2018, 16:25:12 UTC | Normalize messages around TestRunner/Manager more | 08 October 2018, 09:57:53 UTC |
2f50699 | Philip Jägenstedt | 07 October 2018, 13:37:58 UTC | Rename TestManager to TestRunnerManager in documentation/logging (There is no TestManager class.) | 08 October 2018, 09:57:53 UTC |
a772507 | Philip Jägenstedt | 07 October 2018, 13:27:21 UTC | Improve warnings of command left in queue in TestRunnerManager cleanup Part of https://github.com/web-platform-tests/wpt/issues/13407. | 08 October 2018, 09:57:53 UTC |
1cefac0 | Anders Hartvoll Ruud | 08 October 2018, 09:31:26 UTC | [css-properties-values-api] Proper behavior for direct CSSStyleValues. In CSS Typed OM, types that are not yet supported by a specialized CSSStyleValue subclass are represented by "direct" CSSStyleValue objects. These objects are only valid for their associated property. For instance, you may not get a (direct) CSSStyleValue object from 'color', and then set it on 'background-color'; the (direct) CSSStyleValue acquired from 'color' is valid for 'color' only. The same should apply for registered custom properties; a direct CSSStyleValue acquired from '--x' is only valid for '--x'. To do this, blink::CSSUnsupportedStyleValue must be aware of the custom property name. This means the AtomicString with the custom property name must be passed around to almost everything. However, if the property at hand is not a custom property, I want to avoid AtomicString-ing the name of the property. Hence there are DCHECKs to make sure this is avoided. Reminder: this CL would be much less ... grotesque and perhaps less error-prone if dynamic CSSProperties (Ribbon) were in place, but because we like pain, we decided to not give that project a very high priority. R=chrishtr@chromium.org Bug: 641877 Change-Id: I99af4745d1d33365759e9fd4acef5082ab213229 Reviewed-on: https://chromium-review.googlesource.com/c/1257916 Reviewed-by: Chris Harrelson <chrishtr@chromium.org> Commit-Queue: Anders Ruud <andruud@chromium.org> Cr-Commit-Position: refs/heads/master@{#597503} | 08 October 2018, 09:42:19 UTC |
a950440 | James Graham | 19 September 2018, 12:33:05 UTC | Add element_by_selector to SelectorProtocolPart | 08 October 2018, 09:27:05 UTC |
723dcaf | James Graham | 17 September 2018, 12:16:48 UTC | Fix error handling for testdriver tests. Otherwise a protocol error leads to FAIL not ERROR. | 08 October 2018, 09:27:05 UTC |
01dfc0f | James Graham | 04 September 2018, 18:05:27 UTC | [testdriver] Add a builder object for constructing actions Action sequences are not really sensible to construct by hand, so this adds a builder object for constructing sequences of actions. To use it the following script mut be included after testdriver.js: <script src="/resources/testdriver-actions.js"></script> Then actions sequences can be constructed like // Drag the mouse from 0,0 to 100,100 with ctrl held down test_driver.Actions() .pointerMove(0, 0, origin="viewport") .keyDown("\uE009") .pointerDown() .pointerMove(100, 100, origin="viewport") .pointerUp() .keyUp() .send(); | 08 October 2018, 09:27:05 UTC |
f1db6b2 | James Graham | 04 September 2018, 18:04:38 UTC | [testdriver] Support sending actions sequences over WebDriver | 08 October 2018, 09:27:05 UTC |
ceadf71 | Navid Zolghadr | 28 August 2018, 16:20:56 UTC | Add action sequence injection in test_driver Following web driver spec for pointer action sequence and key actions this change adds this API to the test_driver interface. Use action_sequence instead of pointer_action_sequence | 08 October 2018, 09:27:05 UTC |
b0d2d2d | jgraham | 07 October 2018, 20:27:17 UTC | Refactor the testharness/testdriver messaging protocol implementation (#13367) This causes most of the code to be run once when the runner window is opened rather than being run on each resume. This should be a little more efficient. However the main motivation is to have a single long-lived event handler rather than having to ensure that the event handler is correctly cleaned up after each test. Previously this was not done carefully and it was possible for events from one test to affect subsequent tests. | 07 October 2018, 20:27:17 UTC |
e0a939b | Philip Jägenstedt | 07 October 2018, 16:42:48 UTC | Typo: runder → under | 07 October 2018, 17:19:06 UTC |
ba4ca50 | Philip Jägenstedt | 06 October 2018, 21:45:40 UTC | Fix assert for StorageEvent constructor test (#13405) Reported by @Zirro: https://github.com/web-platform-tests/wpt/pull/13368/files/56c3da3f1c6921a9a80433657d49b42bd8c05596#r223192022 | 06 October 2018, 21:45:40 UTC |
21c33fe | Philip Jägenstedt | 06 October 2018, 18:23:59 UTC | Minimize <video> dependency in `test_driver.bless` test (#13403) See diagnosis of Safari Technology Preview failure: https://github.com/web-platform-tests/wpt/issues/12621#issuecomment-427525317 This test does not need to load any media resource or consider the state of the returned promise. | 06 October 2018, 18:23:59 UTC |
c2b30ef | Michael[tm] Smith | 05 October 2018, 23:55:54 UTC | cc Update messages.json | 06 October 2018, 00:10:20 UTC |
c09b800 | Jan-Ivar Bruaroey | 05 October 2018, 18:31:14 UTC | Rewrite MediaStream-MediaElement-srcObject.https.html to use async/await. Differential Revision: https://phabricator.services.mozilla.com/D7568 bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=1493885 gecko-commit: 690cfe65a8c961ec5b2e56f8249498693e26edd0 gecko-integration-branch: autoland gecko-reviewers: pehrsons | 05 October 2018, 23:33:04 UTC |
c8c0f16 | Jan-Ivar Bruaroey | 05 October 2018, 18:27:31 UTC | Update MediaStream-MediaElement-srcObject.https.html to spec, and bring back test that currentTime changes must be ignored. Differential Revision: https://phabricator.services.mozilla.com/D7567 bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=1493885 gecko-commit: 690011b571305eb937172d9d68e5dd4260f6e33e gecko-integration-branch: autoland gecko-reviewers: pehrsons | 05 October 2018, 23:33:04 UTC |
bd5f340 | Philip Jägenstedt | 05 October 2018, 13:16:01 UTC | Consolidate/simplify StorageEvent constructor tests The use of `async_test` was already unnecessary as `dispatchEvent` synchronously invokes listeners, and use of `dispatcEvent` itself is not necessary to test the constructor, so simplify to just `test`s. | 05 October 2018, 19:34:14 UTC |
c4fce5c | Philip Jägenstedt | 04 October 2018, 19:35:05 UTC | Test initStorageEvent() Follows https://github.com/whatwg/html/pull/4063. | 05 October 2018, 19:34:14 UTC |
f159b42 | Michael[tm] Smith | 05 October 2018, 19:10:09 UTC | cc Add some eol newlines to a couple files | 05 October 2018, 19:17:47 UTC |
4628f52 | Mugdha Lakhani | 05 October 2018, 18:44:21 UTC | [Background Fetch] Surface background-fetch permission. "background-fetch" permission was added to the Permissions API with this commit: https://github.com/w3c/permissions/pull/183/commits/096436eb59caaa5b244b0514d84f287392cb069b Chrome uses Automatic Downloads for this permission, so as not to add yet another user facing permission to the UI. This CL: 1. Adds a background_fetch content setting, because that's needed by PermissionContext and allows us to decouple from Automatic Downloads content setting. 2. Adds a BackgroundFetch PermissionContext mapped to this content setting. 3. Makes this PermissionContext decide whether background fetch() is permitted, which we do by: a. Querying DownloadRequestLimiter when there' a top level frame. b. Querying the Automatic Downloads content setting in other cases. For context and details, please see this document: https://docs.google.com/document/d/1rPYSlbzScw_6PLUJ3m96ZLIkxXVAWyBSL75-VDJEMso/edit?usp=sharing Bug: 886896 Change-Id: Id0fcc64d4242290f3782c61ef11babe5717b409f Reviewed-on: https://chromium-review.googlesource.com/c/1233714 Commit-Queue: Mugdha Lakhani <nator@chromium.org> Reviewed-by: Mounir Lamouri <mlamouri@chromium.org> Reviewed-by: Rayan Kanso <rayankans@chromium.org> Reviewed-by: Tobias Sargeant <tobiasjs@chromium.org> Reviewed-by: Raymes Khoury <raymes@chromium.org> Reviewed-by: Alex Moshchuk <alexmos@chromium.org> Reviewed-by: Dominick Ng <dominickn@chromium.org> Reviewed-by: Steven Bennetts <stevenjb@chromium.org> Cr-Commit-Position: refs/heads/master@{#597209} | 05 October 2018, 18:55:13 UTC |
3229936 | James Graham | 05 October 2018, 16:45:53 UTC | Ignore lint failures until we import a rev with them fixed, a=testonly bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=1495168 gecko-commit: f23487d9ab99d486fe3ab7f261e4d4e0bf1172ae gecko-integration-branch: central gecko-reviewers: testonly | 05 October 2018, 18:27:57 UTC |
1fa9850 | Ting-Yu Lin | 05 October 2018, 16:45:40 UTC | Fix wpt test multicol-span-all-margin-bottom-001-ref.xht. Per example in the spec https://drafts.csswg.org/css-multicol-1/#column-span ... the bottom margin of the second spanner does not collapse with the top margin of the subsequent element. Likewise, in the test file multicol-span-all-margin-bottom-001.xht, the bottom margin of the black spanner <h4> should not collapse with the top margin of the orange non-spanner <h4>. Differential Revision: https://phabricator.services.mozilla.com/D7670 bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=1496275 gecko-commit: 2f3ac50f1ce9fb8c9add3390085843975a5be10d gecko-integration-branch: central gecko-reviewers: dbaron | 05 October 2018, 17:35:52 UTC |
8e9df70 | Rayan Kanso | 05 October 2018, 14:56:13 UTC | [Background Fetch] Reject calls to get with invalid IDs. Explicitly check for empty strings and add a WPT test. TBR=peter@chromium.org Change-Id: If969b56382a67419aaee742c47c470efebab95af Reviewed-on: https://chromium-review.googlesource.com/c/1264541 Reviewed-by: Rayan Kanso <rayankans@chromium.org> Reviewed-by: Mugdha Lakhani <nator@chromium.org> Commit-Queue: Rayan Kanso <rayankans@chromium.org> Cr-Commit-Position: refs/heads/master@{#597115} | 05 October 2018, 15:01:56 UTC |
786e201 | Rayan Kanso | 05 October 2018, 14:33:45 UTC | Resubmit cr/1261477 I reproduced the failure locally, and then re-ran with the changes 100 time successfully TBR=peter@chromium.org Change-Id: Ieb60211fcd9c6beb1aec2b152f298335c4d9178b Reviewed-on: https://chromium-review.googlesource.com/c/1264640 Reviewed-by: Mugdha Lakhani <nator@chromium.org> Commit-Queue: Rayan Kanso <rayankans@chromium.org> Cr-Commit-Position: refs/heads/master@{#597106} | 05 October 2018, 14:44:52 UTC |
ed1e187 | Josh Matthews | 05 October 2018, 00:27:50 UTC | Add missing reject argument that is referenced. | 05 October 2018, 13:48:56 UTC |
006e385 | pyup-bot | 28 September 2018, 17:15:19 UTC | Update mozrunner from 7.0.2 to 7.1.0 | 05 October 2018, 09:22:29 UTC |
9efb409 | Ms2ger | 04 October 2018, 08:45:16 UTC | Python 3: Don't decode ReplacementTokenizer results. ReplacementTokenizer was changed to always return text in a4b7841ea1345ffdcbd79e399f7ff1982fab5362, while this decode() call was added around the same time in 27406c9d238ac80c43bb62e87cd0305f59294527. This changes the code to verify that the type of the field local is as expected. | 05 October 2018, 09:19:54 UTC |
13654e2 | Philip Jägenstedt | 05 October 2018, 08:47:17 UTC | Fix typo funtion/funtional Fun as it sounds, this is no place for having fun! (jk) | 05 October 2018, 09:18:39 UTC |
0d90b8b | Morten Stenshorne | 05 October 2018, 08:09:36 UTC | Add elementsFromPoint() tests for non-atomic inlines. Added a comment to LayoutInline, pointing out that LayoutNG is doing the right thing already, while legacy is wrong. Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_layout_ng Change-Id: I026404af70c33b0a94900773636d5e138a725976 Reviewed-on: https://chromium-review.googlesource.com/c/1261436 Reviewed-by: Aleks Totic <atotic@chromium.org> Reviewed-by: Koji Ishii <kojii@chromium.org> Commit-Queue: Morten Stenshorne <mstensho@chromium.org> Cr-Commit-Position: refs/heads/master@{#597027} | 05 October 2018, 08:19:43 UTC |
47cea8c | autofoolip | 05 October 2018, 06:58:21 UTC | Update interfaces/webrtc.idl (#13378) Source: https://github.com/tidoust/reffy-reports/blob/f0a4f8f/whatwg/idl/webrtc.idl Build: https://travis-ci.org/tidoust/reffy-reports/builds/437465234 | 05 October 2018, 06:58:21 UTC |
39e7ba9 | Yoshifumi Inoue | 05 October 2018, 05:44:12 UTC | Make Element#innerText to collapse white spaces after <br> This patch makes Element#innerText to collapse white space after <br> to avoid emitting extra space for improving interop. Bug: 892491 Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_layout_ng Change-Id: I8dd429a8af6ecd2e6d5195d6a6ec7b0214c6345c Reviewed-on: https://chromium-review.googlesource.com/c/1263638 Reviewed-by: Yoichi Osato <yoichio@chromium.org> Commit-Queue: Yoshifumi Inoue <yosin@chromium.org> Cr-Commit-Position: refs/heads/master@{#597004} | 05 October 2018, 05:54:15 UTC |
20d9569 | Ian Clelland | 05 October 2018, 04:22:27 UTC | Enable FP reporting for PIP policy violations. This queues a report through the Reporting API whenever requestPictureInPicture is called in a frame in which the 'picture-in-picture' feature is not allowed according to feature policy. Bug: 867471 Change-Id: I1cede10385f6a5110fd0ddd5bf124ff5d3737dd9 Reviewed-on: https://chromium-review.googlesource.com/c/1262432 Commit-Queue: Ian Clelland <iclelland@chromium.org> Reviewed-by: François Beaufort <beaufort.francois@gmail.com> Reviewed-by: Mounir Lamouri <mlamouri@chromium.org> Cr-Commit-Position: refs/heads/master@{#596988} | 05 October 2018, 04:28:24 UTC |
45fe4bd | Ian Clelland | 05 October 2018, 02:47:16 UTC | Enable FP reporting for EME policy violations. This queues a report through the Reporting API whenever requestMediaKeySystemAccess is called in a frame in which the 'encrypted-media' feature is not allowed according to feature policy. Bug: 867471 Change-Id: I781cf9d9719bcbaffe0b328c2b6bfa5ca197920a Reviewed-on: https://chromium-review.googlesource.com/c/1261921 Reviewed-by: David Dorwin <ddorwin@chromium.org> Commit-Queue: Ian Clelland <iclelland@chromium.org> Cr-Commit-Position: refs/heads/master@{#596965} | 05 October 2018, 02:57:23 UTC |
30c6ad6 | Henrik Boström | 01 October 2018, 08:47:08 UTC | [Unified Plan] Remote MediaStreamTracks should be muted by default. Per-spec, tracks that are created with a receiver (e.g. addTrack or setRemoteDescription) are muted by default. Prior to this CL they were unmuted by default, whether or not they were receiving any packets. A correct implementation should unmute the tracks when RTP packets arrive. We are not quite there yet, this CL assumes that if the receiver becomes active through renegotiation it will unmute. We are careful to make sure that the track is muted on the "ontrack" event so that the application has time to wire up the "onunmute" event. By unmuting as part of processing SDP we fix the Unified Plan bug where a remote track that had previously been muted was not unmuted when becoming active again (transciever.currentDirection == 'sendrecv' or 'recvonly'), https://crbug.com/884023. This CL also makes "ontrack" fire synchronously per-spec, https://crbug.com/788558. Note that some stream events still fire asynchronously, which means they now fire after the "ontrack" event in the Unified Plan case. This is remaining work on that bug. A new file is created to test muting related behaviors, and some helper functions used this and another file are moved to RTCPeerConnection-helper.js. Bug: 884023, 777619, 788558 Change-Id: I8dc3e2adf04e72282f085779639edc73bacfc86b Reviewed-on: https://chromium-review.googlesource.com/1249066 Commit-Queue: Henrik Boström <hbos@chromium.org> Reviewed-by: Guido Urdaneta <guidou@chromium.org> Cr-Commit-Position: refs/heads/master@{#595405} | 05 October 2018, 02:21:50 UTC |
1400d50 | Josh Matthews | 05 October 2018, 00:14:14 UTC | Fix typo in focus-visible selector test. | 05 October 2018, 01:36:45 UTC |
b4c8ec5 | Marijn Kruisselbrink | 04 October 2018, 21:31:22 UTC | Revert "[Background Fetch] Add WPT tests for BackgroundFetchRegistration::abort" This reverts commit c9bd394737bb75c77146870df92e636da8fe1ac6. Reason for revert: test is flaky in https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Mac10.10%20Tests/35061 Original change's description: > [Background Fetch] Add WPT tests for BackgroundFetchRegistration::abort > > Simplify event dispatching code, and also allow for match/matchAll to be > called with the abort event. > > Change-Id: I3a9d17102f176f7c68131ea144e555a44ffd6ab1 > Reviewed-on: https://chromium-review.googlesource.com/c/1261477 > Commit-Queue: Rayan Kanso <rayankans@chromium.org> > Reviewed-by: Peter Beverloo <peter@chromium.org> > Reviewed-by: Mugdha Lakhani <nator@chromium.org> > Cr-Commit-Position: refs/heads/master@{#596723} TBR=peter@chromium.org,nator@chromium.org,rayankans@chromium.org Change-Id: I70bb246625c9d98707c6f61a91dc2dd658b1027f No-Presubmit: true No-Tree-Checks: true No-Try: true Reviewed-on: https://chromium-review.googlesource.com/c/1262965 Reviewed-by: Marijn Kruisselbrink <mek@chromium.org> Commit-Queue: Marijn Kruisselbrink <mek@chromium.org> Cr-Commit-Position: refs/heads/master@{#596854} | 04 October 2018, 21:46:53 UTC |
edee96b | Morten Stenshorne | 04 October 2018, 20:56:10 UTC | [LayoutNG] Correct LayoutText::Quads() for vertical-rl. NG needs to produce a rectangle with the block-axis offset relatively to block-start of the container, just like legacy does. Using purely physical coordinates is wrong. Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_layout_ng Change-Id: I9989628dcc339c71d850818498538aafa142ed29 Reviewed-on: https://chromium-review.googlesource.com/c/1261698 Reviewed-by: Koji Ishii <kojii@chromium.org> Commit-Queue: Morten Stenshorne <mstensho@chromium.org> Cr-Commit-Position: refs/heads/master@{#596747} | 04 October 2018, 21:20:06 UTC |
c93d2c9 | Daniel Vogelheim | 04 October 2018, 20:55:39 UTC | Trusted Types: Implement "trusted-types *" and reserved names. - An empty trusted-types declaration means no policies are allowed. - A declaration of only "*" means all policies are allowed. - Quoted strings are considered reserved (and thus cannot be policy names) This aligns the implementation close with the trusted types polyfill. Bug: 739170 Change-Id: If9fc66e80d60a81ec4228da35ae8c8820ebaa9b5 Reviewed-on: https://chromium-review.googlesource.com/c/1249269 Commit-Queue: Daniel Vogelheim <vogelheim@chromium.org> Reviewed-by: Andy Paicu <andypaicu@chromium.org> Cr-Commit-Position: refs/heads/master@{#596638} | 04 October 2018, 21:20:02 UTC |
23cc124 | Rayan Kanso | 04 October 2018, 20:56:03 UTC | [Background Fetch] Add WPT tests for BackgroundFetchRegistration::abort Simplify event dispatching code, and also allow for match/matchAll to be called with the abort event. Change-Id: I3a9d17102f176f7c68131ea144e555a44ffd6ab1 Reviewed-on: https://chromium-review.googlesource.com/c/1261477 Commit-Queue: Rayan Kanso <rayankans@chromium.org> Reviewed-by: Peter Beverloo <peter@chromium.org> Reviewed-by: Mugdha Lakhani <nator@chromium.org> Cr-Commit-Position: refs/heads/master@{#596723} | 04 October 2018, 21:14:20 UTC |
884ed4b | Mugdha Lakhani | 04 October 2018, 20:55:57 UTC | [Background Fetch] Move CORS preflight check to the controller. Bug: 889888 Change-Id: I2429361627ffea2e1b8d5a11eb0e34b0cbfa1abb Reviewed-on: https://chromium-review.googlesource.com/c/1249265 Reviewed-by: Joshua Bell <jsbell@chromium.org> Reviewed-by: Rayan Kanso <rayankans@chromium.org> Reviewed-by: Peter Beverloo <peter@chromium.org> Commit-Queue: Mugdha Lakhani <nator@chromium.org> Cr-Commit-Position: refs/heads/master@{#596707} | 04 October 2018, 21:14:16 UTC |
2a98839 | Andy Paicu | 04 October 2018, 20:55:51 UTC | Fixed object-src tests As part of checking if the linked bug is still an issue, I have taken the opportunity to fix the current mostly non-sensical tests. Bug: 240058 Change-Id: I716d43d38be6dd161aa0437dbda03f2c77eb6d88 Reviewed-on: https://chromium-review.googlesource.com/c/1225886 Commit-Queue: Andy Paicu <andypaicu@chromium.org> Reviewed-by: Mike West <mkwst@chromium.org> Cr-Commit-Position: refs/heads/master@{#596662} | 04 October 2018, 21:14:13 UTC |
f2ffe05 | Morten Stenshorne | 04 October 2018, 20:55:45 UTC | Add offsetLeft/offsetTop tests for non-atomic inlines. Added a comment to LayoutInline, pointing out that LayoutNG seems to be doing the right thing already, while legacy is wrong. Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_layout_ng Change-Id: I0f906dab6efa7b3d4c72a297926d542f77251052 Reviewed-on: https://chromium-review.googlesource.com/c/1261575 Reviewed-by: Koji Ishii <kojii@chromium.org> Commit-Queue: Morten Stenshorne <mstensho@chromium.org> Cr-Commit-Position: refs/heads/master@{#596657} | 04 October 2018, 21:08:25 UTC |
a13506a | Morten Stenshorne | 04 October 2018, 20:55:33 UTC | [LayoutNG] Correct getClientRects for inlines in vertical-rl. In LayoutInline::GenerateLineBoxRects(), when collecting rectangles, for each rectangle, we need to flip its block axis offset, to make it relative to the block-start of the container, because this is what's expected by the callback. GenerateLineBoxRects() is used to implement Element.getClientRects() (among other things). Added a test that used to fail with LayoutNG, since this obviously lacked test coverage. Also had to fix LayoutText::AbsoluteQuadsForRange(), or fast/writing-mode/flipped-blocks-text-map-local-to-container.html would regress. It used to pass because getClientRects for a text node and getClientRects for an inline (separate code paths) had the same bug. Also had to update NGPhysicalBoxFragment::AddSelfOutlineRects(). Now that GenerateLineBoxRects() (called from LayoutInline::AddOutlineRects()) behaves consistently and always produces rectangles with the block offset relative to the block-start of the container (rather than physical top or left for NG), we need to flip vertical-rl rects to become purely physical on our own. Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_layout_ng Change-Id: I6a5d6ab6012c20b3a2fd943cb9cd3c6ee53bf052 Reviewed-on: https://chromium-review.googlesource.com/c/1257917 Commit-Queue: Morten Stenshorne <mstensho@chromium.org> Reviewed-by: Koji Ishii <kojii@chromium.org> Reviewed-by: Aleks Totic <atotic@chromium.org> Cr-Commit-Position: refs/heads/master@{#596576} | 04 October 2018, 21:08:20 UTC |
e9a0828 | Morten Stenshorne | 04 October 2018, 20:55:26 UTC | [LayoutNG] Correct clip-path reference box calculation. We used coordinates relatively to the line box, while we were expected by the caller to be relative to the containing block. Flipping for writing mode was bogus for NG (but needed by legacy), since NG uses truly physical coordinates. Hardened tests to contain a leading line and padding, and leading content on the first line of the clipped child. Bug: 641907 Change-Id: I2b1b9ff4ea92a6405fcdffcf139842458b46442f Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_layout_ng Reviewed-on: https://chromium-review.googlesource.com/c/1257913 Reviewed-by: Koji Ishii <kojii@chromium.org> Reviewed-by: Fredrik Söderquist <fs@opera.com> Commit-Queue: Morten Stenshorne <mstensho@chromium.org> Cr-Commit-Position: refs/heads/master@{#596554} | 04 October 2018, 21:02:07 UTC |
d77b8fa | Cosmin Sabou | 04 October 2018, 07:02:10 UTC | Backed out changeset 8660ad891a23 (bug 1495372) for causing win2012 bustages. a=backout gecko-commit: 65fc22a1627b7de54303275dbe2422a61080865c gecko-integration-branch: central gecko-reviewers: backout | 04 October 2018, 16:56:31 UTC |
e6d641e | Henrik Skupin | 27 September 2018, 07:00:38 UTC | [wdspec] Add user prompt tests for more WebDriver commands. This adds the user prompt tests for those commands which clearly have to handle the prompt. But it still leaves out those commands where the behavior hasn't been clarified yet. Differential Revision: https://phabricator.services.mozilla.com/D6927 bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=1473814 gecko-commit: 8d9b9e65f06877fd438df9be201abb9b193a08a1 gecko-integration-branch: mozilla-inbound | 04 October 2018, 16:55:41 UTC |
660a694 | jugglinmike | 04 October 2018, 15:17:43 UTC | Merge pull request #12957 from bocoup/csp-cleanup [csp] Remove cookies following test completion Approved by @andypaicu in https://github.com/web-platform-tests/wpt/pull/12957 | 04 October 2018, 15:17:43 UTC |
55ebb91 | Michael[tm] Smith | 04 October 2018, 09:16:34 UTC | cc Update messages.json | 04 October 2018, 09:24:11 UTC |
685c505 | Ms2ger | 03 October 2018, 08:35:43 UTC | Python 3: Fix Content-Length header in slice(). This prevents a http.client.IncompleteRead error in the test. | 04 October 2018, 07:51:13 UTC |
572fdb9 | Mike Pennisi | 04 October 2018, 01:24:21 UTC | [html] Correct file generation logic Previously, a very large asset was removed from WPT in favor of a Python script which was intended to generate the equivalent data dynamically [1]. This script had an error which caused it to generate files that far exceeded the requested size (for example, "1kb" was interpreted as approximately 1.1e1024 bytes instead of 1024 bytes). Replace the generic script with a more targeted one which creates a response that suites the needs of the corresponding test. [1] https://github.com/web-platform-tests/wpt/pull/503 | 04 October 2018, 07:42:50 UTC |
397effa | autofoolip | 04 October 2018, 07:17:16 UTC | Update interfaces/webrtc-dscp.idl (#13353) Source: https://github.com/tidoust/reffy-reports/blob/f128525/whatwg/idl/webrtc-dscp.idl Build: https://travis-ci.org/tidoust/reffy-reports/builds/436975697 | 04 October 2018, 07:17:16 UTC |
f4907bf | autofoolip | 04 October 2018, 07:16:32 UTC | Update interfaces/css-layout-api.idl (#13351) Source: https://github.com/tidoust/reffy-reports/blob/f128525/whatwg/idl/css-layout-api.idl Build: https://travis-ci.org/tidoust/reffy-reports/builds/436975697 | 04 October 2018, 07:16:32 UTC |
a23606c | Ryan Hunt | 03 October 2018, 22:47:03 UTC | Implement multi-position gradient color-stops syntax. This commit adds the multi-position gradient color-stops syntax. GradientItem::parse_comma_separated is extended to attempt to parse a LengthOrPercent after each color stop. If it succeeds, it appends an additional color stop with a duplicate color and the specified position. This change is only to the parsing, serialization is left unchanged as per [1]. [1] https://github.com/w3c/csswg-drafts/issues/2714 Differential Revision: https://phabricator.services.mozilla.com/D7380 bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=1352643 gecko-commit: 984902ba6faf1864c04fef525f6764983439ce1b gecko-integration-branch: central gecko-reviewers: emilio | 04 October 2018, 05:46:08 UTC |
2ab59a5 | James Graham | 03 October 2018, 20:15:46 UTC | Unify wpt manifest download and update The previous code split a lot of logic between the update and download parts, and exposed two different mach commands. In order to simplify things it's better to have a single entry point for manifest download and update, and ensure that's called consistently. Differential Revision: https://phabricator.services.mozilla.com/D7497 bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=1495372 gecko-commit: 8660ad891a23c389512aa8d9d1fb122096df6741 gecko-integration-branch: autoland gecko-reviewers: ato | 04 October 2018, 04:16:23 UTC |
a7d4565 | Yoshifumi Inoue | 04 October 2018, 02:48:20 UTC | Make Element#innerText to not collapse white space around inline-block This patch makes |Element#innerText| not to collapse white space around inline- block for improving interop. Note: The spec doesn't explicitly mention about this. Changes of AX test expectations added missing space after <input>. TBR=dmazzoni@chromium.org Bug: 890020 Cq-Include-Trybots: luci.chromium.try:linux_layout_tests_layout_ng Change-Id: I74a47fd5ba3a22ff17d9c36838a81b4277ac47cc Reviewed-on: https://chromium-review.googlesource.com/c/1250825 Reviewed-by: Xiaocheng Hu <xiaochengh@chromium.org> Commit-Queue: Yoshifumi Inoue <yosin@chromium.org> Cr-Commit-Position: refs/heads/master@{#596485} | 04 October 2018, 02:53:41 UTC |
e29e596 | Geoff Brown | 03 October 2018, 17:55:16 UTC | Fix ca_certificate_path in wpt fennec runner bugzilla-url: https://bugzilla.mozilla.org/show_bug.cgi?id=1495430 gecko-commit: a4b8f37b625d8e9e8d05b64eae60f5adbbccb8ab gecko-integration-branch: mozilla-inbound gecko-reviewers: jgraham | 04 October 2018, 01:03:04 UTC |
e201d80 | Javier Fernandez | 03 October 2018, 22:49:39 UTC | [css-text] A leading white-space should break before handling overflow Leading white-spaces are indeed breaking opportunities that should prevent, if there are no other css properties forcing it, breaking text in the middle of a word when honoring the word-wrap/overflow-wrap CSS property. We are doing so if the leading white-space sequence is longer than 1 character, but when we have a single leading white-space, we are missing that breaking opportunity and we may lead to cases, like the one described in the bug. The root cause of the issue with single leading white-space breaking opportunities is that the RewindToMidWordBreak expects certain width to be committed in order to choose opportunities in previous runs if none of the ones detected by the ICU LazyLineBreakIterator prevents the overflow. However, this breaking opportunity should be considered together with other provided by the word-break CSS property (eg, break-word or break-all), as it was agreed in the discussion [1] with the CSS WG. This CL solves the issue identifying the single leading white-space braking opportunity in a new class field flag, and using it to consider this opportunity inside the mid-word breaking logic, or prevent to run it completely in the cases where 'break-all' is not set. This is basically a reland of 6ea2a2e7f3ef01e0c98424ce272a732ade92ad1a but with some changes to avoid regressions like the one reported in issue #866109. [1] https://github.com/w3c/csswg-drafts/issues/2907 Bug: 854624 Change-Id: I1cc0f55050d54ea1e76c655cf6b3ef8bcc0b0e2c Reviewed-on: https://chromium-review.googlesource.com/c/1209745 Commit-Queue: Javier Fernandez <jfernandez@igalia.com> Reviewed-by: Koji Ishii <kojii@chromium.org> Cr-Commit-Position: refs/heads/master@{#596406} | 03 October 2018, 23:00:24 UTC |
a49e763 | Ms2ger | 03 October 2018, 08:18:49 UTC | Python 3: Encode text content in Response.iter_content(). | 03 October 2018, 20:33:51 UTC |
dc97ad5 | Žan Doberšek | 03 October 2018, 20:06:54 UTC | Switch WebKit browser product to WebDriver executors (#13339) Switch away from Selenium executors to WebDriverTestharnessExecutor and WebDriverRefTestExecutor for the WebKit browser product. The browserVersion value is hard-coded to the 2.20 release series for the GTK+ port of WebKit as that's when the WebDriver functionality was introduced. | 03 October 2018, 20:06:54 UTC |