LayoutTests/editing/spelling/editing-multiple-words-with-markers.html - Issue 906193003: shouldBecomeEqual() behaves as shouldBe() if the testing expression returns the expected value

Side by Side Diff: LayoutTests/editing/spelling/editing-multiple-words-with-markers.html

Issue 906193003: shouldBecomeEqual() behaves as shouldBe() if the testing expression returns the expected value Base URL: https://chromium.googlesource.com/chromium/blink.git@master

Patch Set: Created 5 years, 10 months ago

Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.

Jump to:

View unified diff | Download patch

« no previous file with comments | « LayoutTests/TestExpectations ('k') | LayoutTests/editing/spelling/editing-multiple-words-with-markers-expected.txt » ('j') | LayoutTests/editing/spelling/spellcheck-paste-continuous-disabled.html » ('J')
Toggle Intra-line Diffs ('i') | Expand Comments ('e') | Collapse Comments ('c') | Hide Comments ('s')

OLD	NEW
(Empty)
	1 <html>

	2 <head>

	3 <script src="resources/util.js"></script>

	4 <script src="../editing.js"></script>

	5 <title>Editing multiple words with markers test</title>

	6 </head>

	7 <body>

	8 <textarea id="testElement"></textarea>

	9 <script src="../../resources/js-test.js"></script>

	10 <script>

	11 description('The test verifies if the spelling markers disappear when '

	12 + 'the multiple misspelled words are concatenated by delete command. '

	13 + 'To test manually, type "it\'s a meagesga meagesga ", then '

	14 + 'select and delete "esga meag". The test succeeds '

	15 + 'if the remaining text does not have any underline.');

	16

	17 jsTestIsAsync = true;

	18

	19 if (window.internals) {

	20 internals.settings.setUnifiedTextCheckerEnabled(true);

	21 internals.settings.setAsynchronousSpellCheckingEnabled(true);

	22 }

	23

	24 function resetText() {

	25 var textArea = document.getElementById('testElement');

	26 textArea.value = "";

	27 typeText(textArea, "it's a meagesga meagesga ");
	please use gerrit instead 2015/02/10 22:55:42 This test will be flaky if you don't wait for spel This test will be flaky if you don't wait for spellcheck markers to appear. Use internals.markerCountForNode(). See resources/utils.js for example. grzegorz 2015/02/11 10:54:23 Done. Show quoted text On 2015/02/10 22:55:42, Rouslan Solomakhin wrote: > This test will be flaky if you don't wait for spellcheck markers to appear. Use > internals.markerCountForNode(). See resources/utils.js for example. Done. grzegorz 2015/02/11 11:53:28 The problem here is a little complicated. Waiting Show quoted text On 2015/02/11 10:54:23, grzegorz wrote: > On 2015/02/10 22:55:42, Rouslan Solomakhin wrote: > > This test will be flaky if you don't wait for spellcheck markers to appear. > Use > > internals.markerCountForNode(). See resources/utils.js for example. > > Done. The problem here is a little complicated. Waiting for spelling markers here is fine if we want to simulate a typical user behaviour. Usually it looks like: 1. User types: "it's a meagesga meagesga " 2. Spelling markers appear 3. User selects a range 4. User deletes a range. 5. Spelling marker disappear (caret in the middle of the word) However, if we do it through the script the markers from typing are applied AFTER selection is removed. See https://code.google.com/p/chromium/issues/detail?id=456756 for more details. There is a simple test case as well. My goal here is to show that shouldBecomeEqual has a bug and to prove that I added a spelling test which contains test cases where spelling markers are not expected. Assuming current implementation of shouldBecomeEqual this test passes. However, the bug (https://code.google.com/p/chromium/issues/detail?id=456756) refers to
	28 }

	29

	30 /**

	31 * Those test cases verify if spelling markes disappear when
	please use gerrit instead 2015/02/10 22:55:42 s/markes/markers/ s/markes/markers/ grzegorz 2015/02/11 10:54:23 Done. Show quoted text On 2015/02/10 22:55:42, Rouslan Solomakhin wrote: > s/markes/markers/ Done.
	32 * the multiple misspelled words are concatenated by delete command.

	33 *

	34 * '+' means the position where marker is expected.

	35 * '-' means the position where marker is _not_ expected.

	36 */

	37 var testCases = [

	38

	39 { selectionRange: null, spellingMarker1: [ 0, 4, 'false' ], spellingMa rker2: [ 7, 8, 'true' ]

	40 // it's a meagesga meagesga

	41 // ---- ++++++++

	42 },

	43 { selectionRange: [ 11, 20 ], spellingMarker1: [ 7, 4, 'false' ], spellingMa rker2: [ 11, 4, 'false' ]

	44 // it's a meagesga

	45 // --------

	46 },

	47 { selectionRange: [ 11, 16 ], spellingMarker1: [ 7, 4, 'false' ], spellingMa rker2: [ 11, 8, 'false' ]

	48 // it's a meagmeagesga

	49 // ------------

	50 },

	51 { selectionRange: [ 15, 20 ], spellingMarker1: [ 7, 8, 'false' ], spellingMa rker2: [ 15, 4, 'false' ]

	52 // it's a meagesgaesga

	53 // ------------

	54 }

	55 ];

	56

	57 var marker1, marker2;

	58

	59 function checkSpellingMarkerAfterDeleteingSelection(selectionRange, spellingMark er1, spellingMarker2) {

	60 resetText();

	61 var testElement = document.getElementById('testElement');

	62

	63 if (selectionRange) {

	64 testElement.setSelectionRange(selectionRange[0], selectionRange[1]);

	65 execDeleteCommand();

	66 }

	67

	68 if (!window.internals)

	69 return done();
	please use gerrit instead 2015/02/10 22:55:42 If window.internals is not defined, this statement If window.internals is not defined, this statement will loop through the testCases without doing any work. Why not check for window.internals before calling done(). That way you can avoid looping through testCases if no tests can be performed. grzegorz 2015/02/11 10:54:23 Done. Show quoted text On 2015/02/10 22:55:42, Rouslan Solomakhin wrote: > If window.internals is not defined, this statement will loop through the > testCases without doing any work. Why not check for window.internals before > calling done(). That way you can avoid looping through testCases if no tests can > be performed. Done.
	70

	71 debug(testElement.value);
	please use gerrit instead 2015/02/10 22:55:42 Is debugging necessary here? Is debugging necessary here? grzegorz 2015/02/11 10:54:23 It adds empty line in the output. Show quoted text On 2015/02/10 22:55:42, Rouslan Solomakhin wrote: > Is debugging necessary here? It adds empty line in the output.
	72

	73 marker1 = spellingMarker1;

	74 marker2 = spellingMarker2;

	75

	76 shouldBecomeEqual('internals.hasSpellingMarker(document, marker1[0], marker1 [1])', marker1[2], function() {

	77 shouldBecomeEqual('internals.hasSpellingMarker(document, marker2[0], mar ker2[1])', marker2[2], function() {

	78 debug("");
	please use gerrit instead 2015/02/10 22:55:42 Is debugging necessary here? Is debugging necessary here? grzegorz 2015/02/11 10:54:22 Ditto. Show quoted text On 2015/02/10 22:55:42, Rouslan Solomakhin wrote: > Is debugging necessary here? Ditto.
	79 done();

	80 });

	81 });

	82 }

	83

	84 function done() {

	85 var nextTestCase = testCases.shift();

	86 if (nextTestCase)
	please use gerrit instead 2015/02/10 22:55:42 I would feel better if you put curly braces around I would feel better if you put curly braces around your multiline if statement body. grzegorz 2015/02/11 10:54:23 Done. Show quoted text On 2015/02/10 22:55:42, Rouslan Solomakhin wrote: > I would feel better if you put curly braces around your multiline if statement > body. Done.
	87 return setTimeout(checkSpellingMarkerAfterDeleteingSelection, 0,
	please use gerrit instead 2015/02/10 22:55:42 Why setTimeout(func, 0, args) instead of simply fu Why setTimeout(func, 0, args) instead of simply func(args)? grzegorz 2015/02/11 10:54:23 I can change it if you like. Show quoted text On 2015/02/10 22:55:42, Rouslan Solomakhin wrote: > Why setTimeout(func, 0, args) instead of simply func(args)? I can change it if you like.
	88 nextTestCase.selectionRange,

	89 nextTestCase.spellingMarker1,

	90 nextTestCase.spellingMarker2);

	91

	92 finishJSTest();

	93 }

	94 done();

	95 </script>

	96 </body>

	97 </html>

OLD	NEW