third_party/WebKit/LayoutTests/accessibility/aom-string-properties.html - Issue 2788523002: Finish all string attributes for Accessibility Object Model Phase 1.

Side by Side Diff: third_party/WebKit/LayoutTests/accessibility/aom-string-properties.html

Issue 2788523002: Finish all string attributes for Accessibility Object Model Phase 1. (Closed)

Patch Set: Rebase Created 3 years, 8 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:

« content/shell/test_runner/web_ax_object_proxy.cc ('K') | « content/shell/test_runner/web_ax_object_proxy.cc ('k') | third_party/WebKit/LayoutTests/webexposed/global-interface-listing-expected.txt » ('j') | third_party/WebKit/Source/modules/accessibility/AXNodeObject.cpp » ('J')
Toggle Intra-line Diffs ('i') | Expand Comments ('e') | Collapse Comments ('c') | Hide Comments ('s')

OLD	NEW
(Empty)
	1 <!DOCTYPE HTML>

	2 <script src="../resources/gc.js"></script>

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

	4 <script src="../resources/testharnessreport.js"></script>

	5

	6 <!--

	7

	8 Accessibility Object Model

	9 Explainer: https://github.com/WICG/aom/blob/master/explainer.md

	10 Spec: https://wicg.github.io/aom/spec/

	11

	12 -->

	13

	14 <script>

	15 if (window.internals)

	16 internals.runtimeFlags.accessibilityObjectModelEnabled = true;

	17 </script>

	18

	19 <div role="combobox" id="autocomplete"></div>

	20

	21 <script>

	22 test(function(t) {

	23 var node = document.getElementById("autocomplete");

	24 var axNode = accessibilityController.accessibleElementById("autocomplete");

	25 assert_equals(axNode.autocomplete, "");
	aboxhall 2017/03/31 00:23:38 Is this equivalent to "none"? Per https://www.w3.o Is this equivalent to "none"? Per https://www.w3.org/TR/wai-aria/states_and_properties#aria-autocomplete dmazzoni 2017/03/31 04:34:57 Yes, if we had an enum this would be the same as " Show quoted text On 2017/03/31 00:23:38, aboxhall wrote: > Is this equivalent to "none"? Per > https://www.w3.org/TR/wai-aria/states_and_properties#aria-autocomplete Yes, if we had an enum this would be the same as "none" - I think what's most important for now is that this is consistent with aria-autocomplete
	26 node.accessibleNode.autocomplete = "inline";

	27 assert_equals(axNode.autocomplete, "inline");

	28 }, "AccessibleNode.autocomplete");

	29 </script>

	30

	31 <div role="checkbox" id="checked"></div>

	32

	33 <script>

	34 test(function(t) {

	35 var node = document.getElementById("checked");

	36 var axNode = accessibilityController.accessibleElementById("checked");

	37 assert_equals(axNode.isChecked, false);

	38 node.accessibleNode.checked = "true";
	aboxhall 2017/03/31 00:23:38 Do we want to enable using boolean true/false as a Do we want to enable using boolean true/false as a shorthand for "true"/"false"? I think that would be intuitive. dmazzoni 2017/03/31 04:34:58 Hmmm, not sure I like that since this is actually Show quoted text On 2017/03/31 00:23:38, aboxhall wrote: > Do we want to enable using boolean true/false as a shorthand for "true"/"false"? > I think that would be intuitive. Hmmm, not sure I like that since this is actually a tristate. Setting it to either true or "mixed" seems weird to me. aboxhall 2017/03/31 04:41:54 Yeah, I realise it's a tristate, but "mixed" is so Show quoted text On 2017/03/31 04:34:58, dmazzoni wrote: > On 2017/03/31 00:23:38, aboxhall wrote: > > Do we want to enable using boolean true/false as a shorthand for > "true"/"false"? > > I think that would be intuitive. > > Hmmm, not sure I like that since this is actually a tristate. Setting it to > either true or "mixed" seems weird to me. Yeah, I realise it's a tristate, but "mixed" is so seldom used that I think it'd be a net win to allow setting the "true"/"false" values via boolean values. If we don't do this, I foresee much developer moaning about it :)
	39 assert_equals(axNode.isChecked, true);

	40 node.accessibleNode.checked = "mixed";

	41 assert_equals(axNode.isChecked, false);

	42 assert_equals(axNode.isButtonStateMixed, true);

	43 }, "AccessibleNode.checked");

	44 </script>

	45

	46 <div role="tab" id="current"></div>

	47

	48 <script>

	49 test(function(t) {

	50 var node = document.getElementById("current");

	51 var axNode = accessibilityController.accessibleElementById("current");

	52 assert_equals(axNode.current, "");

	53 node.accessibleNode.current = "page";

	54 assert_equals(axNode.current, "page");

	55 }, "AccessibleNode.current");

	56 </script>

	57

	58 <div role="textbox" id="invalid"></div>

	59

	60 <script>

	61 test(function(t) {

	62 var node = document.getElementById("invalid");

	63 var axNode = accessibilityController.accessibleElementById("invalid");

	64 assert_equals(axNode.invalid, "");

	65 node.accessibleNode.invalid = "spelling";

	66 assert_equals(axNode.invalid, "spelling");

	67 }, "AccessibleNode.invalid");

	68 </script>

	69

	70 <div role="button" id="keyShortcuts"></div>

	71

	72 <script>

	73 test(function(t) {

	74 var node = document.getElementById("keyShortcuts");

	75 var axNode = accessibilityController.accessibleElementById("keyShortcuts");

	76 assert_equals(axNode.keyShortcuts, "");

	77 node.accessibleNode.keyShortcuts = "Ctrl+F";

	78 assert_equals(axNode.keyShortcuts, "Ctrl+F");

	79 }, "AccessibleNode.keyShortcuts");

	80 </script>

	81

	82 <div role="heading" id="label">Inner text</div>

	83

	84 <script>

	85 test(function(t) {

	86 var node = document.getElementById("label");

	87 var axNode = accessibilityController.accessibleElementById("label");

	88 assert_equals(axNode.name, "Inner text");

	89 node.accessibleNode.label = "Label";

	90 assert_equals(axNode.name, "Label");

	91 }, "AccessibleNode.label");

	92 </script>

	93

	94 <div role="banner" id="live"></div>

	95

	96 <script>

	97 test(function(t) {

	98 var node = document.getElementById("live");

	99 var axNode = accessibilityController.accessibleElementById("live");

	100 assert_equals(axNode.live, "");

	101 node.accessibleNode.live = "polite";

	102 assert_equals(axNode.live, "polite");

	103 }, "AccessibleNode.live");

	104 </script>

	105

	106 <div role="slider" id="orientation"></div>

	107

	108 <script>

	109 test(function(t) {

	110 var node = document.getElementById("orientation");

	111 var axNode = accessibilityController.accessibleElementById("orientation");

	112 assert_equals(axNode.orientation, "AXOrientation: AXHorizontalOrientation");

	113 node.accessibleNode.orientation = "vertical";

	114 assert_equals(axNode.orientation, "AXOrientation: AXVerticalOrientation");

	115 }, "AccessibleNode.orientation");

	116 </script>

	117

	118 <input id="placeholder">

	119

	120 <script>

	121 test(function(t) {

	122 var node = document.getElementById("placeholder");

	123 var axNode = accessibilityController.accessibleElementById("placeholder");

	124 assert_equals(axNode.name, "");

	125 node.accessibleNode.placeholder = "Placeholder";

	126 assert_equals(axNode.name, "Placeholder");
	aboxhall 2017/03/31 00:23:39 Why not check axNode.placeholder directly here (as Why not check axNode.placeholder directly here (as well)? dmazzoni 2017/03/31 04:34:57 We actually don't return anything from "placeholde Show quoted text On 2017/03/31 00:23:39, aboxhall wrote: > Why not check axNode.placeholder directly here (as well)? We actually don't return anything from "placeholder" when the placeholder was returned as the name or description. That was one of the nice side effects of the new name calc code. We never duplicate stuff anymore. However, I can assert that nameFrom=="placeholder". Done. aboxhall 2017/04/03 23:20:24 Oh nice :) Show quoted text On 2017/03/31 04:34:57, dmazzoni wrote: > On 2017/03/31 00:23:39, aboxhall wrote: > > Why not check axNode.placeholder directly here (as well)? > > We actually don't return anything from "placeholder" when the placeholder was > returned as the name or description. That was one of the nice side effects of > the new name calc code. We never duplicate stuff anymore. > > However, I can assert that nameFrom=="placeholder". Done. Oh nice :)
	127 }, "AccessibleNode.placeholder");

	128 </script>

	129

	130 <div role="banner" id="relevant"></div>

	131

	132 <script>

	133 test(function(t) {

	134 var node = document.getElementById("relevant");

	135 var axNode = accessibilityController.accessibleElementById("relevant");

	136 assert_equals(axNode.relevant, "additions text");

	137 node.accessibleNode.relevant = "additions";

	138 assert_equals(axNode.relevant, "additions");

	139 }, "AccessibleNode.relevant");

	140 </script>

	141

	142 <div role="button" id="role"></div>

	143

	144 <script>

	145 test(function(t) {

	146 var node = document.getElementById("role");

	147 var axNode = accessibilityController.accessibleElementById("role");

	148 assert_equals(axNode.role, "AXRole: AXButton");

	149 node.accessibleNode.role = "combobox";
	aboxhall 2017/03/31 00:23:38 Test role fallback as well, i.e. node.accessibleNo Test role fallback as well, i.e. node.accessibleNode.role = "banana combobox"? dmazzoni 2017/03/31 04:34:57 Done. Show quoted text On 2017/03/31 00:23:38, aboxhall wrote: > Test role fallback as well, i.e. node.accessibleNode.role = "banana combobox"? Done.
	150 assert_equals(axNode.role, "AXRole: AXComboBox");

	151 }, "AccessibleNode.role");

	152 </script>

	153

	154 <div role="button" id="roleDescription"></div>

	155

	156 <script>

	157 test(function(t) {

	158 var node = document.getElementById("roleDescription");

	159 var axNode = accessibilityController.accessibleElementById("roleDescription" );

	160 assert_equals(axNode.roleDescription, "");

	161 node.accessibleNode.roleDescription = "Push Button";

	162 assert_equals(axNode.roleDescription, "Push Button");

	163 }, "AccessibleNode.roleDescription");

	164 </script>

	165

	166 <div role="columnheader" id="sort"></div>

	167

	168 <script>

	169 test(function(t) {

	170 var node = document.getElementById("sort");

	171 var axNode = accessibilityController.accessibleElementById("sort");

	172 assert_equals(axNode.sort, "");
	aboxhall 2017/03/31 00:23:39 Similarly, should this be explicitly "none"? Maybe Similarly, should this be explicitly "none"? Maybe not necessary now since this is internal only, but for phase 5 we'll need to map this. dmazzoni 2017/03/31 04:34:57 Yeah, this is just supposed to match what we retur Show quoted text On 2017/03/31 00:23:39, aboxhall wrote: > Similarly, should this be explicitly "none"? Maybe not necessary now since this > is internal only, but for phase 5 we'll need to map this. Yeah, this is just supposed to match what we return now. The LayoutTest interface is a real hodgepodge and isn't really supposed to be consistent with anything. Again, what I care about the most for now is if it matches what we get with aria. Also see the next change where I'm working on testing these in content/ - I think that's almost better because what matters the most is what we get in AXNodeData.
	173 node.accessibleNode.sort = "ascending";

	174 assert_equals(axNode.sort, "ascending");

	175 }, "AccessibleNode.sort");

	176 </script>

	177

	178 <div role="slider" id="valueText" aria-valuetext="5"></div>

	179

	180 <script>

	181 test(function(t) {

	182 var node = document.getElementById("valueText");

	183 var axNode = accessibilityController.accessibleElementById("valueText");

	184 assert_equals(axNode.valueDescription, "AXValueDescription: 5");

	185 node.accessibleNode.valueText = "Five";

	186 assert_equals(axNode.valueDescription, "AXValueDescription: Five");

	187 }, "AccessibleNode.valueText");

	188 </script>

OLD	NEW