[meta] missing dijit DOH tests

[Katie Vance]

Doh test cases are needed for dijit.

I attached a simple doh test for dijit.form.SimpleTextarea?.

[Douglas Hays]

Here are some dijit test files for which I was unable to find corresponding automated tests:

_base/test_CrossWindow.html
editor/test_CustomPlugin.html
editor/test_Print.html
editor/test_RichText.html
editor/test_TabIndent.html
editor/test_ToggleDir.html
editor/test_ToggleDir_rtl.html
editor/test_resize.html
form/test_ComboBox_destroy.html
form/test_Form_onsubmit.html
form/test_MultiSelect.html
form/test_SimpleTextarea.html
form/test_validStatePerformance.html
form/test_validationState.html
layout/test_AccordionContainerDestroy.html
layout/test_BorderContainer_complex.html
layout/test_BorderContainer_experimental.html
layout/test_BorderContainer_full.html
layout/test_BorderContainer_nested.html
layout/test_BorderContainer_prog.html
layout/test_ContentPane_prog.html
layout/test_Gui.html
layout/test_LayoutContainer.html
layout/test_SplitContainer.html
layout/test_StackContainer.html
layout/test_StackContainer_code.html
layout/test_TabContainerTitlePane.html
layout/test_TabContainer_noLayout.html
layout/test_TabContainer_prog.html
layout/test_TabContainer_remote.html
layout/test_refreshOnShow.html
test_Declaration.html
test_Dialog_focusDestroy.html
test_Menu_iframe.html
test_ProgressBar.html
test_Tree_Notification_API_Support.html
test_bgIframe.html
test_instantiate.html
tree/test_CustomLabel.html
tree/test_Tree_Programmatic.html
tree/test_Tree_Styling.html
tree/test_Tree_v1.html

[bill]

ksv - thanks for the test case for SimpleTextArea. I've checked one in recently for TextArea but I guess we could use one for SimpleTextArea too. But have you filed a CLA?

About the list above, I wouldn't expect a 1-1 mapping between manual test files and automated test files although I guess that would be ideal. In any case each widget should have an automated test, which is why I've filed many tickets like #10340, #10341, #10342, #10343, #10344, #10366 etc.

In any case, we've already got a generic ticket #7681 for robot test cases (in addition to the widget specific tickets I listed above), so I'm going to use this ticket just for tracking SimpleTextArea. Please open other individual tickets for other widgets that are missing automated tests.

[bill]

Oh nevermind, I see you have a CLA (under IBM) and also that this ticket is assigned to Doug, didn't realize that before.

[Katie Vance]

patch order: 1

[Douglas Hays]

(In [22612]) References #11488. Proxy commit for ksv (CCLA, IBM). Added SimpleTextArea?.html automated test.

[Douglas Hays]

(In [22684]) References #11488. Proxy commit for ksv (CCLA). Add an automated ComboBox? destroy test and remove unnecessary extra test file test_ComboBox_destroy.html.

[Douglas Hays]

(In [22691]) References #11488. Proxy commit for ksv (CCLA). Add an automated Form_onsubmit.html test.

[Katie Vance]

robot test for test_validationState.html

[Douglas Hays]

MultiSelect?.html fails consistently on IE7.

{"easing":["dojo._defaultEasing","_scopeName","_base","registry","getUniqueId","isCollapsed","moveToBookmark","focus","registerIframe","registerWin","_onBlurNode","_onFocusNode","getViewport","_place","placeOnScreenAroundRectangle","placementRegistry","getPopupAroundAlignment","popup","BackgroundIframe","typematic","_XhtmlRoles","getWaiRole","removeWaiRole","getWaiState","removeWaiState","_connectOnUseEventHandler","_Templated","layout","form","_HasDropDown"],"second":["getLastInTabbingOrder"],"select3":["TN","FL"]}
[ Error: test timeout in formSubmit ]
ERROR IN: function()
Total time for GROUP " dijit.form.MultiSelect " is 38.4s
FAILED test: formSubmit 15031 ms

Even though it's a timeout, adding more time won't help. Clicking the submit button printed the form values to the console, but some other event must have been missed that notified the test.
Also I noticed that different items are being selected in IE vs FF during previous tests but those tests passed anyway so the test criteria needs to be tightened to detect that, AND the test should be altered to not select different items on different browsers. IE selected _destroyAll thru _curFocus and FF selected getEnclosingWidget thru _curFocus.

[Douglas Hays]

(In [22721]) References #11488. Proxy commit for ksv (CCLA). Add an automated ValidationState?.html test.

[Katie Vance]

robot test for test_AccordionContainerDestroy.html

[Katie Vance]

robot test for test_BorderContainer_complex.html

[Katie Vance]

robot test for test_BorderContainer_full.html

[Douglas Hays]

(In [22740]) References #11488. Proxy commit for ksv (CCLA). Add an automated MultiSelect?.html test.

[Katie Vance]

robot test for test_BorderContainer_nested.html

[Katie Vance]

Move test_AccordionContainerDestroy to test_AccordionContainer

[Douglas Hays]

BorderContainer_prog should be merged into test_BorderContainer

[Douglas Hays]

(In [22787]) Refs #11488. Proxy commit for ksv (CCLA). Merged test_AccordionContainerDestroy.html into test_AccordionContainer.html and added an automated test.

[Douglas Hays]

(In [22791]) References #11488. Proxy commit for ksv (CCLA). Add an automated BorderContainer_complex.html robot test.

[Douglas Hays]

(In [22792]) References #11488. Proxy commit for ksv (CCLA). Add an automated BorderContainer_full.html robot test.

[Douglas Hays]

(In [22793]) References #11488. Proxy commit for ksv (CCLA). Add an automated BorderContainer_nested.html robot test.

[bill]

(In [22794]) On AccordionContainer.removeChild(child), remove child from AccordionInnerContainer wrapper before destroying wrapper, so that child DOM nodes don't get destroyed along with wrapper. Fixes #11656.

Also fixing the destroy() test in !AccordionContainer_a11y.html (it would never fail even if Accordion was broken), and moving that test to AccordionContainer.html since it works w/out robot. Refs #11488 ([22787]).

[Katie Vance]

robot test for test_BorderContainer_prog.html

[Douglas Hays]

(In [22801]) Refs #11488. Proxy commit for ksv (CCLA). Merged test_BorderContainer_prog.html into test_BorderContainer.html and added an automated test for programmatic creation.

[Katie Vance]

robot test for test_CustomPlugin.html

[Douglas Hays]

(In [22862]) Refs #11488. Proxy commit for ksv (CCLA). Added new automated test for editor/test_CustomPlugin.html. Added a workaround for an IE positioning problem in the unit testcase.

[Katie Vance]

robot test for test_TabIndent.html

[Katie Vance]

robot test for test_ToggleDir_rtl.html

[Katie Vance]

robot test for test_RichText.hml, removed test_resize.html

[Katie Vance]

robot test for test_ToggleDir.html

[Katie Vance]

robot test for test_ContentPane_prog.html

[Douglas Hays]

(In [22923]) Refs #11488. Proxy commit for ksv (CCLA). Added new automated Editor tests.

[Douglas Hays]

(In [22924]) Refs #11488. Handle some Editor tests deletes/renames.

[Katie Vance]

robot test for test_LayoutContainer.html

[Katie Vance]

robot test for test_StackContainer.html and test_StackContainer_code.hml

[Katie Vance]

robot test for test_TabContainer_noLayout.html

[Katie Vance]

robot test for test_TabContainer_prog.html

[Katie Vance]

robot test for test_TabContainer_remote.html

[Katie Vance]

robot test for test_Dialog_focusDestroy.html

[Katie Vance]

robot test for test_Menu_iframe.html

[Douglas Hays]

(In [22996]) Refs #11488. Proxy commit for ksv (CCLA). Added new automated dijit/layout tests.

[Douglas Hays]

(In [22997]) Refs #11488. Proxy commit for ksv (CCLA). Added new automated dijit/Dialog test.

[Douglas Hays]

(In [22998]) Refs #11488. Proxy commit for ksv (CCLA). Added new automated dijit/Menu_iframe test.

[Douglas Hays]

(In [22999]) Refs #11488. Delete obsolete test.

[Douglas Hays]

(In [23016]) Refs #11488. Reorganize dijit.Tree tests under tree/ directory and remove redundant test files. !strict

[Douglas Hays]

Remaining tests to automate:

_base/test_CrossWindow.html (and add both _testCommon.js and claro)
bidi.html (need to test both LTR and RTL tests on same page)
editor/nls_8859-2.html (add _testCommon.js and be mindful that simplistic text editors can munge special characters)
editor/nls_sjis.html (ditto)
editor/nls_utf8.html (ditto)
layout/test_Gui.html
layout/test_refreshOnShow.html
test_ProgressBar.html
test_bgIframe.html (may need the robot to expose its pixel-read api)
test_instantiate.html
tree/test_CustomLabel.html
tree/test_Tree_Styling.html (merge into Tree.html(nonrobot) or test_Tree.html(robot) and delete this)
tree/test_Tree_v1.html

[Katie Vance]

robot test for test_ProgressBar.html

[Katie Vance]

robot test for test_GUI.html

[Douglas Hays]

(In [23056]) Refs #11488. Proxy commit for ksv (CCLA). Add automated ProgressBar?.html test.

[Douglas Hays]

(In [23057]) Refs #11488. Proxy commit for ksv (CCLA). Added automated layout/GUI.html test.

[Katie Vance]

moved test_Tree_Styling.html into Tree.html. No new robot test.

[Douglas Hays]

(In [23058]) Refs #11488. Proxy commit for ksv (CCLA). Merged test_Tree_Styling.html into Tree.html and added an automated test.

[Katie Vance]

Delete refreshOnShow.html and move accordion test to acc test page

[Douglas Hays]

(In [23071]) Refs #11488. Proxy commit for ksv (CCLA). Merged test_refreshOnShow.html into test_AccordionContainer.html and added an automated test.

[Katie Vance]

automated test for test_CustomLabel.html

[Douglas Hays]

(In [23072]) Refs #11488. Proxy commit for ksv (CCLA). Renamed CustomLabel?.html test file and added automated test.

[Douglas Hays]

missing module.js change for CustomLabel?.html

[Katie Vance]

robot test for test_Tree_v1.html

[Douglas Hays]

(In [23076]) Refs #11488. Proxy commit for ksv (CCLA). Added robot test Tree_v1.html.

[Katie Vance]

robot test will have failures due to #11906

[Douglas Hays]

(In [23131]) Refs #11488. Proxy commit for ksv (CCLA). Added robot test Instantiate.html

[Katie Vance]

robot test for test_CrossWindow.html

[Katie Vance]

robot test for test_Bidi.html

[Katie Vance]

automated test

[Katie Vance]

inline automated test for nls_sjis.html

[Katie Vance]

inline automated test for nls_8859-2.html

[Douglas Hays]

(In [23145]) Refs #11488. Proxy commit for ksv (CCLA). Add an automated CrossWindow?.html robot test.

[Douglas Hays]

(In [23146]) Refs #11488. Proxy commit for ksv (CCLA). Add an automated Bidi.html robot test.

[Douglas Hays]

(In [23147]) Refs #11488. Proxy commit for ksv (CCLA). Add automated nls/Editor tests.

[Katie Vance]

The only test left to automate is test_bgIframe.html. I haven't written the test yet because I'm blocked by ticket #11822.

[Katie Vance]

PDF file needed for BgIframe?.patch

[Katie Vance]

BgIframe?.patch is only set to run on IE7 & 8. All other browsers are still blocked by ticket #11822. See that ticket for specific details.

[bill]

I just checked in [23177] which should fix IE6. (The IE6 problem was unrelated to #11822.)

[Douglas Hays]

BgIframe?.patch needs more work. On my IE6 XP box with java 1.6.0_20-b02, the test does not load the applet and just hangs. If I remove the programmatic applet and restore the markup applet, it runs with 0 errors.

[bill]

(In [23187]) Avoid intermittent failures due to race condition where both the robot test and the client HTML file had delays of 4000ms. Refs #11488.

[bill]

Another missing robot test is for CheckBox/RadioButton, see #10340.

[bill]

(In [23277]) Dijit has extensive ContentPane tests, including tests for set("content", ...) and set("href", ...); it doesn't need a special test for programmatic ContentPane creation, certainly not a robot test. This test was probably added for documentation reasons but now we have that on the doc site. Refs #11488.

[bill]

(In [23278]) Rewrite LayoutContainer test.

The main thing to check is that LayoutContainer lays out it's children correctly (in top/bottom/left/right/center alignment). This doesn't require robot.

The other thing to test is tab order, which we usually test via robot keyboard commands, but instead I implemented some test code to search for tabbable nodes programatically.

Refs #11488 !strict.

[haysmark]

Robot jar to fix issues with applets that steal focus.

[haysmark]

Refs #11488. Addresses issues with applets stealing focus.

[Douglas Hays]

(In [23281]) Fixes #11488. Proxy commit for ksv (IBM, CCLA). Add automated bgiframe test. Tweak robot so that 2 applets can exist on the same page in linux.

[bill]

(In [23285]) Don't use robot for testing API's, plain DOH is better. Refs #11488.

[bill]

(In [23286]) Don't use robot for TabContainer remote loading, plain DOH is simpler. Refs #11488.

[bill]

(In [23293]) Intermediate checkin to not use robot for !!test_TabContainer_prog.html tests, refs #11488. Will merge these tests into TabContainer.html test file in next checkin.

[bill]

(In [23294]) Merge non-duplicate !TabContainer_prog.html tests into TabContainer.html, refs #11488.

[bill]

(In [23316]) fixes for [23286], refs #11488.

[bill]

(In [23344]) Fix custom icon path, refs #11488 / [23016].

[bill]

(In [23449]) Get test running on safari. Stop using dojoattachpoint/dojoattachevent to locate nodes since it wasn't working on safari, and also since those attributes/properties may not be set on instantiated templates in the future. Refs #11488.

[bill]

(In [23466]) Fix typo, refs #11488.

[bill]

(In [23484]) Use persist=false flag on Tree rather than manually trying to set TreeNodes to a consistent state. The code to set TreeNodes to a consistent state was incomplete; it didn't adjust grandchildren, and whenever the test was run without clearing the cache, Mexico started out open, causing failures.

Also switching this test to old parser style of dojoType since the test is designed to test backwards compatibility.

Refs #11488.

[bill]

(In [24450]) Make !StackContainer_mouse.html test non-robot, and fix errors, such as calling isHidden()/isVisible() but not checking return value. Refs #11488.

[bill]

In [27121]:

add comments so it's clear which assert failed, refs #11488

[bill]

In [28531]:

fix funny test names, this is a test for TabIndent not ToggleDir, refs #11488

[bill]

In [29984]:

increase timeout, refs #11488.

[bill]

In [30952]:

Prevent spurious test failure. Tab 5 with inner tab 1 selected is the same height or near the same height as Tab 4 with inner tab 2 selected... so don't test for a height difference. Plus which it's silly to test the same behavior 8 times. Refs #11488.

[bill]

In [31417]:

fix the text of all the hints, refs #11488