H

HTML_CodeSniffer

Settings | Report Duplicate

0

I Use This!

Inactive

Commits : Listings

Analyzed 22 days ago. based on code collected 23 days ago.

Commit Message	Contributor	Files Modified	Lines Added	Lines Removed	Code Location	Date
Jan 30, 2024 — Jan 30, 2025 Showing page 12 of 23 Search / Filter on:
Merge pull request #38 from lwright-sq/master	Luke Wright	More...				over 12 years ago
Messages from "_top" sniffs are now saved and replayed when the element is found in the DOM order. This reduces as much as possible (does not eliminate) the amount of bouncing up and down the page - as used to happen when the "_top" sniffs (which could emit messages anywhere) fire, resulting in the first few messages being ordered by sniff (and thus in the WCAG 2.0 case, by Success Criterion). This now saves and clears "_top" messages and replays them into the message list when it is time to handle that element.	Luke Wright	More...				over 12 years ago
Merge pull request #37 from lwright-sq/master	Luke Wright	More...				over 12 years ago
Add ability to ignore certain message codes using ignoreMsgCodes auditor run option. This option is an array containing RegExp objects (or strings interpreted as RegExp's); if any is matched it will be excluded.	Luke Wright	More...				over 12 years ago
Merge pull request #36 from lwright-sq/master	Luke Wright	More...				over 12 years ago
Changing messages and tests for H71 technique, because the technique test itself changed in the 3-Jan-2012 doc update. Check boxes and radio buttons with the same name now emit a warning, not an error, because fieldset is not required if the input labels include "clear instructions and [are] distinct selections".	Luke Wright	More...				over 12 years ago
Merge pull request #35 from lwright-sq/master	Luke Wright	More...				over 12 years ago
Fixed the handling of disabled level switches on the landing page of the Auditor.	Luke Wright	More...				over 12 years ago
Merge pull request #34 from lwright-sq/master	Luke Wright	More...				over 12 years ago
A test for whether an input required a label (eg. if hidden or button) was case sensitive. Although lowercase is best practice, recognise other cases.	Luke Wright	More...				over 12 years ago
Merge pull request #32 from lwright-sq/master	Luke Wright	More...				over 12 years ago
1. Replace workaround for IE8 on the level toggle switches, with something that should work in all browsers. 2. Fix IE8 opacity in the CSS, making sure that disabled options actually look disabled.	Luke Wright	More...				over 12 years ago
Merge pull request #30 from lwright-sq/master	Luke Wright	More...				over 12 years ago
Only try to work out if computed style has blink if computed style actually exists.	Luke Wright	More...				over 12 years ago
Merge pull request #28 from lwright-sq/master	Luke Wright	More...				over 12 years ago
Always throw warnings where a background image is involved, because any solid background behind it may not be an accurate representation of the actual colour the image gives it.	Luke Wright	More...				over 12 years ago
Merge pull request #27 from lwright-sq/master	Luke Wright	More...				over 12 years ago
1. Move the required ratio checks further up the test, so the "has BG image" failures have knowledge of the required ratio. 2. When there is no background image, the value of this style is "none", not a blank string. Fix tests for this.	Luke Wright	More...				over 12 years ago
Elements with background images but no solid background image will now emit a warning, imploring authors to check all colours in the image covered by text for contrast ratio issues.	Luke Wright	More...				over 12 years ago
No longer throw contrast ratio messages where there is no definition on the background colour (ie. in a document fragment).	Luke Wright	More...				over 12 years ago
1. Don't add the extra document fragment div if the body tag is found in the fragment. 2. Add an unlikely ID (that is later removed) for the extra div. This is so we can selectively remove it rather than just pulling out the first div we get.	Luke Wright	More...				over 12 years ago
SC 1.3.1: Non-existent "for" attributes were being thrown as errors regardless of whether it was a full document. Now throw error or warning like SC 2.4.1's bypassable ID check.	Luke Wright	More...				over 12 years ago
1. Source code fragments (not full docs) get placed inside a DIV tag, and the DIV is passed. This is to work around a regression introduced in pull request #23 where SC 2.4.1 treated these as full documents. 2. Fix a trailing comma in array issue which would have bitten HTMLCS when minified.	Luke Wright	More...				over 12 years ago
Failure messages should state "conformance level" instead of just "level" to avoid ambiguity.	Luke Wright	More...				over 12 years ago
1. Fix a regression when trying to determine whether an element is in the current document (ie. it has been replaced by something else dynamically after being run). 2. Moved said code to the isPointable() function, so if it has been replaced, the pointer will be disabled.	Luke Wright	More...				over 12 years ago
Contrast failures now come with a recommendation as to the colour(s) that should be changed to achieve the required contrast ratio (CR). - Recommended colours are determined by tweaking them on a hue/saturation/value model in steps of 0.25% (value in one way, saturation the other) until they reach the required CR and/or white or black. - HTMLCS will first try to change the colour - either background or foreground - furthest away from white/black (according to the sRGB relative luminescence values). If it gets to white or black without the CR being satisfied, it will revert to the original colours then try the other colour. - If the other colour cannot satisfy the CR requirement by the time it gets to white or black (possible with a 7:1 ratio and initial colours both near grey), it will then start to change the first colour while keeping the other colour at white/black. - If the colours are exactly the same distance from a 0.5 relative luminescence, it will change the foreground colour first. - Contrast ratio messages now have the following added: "Recommendation: change text colour to #000, background to #959595" (for instance), deleting as applicable. If both needed to be changed, both will appear.	Luke Wright	More...				over 12 years ago
Fixed bug with pointing to elements that are "position: fixed", by making the pointer fixed itself in these circumstances. This required a lot of unravelling of the pointer code, due to conflicts between a concept of a "global" pointer object, and having one object per document (changes made for multi-frame documents). Now there is no pointer member object; all getting of the pointer should go through getPointer().	Luke Wright	More...				over 12 years ago
Elements hidden off screen will now be considered "hidden" for HTMLCS purposes.	Luke Wright	More...				over 12 years ago
Nodes should be shifted off, not popped off, so the errors come in document order.	Luke Wright	More...				over 12 years ago
Lots of changes, primarily proper colour contrast tests replacing the notices in SC 1.4.3/1.4.6.	Luke Wright	More...				over 12 years ago

←
1
2
…
8
9
10
11
12
13
14
15
16
…
22
23
→