Add support for any attribute #186
Conversation
|
Nice. One thing I'll have to think about is security considerations. Currently someone can enable the no-markup mode and add a callback for changing URLs to filter javascript and other things that could be exploited in user-generated content. This PR pierces hole in those measures. Perhaps this feature should simply be disabled when the no-markup flag is set. Also, is there a reason for wanting a colon as a prefix? Why not simply accept |
|
I added Mainly to be consistent with the id and class attributes, since they are using a prefix as well, but I don't mind changing it. |
|
I'd rather have |
|
I removed the ':' and just checked for a-z as the first character |
| $id = false; | ||
| foreach ($elements as $element) { | ||
| if ($element{0} == '.') { | ||
| $classes[] = substr($element, 1); | ||
| } else if ($element{0} == '#') { | ||
| if ($id === false) $id = substr($element, 1); | ||
| } else if (strpos($element, '=') > 0) { | ||
| $parts = explode('=', $element); | ||
| $attributes[] = $parts[0] . '="' . $parts[1] . '"'; |
There was a problem hiding this comment.
Probably should use explode('=', $element, 2) since a third equal sign shouldn't cause the attribute value to end. At least that's now HTML parses unquoted attributes.
Why not |
|
Common practice is to write attribute in lowercase only, the spec allows uppercase in theory, but all uppercase characters are - internally - converted to lowercase. The data- attribute spec demands lowercase only. |
|
I agree that from an authoring perspective it's best to use lowercase names. But from the parser's perspective, I'm not sure it's a good idea to ignore attributes blocks with attributes set in uppercase. I guess it's rare enough that we can only bother if someone reports the issue at some point. Will merge this shortly. |
|
Thanks |
This PR adds support for any attribute as discussed in #93
Examples:
{:data-group=test}{:rel=nofollow}