Item10230: Non-ASCII chars in HTML tag attributes incorrectly de-encoded (need to stay URI-escaped).

See Item9973. This probably shouldn't be a release blocker for 1.1.3.

-- PaulHarvey - 06 Jan 2011

Technically, it should be perfectly valid to have high-bit chars in HTML tag attribute values which are supposed to contain URIs.

However, the browser will always need to escape them first; and without any guidance, (all?) browsers re-encode the URI as unicode/utf-8, and escape that. So by the time the server sees the URI, if Foswiki has saved all the attachment filenames in the non-utf8 native charset, there can only be 404 not-found errors.

See Item5437 for more.

-- PaulHarvey - 16 Apr 2011

The course of action - for within WysiwygPlugin:

• If we're already using {Site}{CharSet} = 'utf-8', do nothing.
• Otherwise, pre-percent-escape href & src attributes in HTML tags in the original TML2HTML request.

-- PaulHarvey - 16 Apr 2011

Commenting is disabled while not logged in