[dokuwiki] Re: invalid anchors
- From: Andreas Gohr <andi@xxxxxxxxxxxxxx>
- To: dokuwiki@xxxxxxxxxxxxx
- Date: Thu, 09 Feb 2006 12:19:09 +0100
Chris Smith writes:
fyi, the "specification" for names and ids is in html4.01 document,
http://www.w3.org/TR/html401/types.html#h-6.2
they are both specified as CDATA with the following extra constraint
[a-zA-Z][a-zA-Z0-9_:.-]*
*sigh* That's so stupid :-/ I guess even the coders of the W3C validator
found that stupid and didn't implement it.
And add to dokuwiki parser, the ability to understand an actual headline
in a dokuwiki internal link and determine the correct matching fragment
identifier to use in the link. That would allow complete separation of
real fragment ids from headlines making it easy for the parser to
determine if the fragment id is from a pasted link or from a
typed/pasted headline.
Sounds perfect in theory but
a) would require parsing all linked documents for their section headers
b) would not work when linking to remote DokuWikis
How the hell do other Wikis handle this? Can anyone do a quick survey?
Andi
--
DokuWiki mailing list - more info at
http://wiki.splitbrain.org/wiki:mailinglist
Other related posts:
