Thanks for this. It is rather tricky to parse. The only was to correctly parse #NAME..#NAME is to find the second # and the preceding double dots. A NAME can valid syntactically include a double dot, eg #name1..2..#name3
Simon On 9 April 2012 06:11, Patrick R. Michaud <pmich...@pobox.com> wrote: > On Sun, Apr 08, 2012 at 11:14:54PM +1200, Simon wrote: > > Thanks for the very prompt reply. > > I believe that this is a bug then, include does not use the same > definition > > of an anchor token as does the code that parses links and anchors. > > The problem appears to be the definition of $npat on line 1266 of > pwmiki.php: > > $npat = '[[:alpha:]][-\\w*]*'; > > Note that simply adding a dot to the [-\\w]* part of the expression > won't work, as a pair of dots are used to indicate ranges. > It probably needs to be something that can recognize a single dot > in the identifier pattern (i.e., a dot followed by something > other than another dot). > > Also, I suspect the '*' immediately the \\w in the original is a typo. > And there's an extraneous definition of $npat on line 1311 in the > IncludeText() function, probably left over from a previous factoring > of text includes. > > Pm > > -- ____ http://kiwiwiki.co.nz
_______________________________________________ pmwiki-devel mailing list pmwiki-devel@pmichaud.com http://www.pmichaud.com/mailman/listinfo/pmwiki-devel