DSD

[Update 7/7/09: WP-Typogrify has now merged with WP-Hyphenate, and is com­pat­i­ble with cap­tions out of the box.]

Hamish Macpher­son’s WP-Typogrify is one of my favorite Word­Press plu­g­ins. I started using it espe­cially for the Smar­ty­Pants func­tion­al­ity, which fixes “dumb quotes” and poor man’s apos­tro­phes, among other things.^[1]

How­ever, I was dis­ap­pointed to find that this func­tion­al­ity breaks Word­Press cap­tions (intro­duced in WP 2.6), which I’d rather not live with­out. Devel­op­ment on WP-Typogrify seems to have slowed^[2] — there hasn’t been a new ver­sion in a while, so I’ve taken the lib­erty of hack­ing ver­sion 1.6 to fix this incom­pat­i­bil­ity, at least so I can use Smar­ty­Pants until an offi­cial fix comes out. The adjust­ments I made are sim­ple, and I have no idea whether they’re max­i­mally robust. But feel free to

and use at your own risk.

The Changes

It might be use­ful to some to know specif­i­cally how to fix the incom­pat­i­bil­ity. WP-Typogrify already ignores text inside a <code> or <pre> block. All that’s nec­es­sary is to make it rec­og­nize word­press tags like [caption] as things it should ignore too. All the changes are in the file smartypants.php:

1. At the top, where it says “Globals:,” replace

   /* $sp_tags_to_skip = '<(/?)(?:pre|code|kbd|script|math)[\s>]'; */
   $sp_tags_to_skip = '(?:<(/?)(?:pre|code|kbd|script|math)[\s>]|'.
                      '\[(/?)(?:caption|code|php)[\s\]])';
   

2. In the “function SmartyPants” block, in the loop “foreach ($tokens as $cur_token),” replace

   /* $in_pre = isset($matches[2]) && $matches[2]  == '/' ? 0 : 1; */
   $in_pre = (isset($matches[1]) && $matches[1] == '/') ||
             (isset($matches[2]) && $matches[2]  == '/') ? 0 : 1;
   

3. Finally, in the “function _TokenizeHTML($str)” block, replace

   /*
   $match = '(?s:<!(?:--.*?--\s*)+>)|'.   # comment
            '(?s:<\?.*?\?>)|'.   # processing instruction
              # regular tags
            '(?:<[/!$]?[-a-zA-Z0-9:]+\b(?>[^"\'>]+|"[^"]*"|\'[^\']*\')*>)';
   */
   $match = '(?s:<!(?:--.*?--\s*)+>)|'.  # comment
     '(?s:<\?.*?\?>)|'.  # processing instruction
     '(?:<[/!$]?[-a-zA-Z0-9:]+\b(?>[^"\'>]+|"[^"]*"|\'[^\']*\')*>)|'.
     '(?:\[[/!$]?[-a-zA-Z0-9:]+\b(?>[^"\'>]+|"[^"]*"|\'[^\']*\')*\])';
   

That’s all — noth­ing fancy. Happy typography!

1. Of course, it won’t save you from these. [↩]
2. Though it’s appar­ently being taken over by the devel­op­ers of WP-Hyphenate. [↩]