* Don't run tidy with -xml as that fails if the input is not well-formed.

Run it with -asxhtml instead, so it will output well-formed xhtml no
  matter what the input.
* Disable tidy warnings too.
* Add a new format hook, and make tidy use it, since tidy can really only
  operate on and output complete html documents, not the body chunks
  that sanitise gets.

IkiWiki/Plugin/htmltidy.pm

@@ -13,11 +13,11 @@ use IkiWiki;
 use IPC::Open2;
 
 sub import { #{{{
-	IkiWiki::hook(type => "sanitize", id => "tidy", call => \&sanitize);
+	IkiWiki::hook(type => "format", id => "tidy", call => \&format);
 } # }}}
 
-sub sanitize ($) { #{{{
-	open2(*IN, *OUT, 'tidy -quiet -xml -indent -utf8') or return shift;
+sub format ($) { #{{{
+	open2(*IN, *OUT, 'tidy -quiet -asxhtml -indent -utf8 --show-warnings no') or return shift;
 	# open2 doesn't respect "use open ':utf8'"
 	binmode (IN, ':utf8'); 
 	binmode (OUT, ':utf8'); 

IkiWiki/Render.pm

@@ -187,7 +187,13 @@ sub genpage ($$$) { #{{{
 		shift->(page => $page, destpage => $page, template => $template);
 	});
 	
-	return $template->output;
+	$content=$template->output;
+
+	run_hooks(format => sub {
+		$content=shift->($content);
+	});
+
+	return $content;
 } #}}}
 
 sub check_overwrite ($$) { #{{{

debian/changelog

@@ -13,8 +13,15 @@ ikiwiki (1.16) UNRELEASED; urgency=low
   * Also generate rel=bookmark links for permalinks.
   * Fix the htmltidy plugin, which wasn't working due my breaking it when
     I added it..
+  * Don't run tidy with -xml as that fails if the input is not well-formed.
+    Run it with -asxhtml instead, so it will output well-formed xhtml no
+    matter what the input.
+  * Disable tidy warnings too.
+  * Add a new format hook, and make tidy use it, since tidy can really only
+    operate on and output complete html documents, not the body chunks
+    that sanitise gets.
 
- -- Joey Hess <joeyh@debian.org>  Fri,  4 Aug 2006 03:03:19 -0400
+ -- Joey Hess <joeyh@debian.org>  Fri,  4 Aug 2006 03:33:09 -0400
 
 ikiwiki (1.15) unstable; urgency=low
 

doc/plugins/aggregate.mdwn

@@ -13,7 +13,9 @@ aggregated feeds.
 ## setup
 
 Make sure that you have the [[html]] plugin enabled, as the created pages are
-in html format. The [[meta]] and [[tag]] plugins are also recommended.
+in html format. The [[meta]] and [[tag]] plugins are also recommended. The
+[[htmltidy]] plugin is suggested, since feeds can easily contain invalid
+html which tidy can fix.
 
 You will need to run ikiwiki periodically from a cron job, passing it the
 --aggregate parameter, to make it check for new posts. Here's an example

doc/plugins/write.mdwn

@@ -129,10 +129,20 @@ to set it, as setting a variable that's not present is an error.
 	IkiWiki::hook(type => "sanitize", id => "foo", call => \&sanitize);
 
 Use this to implement html sanitization or anything else that needs to
-modify the content of a page after it has been fully converted to html.
+modify the body of a page after it has been fully converted to html.
 The function is passed the page content and should return the sanitized
 content.
 
+## format
+
+	IkiWiki::hook(type => "format", id => "foo", call => \&format);
+
+The function is passed the complete page content and can reformat it
+and return the new content. The difference between format and sanitize is
+that sanitize only acts on the page body, while format can modify the
+entire html page including the header and footer inserted by ikiwiki, the
+html document type, etc.
+
 ## delete
 
 	IkiWiki::hook(type => "delete", id => "foo", call => \&delete);