path: root/doc/todo/format_escape.mdwn

Since some preprocessor directives insert raw HTML, it would be good to 
specify, per-format, how to pass HTML so that it goes through the format 
OK. With Markdown we cross our fingers; with reST we use the "raw" 
directive.

I added an extra named parameter to the htmlize hook, which feels sort of
wrong, since none of the other hooks take parameters. Let me know what 
you think. --Ethan

Seems fairly reasonable, actually. Shouldn't the `$type` come from `$page`
instead of `$destpage` though? Only other obvious change is to make the
escape parameter optional, and only call it if set. --[[Joey]]

> I couldn't figure out what to make it from, but thinking it through, 
> yeah, it should be $page. Revised patch follows. --Ethan

>> I've updated the patch some more, but I think it's incomplete. ikiwiki
>> emits raw html when expanding WikiLinks too, and it would need to escape
>> those. Assuming that escaping html embedded in the middle of a sentence
>> works.. --[[Joey]]

>>> Revised again. I get around this by making another hook, htmlescapelink,
>>> which is called to generate links in whatever language. In addition, it 
>>> doesn't (can't?) generate
>>> spans, and it doesn't handle inlineable image links. If these were 
>>> desired, the approach to take would probably be to use substitution
>>> definitions, which would require generating two bits of code for each
>>> link/html snippet, and putting one at the end of the paragraph (or maybe
>>> the document?).
>>> To specify that (for example) Discussion links are meant to be HTML and
>>> not rst or whatever, I added a "genhtml" parameter to htmllink. It seems
>>> to work -- see <http://ikidev.betacantrips.com/blah.html> for an example.
>>> --Ethan

## Alternative solution

[Here](http://www.jk.fr.eu.org/ikiwiki/format-escapes-2.diff) is a patch
largely inspired from the one below, which is up to date and written with
[[todo/multiple_output_formats]] in mind. "htmlize" hooks are generalized
to "convert" ones, which can be registered for any pair of filename
extensions.

Preprocessor directives are allowed to return the content to be inserted
as a hash, in any format they want, provided they provide htmlize hooks for it.
Pseudo filename extensions (such as `"_link"`) can also be introduced,
which aren't used as real extensions but provide useful intermediate types.

--[[JeremieKoenig]]

> Wow, this is in many ways a beautiful patch. I did notice one problem,
> if a link is converted to rst and then from there to a hyperlink, the
> styling info usially added to such a link is lost. I wonder if it would
> be better to lose _link stuff and just create link html that is fed into
> the rst,html converter. Other advantage to doing that is that link
> creation has a rather complex interface, with selflink, attrs, url, and
> content parameters.
> 
> --[[Joey]]

>> Thanks for the compliment. I must confess that I'm not too familiar with
>> rst. I am using this todo item somewhat as a pretext to get the conversion
>> stuff in, which I need to implement some other stuff. As a result I was
>> less careful with the rst plugin than with the rest of the patch.

>> This being said, as I understand it rst cannot embed raw html in
>> the middle of a paragraph. I just found with more tests that even
>> links are a bit tricky, and won't work if they're not surrounded by
>> whitespace; the problem is that if we add this space, links
>> and preprocessor directives at the beginning of a line will be indented,
>> and this means something to rst. Also, rst complains about "?"
>> being used multiple times when the page contains more than one broken link,
>> apparently it uses it as a name for the reference as well as the link text.

>> The idea behind _link and other "intermediate
>> forms" was also that, when we can use rst's ability to target other output
>> formats, raw html won't be included in this process, and that
>> complications will happen with all markup languages if html continues
>> to be used as the language for preprocessor directive output.
>> Of course this could have been postponed until we actually need it,
>> but since we do... :-)

>> I think I will document the limitations, and tune the bugs of the
>> rst plugin code to do the most sensible thing after some more reading
>> of the rst docs. Expect an updated patch in the next few days, and feel
>> free to ask for other adjustments in the meantime.

>> Beyond being buggy in the least horrible way, I'm afraid I won't have
>> much time for ikiwiki in the next two or three weeks (exams),
>> but I think that ultimately these limitations could be worked around.
>> I'm not sure it is desirable for ikiwiki to know too much about the
>> syntax of its markup languages. Maybe the tricky "format" stuff
>> the toc plugin does could be used; maybe we need to think about more
>> generic ways to put "marks" in the various types of pages, which could
>> be expanded afer htmlization, and maybe the convert stuff could be used
>> to do this in an elegant way;
>> but then this is not very [[multiple_output_formats]] friendly either.
>> What do you think?

>> --[[JeremieKoenig]]

## Original patch
[[tag patch]]

<pre>
Index: debian/changelog
===================================================================
--- debian/changelog	(revision 3197)
+++ debian/changelog	(working copy)
@@ -24,6 +24,9 @@
     than just a suggests, since OpenID is enabled by default.
   * Fix a bug that caused link(foo) to succeed if page foo did not exist.
   * Fix tags to page names that contain special characters.
+  * Based on a patch by Ethan, add a new htmlescape hook, that is called
+    when a preprocssor directive emits inline html. The rst plugin uses this
+    hook to support inlined raw html.
 
   [ Josh Triplett ]
   * Use pngcrush and optipng on all PNG files.
Index: IkiWiki/Render.pm
===================================================================
--- IkiWiki/Render.pm	(revision 3197)
+++ IkiWiki/Render.pm	(working copy)
@@ -96,7 +96,7 @@
 		if ($page !~ /.*\/\Q$discussionlink\E$/ &&
 		   (length $config{cgiurl} ||
 		    exists $links{$page."/".$discussionlink})) {
-			$template->param(discussionlink => htmllink($page, $page, gettext("Discussion"), noimageinline => 1, forcesubpage => 1));
+			$template->param(discussionlink => htmllink($page, $page, gettext("Discussion"), noimageinline => 1, forcesubpage => 1, genhtml => 1));
 			$actions++;
 		}
 	}
Index: IkiWiki/Plugin/rst.pm
===================================================================
--- IkiWiki/Plugin/rst.pm	(revision 3197)
+++ IkiWiki/Plugin/rst.pm	(working copy)
@@ -30,15 +30,36 @@
 html = publish_string(stdin.read(), writer_name='html', 
        settings_overrides = { 'halt_level': 6, 
                               'file_insertion_enabled': 0,
-                              'raw_enabled': 0 }
+                              'raw_enabled': 1 }
 );
 print html[html.find('<body>')+6:html.find('</body>')].strip();
 ";
 
 sub import { #{{{
 	hook(type => "htmlize", id => "rst", call => \&htmlize);
+	hook(type => "htmlescape", id => "rst", call => \&htmlescape);
+	hook(type => "htmlescapelink", id => "rst", call => \&htmlescapelink);
 } # }}}
 
+sub htmlescapelink ($$;@) { #{{{
+	my $url = shift;
+	my $text = shift;
+	my %params = @_;
+
+	if ($params{broken}){
+		return "`? <$url>`_\ $text";
+	}
+	else {
+		return "`$text <$url>`_";
+	}
+} # }}}
+
+sub htmlescape ($) { #{{{
+	my $html=shift;
+	$html=~s/^/  /mg;
+	return ".. raw:: html\n\n".$html;
+} # }}}
+
 sub htmlize (@) { #{{{
 	my %params=@_;
 	my $content=$params{content};
Index: doc/plugins/write.mdwn
===================================================================
--- doc/plugins/write.mdwn	(revision 3197)
+++ doc/plugins/write.mdwn	(working copy)
@@ -121,6 +121,26 @@
 The function is passed named parameters: "page" and "content" and should
 return the htmlized content.
 
+### htmlescape
+
+	hook(type => "htmlescape", id => "ext", call => \&htmlescape);
+
+Some markup languages do not allow raw html to be mixed in with the markup
+language, and need it to be escaped in some way. This hook is a companion
+to the htmlize hook, and is called when ikiwiki detects that a preprocessor
+directive is inserting raw html. It is passed the chunk of html in
+question, and should return the escaped chunk.
+
+### htmlescapelink
+
+	hook(type => "htmlescapelink", id => "ext", call => \&htmlescapelink);
+
+Some markup languages have special syntax to link to other pages. This hook
+is a companion to the htmlize and htmlescape hooks, and it is called when a
+link is inserted. It is passed the target of the link and the text of the 
+link, and an optional named parameter "broken" if a broken link is being
+generated. It should return the correctly-formatted link.
+
 ### pagetemplate
 
 	hook(type => "pagetemplate", id => "foo", call => \&pagetemplate);
@@ -355,6 +375,7 @@
 * forcesubpage  - set to force a link to a subpage
 * linktext - set to force the link text to something
 * anchor - set to make the link include an anchor
+* genhtml - set to generate HTML and not escape for correct format
 
 #### `readfile($;$)`
 
Index: doc/plugins/rst.mdwn
===================================================================
--- doc/plugins/rst.mdwn	(revision 3197)
+++ doc/plugins/rst.mdwn	(working copy)
@@ -10,10 +10,8 @@
 Note that this plugin does not interoperate very well with the rest of
 ikiwiki. Limitations include:
 
-* reStructuredText does not allow raw html to be inserted into
-  documents, but ikiwiki does so in many cases, including
-  [[WikiLinks|WikiLink]] and many
-  [[PreprocessorDirectives|PreprocessorDirective]].
+* Some bits of ikiwiki may still assume that markdown is used or embed html
+  in ways that break reStructuredText. (Report bugs if you find any.)
 * It's slow; it forks a copy of python for each page. While there is a
   perl version of the reStructuredText processor, it is not being kept in
   sync with the standard version, so is not used.
Index: IkiWiki.pm
===================================================================
--- IkiWiki.pm	(revision 3197)
+++ IkiWiki.pm	(working copy)
@@ -469,6 +469,10 @@
 	my $page=shift; # the page that will contain the link (different for inline)
 	my $link=shift;
 	my %opts=@_;
+	# we are processing $lpage and so we need to format things in accordance
+	# with the formatting language of $lpage. inline generates HTML so links
+	# will be escaped seperately.
+	my $type=pagetype($pagesources{$lpage});
 
 	my $bestlink;
 	if (! $opts{forcesubpage}) {
@@ -494,12 +498,17 @@
 	}
 	if (! grep { $_ eq $bestlink } map { @{$_} } values %renderedfiles) {
 		return $linktext unless length $config{cgiurl};
-		return "<span><a href=\"".
-			cgiurl(
-				do => "create",
-				page => pagetitle(lc($link), 1),
-				from => $lpage
-			).
+		my $url = cgiurl(
+				 do => "create",
+				 page => pagetitle(lc($link), 1),
+				 from => $lpage
+				);
+
+		if ($hooks{htmlescapelink}{$type} && ! $opts{genhtml}){
+			return $hooks{htmlescapelink}{$type}{call}->($url, $linktext,
+							       broken => 1);
+		}
+		return "<span><a href=\"". $url.
 			"\">?</a>$linktext</span>"
 	}
 	
@@ -514,6 +523,9 @@
 		$bestlink.="#".$opts{anchor};
 	}
 
+	if ($hooks{htmlescapelink}{$type} && !$opts{genhtml}) {
+	  return $hooks{htmlescapelink}{$type}{call}->($bestlink, $linktext);
+	}
 	return "<a href=\"$bestlink\">$linktext</a>";
 } #}}}
 
@@ -628,6 +640,14 @@
 				preview => $preprocess_preview,
 			);
 			$preprocessing{$page}--;
+
+			# Handle escaping html if the htmlizer needs it.
+			if ($ret =~ /[<>]/ && $pagesources{$page}) {
+				my $type=pagetype($pagesources{$page});
+				if ($hooks{htmlescape}{$type}) {
+					return $hooks{htmlescape}{$type}{call}->($ret);
+				}
+			}
 			return $ret;
 		}
 		else {
</pre>