Markdown Discuss - Jan 2008 - Possible bug in handling of HTML comments

It appears that Markdown will process ampersands contained within HTML  
comments if that comment is part of a markdown paragraph, but will not  
when that paragraph is contained within an HTML block of it's own (and  
therefore ignored by Markdown.)

For example:

This is a markdown paragraph with a comment that *will* be processed  
<!-- This & *will* be converted -->

<p><!-- This & will *not* be converted --></p>

It seems to me that the & within a comment should be ignored by  
Markdown and not converted to &amp;


The reason this is important for me, is that I am experimenting with  
using HTML comments as a means to pass text that should be ignored by  
MultiMarkdown (e.g. to allow raw LaTeX, etc).  But some of the  
comments *are* being processed, which screws them up.


Thanks for any insight, and ideas on how to fix this!


Fletcher Penney


-- 
Fletcher T. Penney
fletcher at fletcherpenney.net

Without question, the greatest invention in the history of mankind
is beer. Oh, I grant you that the wheel was also a fine invention,
but the wheel does not go nearly as well with pizza.
	- Dave Barry

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2437 bytes
Desc: not available
Url :
<http://six.pairlist.net/pipermail/markdown-discuss/attachments/20080113/f213fb1b/attachment.bin>

So, another example:

	Test & <  >  <!-- Test & < sd > --> Test
	
	<p>Test & < > <!-- Test & < sd > -->
Test</p>

produces:

	p>Test &amp; &lt; > <!-- Test &amp; &lt; sd >
--> Test</p>
	
	<p>Test & < > <!-- Test & < sd > -->
Test</p>

It appears that the _EncodeAmpsAndAngles routine is being run over the  
comments, converting & and < to the HTML markup.

I tried adding & and < to $g_escape_table, and then protecting them.   
This seemed to work, but caused other problems and is clearly not a  
viable solution.

So it looks like there would have to be a specific mechanism in place  
for leaving comments alone, am I correct?


Thanks!

Fletcher


On Jan 13, 2008, at 8:06 PM, Fletcher T. Penney wrote:
> It appears that Markdown will process ampersands contained within  
> HTML comments if that comment is part of a markdown paragraph, but  
> will not when that paragraph is contained within an HTML block of  
> it's own (and therefore ignored by Markdown.)
>
> For example:
>
> This is a markdown paragraph with a comment that *will* be processed  
> <!-- This & *will* be converted -->
>
> <p><!-- This & will *not* be converted --></p>
>
> It seems to me that the & within a comment should be ignored by  
> Markdown and not converted to &amp;
>
>
> The reason this is important for me, is that I am experimenting with  
> using HTML comments as a means to pass text that should be ignored  
> by MultiMarkdown (e.g. to allow raw LaTeX, etc).  But some of the  
> comments *are* being processed, which screws them up.
>
>
> Thanks for any insight, and ideas on how to fix this!
>
>
> Fletcher Penney
>
>
> -- 
> Fletcher T. Penney
> fletcher at fletcherpenney.net
>
> Without question, the greatest invention in the history of mankind
> is beer. Oh, I grant you that the wheel was also a fine invention,
> but the wheel does not go nearly as well with pizza.
> 	- Dave Barry
>


-- 
Fletcher T. Penney
fletcher at fletcherpenney.net

Without question, the greatest invention in the history of mankind
is beer. Oh, I grant you that the wheel was also a fine invention,
but the wheel does not go nearly as well with pizza.
	- Dave Barry

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2437 bytes
Desc: not available
Url :
<http://six.pairlist.net/pipermail/markdown-discuss/attachments/20080113/3155e2e5/attachment.bin>

Le 2008-01-13 ? 21:08, Fletcher T. Penney a ?crit :
> It appears that the _EncodeAmpsAndAngles routine is being run over  
> the comments, converting & and < to the HTML markup.
>
> I tried adding & and < to $g_escape_table, and then protecting  
> them.  This seemed to work, but caused other problems and is  
> clearly not a viable solution.
>
> So it looks like there would have to be a specific mechanism in  
> place for leaving comments alone, am I correct?

PHP Markdown deals with this correctly since I added span-level  
hashes and use them to protect tags, comments, escapes, and  
everything that must be left alone. Perhaps someone ought to backport  
that system to Perl.


Michel Fortin
michel.fortin at michelf.com
http://michelf.com/