Hi, I'am writing some extensions an use usually a scheme like that:
<url src="http://www.destatis.de/basis/d/biwiku/hochtab1.php">/<!-- Content Begin -->(.*?)<!-- Content End -->/s</url>
I rely heavily on the fact that in my render function "renderUrl ($input, $params)" $input contains the unrendered code between <url>...</url>. But if I look to the content (after updating from 1.5.5 to 1.6.5) $input contains:
/UNIQ197a31dbac4f402-HTMLCommentStrip27fc8dec69226a9200000001(.*?)UNIQ197a31dbac4f402-HTMLCommentStrip27fc8dec69226a9200000002/s
instead of
/<!-- Content Begin -->(.*?)<!-- Content End -->/s
Obviously some rendering has been done. I remember I have seen somewhere (but where ?, can't find it anymore) how to access the unrendered text. Can someone help me ?
Thanks in advance
Sigbert
On 17/05/06, Sigbert Klinke sigbert@wiwi.hu-berlin.de wrote:
Hi, I'am writing some extensions an use usually a scheme like that:
<url src="http://www.destatis.de/basis/d/biwiku/hochtab1.php">/<!-- Content Begin -->(.*?)<!-- Content End -->/s</url>
I rely heavily on the fact that in my render function "renderUrl ($input, $params)" $input contains the unrendered code between <url>...</url>. But if I look to the content (after updating from 1.5.5 to 1.6.5) $input contains:
/UNIQ197a31dbac4f402-HTMLCommentStrip27fc8dec69226a9200000001(.*?)UNIQ197a31dbac4f402-HTMLCommentStrip27fc8dec69226a9200000002/s
instead of
/<!-- Content Begin -->(.*?)<!-- Content End -->/s
Obviously some rendering has been done. I remember I have seen somewhere (but where ?, can't find it anymore) how to access the unrendered text. Can someone help me ?
It's not rendering, it's "armouring" - that is, hiding something from the parser, in this case, comments. Obviously the parser was changed to strip out comments from 1.6.0 onwards - sensible move, I feel.
Rob Church
Hi,
It's not rendering, it's "armouring" - that is, hiding something from the parser, in this case, comments. Obviously the parser was changed to strip out comments from 1.6.0 onwards - sensible move, I feel.
Not really, because
1.) usually I embed between two tags <tag>...</tag> source code for a server software which delivers HTML code back. If by chance a combination of operators becomes a HTML comment or, whatever else is armoured, the program code won't execute anymore. 2.) The same behaviour appears if I use a construct like <url src="...." expr="/<!-- Content Begin -->(.*?)<!-- Content End -->/s">, the contents of params['expr'] is also /UNIQ197a31dba... Why the attribute of tag is "armoured" as well ?
Thanks for the answer nevertheless Sigbert
On 18/05/06, Sigbert Klinke sigbert@wiwi.hu-berlin.de wrote:
1.) usually I embed between two tags <tag>...</tag> source code for a server software which delivers HTML code back. If by chance a combination of operators becomes a HTML comment or, whatever else is armoured, the program code won't execute anymore.
The intent of a comment is usually to hide text and/or prevent markup from having an effect.
2.) The same behaviour appears if I use a construct like <url src="...." expr="/<!-- Content Begin -->(.*?)<!-- Content End -->/s">, the contents of params['expr'] is also /UNIQ197a31dba... Why the attribute of tag is "armoured" as well ?
I don't *think* the parser makes that distinction at the time it does the stuff.
Rob Church
mediawiki-l@lists.wikimedia.org