Re: [xml] xmlRecoverDoc should take a const xmlChar* ... (?)

[Martin Trappel <0xCDCDCDCD gmx at>]

Daniel Veillard wrote:
> On Wed, Jul 22, 2009 at 05:47:10PM +0200, Martin Trappel wrote:
>> Hi all!
>>
>> xmlDocPtr xmlRecoverDoc (xmlChar * cur) ... takes a non-const xmlChar*
>>
>> But looking at parser.c (2.6.27) shows that it just calls
>> xmlDocPtr
>> xmlRecoverDoc(xmlChar *cur) {
>>     return(xmlSAXParseDoc(NULL, cur, 1));
>> }
>>
>> and xmlSAXParseDoc takes a const xmlChar*
>>
>>
>> Maybe this should be fixed ...
>
>   Actually the true fix would be to drop xmlRecoverDoc() since it's
> contrary to the spec
>     http://www.w3.org/TR/REC-xml/#dt-fatal
> and people seems to abuse it despite warnings
>   http://www.advogato.org/person/DV/diary/224.html
> a broken RSS feed must be dropped at the place where the well-formedness
> happens and feedback should be provided that said feed is broken.
>

OK, cheers.

I will double-check the usage of the function and see if we can use a
different approach. (As far as I was told, we're not actually using it
for invalid data, but to parse an incomplete node or something ...)

br,
Martin


(p.s. sorry for double posts, messed up the reply clickies)