Re: [xml] encoding question

[Marcelo Elias Del Valle <marcelo valle thales-is com br>]

Hello,

   And is there a way to specify to libxml that all strings I insert 
should be in some encoding, so I don´t need to do it manually? For 
example, I say to libxml that and node I insert should be converted from 
latin1 to utf8, automatically...
   But what I don´t understand is: I parsed a latin1 xml. So, libxml 
parser (version 2.6.9) converted all node contents to utf-8, right? So I 
have a utf-8 xml on memory... But when I dump the xml to a file whitout 
specifying encoding, it dumps using latin1... and when I add the root 
node of this tree as a child of a second tree created by hand and I try 
to dump as latin1, libxml gives me an error saying there are latin1 
caracters and those cannot be converted. So, it doesn´t appear libxml is 
storing the content of a parsed latin1 xml file in UTF-8...
   I will check it tomorow and examine better my source code. I will 
take a look to xml source code too. In the case I nee to send some 
patch, how should I proceed? I am using a stable version of libxml and I 
will need to run it on a Solaris... Do you know if cvs version is 
working on Solaris?

Thank you very much,

Marcelo.

Kaarle Kaila wrote:

> All data in the memory structures in libxml are in UTF-8 encoding. The
> encoding
> is used when reading data but is lost after that. You can specify what
> encoding
> to use when serializing the structure again.
>
> So if I understand your question and libxml right you cannot do what you
> want to do.
> I think the same goes for the other xml-software packages too.
>
> regards
> Kaarle
>