Re: [xml] Problem parsing a string

[Igor Zlatkovic <igor zlatkovic com>]

On 19.05.2004 13:55, Juan David Palomino wrote:

> I don't know how parse a string with some \n inside.
> I have this string:
>  
>  res = "<?xml version=\"1.0\" encoding=\"iso-8859-1\"?>";
>  res += "<Message>";
>  res += "<ResponsePingATM Type=\"Status\" Order=\"1\" PhoneNumber=\"899999999\" Response=\"1\">";
>  res += "</ResponsePingATM>";
>  res += "</Message>";
>
> string without end of lines;
> When i do 
> doc = xmlParseMemory( res.c_str(), res.length());
> cur = xmlDocGetRootElement(doc);
>     ** cur->name is Message
> cur = cur->xmlChildrenNode;
>     ** cur->name is ResponsePingATM
>
> The problem is that i receive the following string
>
>  res = "<?xml version=\"1.0\" encoding=\"iso-8859-1\"?>\n";
>  res += "<Message>\n";
>  res += "<ResponsePingATM Type=\"Status\" Order=\"1\" PhoneNumber=\"899999999\" Response=\"1\">\n";
>  res += "</ResponsePingATM>\n";
>  res += "</Message>";
>
> string with end of lines;
> When i do 
> doc = xmlParseMemory( res.c_str(), res.length());
> cur = xmlDocGetRootElement(doc);
>     ** cur->name is Message
> cur = cur->xmlChildrenNode;
>     ** cur->name is text
>
> Are there other ways to parse this string (cur->next not allowed) that ignores \n?
> Thnak you to all.

No, sorry. The parser will never ignore anything. Doing so would break 
the XML specification.

Also, if you cannot do a cur-next, then something in your program is 
seriously broken. You cannot use DOM without full tree walk capability.

You can use XPath to select the nodes you need from the DOM. To avoid 
using cur->next in the result set, you will have to construct an extra 
XPath expression for every node you need. This means an XPath evaluation 
instead of doing a cur->next and checking the node type, that will bring 
the performance down.

Ciao,
Igor