Re: [xml] XML Catalog problem

[Chris Wareham <chris wareham pa press net>]

Daniel Veillard wrote:
> On Mon, Feb 13, 2006 at 01:10:38PM +0000, Chris Wareham wrote:
> >     if (xmlValidateDocument(&valid_ctxt, parser->doc) == 0) {
>
>   Very first thing is to not use xmlValidateDocument, it's a post
> parsing validation mechanism, not the real one and for example won't
> process the internal subset so won't work in this case. Use the real
> validation as defined in the XML-1.0 spec i.e. as part of parsing.
>   Use xmlReadDoc with the XML_PARSE_DTDVALID like xmllint --valid 
> would do.
>
> Daniel

Hi Daniel,

I amended my function to the following:

int
newsmlparser_set_xml(NewsMLParser *parser, const char *xml)
{
    parser->doc = xmlReadFile(xml, "UTF-8", XML_PARSE_DTDVALID);

    if (parser->doc == NULL) {
        parser->status = NEWSMLPARSER_MALFORMED_XML;
        return 1;
    }

    return 0;
}

This now appears to validate the NewsML file with embedded NITF:

Resolve: sysID http://www.customer.press.net/resources/nitf.dtd
0 Parsing catalog newsmlcatalog
newsmlcatalog added to file hash
Found system match http://www.customer.press.net/resources/nitf.dtd
Resolve: pubID urn:newsml:iptc.org:20001006:NewsMLv1.0.dtd:1 sysID 
http://www.customer.press.net/resources/NewsMLv1.0.dtd
Found system match http://www.customer.press.net/resources/NewsMLv1.0.dtd

Before running my program, I disconnected my development machine from
the network as it has a version of libxml2 installed that includes HTTP
and FTP support. The NewsML file was validated, so my program must be
using the local DTD's referenced in the catalog.

If I deliberately break my NewsML file by adding an invalid element, I
get an error message on the console, but the xmlReadFile function still
returns a non-NULL xmlDocPtr. How should I test the validity of the XML
after a call to xmlReadFile?

Regards,

Chris