Re: [xml] libxml2 does xmlSchemaValidateDoc() support xml schema extensions?

Hi Piortr

I still can not validate xml that uses xml schema extensions. I think I figured out what the problem might be, but do not know how to resolve i. I wonder if the problem is with the way I set up my xml namespace?

I am setting up the xml schema namespace the same way eclipse does when you create a new XSD or XML file. I entered in the address bar of my browsers and got quite a surprise!
  <xs:documentation><p>This schema should never be used as such:
                    <a href="">the XML
                    Schema Recommendation</a> forbids the declaration of
                    attributes in this namespace</p>
 </xs:annotation> is very confusing. 

where does libxml get the schema lang schema/dtd? As a test I put some typos in my xsd file. libxml generated an error when I tried to validate the xsd file as expected?

what URL should I be using?

Do I need to make a special call to cause the libxml to use this other location?

My app will be running on a private network and not have access to, Do I need to copy a dtd or something to my local env?

thanks in advance


Attachment: ZooRequest.xsd
Description: Binary data

p.s. Bellow is a long email I was going to send before I got the surprise I mentioned above. There is a nice code example of how how to validate. Its much simpler than the other examples I have seen posted in the past

On Dec 13, 2011, at 7:19 AM, Piotr Sipika wrote:

I wrote a sample program (in C) to attempt to mimic what you have, and
was able to successfully validate the instance against the schema you

Would you be willing to send my your sample C program? 

I must be missing a function call somewhere?  My XML/XSD files that do not use extensions validate so I think my basic code structure is correct. In your original post you mentioned "xmllint". I did not know about that tool. All of my files validate according to xmllint.I stepped through xmllint to see how it worked. Its very complicated. I must have missed something. Looking at the xmllint source I was able to simplify my code and get rid some some memory leaks.

Here is my simplified code listing based on the xmllint source

   const char *schemaFileNameStr = ...;


    xmlSchemaParserCtxtPtr ctxt = xmlSchemaNewParserCtxt(schemaFileNameStr);


                             (xmlSchemaValidityErrorFunc) fprintf,
                             (xmlSchemaValidityWarningFunc) fprintf,


xmlSchemaPtr wxschemas = xmlSchemaParse(ctxt);
if (wxschemas == NULL) {
    NSLog(@"******* ERROR schema %@ failed to compile\n", xmlSchemaFilePath);
    return ret; 

for (all the xml files) {
    xmlSchemaValidCtxtPtr vctxt = xmlSchemaNewValidCtxt(wxschemas);


                            (xmlSchemaValidityErrorFunc) fprintf,
                            (xmlSchemaValidityWarningFunc) fprintf,


int error = xmlSchemaValidateDoc(vctxt, doc);
if (error == 0) {
        ret = YES;
} else if (error > 0) {
        ret = NO;
    fprintf(stderr, "%s fails to validate\n", filename);
} else {
        ret = NO; // validation generated an internal error



<?xml version="1.0" encoding="UTF-8"?>
  <animal xsi:type="zoo:Fish">
    <name>Blue Fin Tuna</name>

element animal: Schemas validity error : Element 'animal', attribute 'xsi:type': The attribute 'xsi:type' is not allowed.
element numberOfFins: Schemas validity error : Element 'numberOfFins': This element is not expected.

  ELEMENT zoo:cageRequest
    namespace zoo href="">
    namespace xsi href="">
    ELEMENT animal
      ATTRIBUTE xsi:type
      ELEMENT name
          content=Blue Fin Tuna
      ELEMENT numberOfFins

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]