Re: Failure to assign/relate an in-memory schema to an in-memory xml

[Alberto Massari <am...@datadirect.com>]

Hi Frank,
whenever I need to validate an XML file against a schema, and the XML 
doesn't have a reference to that schema, I do the following:

- create a parser
- call useCachedGrammarInParse(true) or 
setFeature(fgXercesCacheGrammarFromParse, true)
- call loadGrammar(MemBufInputSource, Grammar::SchemaGrammarType, true) 
[the last argument tells the parser to cache the grammar)
- call setDoValidation(true), setValidationScheme(Val_Always), 
setDoNamespaces(true)
- call parse(MemBufInputSource) to load the XML

Hope this helps,
Alberto

At 17.45 18/05/2005 +0200, Gierschner, Frank wrote:
>Hi all.
>
>I have some experience with XML in general but I am new to Xerces, so 
>please provide some hints.
>
>I wish to schema-validate an in-memory xml-string, coming in from a CORBA 
>client, against an in-memory schema. But the parser (DOMBuilder) does not 
>seem to associate the xml-data with the provided schema. I provide both 
>from within MemBufInputSource and I am unaware of the real functionalty of 
>the so-called 'fake system id'(argument 3: bufId) and if it has something 
>to do with the failure.
>
>I even considered to write the schema temporarily to a xsd-file, build a 
>DOM from the incoming xml-string, adding a schemaLocation attribute to the 
>still unvalidated DOM, serializing this DOM into a new string and the 
>reparsing and validating against the xsd-file but that seems error-prone, 
>cumbersome and does not work yet, too.
>
>I provided some test results I encountered when dealing with files and the 
>interesting part of the code below.
>
>I would appreciate your help.
>  Thanks in advance.
>    Frank Gierschner
>
>
>// small excerpt on dealing with files now
>In order to get track of the problem I modified the SCMPrint example 
>around line 335 to be:
><code>
>    parser->loadGrammar(xsdFile, Grammar::SchemaGrammarType, true);
>    if (handler.getSawErrors())
>    {
>       handler.resetErrors();
>    }
>    else
>    {
>       parsedOneSchemaOkay = true;
>    }
>
>    parser->parse(xmlFile); // NEW
></code>
>where xmlFile denotes a file which is a XML-Document having a 
>schemaLocation attribute directing to xsdFile. If in this case xmlFile 
>contains an invalid document the previously applied ErrorHandler will see 
>fatal errors - that is correct. But when I remove the schemaLocation 
>attribute from xmlFile the same procedure results in errors like 'unknown 
>element ...' which reveal that xml-data and schema seem to be unrelated. I 
>guess the latter is the same problem occuring with the above stated 
>in-memory procedure.
>// end of excerpt
>
>
>The interesting part of the code I use for the in-memory variant looks 
>somewhat like the following:
>
><code>
>// the in memory schema
>static const char gMySchema[] =
>"<?xml version=\"1.0\" encoding=\"utf-8\" ?> \
><xs:schema xmlns:xs=\"http://www.w3.org/2001/XMLSchema\" 
>targetNamespace=\"urn:rtc:tgam:MNCfgTmpl\" \
>elementFormDefault=\"qualified\" attributeFormDefault=\"unqualified\"> \
><xs:element name=\"root\"> \
><xs:complexType> \
><xs:sequence> \
><xs:element name=\"FirstElement\" type=\"xs:string\" /> \
><xs:element name=\"SecondElement\" minOccurs=\"0\" type=\"xs:string\" /> \
><xs:element name=\"ShirdElement\" type=\"xs:string\" /> \
></xs:sequence> \
></xs:complexType> \
></xs:element> \
></xs:schema> \
>";
>
>... -- omitted other stuff --
>
>static DOMBuilder * getDOMBuilder(MyDOMErrorHandler *& myErrorHandler, 
>const char pSchema2use[], Grammar *& grammar)
>{
>    DOMImplementation *impl = 
> DOMImplementationRegistry::getDOMImplementation(L"LS");
>    DOMBuilder* theBuilder = 
> ((DOMImplementationLS*)impl)->createDOMBuilder(DOMImplementationLS::MODE_SYNCHRONOUS, 
> L"http://www.w3.org/2001/XMLSchema");
>
>    if (theBuilder->canSetFeature(XMLUni::fgDOMNamespaces, true))
>       theBuilder->setFeature(XMLUni::fgDOMNamespaces, true);
>
>    if (theBuilder->canSetFeature(XMLUni::fgDOMValidation, true))
>       theBuilder->setFeature(XMLUni::fgDOMValidation, true);
>
>    if (theBuilder->canSetFeature(XMLUni::fgDOMDatatypeNormalization, true))
>       theBuilder->setFeature(XMLUni::fgDOMDatatypeNormalization, true);
>
>    if (theBuilder->canSetFeature(XMLUni::fgXercesSchemaFullChecking, true))
>       theBuilder->setFeature(XMLUni::fgXercesSchemaFullChecking, true);
>
>    myErrorHandler = new MyDOMErrorHandler();
>    theBuilder->setErrorHandler(myErrorHandler);
>
>    if (pSchema2use)
>    {
>       grammar = NULL;
>       MemBufInputSource *myXSDInputSource = new MemBufInputSource((const 
> XMLByte *const)gMySchema, (unsigned int) strlen(gMySchema), 
> L"http://www.w3.org/2001/XMLSchema", false); // unaware about 
> functionality of argument 3
>       Wrapper4InputSource wrapper(myXSDInputSource); // will adopt 
> myXSDInputSource
>       grammar = theBuilder->loadGrammar(wrapper, 
> Grammar::SchemaGrammarType);
>    }
>
>    return theBuilder;
>}
>
>... -- omitted other stuff --
>
>static int buildDOMDocument(const XMLByte* const buf, const unsigned int 
>uiSize, DOMDocument* &doc)
>{
>    DOMBuilder* theBuilder = NULL;
>    MyDOMErrorHandler* myErrorHandler = NULL;
>    Grammar *grammar = NULL;
>    theBuilder = getDOMBuilder(myErrorHandler, gMySchema, grammar); // see 
> above
>
>
>    MemBufInputSource *myInputSource = new MemBufInputSource(buf, uiSize, 
> L"TEST", false); // unaware about functionality of argument 3
>
>    Wrapper4InputSource wrapper(myInputSource); // will adopt myInputSource
>
>    int retVal = -1;
>    try
>    {
>       DOMDocument* newDoc = theBuilder->parse(wrapper);
>       retVal = (myErrorHandler->sawFatals() ? -9 : 
> (myErrorHandler->sawErrors() ? -5 : 0));
>
>       if (newDoc)
>          doc = (DOMDocument*) newDoc->cloneNode(true);
>    }
>    catch (const XMLException& toCatch)
>    {
>       cout << "Exception message is: \n" << X2c(toCatch.getMessage()) << 
> "\n";
>    }
>    catch (const DOMException& toCatch)
>    {
>       cout << "Exception message is: \n" << X2c(toCatch.msg) << "\n";
>    }
>    catch (...)
>    {
>       cout << "Unexpected Exception \n" ;
>    }
>
>    theBuilder->release();
>
>    delete myErrorHandler;
>
>    return retVal;
>}
>
>// the in-memory xml created previously via ...
>         ...
>       DOMImplementation* impl = 
> DOMImplementationRegistry::getDOMImplementation(L"Range");
>       DOMDocumentType* doctyp= 0; // impl->createDocumentType(L"root", 
> NULL, L"TEST");
>       DOMDocument* doc = impl->createDocument(L"urn:rtc:tgam:MNCfgTmpl", 
> L"root", doctyp);
>       doc->setEncoding(L"UTF-8");
>       // 
> doc->setDocumentURI(L"http://www.w3.org/2001/XMLSchema-instance");// 
> xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" ???
>       DOMElement* root = doc->getDocumentElement();
>         ...
>// ... and generated via serialization by DOMWriter::writeNode() from this 
>DOMDocument (from an example) looks like this
>
><?xml version="1.0" encoding="UTF-8" standalone="no" ?>
><root xmlns="urn:rtc:tgam:MNCfgTmpl">
>   <FirstElement>aTextNode</FirstElement>
>   <ThirdElement>aTextNode2</ThirdElement>
></root>
>
></code>