[xml] encoding

["Thomas Dold" <thomas dold gmx de>]

Another two question to the encoding:

First:
In the following program I set the property "town" which the value
"München".
Before I call the xmlSetProp-Function, I transform the value-string with the
function isolat1ToUTF8.
After setting the property, I read it again with the function xmlGetProp.
The returned string is in UTF-8, so I use the function UTF8Toisolat1.
But when I then print the string, I only get "Mnchen".

Then I save the document in a file (xmlSaveFile), and parse it again
(xmlParseFile).
Then I read once more the property on the way described above, but now I
get the strin "München".

And second:
As you can see, I try to save the file in different encodings. Only the
saving
without encoding has no error, and so it's in UTF-8 (or not?).
Why do I get the error, that the document is not in UTF-8, when I try to
save in
in another encoding?

Can someone help me? What's wrong?

Thanks.

Thomas

The Program:

#include <stdio.h>
#include "libxml/parser.h"
#include "libxml/tree.h"
#include "libxml/xmlmemory.h"
#include "libxml/xpath.h"
#include "libxml/encoding.h"

int
main(int argc, char **argv)
{
        xmlDocPtr doc;
        xmlNodePtr root, node;
        int i, rc;
        xmlXPathContextPtr ctx;
        xmlXPathObjectPtr path;
   const xmlChar *prop;
   xmlChar stringISO[50];
   xmlChar stringUTF[50];
   xmlChar* pStringUTF;
   xmlChar* pStringISO;
   int lenISO;
   int lenUTF;


        /* Test One */
        printf("Test1...\n");
        /* build the document */
   doc  = xmlNewDoc("1.0");
   root = xmlNewDocNode(doc, NULL, "root", NULL);
   node = xmlDocSetRootElement(doc, root);
   node = xmlNewChild(root, NULL,"node", NULL);
   /* set a property */
   strcpy(stringISO, "München");
   lenISO = strlen(stringISO);
   lenUTF = sizeof(stringUTF);
   memset(stringUTF, 0, lenUTF);
   rc = isolat1ToUTF8(stringUTF, &lenUTF, stringISO, &lenISO);
   xmlSetProp(node, "town", stringUTF);
   /* read the property */
   pStringUTF = xmlGetProp(node, (xmlChar*)"town");
   lenUTF = strlen(pStringUTF);
   lenISO = sizeof(stringISO);
   memset(stringISO, 0, lenISO);
   rc = UTF8Toisolat1(stringISO, &lenISO, pStringUTF, &lenUTF);
        printf("...with property town=%s\n",stringISO);

   printf("Save in UTF8: ");
   rc = xmlSaveFile("test1.xml", doc);
   if (rc > 0) printf(" OK\n"); else printf(" not OK, rc=%i\n",rc);

   printf("Save in ISO-Latin-1: ");
   rc = xmlSaveFileEnc("test1-ISO.xml", doc, "ISO-Latin-1");
   if (rc > 0) printf(" OK\n"); else printf(" not OK, rc=%i\n",rc);

   printf("Save in ISO-8859-1: ");
   rc = xmlSaveFileEnc("test1-ISO2.xml", doc, "ISO-8859-1");
   if (rc > 0) printf(" OK\n"); else printf(" not OK, rc=%i\n",rc);

   xmlFreeDoc(doc);
        printf("Test1-Ende...\n");

   /*-------------------------------------------------------------------*/

        printf("Test2...\n");
        /* parse the file */
        doc = xmlParseFile("test1.xml");

        xmlXPathInit();
        ctx = xmlXPathNewContext(doc);
        path = xmlXPathEvalExpression((const xmlChar *)"/child::*/child::*", ctx);
        if (path)
        {
                if (path->nodesetval->nodeNr > 0)
                {
                        node = path->nodesetval->nodeTab[0];
                        pStringUTF = xmlGetProp(node, (xmlChar*)"town");
         lenUTF = strlen(pStringUTF);
         lenISO = sizeof(stringISO);
         memset(stringISO, 0, lenISO);
         rc = UTF8Toisolat1(stringISO, &lenISO, pStringUTF, &lenUTF);
                        printf("...with property town=%s\n",stringISO);
                }
                xmlXPathFreeObject(path);
        }
        xmlXPathFreeContext(ctx);

   printf("Save in ISO-8859-1: ");
   rc = xmlSaveFileEnc("test2out.xml", doc, "ISO-8859-1");
   if (rc > 0) printf("OK\n");
        printf("Test2-Ende...\n");

   xmlFreeDoc(doc);

   return (0);
}

The Output:

Test1...
...with property town=Mnchen
Save in UTF8:  OK
xmlSaveFileEnc: document not in UTF8
Save in ISO-Latin-1:  not OK, rc=-1
xmlSaveFileEnc: document not in UTF8
Save in ISO-8859-1:  not OK, rc=-1
Test1-Ende...
Test2...
...with property town=München
Save in ISO-8859-1: OK
Test2-Ende...