Ignoring non-ASCII characters in XML file.

Question

Hello all who read this,
I got the following problem. Trying to read an XML file which contains UTF-8 characters. My database doesnt support UTF-8 char set .... obviouselly.
Below you will find my procedure for parsing XML file ... now i was wondering if there is a way (and how ) to ignore those "special" characters in my XML file.
Thank you for you help!
PROCEDURE xml_import is
BEGIN
DECLARE
  l_bfile   BFILE;
  l_clob    CLOB;
  l_parser  dbms_xmlparser.Parser;
  l_doc     dbms_xmldom.DOMDocument;
  l_nl      dbms_xmldom.DOMNodeList;
  l_n       dbms_xmldom.DOMNode;
  l_temp    VARCHAR2(1000);
  FILER     VARCHAR2(1000) := 'xml.xml';  
  TYPE tab_type IS TABLE OF X_XML%ROWTYPE;
  t_tab  tab_type := tab_type();
BEGIN
  l_bfile := BFileName('XML_DIR',FILER);
  dbms_lob.createtemporary(l_clob, cache=>FALSE);
  dbms_lob.open(l_bfile, dbms_lob.lob_readonly);
  dbms_lob.loadFromFile(dest_lob => l_clob,
                        src_lob  => l_bfile,
                        amount   => dbms_lob.getLength(l_bfile));
  dbms_lob.close(l_bfile);
  dbms_session.set_nls('NLS_DATE_FORMAT','''DD-MON-YYYY''');
  l_parser := dbms_xmlparser.newParser;
  dbms_xmlparser.parseClob(l_parser, l_clob);
  l_doc := dbms_xmlparser.getDocument(l_parser);
  dbms_lob.freetemporary(l_clob);
  dbms_xmlparser.freeParser(l_parser);
  l_nl :=  dbms_xslprocessor.selectNodes(dbms_xmldom.makeNode(l_doc),'/Doc/Upk');
  FOR cur_emp IN 0 .. dbms_xmldom.getLength(l_nl) - 1 LOOP
    l_n := dbms_xmldom.item(l_nl,cur_emp);
    t_tab.extend;
    dbms_xslprocessor.valueOf(l_n,'@id',t_tab(t_tab.last).id);
    dbms_xslprocessor.valueOf(l_n,'line1',t_tab(t_tab.last).vr);
    dbms_xslprocessor.valueOf(l_n,'line2',t_tab(t_tab.last).used);
    dbms_xslprocessor.valueOf(l_n,'line3',t_tab(t_tab.last).got);
  END LOOP;
  FORALL i IN t_tab.first .. t_tab.last
    INSERT INTO X_XML VALUES t_tab(i);
  COMMIT;
  dbms_xmldom.freeDocument(l_doc);
--EXCEPTION
--  WHEN OTHERS THEN
--    dbms_lob.freetemporary(l_clob);
--    dbms_xmlparser.freeParser(l_parser);
--    dbms_xmldom.freeDocument(l_doc);
END;
END;
XML file structure:
<?xml version="1.0" encoding="UTF-8" ?>
- <AjpesDokument>
- <Ident vrsta="KOMP">
  <Datum>2011-06-17</Datum>
  <Ura>09:33:10</Ura>
  </Ident>
- <Dolznik id="88888888">
  <DatumObdelave>2011-06-10</DatumObdelave>
  <Trr>56546654545</Trr>
  <PopolnoIme>Firm d.o.o.</PopolnoIme>
  <Ulica>Kičečeva ul</Ulica>
  <HisnaStev>026</HisnaStev>
  <PostaId>5421</PostaId>
  <Posta>MELJE</Posta>
  <IzpSif>00</IzpSif>
  <IzpNaziv>MELJE</IzpNaziv>
  <IzpNaslov>ČUPAVA</IzpNaslov>
  <IzpNaslov2>MELJE</IzpNaslov2>
- <Upnik davcna="10232293">
  <Naziv>CMC GALVANIKA d.o.o.</Naziv>
  <Kraj>LEŠČE</Kraj>
  <VrstaPobota>2</VrstaPobota>
  <Prijavljeno>3227.12</Prijavljeno>
  <Pobotano>3227.12</Pobotano>
  </Upnik>
- <Upnik davcna="11212191">
  <Naziv>AŽERO  d.o.o.</Naziv>
  <Kraj>MELJE</Kraj>
  <VrstaPobota>2</VrstaPobota>
  <Prijavljeno>219.30</Prijavljeno>
  <Pobotano>219.30</Pobotano>
  </Upnik>
  </Dolznik>
</AjpesDokument>
Edited by: 869620 on 1.7.2011 2:04

Answer 1

869620 wrote:
Hello all who read this,
I got the following problem. Trying to read an XML file which contains UTF-8 characters. My database doesnt support UTF-8 char set .... obviouselly.It's not obvious as Oracle can certainly be configured for UTF-8.
How about giving us some example XML and what you expect as output as there can be easier ways of extracting data from XML than using the dbms_xmldom package.
The dbms_xmldom package is one way, but it's not necessarily the best, as I would tend to use that for processing XML based structure that I don't actually know (I have used it myself, for example, for parsing XML Schemas (XSD) to auto-generate code for producing XML to match that schema), but for extracting date from a known XML structure, I would probably use SQL's XMLTABLE functionality.
And when you post code/data, please use the {noformat}{noformat} tags as detailed in the FAQ: {message:id=9360002}