Building XPath with the XML having the Namespace using PL SQL

Question

While trying to build the path of each node, when the XML has no namespace in it , the below code works fine providing the result
1~/
2~/person/
3~/person/age/
4~/person/homecity/
5~/person/name/
6~/person/homecity/lat/
7~/person/homecity/name/
8~/person/homecity/long/
But when the xml is changed to
<person xmlns="urn:person" xmlns:lat="urn:lat">
<name>Rob</name>
<age>37</age>
<homecity>
    <name>London</name>
    <lat>123.000</lat>
    <long>0.00</long>
</homecity>
</person>"
The result on executing the below code is resulting into only into below result
1~/
2~/person/
In the XML provided above, XML name space is not constant and it may vary for every XML. My requirement is to parse the complete XML, where i can read the XML with namespace and get the result as mentioned below
1~/
2~/person/
3~/person/age/
4~/person/homecity/
5~/person/name/
6~/person/homecity/lat:lat/
7~/person/homecity/name/
8~/person/homecity/long/
Can you please help me resolving the issue mentioned. Thanks in advance. --Code Snippet below :
DECLARE
  l_File VARCHAR2(32000) := '<person>
<name>Rob</name>
<age>37</age>
<homecity>
    <name>London</name>
    <lat>123.000</lat>
    <long>0.00</long>
</homecity>
</person>';
l_Where_Clause VARCHAR2(100) := '/*';
l_Append_Var   VARCHAR2(100) := '/';
TYPE Ty_Paths IS TABLE OF VARCHAR2(1000) INDEX BY PLS_INTEGER;
l_Ty_Paths      Ty_Paths;
l_Ty_Paths_Temp Ty_Paths;
TYPE Ty_Verifier IS TABLE OF VARCHAR2(1000) INDEX BY VARCHAR2(1000);
l_Ty_Varifier Ty_Verifier;
l_Prev_Query_Rec VARCHAR2(100);
l_Index_Num      NUMBER := 0;
l_Cur_Exec_Row   NUMBER := 0;
BEGIN
l_Ty_Paths(Nvl(l_Ty_Paths.COUNT, 0) + 1) := l_Append_Var;
l_Cur_Exec_Row := 1;
--Dbms_Output.put_line('Before entering the loop');
LOOP
   l_Ty_Paths_Temp.DELETE;
  SELECT DISTINCT REPLACE(l_Append_Var || '/' || t.Xml || '/', '//', '/') BULK COLLECT
   INTO   l_Ty_Paths_Temp
  FROM   (SELECT Xmltype(Extract(VALUE(e), '/').Getstringval()) .Getrootelement() AS Xml
           FROM   TABLE(Xmlsequence(Extract(Xmltype(l_File), l_Where_Clause))) e) t;
  l_Ty_Varifier(Nvl(l_Ty_Varifier.COUNT, 0) + 1) := l_Append_Var;
  --Dbms_Output.put_line('L_TY_PATHS_TEMP.Count::'||L_TY_PATHS_TEMP.Count);
  IF l_Ty_Paths_Temp.COUNT > 0 THEN
     l_Index_Num := Nvl(l_Ty_Paths.COUNT, 0) + 1;
     FOR i IN l_Ty_Paths_Temp.FIRST .. l_Ty_Paths_Temp.LAST LOOP
        l_Ty_Paths(l_Index_Num) := l_Ty_Paths_Temp(i);
        --Dbms_Output.put_line('L_INDEX_NUM::'||L_INDEX_NUM);
        --Dbms_Output.put_line('L_TY_PATHS(L_INDEX_NUM)::'||L_TY_PATHS(L_INDEX_NUM));
        l_Index_Num := l_Index_Num + 1;
     END LOOP;
  END IF;
  --Dbms_Output.put_line('L_TY_PATHS.Count::'||L_TY_PATHS.Count);
  --Dbms_Output.put_line('L_TY_PATHS.Count::'||L_CUR_EXEC_ROW);
  IF (NOT l_Ty_Paths.EXISTS(l_Cur_Exec_Row + 1)) OR (l_Cur_Exec_Row = l_Ty_Paths.COUNT) THEN
     --Dbms_Output.put_line('Exiting');
     EXIT;
  ELSE
     --Dbms_Output.put_line('Inside the Else part');
     l_Cur_Exec_Row := l_Cur_Exec_Row + 1;
     l_Append_Var   := l_Ty_Paths(l_Cur_Exec_Row);
     l_Where_Clause := l_Ty_Paths(l_Cur_Exec_Row) || '*';
  END IF;
  --To Display the record:
     --Dbms_Output.put_line(L_TY_PATHS.Count);
  END LOOP;
  IF l_Ty_Paths.COUNT > 0 THEN
    FOR i IN l_Ty_Paths.FIRST .. l_Ty_Paths.LAST LOOP
      Dbms_Output.Put_Line(i || ' record is ' || l_Ty_Paths(i));
   END LOOP;
END IF;
END;

Answer 1

Hi,
Thanks for the reply.
Please find the details :
1) What's your database version ?
Database version  :
Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - 64bit Production
2) What's the practical use of extracting all the paths like this, what are you trying to achieve, besides a purely academic exercise ?
The XML provided was for sample XML.
The practical use being ,I wanted to parse the SEPA messages using the below code snippet dynamically.
create table data_table (
     data xmltype
    xmltype column data store as securefile binary xml
insert into data_table values (
    xmltype('<?xml version="1.0" encoding="UTF-8"?>
<SCTScfBlkCredTrf xmlns="urn:S2SCTScf:xsd:$SCTScfBlkCredTrf">
  <SndgInst>CMCIFRPPXXX</SndgInst>
  <RcvgInst>RLBBAT2E083</RcvgInst>
  <FileRef>006FRID2516PH712</FileRef>
  <SrvcId>SCT</SrvcId>
  <TstCode>T</TstCode>
  <FType>SCF</FType>
  <FDtTm>2012-11-01T01:12:22</FDtTm>
  <NumCTBlk>1</NumCTBlk>
  <NumPCRBlk>0</NumPCRBlk>
  <NumRFRBlk>0</NumRFRBlk>
  <NumROIBlk>0</NumROIBlk>
<FIToFICstmrCdtTrf xmlns="urn:iso:std:iso:20022:tech:xsd:sct:pacs.008.001.02">
  <GrpHdr>
    <MsgId>006MSID12511PH712</MsgId>
    <CreDtTm>2012-11-01T08:09:14</CreDtTm>
    <NbOfTxs>1</NbOfTxs>
    <TtlIntrBkSttlmAmt Ccy="EUR">20000</TtlIntrBkSttlmAmt>
    <IntrBkSttlmDt>2012-11-01</IntrBkSttlmDt>
    <SttlmInf>
      <SttlmMtd>INGA</SttlmMtd>
      <ClrSys>
        <Prtry></Prtry>
      </ClrSys>
    </SttlmInf>
    <InstgAgt>
      <FinInstnId>
        <BIC>RLBBAT2E083</BIC>
      </FinInstnId>
    </InstgAgt>
    <InstdAgt>
      <FinInstnId>
        <BIC>HELADEF1XXX</BIC>
      </FinInstnId>
    </InstdAgt>
  </GrpHdr>
  <CdtTrfTxInf>
    <PmtId>
      <EndToEndId>006SEOP23END1712</EndToEndId>
      <TxId>006SEOP231PH1712</TxId>
    </PmtId>
    <PmtTpInf>
      <SvcLvl>
        <Cd>SEPA</Cd>
      </SvcLvl>
    </PmtTpInf>
    <IntrBkSttlmAmt Ccy="EUR">20000</IntrBkSttlmAmt>
    <AccptncDtTm>2012-11-01T12:00:00</AccptncDtTm>
    <ChrgBr>SLEV</ChrgBr>
    <Dbtr>
      <Nm>Mrinmoy Sahu</Nm>
      <PstlAdr>
        <Ctry>FR</Ctry>     
        <AdrLine>6</AdrLine>
        <AdrLine>Bangalore</AdrLine>
      </PstlAdr>
      <Id>
        <PrvtId>
          <Othr>
            <Id>E20809</Id>
            <SchmeNm></SchmeNm>
            <Issr>ORACLE CORP</Issr>
          </Othr>
        </PrvtId>
      </Id>
    </Dbtr>
    <DbtrAcct>
      <Id>
        <IBAN>FR7030087330086000000000591</IBAN>
      </Id>
    </DbtrAcct>
    <DbtrAgt>
      <FinInstnId>
        <BIC>CMCIFRPPXXX</BIC>
      </FinInstnId>
    </DbtrAgt>
    <CdtrAgt>
      <FinInstnId>
        <BIC>RLBBAT2E083</BIC>
      </FinInstnId>
    </CdtrAgt>
    <Cdtr>
      <Nm>Mrinmoy Sahu</Nm>
      <PstlAdr>
        <Ctry>FR</Ctry>     
        <AdrLine>22 </AdrLine>
      </PstlAdr>
      <Id>
        <PrvtId>
          <Othr>
            <Id>F676869</Id>
            <SchmeNm></SchmeNm>
            <Issr>GOVT OF INDIA</Issr>
          </Othr>
        </PrvtId>
      </Id>
    </Cdtr>
    <CdtrAcct>
      <Id>
        <IBAN>FR7630087330086000000004266</IBAN>
      </Id>
    </CdtrAcct>
    <RmtInf>
        <Ustrd>abc</Ustrd>
      </RmtInf>
  </CdtTrfTxInf>   
  </FIToFICstmrCdtTrf>
</SCTScfBlkCredTrf>
alter session set nls_numeric_characters = ".,";
      SELECT Msgid, Msgid1, Sttlmmtd
   FROM   data_table t,Xmltable(Xmlnamespaces('urn:S2SCTScf:xsd:$SCTScfBlkCredTrf' AS "A", 'urn:iso:std:iso:20022:tech:xsd:sct:pacs.008.001.02' AS "B")
                    ,'/A:SCTScfBlkCredTrf/B:FIToFICstmrCdtTrf' Passing t.data Columns
                    Msgid Path 'A:SndgInst'
                    ,Msgid1 Path 'B:GrpHdr/B:MsgId'
                    ,Sttlmmtd Path 'B:GrpHdr/B:SttlmInf/B:SttlmMtd');
MSGID             ;MSGID1           ;STTLMMTD         ;
                 ;006MSID12511PH712;INGA             ;
The idea was to :
1). Map the Column Names for the XPath built using the previous code.
2). Build the Select statement shown above dynamically based on the Xpath built in the previous code using a PLSQL block, and to process the data by doing a bulk collect on the XML.
The above XML may contain multiple <FIToFICstmrCdtTrf></FIToFICstmrCdtTrf> nodes.
Since the name space in the XML varies for each version of the SEPA messages , i need to get the names spaces ALSO to be picked up dynamically and parse the XML data based on the select statement as shown above.
Could you please guide me with
1). Is the approach taken appropriate?
2). Any alternative approach should be looked for ?
3). How to extract the name spaces available in the XML?
Thanks.