Output from xml column

Question

I have a table
message with a column xml_receive
I want to query the text from the xml_receive column
XML_RECEIVE (data):
<?xml version='1.0' encoding='UTF-8'?>
<soapenv:Envelope xmlns:soapenv='http://schemas.xmlsoap.org/soap/envelope/'>
<soapenv:Header/>
  <soapenv:Body>
   <RechargeMSISDN>  
    <Timestamp>2013-11-21T05:33:39.000Z</Timestamp>
    <Transactionid>24</Transactionid>         
    <ResponseCode>0</ResponseCode>  
  </RechargeMSISDN>
</soapenv:Body>
</soapenv:Envelope>
sql query:
SELECT
EXTRACTVALUE(XMLTYPE(XML_RECEIVE), '//Timestamp') AS datum,
EXTRACTVALUE(XMLTYPE(XML_RECEIVE), '//ResponseCode') AS code,
EXTRACTVALUE(XMLTYPE(XML_RECEIVE), '//Transactionid') AS id,
EXTRACTVALUE(XMLTYPE(XML_RECEIVE), '//ClientAccountBalance') AS balans
FROM message
Can anyone tell me what is wong with my query?

Answer 1

The root cause of the error is probably that when you insert data to XML_RECEIVE, the data is not clean. It is well known that the prolog of a well-formed xml (if exists) must be at the very start of it.
The error message:
LPX-00209: PI names starting with XML are reserved
Error at line 2
means there is a dubious line before the prolog.
You sure would be better off in the long run to clean the column up. In the meantime, you can use a regexp_replace to preprocess the varchar2 type of XML_RECEIVE.
Instead of this:
EXTRACTVALUE(XMLTYPE(XML_RECEIVE), '//Timestamp') AS datum,
script it like this:
EXTRACTVALUE(XMLTYPE(regexp_replace(XML_RECEIVE, '.*<\?xml','<?xml',1,1,'n')), '//Timestamp') AS datum,
etc,,, (the same for any other occurrences of XML_RECEIVE). I avoid more sophisticated pattern as the support is doubtful. The pattern uses some basic construction that the support is solid. Also the pattern actually allowed multiple line of 'dirty' junk data in front of the prolog, not necessarily whitespaces nor just a single line.