|
Archives of the TeradataForumMessage Posted: Fri, 14 Jan 2005 @ 11:22:28 GMT
Here's a couple of other approaches. All of the ETL/EAI product support XML-Relational mapping. You can always write code--it would use a SAX or DOM based parser to process the document and would build and execute SQL (ODBC/JDBC) or output flat records (e.g pipe to a TPump). On Windows, OLE-DB can map XML content to tables/columns from a properly formatted source document (perhaps use XSLT to convert the source to this format). There was an XPath UDF (prototype, not product) shown at Partners--this allows xpath expressions e.g. in an insert/select. You could insert the XML into a CLOB or Varchar column and then pick out the interesting elements/attributes via xpath expressions in the select list (e.g. .. Select .. XPathUDF(xmlsourcecol,'//Order/BillTo/') One interesting approach is use of XSLT. It is an XML transformation language that can read XML and output flat records, e.g. delimited vartext suitable for TPump, BTEQ, Multiload, Fastload or table function return arguments. There are many available XSLT processors and XSLT Design tools that can help with syntax, debugging, testing.
| ||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||
Copyright 2016 - All Rights Reserved | ||||||||||||||||||||||||||||||||||||||||||||||||
Last Modified: 15 Jun 2023 | ||||||||||||||||||||||||||||||||||||||||||||||||