XSU and multiple tables

Question

Hi,
So far I have success using XSU to store an XML file to a single table datawise where each element of the XML correspond to the field name of the table.
But I have couple of questions.
1. I am still using ROWSET and ROW tags in my XML. Can i change it to something meaningful say...PRODUCTS and PRODUCT respectively? If so how can i do that?
2. How do i stored an XML onto more than one table. Each PRODUCT have some linked information to be stored in another table. How can i do this?
Basically this time the XML will look somewhat like this:
<?xml version="1.0" ?>
<ROWSET>
<ROW num="0">
<BRAND>brand name</BRAND>
<CATEGORY>category</CATEGORY>
<ROWSET1>
<ROW1>
<MERCHANT>merchant name</MERCHANT>
</ROW1>
</ROWSET1>
</ROW>
</ROW>
</ROWSET>
Please help me. Thank you.
Regards,
Abdul Rasheed

Answer 1

Hi,
search this xml form for 'multiple tables' and you will learn a lot about it. it seems that using XSU you can only to save to one table or a view. Possible solutions are:
1. create a view over those tables. If your view is complex and not
updatable then you can use insteadof triggers with your
non-updatable view"
2. split your mutl-level XML to several single level XML files and save each xml seperately.
3. And another person was nice enough to mail me the following solution, which might be of help. i copy the mail here...
I saw your message in the XML oracle forum, and i've just finished a small project for my company in doing things like saving an XML document to a DB table.
I think that saving the XML document in one or many tables is not different because what you need is to build dynamic SQL queries to update your tables so you need to guess the setter part of the querry.
UPDATE my_table SET column1=value1,column2=value2... WHERE table_key=key_value.
the best way to do the job is to use an XSLT processor, you give it your XML document and an XSL document for filtering the output.
the output of the XSLT processor could look like:
<?xml version="1.0" ?>
<ROOT>
<TABLEROW table="table1">
<COLUMN1_1>value1_1</COLUMN>
</TABLEROW>
<TABLEROW table="table2">
<COLUMN1_2>value1_2</COLUMN>
</TABLEROW>
</ROOT>
now it's easier to guess the SQL queries all you have to do is to parse this output document and say in you code something like:
String[NBR_OF_TABLES] Queries;
int i=0;
-- Loop inside the xml document {
Queries="UPDATE ";
Queries[i]=+(curent_node.getAttribute("table")).getNodeValue();
-- Loop inside the curent node {
column_name=current_node.getNodeName();
column_value=curent_node.getChild().getNodeValue();
Queries[i]=+" SET "+column_name+"="+column_value; //FIRST TIME
Queries[i]=+","+column_name+"="+column_value; //2,3,4...
you may need some type conversion, from string to numbers,and dates, and the WHERE part of the query is up to your indexation mecanism
of course you can put the tables primary keys values and names rigth in the yout XML document.
and last you simply execute the Array of queries you have:
for (int i=0i<NBR_OF_TABLES;i++)
execute(Queries[i]) .
hope this helps.
abdul rasheed