How to delete a node from Oracle XMLTYPE based on a condition?

Question

I have a XML data stored in a CLOB column and I would like to delete some nodes based on a specific condition.

Example XML Data :

<?xml version="1.0" encoding="UTF-8"?>
<payment>
  <person>
    <surname>Marco</surname>
    <name>Gralike</name>
    <salary>2345</salary>
  </person>
  <person>
    <surname>ABC</surname>
    <name>TEST</name>
    <salary>1234</salary>
    <person>
    <surname>Tiger</surname>
    <name>Scott</name>
    <salary>2222</salary>
  </person>
  </person>
 </payment>
 <payment>
  <person>
    <surname>BertJan</surname>
    <name>Meinders</name>
    <salary>3456</salary>
    <salary>125</salary>
  </person>
  <person>
    <surname>XYZ</surname>
    <name>TEST</name>
    <salary>1234</salary>
  </person>
 </payment>
 <payment>
  <person>
    <surname>Chris</surname>
    <name>Gralike</name>
    <salary>4567</salary>
  </person>
  <person>
    <surname>LMN</surname>
    <name>TEST</name>
    <salary>1234</salary>
  </person>
 </payment>

I need a Oracle PLSQL script to delete all the person tags if it contains TEST.

Final output would be :

<?xml version="1.0" encoding="UTF-8"?>
<payment>
  <person>
    <surname>Marco</surname>
    <name>Gralike</name>
    <salary>2345</salary>
  </person>
 </payment>
 <payment>
  <person>
    <surname>BertJan</surname>
    <name>Meinders</name>
    <salary>3456</salary>
    <salary>125</salary>
  </person>
 </payment>
 <payment>
  <person>
    <surname>Chris</surname>
    <name>Gralike</name>
    <salary>4567</salary>
  </person>
 </payment>

Thanks in advance.

Answer 1

Your provided XML doesn't have a root and can't be parsed by XML parser.

Assuming it does (say payments) as shown below:

create table t(txt clob);
insert into t values('<?xml version="1.0" encoding="UTF-8"?>
<payments>
    <payment>
        <person>
            <surname>Marco</surname>
            <name>Gralike</name>
            <salary>2345</salary>
        </person>
        <person>
            <surname>ABC</surname>
            <name>TEST</name>
            <salary>1234</salary>
            <person>
                <surname>Tiger</surname>
                <name>Scott</name>
                <salary>2222</salary>
            </person>
        </person>
    </payment>
    <payment>
        <person>
            <surname>BertJan</surname>
            <name>Meinders</name>
            <salary>3456</salary>
            <salary>125</salary>
        </person>
        <person>
            <surname>XYZ</surname>
            <name>TEST</name>
            <salary>1234</salary>
        </person>
    </payment>
    <payment>
        <person>
            <surname>Chris</surname>
            <name>Gralike</name>
            <salary>4567</salary>
        </person>
        <person>
            <surname>LMN</surname>
            <name>TEST</name>
            <salary>1234</salary>
        </person>
    </payment>
</payments>');

You can use this:

update t
set txt = to_clob(deletexml(
  xmltype(t.txt),
  '//payment/person[./name[text()="TEST"]]'
));

Produces:

<?xml version="1.0" encoding="UTF-8"?>
<payments>
    <payment>
        <person>
            <surname>Marco</surname>
            <name>Gralike</name>
            <salary>2345</salary>
        </person>
    </payment>
    <payment>
        <person>
            <surname>BertJan</surname>
            <name>Meinders</name>
            <salary>3456</salary>
            <salary>125</salary>
        </person>
    </payment>
    <payment>
        <person>
            <surname>Chris</surname>
            <name>Gralike</name>
            <salary>4567</salary>
        </person>
    </payment>
</payments>

EDIT:

If you want to delete a node that doesn't have a given child, use this:

update t
set txt = to_clob(deletexml(
  xmltype(t.txt),
  '//payment[not(./person)]'
));

It'll delete all the payment tags that don't have a person in it.