Zorba - The XQuery Processor

Hi,

We're working up to trying Zorba indexes. The data set that most
urgently needs indexing goes something like this:

<records>
   <record id="1">
      <tag>foo</tag>
      <tag>baa</tag>
     [...]
   </record>
   <record id="2">
       <tag>foo</tag>
       <tag>frob</tag>
   </record>
</records>

ie each record can contain an arbitrary number of tags. We're looking
for an efficient alternative to

/records/record[tag[text() = 'foo']]

without creating one index per tag value. Any ideas?

Regards,

Mark