Thanks a lot! I have tried the examples you provided.
It's good to know how it works.

Leszek

"Joe Fawcett" <joefawcett@n...> wrote in message 
news:%237xsTURBIHA.464@T......
> "TarTar" <spam.spam@l...> wrote in message 
> news:%238uJJxPBIHA.4836@T......
>> Hi Dimitre,
>>
>> Thanks a lot! Now the code works like a charm.
>>
>> BTW
>> Could you explain (or point to a resource) what's the differenece between 
>> //name[position() > 10] and (//name)[position() > 10] ? The former one 
>> (with brackets) works perfectly - but why?
>>
>> Leszek
>>
> The first retrieves name elements when they have a position() greater than 
> ten within their siblings.
> The second one retrieves all the name elements as one node-set and then 
> filters them, removing the first ten.
> An example:
>
> <data>
> <names>
> <name att="1" block="1"/>
> <name att="2" block="1"/>
> <name att="3" block="1"/>
> <name att="4" block="1"/>
> <name att="5" block="1"/>
> </names>
> <names>
> <name att="1" block="2"/>
> <name att="2" block="2"/>
> <name att="3" block="2"/>
> <name att="4" block="2"/>
> <name att="5" block="2"/>
> </names>
> <names>
> <name att="1" block="3"/>
> <name att="2" block="3"/>
> <name att="3" block="3"/>
> <name att="4" block="3"/>
> <name att="5" block="3"/>
> </names>
> </data>
>
> //name[position() > 3] gives:
> <name att="4" block="1"/>
> <name att="5" block="1"/>
> <name att="4" block="2"/>
> <name att="5" block="2"/>
> <name att="4" block="3"/>
> <name att="5" block="3"/>
>
> (//name)[position() > 3] gives:
> <name att="4" block="1"/>
> <name att="5" block="1"/>
> <name att="1" block="2"/>
> <name att="2" block="2"/>
> <name att="3" block="2"/>
> <name att="4" block="2"/>
> <name att="5" block="2"/>
> <name att="1" block="3"/>
> <name att="2" block="3"/>
> <name att="3" block="3"/>
> <name att="4" block="3"/>
> <name att="5" block="3"/>
>
> so if all the names are siblings anyway the two expressions give the same 
> result.
>
> -- 
>
> Joe Fawcett (MVP - XML)
>
> http://joe.fawcett.name
>
>