Hi Kelly,

It was indeed the solution to my problem.

Thanks,
Majid

On 28 June 2011 13:01, Kelly Stirman <[email protected]> wrote:

> If you build range indexes on foo and bar, the entries will contain values
> equivalent to the serialization of data(foo), and data(bar). If all you have
> is a child element string wrapped around the value, you may be fine. But if
> there are multiple elements, you can have results you don't like:
>
> <name>
>  <first>Jim</first>
>  <last>Smith</last>
> </name>
>
> data(/name)
> --> JimSmith
>
> I would try it on a sample and see if it yields what you want. Based on
> your example, I think it will.
>
> Kelly
>
>
> Message: 6
> Date: Tue, 28 Jun 2011 11:36:11 -0400
> From: Majid Valipour <[email protected]>
> Subject: Re: [MarkLogic Dev General] Facet on Field
> To: General MarkLogic Developer Discussion
>        <[email protected]>
> Message-ID: <[email protected]>
> Content-Type: text/plain; charset="utf-8"
>
> Hi Abhishek,
>
> Actually my problem is the reverse. I need two facets for two elements that
> have child elements with the same QName i.e., e.g., <foo/> and <bar/> that
> have the <string> child from my example.
> I cannot create range index on <foo/> and <bar/> because the text node is
> not their direct child but rather a children of the <string/> element. I
> suppose restructuring the content will fix this problem too but I was hoping
> to avoid doing that. We are able to create word constraint without a content
> reconstruction by defining two fields but as it turned out this solution
> does not work for facets.
>
> Regrads,
> Majid
> _______________________________________________
> General mailing list
> [email protected]
> http://developer.marklogic.com/mailman/listinfo/general
>
_______________________________________________
General mailing list
[email protected]
http://developer.marklogic.com/mailman/listinfo/general

Reply via email to