@Danny,

No descendants, good one..

@Mike,

Hmm, attribute existence *without* taking parent into account, something
like cts:attribute-query(...) and cts:attribute-values(). That could be
interesting though for backing XPaths like //@xlink:href. :-)

Cheers,
Geert

> -----Oorspronkelijk bericht-----
> Van: [email protected] [mailto:general-
> [email protected]] Namens Michael Blakeley
> Verzonden: donderdag 30 januari 2014 22:53
> Aan: MarkLogic Developer Discussion
> Onderwerp: Re: [MarkLogic Dev General] No cts:element-attribute-query?
>
> There is an element-exists termlist entry, so cts:element-
> query(xs:QName('p'), cts:and-query(())) is fast.
>
> Element-attribute values and element-attribute words are indexed by
> default. But attribute values and words are indexed as dependents on a
> parent element. I don't think there's an existence termlist entry for
> attributes, with or without a parent element. So
cts:element-attribute-value-
> query(xs:QName('p'), xs:QName('id'), 'fubar') is fast. But there's no
default
> index for a cts:element-attribute-query(xs:QName('p'), xs:QName('id'),
> cts:and-query(())) to use.
>
> In theory I think a cts:element-attribute-query term could use a range
index
> on that element-attribute if it existed, and simply ignore the values.
But
> there might be a technical obstacle, and in any case I haven't heard
much
> demand for it.
>
> -- Mike
>
> On 30 Jan 2014, at 13:36 , Geert J. <[email protected]> wrote:
>
> > Hi,
> >
> > http://stackoverflow.com/questions/21434099/checking-presence-of-
> attribute
> > -with-ctsquery
> >
> > Was looking at this stackoverflow question, and couldn't help
wondering
> > why isn't there a cts:element-attribute-query function. Because
attributes
> > are not indexed by default?
> >
> > Cheers
> > _______________________________________________
> > General mailing list
> > [email protected]
> > http://developer.marklogic.com/mailman/listinfo/general
> >
>
> _______________________________________________
> 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