Re: [Exist-open] Limit kwic summary to a single XML element

[Kemal P. <ke...@pa...>]

Thank you for your reply Ron, it certainly was of great help.

The code you supplied worked very well except in cases where a match was
found more than once within the same XML element (for instance if the
actor’s name was “Christian Bale Christian” :) ). In this case
$ancestor//exist:match has two elements which clashes with the cardinality
of a function called inside kwic:get-summary. As I see it I have two
options here, neither of which is optimal:

1. I can simply call kwic:get-summary with just one of the matches like
$ancestor//exist:match[1]
    This means that I will only highlight the one match.
2. I can use an nested for loop to include all of the matches like for
$match in $ancestor//exist:match
    However this will return a context for each of the matches and they
will often even overlap. This gives me a challenge for when I want to
display the context and highlight the matches, because ideally I only want
to show the text once with all the matches highlighted.

2012/1/7 Ron Van den Branden <ron...@ka...>

> Hi,
>
>
>  Is it possible to limit the summary returned by kwic:summarize to just
>> the XML element where the match was found.
>>
>
> For this, you'll need the more advanced kwic:get-summary() function (<
> http://demo.exist-db.org/**exist/functions/kwic/get-**summary<http://demo.exist-db.org/exist/functions/kwic/get-summary>
> >).
>
>
>  Consider this: <Film> <Title>The Dark Knight</Title>
>> <Director>Christopher Nolan</Director> <Star>Christian Bale</Star>
>> <Star>Heath Ledger</Star> <Star>Aaron Eckhart</Star> </Film>
>>
>> Now, if I run this query //Film[ft:query(Star, "Christian")] I would like
>> kwic to return <p> <span class="previous"></span> <span
>> class="hi">Christian</span> <span class="following"> Bale</span> </p> and
>> not give me a chunk of the Director or the next Star elements.
>>
>>
> You could try something like:
>
>  import module namespace kwic="http://exist-db.org/**xquery/kwic<http://exist-db.org/xquery/kwic>
> ";
>  let $nodes := //Film[ft:query(Star, 'Christian')]
>  let $matches := util:expand($nodes)//exist:**match
>  for $ancestor in $matches/ancestor::*[1]
>  return kwic:get-summary($ancestor, $ancestor//exist:match,<config
> width="40"/>, ())
>
> (based on the example at the top of the KWIC module function documentation
> at <http://demo.exist-db.org/**exist/functions/kwic/<http://demo.exist-db.org/exist/functions/kwic/>>,
> actually)
>
>
>  Please notice that this XML example is purely illustrative and my actual
>> database contains many different types of elements so I cannot really do
>> any element-specific (i.e. by a specific element-name) filtering. Also
>> notice that I am actually interested in the Film element, not Star (and
>> indeed may not even know by which sub-element Film was found).
>>
>>
> Hard to tell without knowing about your index configuration and maybe a
> short example, but in general, I think the approach in above query should
> work: first, find the matching nodes ($nodes), next find all matching text
> fragments ($matches), determine the context cutoff point $ancestor, and run
> kwic:get-summary().
>
> Hope this helps,
>
> Ron
>
> --
> Ron Van den Branden
> Wetenschappelijk attaché / Senior Researcher
> Reviews Editor LLC. The Journal of Digital Scholarship in the Humanities
>
> Centrum voor Teksteditie en Bronnenstudie - CTB (KANTL)
> Centre for Scholarly Editing and Document Studies
> Koninklijke Academie voor Nederlandse Taal- en Letterkunde
> Royal Academy of Dutch Language and Literature
> Koningstraat 18 / b-9000 Gent / Belgium
> tel: +32 9 265 93 51 / fax: +32 9 265 93 49
> E-mail : ron...@ka...
> http://www.kantl.be/ctb
>
>