Thanks, Mike and Geert -- then the empty sequence is more efficient; Glad to know, since that's what I've been using in my code.

On Jun 1, 2009, at 3:59 PM, Michael Blakeley wrote:

It's just an extra constructor.

empty( cts:query( () ) )
=> true

However, that can take up valuable cycles:

for $i in 1 to 1000 * 1000
return cts:query( () )
, data( xdmp:query-meters()/*[1] )
=> PT0.825918S

for $i in 1 to 1000 * 1000
return ()
, data( xdmp:query-meters()/*[1] )
=> PT0.509286S

Still, it's highly unlikely to be the bottleneck in any real-world query.

-- Mike

On 2009-06-01 12:44, Shannon wrote:
I'm watching to see if ML has an opinion on your () vs. cts:query(()) question! As to the modularized approach, I realize your example didn't call for it, and you're probably already using that strategy in your production code, but I thought I'd mention it in case it would help somebody watching the discussion.

On Jun 1, 2009, at 3:35 PM, Tim Meagher wrote:

Hi Shannon,

Thanks, and yes, I like the modularized approach too!

Tim

________________________________
From: [email protected]<mailto:[email protected] > [mailto:[email protected]] On Behalf Of Shannon
Sent: Monday, June 01, 2009 3:32 PM
To: General Mark Logic Developer Discussion
Subject: Re: [MarkLogic Dev General] Building cts:search queries withconditional logic

Hi, Tim,

I've always had great results using simply the empty sequence. I'd be interested to know if there is any advantage to using cts:query(()).

Also, since this is not reflected in your example, just in case, I'd like to suggest modularizing it a bit, for far easier reading and maintenance--After initializing your variables with a series of a let assignments, create a combined cts:query by calling other functions, like so:

let $combined-query := cts:and-query((
            if ($foo) then search:qry1(...) else (),
            if ($bar) then search:qry2(...) else (),
...
))

The search:qry* functions would return a cts:query type -- basically breaking down each field or fieldset of your form, or some other logical grouping, into separate functions. Then call cts:search(input(), cts:combined-query)

Best,

On Jun 1, 2009, at 3:02 PM, Tim Meagher wrote:



Hi Folks,



Is there a preferred way to build a series of cts:search constructor functions when using conditional (and/or) logic? In the following example I insert a cts:query() if an value is blank, in the other I insert an empty sequence:



cts:search(

  //Record,

  cts:and-query ((

    if ($publisher ne "") then

      cts:element-value-query(xs:QName("Publisher"), $publisher,

("case-insensitive", "diacritic-insensitive", "punctuation- insensitive")

      )

    else cts:query(),

    if ($title ne "") then

      cts:element-value-query(xs:QName("Title"), $title,

("case-insensitive", "diacritic-insensitive", "punctuation- insensitive")

      )

    else ()

  ))

)



Is there a rule for using cts:query() or the empty sequence () for dynamically building cts:search constructors?



Thanks again,



Tim


_______________________________________________
General mailing list
[email protected]<mailto:[email protected] >
http://xqzone.com/mailman/listinfo/general


--
Shannon Scott Shiflett, XML Programmer
ROTUNDA, The University of Virginia Press
PO Box 801079, Charlottesville, VA 22904-4318 USA
Courier: 310 Old Ivy Way, Suite 302, Charlottesville VA 22903
Email: [email protected]<mailto:[email protected]> Tel: +1 434 924 4495
Web: http://rotunda.upress.virginia.edu/

_______________________________________________
General mailing list
[email protected]<mailto:[email protected] >
http://xqzone.com/mailman/listinfo/general

--
Shannon Scott Shiflett, XML Programmer
ROTUNDA, The University of Virginia Press
PO Box 801079, Charlottesville, VA 22904-4318 USA
Courier: 310 Old Ivy Way, Suite 302, Charlottesville VA 22903
Email: [email protected]<mailto:[email protected]> Tel: +1 434 924 4495
Web: http://rotunda.upress.virginia.edu/



_______________________________________________
General mailing list
[email protected]
http://xqzone.com/mailman/listinfo/general

--
Shannon Scott Shiflett, XML Programmer
ROTUNDA, The University of Virginia Press
PO Box 801079, Charlottesville, VA 22904-4318 USA
Courier: 310 Old Ivy Way, Suite 302, Charlottesville VA 22903
Email: [email protected]   Tel: +1 434 924 4495
Web: http://rotunda.upress.virginia.edu/

_______________________________________________
General mailing list
[email protected]
http://xqzone.com/mailman/listinfo/general

Reply via email to