====== Croatica et Tyrolensia: XQuery scripts ======
[[https://commons.wikimedia.org/wiki/File%3ACostumes_de_Differents_Pays%2C_'Femme_Morlaque_du_Canton_de_Segna'_LACMA_M.83.190.48.jpg |{{ http://upload.wikimedia.org/wikipedia/commons/thumb/b/b3/Costumes_de_Differents_Pays%2C_%27Femme_Morlaque_du_Canton_de_Segna%27_LACMA_M.83.190.48.jpg/256px-Costumes_de_Differents_Pays%2C_%27Femme_Morlaque_du_Canton_de_Segna%27_LACMA_M.83.190.48.jpg}}]] 
===== Adding a function to parse wordforms externally =====
What we need:
  - a working [[http://basex.org/|BaseX]] installation; see [[z:croat-tyrol-basex-deploy|The easiest way to deploy BaseX]]
  - permissions and ways to edit a ''.xqm'' XQuery module file (such as the ''restxq.xqm'' supplied with BaseX)
  - address of a parsing service (e. g. [[http://services.perseids.org/bsp/morphologyservice/analysis/word?word=homines&lang=lat&engine=morpheuslat]] -- check this out to see the kinds of results that will be provided)



First we edit the ''form'' part, where users submit sentences (or words) to be parsed:
<code html>
<h3>Parse words from Latin sentence</h3>
 <p>Enter a Latin sentence which you want to parse. 
    The result will be an XML file with lots of grammatical
    information for each word.</p>
      <form method="post" action="homon">
        <p>Sententia:<br />
        <input name="message" size="50"></input>
        <input type="submit" /></p>
      </form>
</code>
Then we prepare the function (e. g. by adding this to the end of the ''restxq.xqm'' module file). //Caveat//: the service address here is not real, it is a placeholder!
<code lisp>
(:~
 : A comment, to help us know what we're doing:
 : this function retrieves lexical information 
 : from an external service 
 : for words in a form request.
 : @param  $message  message to be included in the response
 : @return response element 
 :)
declare
  %rest:path("/homon")
  %rest:POST
  %rest:form-param("query","{$query}", "(no query)")
  function page:homon-postman(
    $message as xs:string)
    { 
let $url := ("http://some.parsing.service.org/somemorphologyservice/someword?word=SOMEWORD&amp;lang=latamp;engine=morpheuslat")
(: convert string into sequence :)
let $rijeci := tokenize($query, "\W+")
(: for each item in sequence :)
return element w { 
for $r in $rijeci
let $parsed := (doc(replace($url,'SOMEWORD',$r)))
return element ana { $parsed } 
                  }
    };
</code>