Example SNAT and source code


Prev		Next

A small example will illustrate the use of the code generator. Let the goal again be the construction of a publications document, but this time using an RDF data source. The following two listings show a SNAT document and the source code generated from it.

Example 5. SNAT document, defining the transformation of RDF data into an XML document.

<z:snats xmlns:z="http://www.xsdplus.org/ns/structure">
  <z:prolog>
    <z:nsMap>
      <z:ns prefix="lib" uri="http://www.example.org/ns/resources/library#"/>
    </z:nsMap>
  </z:prolog>    
  <z:snat>
    <publications ctxt="$libs"
      <z:_attributes_>
        <updatedAt src="lib:lastUpdate" alt=""/>
      </z:_attributes_>
      <publication for-each="descendant::lib:book">
        <z:_attributes_>
          <publicationYear src="lib:py" alt=""/>
        </z:_attributes_>
        <isbn src="lib:isbn" dflt="'#MISSING'"/>
        <title src="lib:title" dflt=""/>
        <creator for-each="lib:author">
          <creatorRole src="'Author'" dflt=""/>
          <creatorName src="." dflt=""/>
        </creator>
      </publication>
    </publications>
  </z:snat>
</z:snats>

Example 6. XQuery code, generated from the SNAT document shown in the preceding listing.

declare variable $dataSources as element(rx:dataSources) external;
declare variable $sparqlLib :=                    
<sparqlLib>
  <sparql key="$libs" initialContext="$libs"/>
  <sparql key="descendant::lib:book">...</sparql>
  <sparql key="lib:author">...</sparql>
  <sparql key="lib:books/lib:lastUpdate">...</sparql>
  <sparql key="lib:isbn">...</sparql>  
  <sparql key="lib:py">...</sparql>  
  <sparql key="lib:title">...</sparql>  
<sparqlLib>;  

declare function f:updContext($dataSources as element(rx:dataSources),
                              $context as xs:string*, 
                              $queryKey as xs:string) as xs:string* {...};

let $c := f:updContext($dataSources, (), "$libs")
return   
<publications>{
    let $v := f:updContext($dataSources, $c, "lib:books/lib:lastUpdate")
    return
        if (empty($v)) then () else
        attribute <updatedAt {$v},
        
    for $c in f:updContext($dataSources, $c, "descendant::lib:book")
    return
        <publication>{
            let $v := f:updContext($dataSources, $c, "lib:py")
            return
               if (empty($v)) then () else
               attribute publicationYear {$v},
            <isbn>{
               let $v := f:updContext($dataSources, $c, "lib:isbn")               
               return
                  if (exists($v)) then $v else '#MISSING'
            }</isbn>,            
            <title>{f:updContext($dataSources, $c, "lib:title")}</title>,
            
            for $c in f:updContext($dataSources, $c, "lib:author")
            return
               <creator>{
                  <creatorRole>Author</creatorRole>,
                  <creatorName>{$c}</creatorName>
               }</creator>
        }</publication>    
}</publications>

The XQuery source code has the same structure as the code generated for XML data sources (compare example "XQuery code generated for XML data source". Note the difference - using an XML data source, navigation is accomplished by path expressions, whereas using an RDF data source, navigation is accomplished by submitting SPARQL queries referenced by query keys to a generic evaluator. The original RPath expressions are used as query keys, which enhances the readability of the code.

The input of the transformation is an XML document identifying the RDF graph and a set of resources bound to user-defined names. Example:

<rx:dataSources xmlns:rx="https://www.xsdplus.org/ns/...">
  <rx:namespace prefix="lib" uri="http://www.example.org/ns/..."/>
  <rx:graph name="main" uri="localhost:9093//...">
    <rx:resource uri="lib:rwth" name="libs"/>
    <rx:resource uri="lib:wage" name="libs"/>
    <rx:resource uri="lib:jota" name="libs"/>
  </rx:graph>
</rx:dataSources>

The metadata expressions can reference the resources via variable references (e.g. ctxt="$libs" and thus use them as starting points of navigation.