Commit a4167e47 authored by Steinberg, Jan's avatar Steinberg, Jan
Browse files

minnor ppn changes

parent 60a170de

src/test/java/org/gesis/dda/transformer/impl/XsltMetadataTransformerTest.java

+43 −0
Original line number Diff line number Diff line
@@ -73,6 +73,49 @@ public class XsltMetadataTransformerTest { 
        assertEquals(18, transformedMetadata.size() );

    }

    

    @Test

    public void testAnotherPicaxmlTransformation() throws IOException {

        MetadataTransformer underTest = new XsltMetadataTransformer();

        URL url = Resources.getResource("xslt/normalized-picaxml-to-metadataxml.xml");

        String configuration= Resources.toString(url, Charsets.UTF_8);

        LOG.info("configuration={}", configuration);

        underTest.configure(configuration);

        

        // from file DSA_2017_07_Teil-2-huebbegt.xlsx, "Traditionelle Boote in Deutschland. T. 1, Die Ruhrfähre von Oefte"

        BundlesStreamSource source = new UnapiBundlesStreamSource("http://unapi.gbv.de", "picaxml", ImmutableList.of(), ImmutableList.of() );

        Set<Metadatum> originalMetadata = source.getAllMetadata("gvk:ppn:729077314");

        

        LOG.info("--------- BEGIN original metadata ------------");

        originalMetadata.stream().map(Object::toString).forEach(LOG::info);

        LOG.info("--------- END original metadata ------------");

        

        Set<Metadatum> transformedMetadata = underTest.transformMetadata(originalMetadata);

        LOG.info("--------- BEGIN transformed metadata ------------");

        transformedMetadata.stream().map(Object::toString).forEach(LOG::info);

        LOG.info("--------- END transformed metadata ------------");

        

        assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("dc.contributor.author", "Ancken, Rüdiger von") )  )   );

        // assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("dc.contributor.author", "Hanke, Fred") )  )   );

        // assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("dc.source.journal", "Deutsches Schiffahrtsarchiv") )  )   );

        assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("dc.language", "de") )  )   );

        assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("dc.title", "Traditionelle Boote in Deutschland , Teil 1: Die Ruhrfähre von Oefte") )  )   );

        // assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("dc.source.volume", "16") )  )   );

        // assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("dc.source.pageinfo", "211-228") )  )   );

        // assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("dc.rights.licence", "3") )  )   );

        // assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("dc.type.document", "32") )  )   );

        // assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("dc.type.stock", "article") )  )   );

        // assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("ssoar.wgl.collection", "true") )  )   );

        // assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("ssoar.contributor.institution", "DSM") )  )   );

        // assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("dc.date.issued", "1993") )  )   );

        // assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("dc.description.review", "3") )  )   );

        // assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("dc.description.pubstatus", "1") )  )   );

        // assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("dc.identifier.url", "http://nbn-resolving.de/urn:nbn:de:0168-ssoar-52582-3") )  )   );

        // assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("internal.dda.reference", "http://unapi.gbv.de@@gvk:ppn:882380753") )  )   );

        // assertTrue( transformedMetadata.stream().anyMatch(  m -> m.equals( new SimpleMetadatum("dc.identifier.issn", "0343-3668") )  )   );

        

        // assertEquals(18, transformedMetadata.size() );

    }

    

    @Test

    public void testTransformationWithAdditionToTitle() throws IOException {

        MetadataTransformer underTest = new XsltMetadataTransformer();

src/test/resources/xslt/normalized-picaxml-to-metadataxml.xml

+74 −115
Original line number Diff line number Diff line 
<?xml version="1.0" encoding="UTF-8"?>

<xsl:stylesheet version="2.0"

  xmlns:xsl="http://www.w3.org/1999/XSL/Transform"

  xmlns:pica="info:srw/schema/5/picaXML-v1.0"

  exclude-result-prefixes="pica"

  >

  

<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">

  <xsl:output indent="yes"/>

  <xsl:strip-space elements="*"/>

  <!-- <xsl:message terminate="no">here is the tree at the current node <xsl:copy-of select="."/>  </xsl:message> -->

  <!-- <xsl:message terminate="no">here is some message for stderr, e.g. <xsl:value-of select='text()' /> </xsl:message> -->

  <!-- 

       <xsl:for-each select="./*">

         <xsl:message terminate="no">a child=<xsl:value-of select="name()"/></xsl:message>

       </xsl:for-each>

     -->

  

  

  <xsl:template match="/">

  <xsl:template match="/record">

    <publication>

      <ssoar>

        <wgl>
@@ -27,11 +12,11 @@ 
        </contributor>

      </ssoar>

      <dc>

        <!--source>

        <source>

          <journal>

            Deutsches Schiffahrtsarchiv

          </journal>

        </source-->

        </source>

        <language>de</language>

        <rights>

          <licence>3</licence>
@@ -47,103 +32,77 @@ 
        <identifier>

          <issn>0343-3668</issn>

        </identifier>

        <xsl:for-each select="record/datafield">

            <xsl:if test="./tag = '31A'">

            <source>

            <volume>

                <xsl:value-of select="./subfield/content"/>

                </volume>

                </source>

            </xsl:if>

        </xsl:for-each>

         

        <xsl:apply-templates select="pica:record" />

      </dc>

      

    </publication>

  </xsl:template>

  <xsl:template match="pica:record">

    <!--xsl:apply-templates select="pica:datafield[@tag='021A']"/-->

    <xsl:apply-templates select="pica:datafield[@tag='031A']"/>

    <xsl:apply-templates select="pica:datafield[@tag='028A']"/>

    <xsl:apply-templates select="pica:datafield[@tag='027D']"/>

    <xsl:apply-templates select="pica:datafield[@tag='028C']"/>

    <xsl:apply-templates select="pica:datafield[@tag='009P']"/>

  </xsl:template>

      <xsl:variable name="preferredMainTitle" select="datafield[tag='039I']/subfield[code='t']/content" />

      <xsl:variable name="mainTitle">

        <xsl:choose>

          <xsl:when test="$preferredMainTitle != ''">

            <xsl:value-of select="$preferredMainTitle" />

          </xsl:when>

          <xsl:otherwise>

            <xsl:value-of select="datafield[tag='021A']/subfield[code='a']/content" />

          </xsl:otherwise>

        </xsl:choose>

      </xsl:variable>

    

  <!--xsl:template match="pica:datafield[@tag='021A']">

      <xsl:for-each select="datafield[tag='021A']">

        <dc>

          <title>

        

              <xsl:value-of select="$mainTitle"/><xsl:if test="subfield[code='d']">: <xsl:value-of select="subfield[code='d']/content"/></xsl:if>

          </title>

        </dc>

  </xsl:template-->

  

      </xsl:for-each>

      

  <xsl:template match="pica:datafield[@tag='031A']">

      <xsl:for-each select="datafield[tag='031A']">

        <dc>

          <source>

      <xsl:if test="pica:subfield[@code='d']/text()">

            <volume>

          <xsl:value-of select="pica:subfield[@code='d']/text()"/>

              <xsl:value-of select="subfield[code='d']/content"/>

            </volume>

      </xsl:if>

      <xsl:if test="pica:subfield[@code='h']/text()">

            <pageinfo>

          <xsl:value-of select="pica:subfield[@code='h']/text()"/>

              <xsl:value-of select="subfield[code='h']/content"/>

            </pageinfo>

      </xsl:if>

          </source>

    <xsl:if test="pica:subfield[@code='j']/text()">

          <date>

            <issued>

          <xsl:value-of select="pica:subfield[@code='j']/text()"/>

              <xsl:value-of select="subfield[code='j']/content"/>

            </issued>

          </date>

    </xsl:if>

    <xsl:choose>

      <xsl:when test="string-length(/pica:record/pica:datafield[@tag='039I']/pica:subfield[@code='t']/text()) > 0">

        <title>

          <xsl:value-of select="/pica:record/pica:datafield[@tag='039I']/pica:subfield[@code='t']/text()"/>

        </title>

      </xsl:when>

      <xsl:otherwise>

        <title>

          <xsl:value-of select="/pica:record/pica:datafield[@tag='021A']/pica:subfield[@code='a']/text()"/>

          <xsl:if test="/pica:record/pica:datafield[@tag='021A']/pica:subfield[@code='d']">

            <xsl:text>: </xsl:text>

            <xsl:value-of select="/pica:record/pica:datafield[@tag='021A']/pica:subfield[@code='d']/text()"/>

          </xsl:if>

        </title>

      </xsl:otherwise>

    </xsl:choose>

  </xsl:template>

        </dc>

      </xsl:for-each>

      

  <xsl:template match="pica:datafield[@tag='028A']">

      <xsl:for-each select="datafield[tag='028A']">

        <dc>

          <contributor>

            <author>

        <xsl:value-of select="pica:subfield[@code='a'][1]/text()"/>, <xsl:value-of select="pica:subfield[@code='d'][1]/text()"/>

              <xsl:value-of select="subfield[code='a']/content"/>, <xsl:value-of select="subfield[code='d']/content"/>

              <xsl:if test="subfield[code='d']/content"><xsl:text> </xsl:text><xsl:value-of select="subfield[code='c']/content"/></xsl:if>

            </author>

          </contributor>

  </xsl:template>

  <xsl:template match="pica:datafield[@tag='027D']">

    <source>

      <journal>

        <xsl:value-of select="pica:subfield[@code='a']/text()"/>

      </journal>

    </source>

  </xsl:template>

  <xsl:template match="pica:datafield[@tag='028C']">

        </dc>

      </xsl:for-each>

      

      <xsl:for-each select="datafield[tag='028C']">

        <dc>

          <contributor>

            <author>

        <xsl:value-of select="pica:subfield[@code='a'][1]/text()"/>, <xsl:value-of select="pica:subfield[@code='d'][1]/text()"/>

              <xsl:value-of select="subfield[code='a']/content"/>, <xsl:value-of select="subfield[code='d']/content"/>

            </author>

          </contributor>

  </xsl:template>

  <xsl:template match="pica:datafield[@tag='009P']">

        </dc>

      </xsl:for-each>

      

      <xsl:for-each select="datafield[tag='009P']">

        <dc>

          <identifier>

            <url>

        <xsl:value-of select="pica:subfield[@code='a']/text()"/>

              <xsl:value-of select="subfield[code='a']/content"/>

            </url>

          </identifier>

        </dc>

      </xsl:for-each>

    </publication>

  </xsl:template>



</xsl:stylesheet>
 
 No newline at end of file

src/test/resources/xslt/normalized-picaxml-to-metadataxml.xsl

0 → 100755
+103 −0
Original line number Diff line number Diff line 
<?xml version="1.0" encoding="UTF-8"?>

<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">

  <xsl:output indent="yes"/>

  <xsl:template match="/record">

    <publication>

      <ssoar>

        <wgl>

          <collection>true</collection>

        </wgl>

        <contributor>

          <institution>DSM</institution>

        </contributor>

      </ssoar>

      <dc>

        <source>

          <journal>

            Deutsches Schiffahrtsarchiv

          </journal>

        </source>

        <language>de</language>

        <rights>

          <licence>3</licence>

        </rights>

        <type>

          <document>32</document>

          <stock>article</stock>

        </type>

        <description>

          <review>3</review>

          <pubstatus>1</pubstatus>

        </description>

        <identifier>

          <issn>0343-3668</issn>

        </identifier>

      </dc>

     

      <xsl:for-each select="datafield[tag='021A']">

        <dc>

          <title>

            <xsl:choose>

              <xsl:when test="../datafield[tag='039I']/subfield[code='t']/content">

                <xsl:value-of select="../datafield[tag='039I']/subfield[code='t']/content"/>

              </xsl:when>

              <xsl:otherwise>

                <xsl:value-of select="subfield[code='a']/content" />

              </xsl:otherwise>

            </xsl:choose>

            <xsl:if test="subfield[code='d']"><xsl:text>: </xsl:text><xsl:value-of select="subfield[code='d']/content"/></xsl:if>

          </title>

        </dc>

      </xsl:for-each>

      

      <xsl:for-each select="datafield[tag='031A']">

        <dc>

          <source>

            <volume>

              <xsl:value-of select="subfield[code='d']/content"/>

            </volume>

            <pageinfo>

              <xsl:value-of select="subfield[code='h']/content"/>

            </pageinfo>

          </source>

          <date>

            <issued>

              <xsl:value-of select="subfield[code='j']/content"/>

            </issued>

          </date>

        </dc>

      </xsl:for-each>

      

      <xsl:for-each select="datafield[tag='028A']">

        <dc>

          <contributor>

            <author>

              <xsl:value-of select="subfield[code='a']/content"/>, <xsl:value-of select="subfield[code='d']/content"/>

            </author>

          </contributor>

        </dc>

      </xsl:for-each>

      

      <xsl:for-each select="datafield[tag='028C']">

        <dc>

          <contributor>

            <author>

              <xsl:value-of select="subfield[code='a']/content"/>, <xsl:value-of select="subfield[code='d']/content"/>

            </author>

          </contributor>

        </dc>

      </xsl:for-each>

      

      <xsl:for-each select="datafield[tag='009P']">

        <dc>

          <identifier>

            <url>

              <xsl:value-of select="subfield[code='a']/content"/>

            </url>

          </identifier>

        </dc>

      </xsl:for-each>

    </publication>

  </xsl:template>



</xsl:stylesheet>
 
 No newline at end of file