Creating Indices at the VO Level

From Gcube Wiki
Revision as of 20:52, 30 September 2011 by Vassilis.verroios (Talk | contribs) (→‎Creating a Full Text and a Forward Index for a collection with two views)

Jump to: navigation, search

Indexing Procedure

The Indexing procedure refers to the creation of indices for the collections imported in a Virtual Organization. It consists of three steps:

  • Creation of the Rowset XSLT generic resources, that transform collection data into data that can be fed to an Index.
  • Creation of the Index type generic resources, that define the Index configuration.
  • Definition of an IRBootstrapper job that will perform the steps required to create the Indices.

In the first two steps we create generic resources for the Rowset XSLTs and Index Types through the Resource Management portlet . You can find detailed descriptions for the Rowset data (the output of the Rowset XSLT transformation) in the following sections:

You can find detailed descriptions for the Index Type definition here:

For the third step, a definition of an IRBootstrapper job is required. You can find the details for defining such a job in the IR Bootstrapper section. The two examples that follow will clarify the three steps.

Creating a Full Text and a Forward Index for a collection with two views

In this example we will create a Full Text and a Forward Index for collection that has one view for the data in ES schema and one view for the data in DC schema. First of all let's create a generic resource for the Full Text Rowset XSLT for the collection data in ES schema: 

<Resource version="0.4.x">
 
   <ID>5d10db30-f372-11dd-8807-8f9f04747f1a</ID>
 
   <Type>GenericResource</Type>
 
   <Scopes>
 
      <Scope>/d4science.research-infrastructures.eu/Ecosystem</Scope>
 
   </Scopes>
 
   <Profile>
 
      <SecondaryType>MetadataBrokerXSLT</SecondaryType>
 
      <Name>BrokerXSLT_es_anylanguage_to_ftRowset_anylanguage</Name>
 
      <Description>XSLT to transform from es schema to full text index rowset schema</Description>
 
      <Body>
 
         <xsl:stylesheet xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="1.0">
 
            <xsl:output xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" indent="yes" method="xml" omit-xml-declaration="yes" />
 
            <xsl:template xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" match="/">
 
               <ROWSET xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
 
                  <xsl:apply-templates xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="//esObject" />
 
               </ROWSET>
 
            </xsl:template>
 
            <xsl:template xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" match="//esObject">
 
               <ROW xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
 
                  <FIELD xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" name="title">
 
                     <xsl:value-of xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="title" />
 
                  </FIELD>
 
                  <FIELD xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" name="description">
 
                     <xsl:value-of xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="description" />
 
                  </FIELD>
 
                  <FIELD xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" name="purpose">
 
                     <xsl:value-of xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="purpose" />
 
                  </FIELD>
 
                  <xsl:for-each xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="provenance/creator">
 
                     <FIELD xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" name="creator">
 
                        <xsl:value-of xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="." />
 
                     </FIELD>
 
                  </xsl:for-each>
 
                  <xsl:for-each xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="descKeys/keyword">
 
                     <FIELD xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" name="keyword">
 
                        <xsl:value-of xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="." />
 
                     </FIELD>
 
                  </xsl:for-each>
 
                  <FIELD xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" name="date">
 
                     <xsl:value-of xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="timeFrame/end" />
 
                  </FIELD>
 
                  <FIELD xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" name="suppInfo">
 
                     <xsl:value-of xmlns:dc="http://dublincore.org/documents/dcmi-terms/" xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="suppInfo" />
 
                  </FIELD>
 
               </ROW>
 
            </xsl:template>
 
         </xsl:stylesheet>
 
      </Body>
 
   </Profile>
 
</Resource>

Note that this XSLT extracts a number of fields from the collection payload. In the same fashion we create the Full Text Rowset XSLT for the DC schema: 

<Resource version="0.4.x">
 
   <ID>5b31baf0-f372-11dd-8807-8f9f04747f1a</ID>
 
   <Type>GenericResource</Type>
 
   <Scopes>
 
      <Scope>/d4science.research-infrastructures.eu/Ecosystem</Scope>
 
   </Scopes>
 
   <Profile>
 
      <SecondaryType>MetadataBrokerXSLT</SecondaryType>
 
      <Name>BrokerXSLT_dc_anylanguage_to_ftRowset_anylanguage</Name>
 
      <Description>XSLT to transform from DC schema to FTS rowset schema (suitable for feeding the full text index).</Description>
 
      <Body>
 
         <xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="1.0">
 
            <xsl:output xmlns:xsl="http://www.w3.org/1999/XSL/Transform" indent="yes" method="xml" omit-xml-declaration="yes" />
 
            <xsl:template xmlns:xsl="http://www.w3.org/1999/XSL/Transform" match="/">
 
               <ROWSET xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
 
                  <ROW xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
 
                     <FIELD xmlns:xsl="http://www.w3.org/1999/XSL/Transform" name="description.abstract">
 
                        <xsl:value-of xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="//*[local-name()='description.abstract']" />
 
                     </FIELD>
 
                     <xsl:for-each xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="//*[local-name()='subject']">
 
                        <xsl:if xmlns:xsl="http://www.w3.org/1999/XSL/Transform" test="normalize-space(.)">
 
                           <FIELD xmlns:xsl="http://www.w3.org/1999/XSL/Transform" name="subject">
 
                              <xsl:value-of xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="normalize-space(.)" />
 
                           </FIELD>
 
                        </xsl:if>
 
                     </xsl:for-each>
 
                     <FIELD xmlns:xsl="http://www.w3.org/1999/XSL/Transform" name="type">
 
                        <xsl:value-of xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="//*[local-name()='type']" />
 
                     </FIELD>
 
                     <FIELD xmlns:xsl="http://www.w3.org/1999/XSL/Transform" name="identifier">
 
                        <xsl:value-of xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="//*[local-name()='identifier']" />
 
                     </FIELD>
 
                     <FIELD xmlns:xsl="http://www.w3.org/1999/XSL/Transform" name="language">
 
                        <xsl:value-of xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="//*[local-name()='language']" />
 
                     </FIELD>
 
                     <FIELD xmlns:xsl="http://www.w3.org/1999/XSL/Transform" name="publisher">
 
                        <xsl:value-of xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="//*[local-name()='publisher']" />
 
                     </FIELD>
 
                     <FIELD xmlns:xsl="http://www.w3.org/1999/XSL/Transform" name="contributor">
 
                        <xsl:value-of xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="//*[local-name()='contributor']" />
 
                     </FIELD>
 
                     <FIELD xmlns:xsl="http://www.w3.org/1999/XSL/Transform" name="source">
 
                        <xsl:value-of xmlns:xsl="http://www.w3.org/1999/XSL/Transform" select="//*[local-name()='source']" />
 
                     </FIELD>
 
                  </ROW>
 
               </ROWSET>
 
            </xsl:template>
 
         </xsl:stylesheet>
 
      </Body>
 
   </Profile>
 
</Resource>

Additionally we must create the corresponding Full Text Index Type for the fields extracted from the ES and DC payload: 

<Resource version="0.4.x">
 
   <ID>06752440-bdb7-11e0-8e2a-d27a0ad33f7c</ID>
 
   <Type>GenericResource</Type>
 
   <Scopes>
 
      <Scope>/d4science.research-infrastructures.eu/Ecosystem</Scope>
 
   </Scopes>
 
   <Profile>
 
      <SecondaryType>FullTextIndexType</SecondaryType>
 
      <Name>IndexType_ft_es_dc_2.0</Name>
 
      <Description>Definition of the fulltext index type for the 'es' + 'dc' schema</Description>
 
      <Body>
 
         <index-type name="default">
 
            <field-list sort-xnear-stop-word-threshold="2E8">
 
               <field name="title">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="description">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="purpose">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="creator">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="keyword">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="date">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="suppInfo">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="description.abstract">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="subject">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="type">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="identifier">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="language">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="publisher">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="contributor">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="source">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="gDocCollectionID">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
               <field name="gDocCollectionLang">
 
                  <index>yes</index>
 
                  <store>yes</store>
 
                  <return>yes</return>
 
                  <tokenize>yes</tokenize>
 
                  <sort>no</sort>
 
                  <boost>1.0</boost>
 
               </field>
 
            </field-list>
 
         </index-type>
 
      </Body>
 
   </Profile>
 
</Resource>

Note that we also need to define the default "gDocCollectionID" and "gDocCollectionLang" fields, although there are not referenced in the XSLTs(the are automatically added by the DTS ) 

<job jobtype="IndexCollection" name="IndexEsDcCollection">
 
                  <initialization>
 
                     <assign to="%IndexCollection.input.ColName" value="European Environment Agency: Topic Reports" />
 
                     <assign to="%Create_MC_ft_index.FullTextIndexGenerationTask.IndexTypeID" value="ft_es_dc_2.0" />
 
                     <assign to="%Create_MC_ft_index.FullTextIndexGenerationTask.TransformationXSLTID" value="$BrokerXSLT_wrapperFT" />
 
                     <assign to="%Create_MC_ft_index.FullTextIndexGenerationTask.Views" value="[ 5ec68209-41f1-4c1c-9359-ff055a38d981, 55404922-12ee-4808-9cc7-ac1ab38c4a2e ]" />
 
                     <assign to="%Create_MC_ft_index.FullTextIndexGenerationTask.ViewsXSLTSIDs" value="[ $BrokerXSLT_es_anylanguage_to_ftRowset_anylanguage, $BrokerXSLT_dc_anylanguage_to_ftRowset_anylanguage ]" />
 
                     <assign to="%Create_MC_ft_index.FullTextIndexGenerationTask.IdOfIndexManagerToAppend" userInputLabel="ID of FT index management resource to append" value="%userInput" />
 
                     <assign to="%Create_MC_fwd_index.ForwardIndexGenerationTask.TransformationXSLTID" value="$BrokerXSLT_wrapperFWD" />
 
                     <assign to="%Create_MC_fwd_index.ForwardIndexGenerationTask.IndexedKeyNames" value="[ ObjectID, gDocCollectionID, gDocCollectionLang, title, creator ]" />
 
                     <assign to="%Create_MC_fwd_index.ForwardIndexGenerationTask.IndexedKeyTypes" value="[ fwd_string_string, fwd_string_string, fwd_string_string, fwd_string_string, fwd_string_string ]" />
 
                     <assign to="%Create_MC_fwd_index.ForwardIndexGenerationTask.Views" value="[ 5ec68209-41f1-4c1c-9359-ff055a38d981 ]" />
 
                     <assign to="%Create_MC_fwd_index.ForwardIndexGenerationTask.ViewsXSLTSIDs" value="[ $BrokerXSLT_es_anylanguage_to_fwRowset_anylanguage_title_creator ]" />
 
                     <assign to="%Create_MC_fwd_index.ForwardIndexGenerationTask.IdOfIndexManagerToAppend" userInputLabel="ID of FWD index management resource to append" value="%userInput" />
 
                  </initialization>
 
</job>

For the jobtype: 

<jobtype description="Creates the required FT-FWD indices for a collection." name="IndexCollection">
 
                  <input type="GCUBECollection" />
 
                  <jobDefinition>
 
                     <sequential>
 
                        <sequential>
 
                           <assign to="%Create_MC_ft_index.input" value="%IndexCollection.input" />
 
                           <assign to="%Create_MC_ft_index.output.IndexedCollectionID" value="%Create_MC_ft_index.input.ColID" />
 
                           <assign to="%Create_MC_ft_index.FullTextIndexGenerationTask.NumberOfLookups" value="2" />
 
                           <task name="Create_MC_ft_index" tasktype="FullTextIndexGenerationTask" />
 
                        </sequential>
 
                        <sequential>
 
                           <assign to="%Create_MC_fwd_index.input" value="%IndexCollection.input" />
 
                           <assign to="%Create_MC_fwd_index.output.IndexedCollectionID" value="%Create_MC_fwd_index.input.ColID" />
 
                           <assign to="%Create_MC_fwd_index.ForwardIndexGenerationTask.NumberOfLookups" value="2" />
 
                           <task name="Create_MC_fwd_index" tasktype="ForwardIndexGenerationTask" />
 
                        </sequential>
 
                     </sequential>
 
                  </jobDefinition>
 
</jobtype>
Retrieved from "https://wiki.gcube-system.org/index.php?title=Creating_Indices_at_the_VO_Level&oldid=12619"