Creating an Index

Index services have been configured in your system. These can be services for TREX indexes that were contained in the delivery, or index services implemented for other search engines.

You start off in the Index Administration iView.

1. Choose Create .

   The New Index screen appears.

   Field	Description
   ID (obligatory)	Unique ID
   Name	Meaningful name This name is displayed in the search options (choose DisplayIndexes ).
   Group	An index group includes several indexes. This name is displayed in the search options. For more information, see Index Groups .
   Crawler Parameters	The crawler parameters that you select are valid by default for all data sources for this index. However, you can override this setting for individual data sources. Crawler parameters are defined in the configuration of Content Management. For more information, see Crawlers and Crawler Parameters .
   Service (obligatory)	Select an index service. By default, you have the choice between a pure search index, a pure classification index, or a combined search and classification index.
   Items to Index	Specifies which items are to be indexed. Only documents (items that are not folders) Only folders All items You cannot change this setting after saving an index. You index folders so that they can be included in searches. Folders whose properties match the search term can be displayed in the results list. Folders are not classified in taxonomies. The system indexes folders by their properties, not their content. However, you can enter a content specification for a folder in the Description property.
   Custom Properties	See table below.

   Property Name	Permitted Values	Comments
   fuzziness	Number between 0 and 1	Specifies the degree of fuzziness for the fuzzy search. The smaller the value, the more fuzzy the search. The system creates this property with a default value when you create a TREX search index or search and classification index.
   IndexInternalLinks	true or false	If you set this property to true , the system also indexes internal links. In this case, the links properties are indexed, but not the link target.
   searchhidden	true or false	You can use this additional property to avoid the indexed documents being displayed in the search results list. You require this additional property when configuring an index for sponsored links. For more information, see Configuration of the Sponsored Links Function .
   indexContentOfExternalLink	true or false	The target objects of external links are not indexed by default. If you set this property to true , the system indexes not only the link properties for external links but also the content of the target object. This means that users can find these external links when they search for terms contained in the content of the target objects. Since indexes without assigned data sources are not taken into account by default during searches, you should add the showWithoutDatasource custom property with the value true .
   showWithoutDatasource	true or false	In the standard search, the system searches only in indexes to which at least one data source is assigned. The advanced search displays only indexes to which at least one data source is assigned. If you set this value to true , the index is searched in the standard search and displayed in the advanced search even if no data source is assigned to it.
   QueryExtender	Path to the Java class	You can use this property to link the user's search query with another search query that you have defined in a Java class. The user's search query and the search query in the class are linked using AND. Your Java class must be derived from the following class: com.sapportals.wcm.service. indexmanagement.retrieval. search.DefaultQueryExtender You can find an example in Extending the Search Query . If you want to execute a search of a large number of search indexes and a query extender is configured for at least one search index, the search query can become very complex and extend the duration of the search. To switch off the use of query extenders in a search iView, you have to enhance the specification in the query parameter in the iView properties of the search iView in question. Add the specification &disQExtender=true . This means that no query extenders are supported for future searches using this search iView. However, all other search iViews still use the query extenders.
   queueFlushTime	See: Start Condition	The start condition determines when the queue server triggers indexing of the documents collected in the queue.
   langPropId	ID for a property with type String that is predefined in the property metadata service and that can take only one value (single-valued). Enter the property in the following format. <namespacealias>: <propertyId>	Specify a document property that allows authors to specify the language of a document. The allowed values are defined in the index management service ( Supported Languages parameter). Enter them into the document property configuration in the Allowed Values parameter. During indexing, the system checks whether or not the specified document property contains one of the allowed values. If the answer is yes, the system indexes the document in the corresponding language index. If the answer is no, TREX runs an automatic language identification check on the document before indexing it in the corresponding language index. Example: langPropId=mynamespace:myLanguage
   contentUrlProp	ID for a property with type String that is predefined in the property metadata service and that can take only one value (single-valued). Enter the property in the following format. <namespacealias>: <propertyId>	You can specify a URL in a document property from which TREX is to retrieve the content to be indexed. For all documents for which this property is set, TREX gets the content to be indexed from the URL that is specified in the property value.
   indexSearchGroup	Any character string	If you carry out a search across several TREX indexes, the system calls all TREX indexes with a single search query by default. Especially if you use query extenders, this can result in extremely complex search queries that can negatively affect TREX performance. To improve performance, you can specify different values in the indexSearchGroup custom property. The system starts a separate TREX search query in a parallel thread for each group. This can significantly reduce the complexity of search queries.

2. Choose CreateIndex .

The system creates the index. You can now assign data sources to the index.