Tuesday 05 June 2012 10:03:05 am - 2 replies
I'm having an issue with eZ Find. When I run the following fetch:
fetch( ezfind, search, hash( 'query', 'page', 'subtree_array', $search_subtree_array, 'offset', $view_parameters.offset, 'limit', $page_limit, 'sort_by', hash( 'score', 'desc' ), 'facet', $defaultSearchFacets, 'filter', $filterParameters, 'publish_date', $dateFilter, 'spell_check', array( true() ) ))
The results only contain objects with the exact query: 'page', but not 'webpage' for example. How do I change eZ Find behaviour in a way that it returns objects with 'webpage' in the content as well? I.e. have it always use wildcard? What is the best practise? Use concat with *'s?
Thanks for your help!
Modified on Tuesday 05 June 2012 10:04:17 am by Sander van den Akker
Tuesday 05 June 2012 2:40:40 pm
Solr tokenizes strings using some predefined rules. Tokens are later used for searching. There is no way that solr will treat as the same word "webpage" and "page". Unless you proclaim that "web" is delimiter
If you don't have much of those pairs you can add them to similar tokens file.
Or use "*" but placing it in front is a performance killer.
Tuesday 05 June 2012 2:43:52 pm
If the configuration in ezfind.ini is the default one, then eZ Find will use the heuristic search handler to determine how to perform searches (with wildcard support, or without, for example). If the default search handler is indeed heuristic in your ezfind.ini, you are able to do wildcard searches as you see fit. If wildcard is specified, then wildcard search is performed, if not, only objects containing the exact word are returned.
You must be logged in to post messages in this topic!