[LUCENE-6775] Improve MorfologikFilterFactory to allow arbitrary dictionaries from ResourceLoader

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 5.4, 6.0
• Component/s: modules/analysis
• Lucene Fields: New
• Assignee: Uwe Schindler
• Reporter: Uwe Schindler
• Created: 02/Sep/15 21:02
• Updated: 28/Aug/22 14:42
• Resolved: 09/Sep/15 21:53

Description

Followup issue for LUCENE-6774: The filter ctor already allows to pass any dictionary to the filter, but you have no chance to configure this through the Factory (CustomAnalyzer/Solr/Elasticsearch/...). This will add 2 parameters to the factory (exclusive with the dictionary string specifying language, default "pl"), to load FSA (dictionary) and corresponding property file (metadata/featureData). This dictionary could be placed, e.g. in Solr's conf dir and loaded, because this would be done via ResourceLoader.
Alternatively the language could still be passed, but must be part of JAR file distribution. Currently this defaults to "pl" at the moment and plain Lucene does not allow more, unless you add own JAR files. So practically, the parameter is useless for a pure, uncustomized Lucene-Impl.

Issue Links

• Is contained by: LUCENE-6833 Upgrade morfologik to version 2.0.1, simplify MorfologikFilter's dictionary lookup
• is related to: LUCENE-6774 Remove solr hack in MorfologikFilter