Hi Swan, Yes, I checked in solrconfig.xml and there was the definition of <luceneMatchVersion>9.4</luceneMatchVersion>
I tried to modify it to 9.4.2 but it had no impact, I got the same error that "Error instantiating stemmer for language Hungarianfrom class null". I ran out of ideas about what goes wrong. The jar file is there. The jar file includes org/tartarus/snowball/ext/HungarianStemmer.class Cheers, Roland Shawn Heisey <apa...@elyograg.org> ezt írta (időpont: 2023. júl. 3., H, 21:23): > On 7/3/23 08:28, Szűcs Roland wrote: > > What strange is that if I go to schema menu on admin UI and check the > query > > analyzer it says the following > > Tokenizer: > > org.apache.lucene.analysis.standard.StandardTokenizerFactory > > luceneMatchVersion: 9.4.0 > > name: standard > > > > Token Filters: > > org.apache.lucene.analysis.core.LowerCaseFilterFactory > > luceneMatchVersion: 9.4.0 > > name: lowercase > > org.apache.lucene.analysis.snowball.SnowballPorterFilterFactory > > luceneMatchVersion: 9.4.0 > > language: Hungarian > > name: snowballPorter > > > > Why admin shows 9.4.0 version if my jar has 9.4.2 version? Can it be a > > problem? > > What is being reported there is your luceneMatchVersion setting. This > is a configuration item in your solrconfig.xml file, it is not being > pulled from the jar. > > luceneMatchVersion is a configuration item that can in some cases > restore older analysis component behavior. There is no guarantee that > the setting will do anything at all, but occasionally some analysis > components will behave differently with an older match version. That > setting does NOT affect the file format of the Lucene index. > > Thanks, > Shawn >
