kotman12 commented on PR #11955: URL: https://github.com/apache/lucene/pull/11955#issuecomment-1321176867
> I will take a look at the test locally, thanks. > > > There is only one thread/tokenstream for *Op. > > I think we are talking about different things. The mutable members within *Op are not my concern, although I could see how you would think that because at first glance that is what those synch methods may have been guarding. I am more worried about the lazy initialization of members at the **NLP** Factory level which appear to be singletons and are indirectly called from lucene's `FilterFactory::create`, which presumably happens on multiple threads at once (and even today is unsynchronized). Please c`rrect me if my thinking is incorrect but when I saw that I started to question the open-nlp library at a deeper level. > > I'll reiterate that I see no immediate problem with this change since I can't prove that these synchronized blocks were even doing anything. But I am questioning the thread safety of the integration as it currently stands, irrespective of this change. I might raise a separate issue to not confuse this thread anymore... what I am talking about is really only tangentially related to this current PR but in my mind it puts into question the thread safety of the integration as it stands now. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For additional commands, e-mail: issues-h...@lucene.apache.org
