#APACHE LUCENE TUTORIAL JAVA PLUS#

I am using the most up-to-date release - which at the time of writing is 8.3.0 (released November 2019). Now we know how it ends, let’s look at the rest of the story. Spoiler alert: I ended up with a small stand-alone application which uses typeahead to search all 6 million titles in my database: I’ve never used Lucene’s API before - so I wanted to give it a try. Term vectors - (still not entrely sure when it’s best to use these).Which brings us to the need for text searching. That may be necessary - but it’s probably not sufficient. And then fetch the next chunk of data only if needed.

#APACHE LUCENE TUTORIAL JAVA PLUS#

One solution already hinted at would be to introduce server-side paging: Fetch as much data as a user can see at one time, plus perhaps a little more. But it’s not going to be practical in the real world, except for modestly sized data sets. This is OK, given the purposes of my web app - to explore various technologies such as Javalin and Thymeleaf. But when that table is created, all 5,000 titles are sent from the server. Within my web app, the 5,000 titles it does contain are all displayed in a single HTML table (albeit with client-side paging, courtesy of DataTables). Similarly, my IMDb data set contains over 9.6 million people (actors, directors, producers, etc.), and almost 36 million records to describe which people appear in which titles. But my demo only handles a modest 5,000 of these, as highlighted above. The IMDb data set I am using contains over 6 million title records (movies, TV episodes, etc).

My demo web application, which I describe here (with sources here), displays a core set of IMDb data:īut the application has a fatal flaw (maybe it has several, but I’m going to focus on one):

Apache lucene tutorial java