There are a few deployment issues involved with trying to use Solr for search. Moreover, it doesn't really seem necessary to use Solr in the first place -- the compendium isn't very large, and is unlikely to become large enough to require something as high-powered as Solr anytime soon. A third issue is that it's hard to perform unit testing with Solr.
~Most likely, the solution is that we should have some kind of purely Python-based search engine on the backend to perform search for us. With a small compendium it ought to be fast enough to avoid performance issues.~
The experiments on the static site show that in fact a purely JavaScript-based solution, such as using Fuse.js, is probably sufficient (and would certainly be more efficient for the server-side).
There are a few deployment issues involved with trying to use Solr for search. Moreover, it doesn't really seem necessary to use Solr in the first place -- the compendium isn't very large, and is unlikely to become large enough to require something as high-powered as Solr anytime soon. A third issue is that it's hard to perform unit testing with Solr.
~Most likely, the solution is that we should have some kind of purely Python-based search engine on the backend to perform search for us. With a small compendium it ought to be fast enough to avoid performance issues.~
The experiments on the static site show that in fact a purely JavaScript-based solution, such as using Fuse.js, is probably sufficient (and would certainly be more efficient for the server-side).