I basically spent the whole week working on implementing the Explore page with ReactJS allowing for browsing Artists only just as my mentor requested. By yesterday I had the work almost completed with tabular results showing up for the browse queries generated through the UI. Only thing that was missing was pagination of results and I am working on it now.
The work I did was pushed to Bitbucket and I plan on improving it and working on the fixes suggested by my mentor in the next couple of days.
This week I spent most of my time reading through Andrea Gazzarini's Apache Solr Essentials (tremendous instructive resource) and working through mbsssss and each core's schema.xml and solrconfig.xml files. After struggling to figure out how the current search server's org.musicbrainz.search.index package was going to work in Solr, I realized that most of that is just going to translate into defining field types and analysis chains in the Solr configuration.
I also discovered that some of the configuration will have to change. For example we can't use analyzers on StrField types, Solr only allows them on TextField types. Also, some of the analysis chains in the previous setup look like they will create duplicate documents in the Solr index. We might also consider changing some fieldtypes (currently all of our field types are string and text) to aid future improvements (faceting, "more like this", stats for numeric fields, etc.) but I'm focused on just matching the current functionality for now.
This week I hope to finish up with mbsssss and get started with implementing support for the last few search fields in sir: