Closed stefandesu closed 1 year ago
I think we will use a database in the future for #29, #36, and #65 but these will better be another kind of database. Actually I just started with SQLite as simple cache to not call coli-ana twice with same notation. For more sophisticated queries we should better experiment with Neo4J and/or Solr to include full DDC hierarchy, DDC analysis and K10plus titles indexed with DDC. In any case this will first be an independent repository that uses coli-ana as backend (and/or the reverse) but not part of coli-ana.
Alright, sounds reasonable. 👍
@nichtich Could you check the commit where I removed the database-related stuff? The only things I left were a reference in the README (with a note that the database backend is not used anymore) as well as the error type - although I added a TODO there that we should add proper error types that are used throughout the application.
In particular, there was still a reference to GET /analyze?member=members
which I could not find in the code. Was this removed before, when we decided to remove the database backend?
@nichtich Can we remove anything related to database access, or do you think it'll be necessary to use a database again in the future?