Replace csv files with a SQL database

[jack89roberts]

See https://github.com/alan-turing-institute/WimbledonPlanner/blob/master/notes/Creating_SQL_Database.md

[jack89roberts]

Needs to be implemented in Azure app.

• Create database on Azure.

• Copy forecast/harvest data model initially (at least relevant parts)

  • But convert hacked tags like GitHub issue numbers, REG association types etc. to proper tables/columns.

• Create SQLAlchemy representation of model.

• Implement upserts - i.e. update if already present, else append.

• Schedule database to regularly update from harvest/forecast APIs

• Think about snapshots, proposed but not final changes etc.

  • This sounds a lot like needing git for databases, e.g. branches, pull requests etc. Maybe some things out there, e.g. https://flywaydb.org/

• Users, authentication etc.

[jack89roberts]

Database now fully integrated. Still plenty that can be done r.e. data model, scheduled updates and snapshots, but closing this issue.