Client --HTTP-> Landtable --SQL/HTTP/BSON-> your database
For this section, it is assumed that you understand basic concepts about a database.
A workspace is a collection of tables. It contains:
- database backend connection information
- schema migration strategies
- aliases
A table contains your data. It contains:
- the schema (the fields)
- additional information for database backends (for example, the corresponding database column for this table)
Database backends tell Landtable how to fetch your data. To write a new one, use Python package entry points. For a pyproject.toml example:
[project.entry-points."landtable.backends"]
postgres = "landtable.backends.postgres_backend:PostgresBackend"See landtable/backends/abstract.py for a guide on how to write your
own database backend.
An authentication plugin is something that tells Landtable who you are. Authentication plugins provide authentication contexts.
Examples of authentication plugins include:
- a JWT plugin that provides contexts based on its claims
- an OpenID Connect plugin (for Slack)
In Landtable, whether a user is allowed to do something is checked at multiple points inside the program. This includes:
- at the request handler
- at the database backend level, where data is actually handled
- at the configuration level
An authentication context is an object that answers the fundamental question of "is (subject) allowed to (action) on (resource)?".
While authentication contexts provide additional security, you are responsible for ensuring that your authentication plugins are configured correctly.
Authentication contexts provide another layer as part of a defense in depth strategy.
Database strategies manage database tables and Landtable tables on your behalf.
For example, a database strategy would tell Landtable how to delete a Postgres database column when you delete a field in a table.