datasette-graphql VS pluralize

Impressive how little code is involved here! This is really neat.

The biggest feature I can see that's missing is pagination - it looks like this doesn't have a way to retrieve e.g. ten results, then pass a next token to get back the next set.

Here's how I implemented pagination in my similar datasette-graphql plugin (which also gives you a GraphQL API for an existing SQLite database): https://github.com/simonw/datasette-graphql#pagination

My source code is here, but it likely won't be much use to you - it's full of weird introspection because I'm trying to make this work against any database table: https://github.com/simonw/datasette-graphql/blob/1.4/datasette_graphql/utils.py

> The primary key column is named id or thing_id or thingId, where thing is the singular form of the table name.

Is it naming relations in plural a common thing in practice?

I thought best-practice was to name relations either singular (as each tuple represents one entry) or uninflected (still singular for most words), specially when you're not a fluent speaker of the language being used to name the relations of the database.

Plurals are often irregular for commonly used words, and the fact that this requires a external dependency ( https://github.com/plurals/pluralize ) to cover for some "common plurals" is telling that supporting this feature is a complex thing indeed - that would not be required in the first place with singular everywhere.