chewselene

In the end `DISTINCT` was unnecessary (query 5) since each array aggregation will occur in a sub-select statement. It was join an intermediate thing I tried since the first query...

Here's the plan for how to handle filters and `_x_count`: https://www.db-fiddle.com/f/4EXkttLVoWTdbMZQBsNDDB/1 The basic idea is to have `@fold` and `_x_count` be a part of a sub-`SELECT` statement that is joined...

Yeah, that's my bad in the explanation; this part > which will require that the query have a GROUP BY clause on all fields not used in a @filter. is...

Here's a new fiddle for how traversals inside a fold would work: https://www.db-fiddle.com/f/d5fdweFyvHQKCdVDq8BeWe/2 I ended up using array_remove. Not sure if it is import to not use that since we...

Good point, @obi1kenobi ! Updated version here: https://www.db-fiddle.com/f/d5fdweFyvHQKCdVDq8BeWe/3 Let me know what you think.

I don't think `HAVING` is what we want because we're not using an aggregate to filter on, but you are right that it was needlessly complex - see example 5...

Oh nice, but I think we should communicate that better on our home page https://graphql-compiler.readthedocs.io/en/latest/index.html. Right now it just shows the OrientDB stuff and isn't intuitive. Might be better to...