Mike Perrone
Mike Perrone
I don't think I will be able to make another iteration on this for several more weeks
I can't think of any _database_ types that can't be represented in recap. The only context I can imagine a type that Recap can't represent is [Unknown in the TypeScript...
https://github.com/recap-build/recap/pull/430 https://github.com/recap-build/recap-website/pull/16 Still needs some work but here's a start
I think building out a test suite like that will be very valuable. There are a lot of good ideas expressed in this. It was a little hard for me...
tagging @mroeschke as he implemented filesystem for [orc](https://github.com/pandas-dev/pandas/pull/51623) and [parquet](https://github.com/pandas-dev/pandas/pull/51609). @jorisvandenbossche as [author](https://github.com/pandas-dev/pandas/pull/41194) of initial implementation of parquet
I was also able to use this workaround for specifying array values. It is not very readable as is
@thdxr if you let me know what additional steps I need to take to make this change fully valid, I would be willing to do it to help out
@thdxr @jayair anything I can do to help this get merged?
FWIW, AWS Lambda will be supporting Node22 as a lambda runtime [in November](https://docs.aws.amazon.com/lambda/latest/dg/lambda-runtimes.html#runtimes-future), so this will be more important at that time.