not sure I follow, what would be the type I would use in e.g. Postgresql? I was thinking string so that I can do indexing etc, you mean converting it manually on writing/reading?

I was wondering if there was a way to automatically convert a given column , but I guess doing the conversion when reading/writing does also work

https://cljdoc.org/d/seancorfield/next.jdbc/1.2.659/doc/getting-started/tips-tricks#working-with-json-and-jsonb

that will convert to/from json for you. I'm sure there's a way to like make a custom type in postgres and add automatic conversion with edn but really that seems like a huge pain

I was thinking on primite type colums (a string in DB that is a keyword on clojure, or enum in pg maybe?), but I’ll also need that for json/edn interop, thanks

hmm enums like described here might be what I was looking for: https://www.bevuta.com/en/blog/using-postgresql-enums-in-clojure/

that jdbc/IResultSetReadColumn looks like it could potentially cause some issues

if any other string column gets one of the enum values it would automatically convert that, too

also that looks like it's using the old version of jdbc for clojure

specifically clojure.java.jdbc instead of the newer next.jdbc

I'd just accept that it's going to be a string, myself

yes you’re right, I think I’d change the implementation, though going for enums might be a reasonable approach (as the kw values I was thinking of are specced as sets of possible values in clojure)

the easier way seems to be to transform directly on clojure and have it as string

absolutely

I’ll investigate a bit more but it looks like the solution will come from one of those two approaches

thanks for the help!

👍:skin-tone-2: