Skip to content

Pg array improvement - #28

Merged
antialize merged 4 commits into
mainfrom
pg_array_improvement
Aug 4, 2026
Merged

antialize merged 4 commits into
mainfrom
pg_array_improvement

Conversation

@antialize

Copy link
Copy Markdown
Owner

No description provided.

Type row constructor lhs of IN element-wise against a subquery's columns
(e.g. UNNEST($1::t[], $2::t[], ...)) or another row constructor's elements,
instead of hitting a not-yet-implemented error and requiring the subquery to
yield exactly one column.
Thread the resolved lhs type of a comparison down as context for the rhs,
and use that context to constrain a bare, uncast $N argument passed to
ANY(...) to an array of the lhs's type, instead of leaving it untyped.
Add an ArgIn<Any> impl for Vec<&[u8]> (borrowed byte slices) alongside the
existing owned Vec<Vec<u8>> mapping, so a bytea[] arg bound with borrowed
slices doesn't fail with an unsatisfied trait bound at the query!() call
site. Add a qusql-type test confirming an INSERT ... SELECT * FROM UNNEST(...)
with a bytea[] array type-checks, and a live-DB integration test in
qusql-sqlx-type-test exercising the fixed Vec<&[u8]> binding.
@antialize
antialize force-pushed the pg_array_improvement branch from 17d05f9 to 4e8845a Compare August 4, 2026 12:50
@antialize
antialize merged commit a5b5756 into main Aug 4, 2026
6 of 7 checks passed
@antialize
antialize deleted the pg_array_improvement branch August 4, 2026 12:53
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant