[jabrailkhalil]

Rationale

Closes #8628.

Previously VariantArray could be collected only from Option<Variant> items:

let arr: VariantArray = values.iter().map(|v| Some(Variant::from(*v))).collect();

With this change, any type that converts cleanly into Variant can be collected directly:

let arr: VariantArray = vec![Some(1_i64), None, Some(42_i64)].into_iter().collect();

let arr: VariantArray = vec![Some("hello"), Some("a very long string exceeding the short string limit")].into_iter().collect();

This follows the suggestion made in #8625 (comment).

Changes

• parquet-variant-compute/src/variant_array.rs: generalize FromIterator<Option<Variant<'m, 'v>>> to impl<'m, 'v, V: Into<Variant<'m, 'v>>> FromIterator<Option<V>>, converting each item through Into::into. Existing code that collects Option<Variant> items keeps working (identity Into).
• Added regression tests: i64 items with nulls, &str items covering both short-string (ShortString) and long-string (String) storage.

Verification

• cargo test -p parquet-variant-compute --lib — 366 tests passed (364 existing + 2 new), including the new test_from_option_into_variants_into_variant_array and test_from_option_str_into_variant_array.
• cargo clippy -p parquet-variant-compute --all-targets -- -D warnings — clean.
• cargo fmt --all -- --check — clean.

Note: local verification used the x86_64-pc-windows-gnu target; upstream CI covers the full matrix.

AI disclosure

Per the Arrow project's AI-generated submissions guidance: this change was implemented with the assistance of an AI coding agent. The author reviewed the final diff, verified it compiles and passes tests/clippy/fmt locally, and takes full ownership of the change.

The ASF ICLA may need to be on file before merge (asf-clabot will report the status).

[sdf-jkl]

Thanks @jabrailkhalil. Everything looks great.

I'd add examples of building a VariantArray::from_iter.

This change also introduces a breaking API change with None handling. Earlier, None could only mean None::<Variant>, but now it can be any type coercible into Variant.

[None, Some(42_i64)] // Element type inferred as Option<i64>
[None, None]         // Element type unknown

We should add an example of building a correct Variant::Null array to the docs addition.

let null_rows = VariantArray::from_iter(vec![None::<Variant>; 3]);

[jabrailkhalil]

@sdf-jkl Addressed in 75e9e43. The VariantArray docs now show direct collection from Option, explain the distinction between None and Some(Variant::Null), and include the typed None:: all-null case you requested. cargo fmt --check passes locally. I also attempted the targeted crate test, but this Windows environment has no MSVC link.exe; CI will run the Rust test suite.

[sdf-jkl]

Thanks @jabrailkhalil this is great. This should also let us to simplify some of the VariantArray generations in tests where arrays are homogeneous. LMK if you want to add the simplifications here or in a separate PR.

[alamb]

Thanks @jabrailkhalil this is great. This should also let us to simplify some of the VariantArray generations in tests where arrays are homogeneous. LMK if you want to add the simplifications here or in a separate PR.

@jabrailkhalil will you have time to address @sdf-jkl 's comments? Or shall we merge this PR and do the simplications as a follow on?

[sdf-jkl]

[alamb]

🚢 🇮🇹

[jabrailkhalil]

Yes, I will continue this work. The documentation comments from the original PR were addressed before merge. I have now opened the small follow-up #11320 to simplify five homogeneous VariantArray doctest fixtures using the newly merged API. The assertions and resulting values are preserved; 369 crate unit tests and all 15 doctests pass, and formatting passes.