GH-49114: [C++][Parquet] Fix converting schema failure with deep nested two-level encoding list structure

[HuaHuaY]

Rationale for this change

Fix the failure when converting parquet schema with deep nested two-level encoding list structure to arrow schema.

What changes are included in this PR?

Modified the ListToSchemaField and MapToSchemaField methods.

Are these changes tested?

Yes.

Are there any user-facing changes?

It allows some legal schemas that were once considered illegal.

• GitHub Issue: [C++][Parquet] Failed to read deep nested two-level encoding list structure #49114

[HuaHuaY]

@pitrou @emkornfield @wgtmac Please take a look.

[Copilot]

Pull request overview

Fixes Parquet-to-Arrow schema conversion for deep nested two-level encoding LIST structures (and related MAP-in-LIST cases) that previously errored as illegal.

Changes:

• Refactors list element resolution into a new ResolveList() helper to better handle nested two-level LIST encodings.
• Relaxes repetition constraints to allow LIST/MAP nodes to be REPEATED when they are elements of an enclosing LIST.
• Adds/updates unit tests to cover deep nested two-level LISTs and two-level LIST-of-MAP schemas, and adjusts expected error text.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
cpp/src/parquet/arrow/schema.cc	Adjusts LIST/MAP schema conversion logic and introduces ResolveList() for nested/two-level handling.
cpp/src/parquet/arrow/arrow_schema_test.cc	Adds coverage for deep nested two-level LIST and allows previously-failing LIST-of-MAP case; updates an error expectation.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated no new comments.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[wgtmac]

Mostly look good to me. Thanks for fixing this!

[wgtmac]

Do we need to validate that list_group has no logical type annotation?

[HuaHuaY]

I think this is not in backward-compatibility rules.

[wgtmac]

Do we want to add a case for the below which is still produced by parquet-java by default:

    // optional group my_list (LIST) {
    //   repeated group array (MAP) {
    //     repeated group key_value (MAP_KEY_VALUE) {
    //       required binary key (STRING);
    //       optional binary value (STRING);
    //     }
    //   }
    // }

[HuaHuaY]

I think it's not in backward-compatibility rules too. We may add it into document first.

[wgtmac]

Do you want to take a look? @pitrou @emkornfield @mapleFU

[conbench-apache-arrow]

After merging your PR, Conbench analyzed the 3 benchmarking runs that have been run so far on merge-commit 0cf32b2.

There were no benchmark performance regressions. 🎉

The full Conbench report has more details. It also includes information about 1 possible false positive for unstable benchmarks that are known to sometimes produce them.