[SPARK-57210][SQL] Skip statically-dead null-write branch in CreateNamedStruct codegen for non-nullable fields

[LuciferYang]

What changes were proposed in this pull request?

CreateNamedStruct.doGenCode emitted, for every value child, an if (isNull) values[i] = null; else values[i] = value; block. For a non-nullable child the isNull branch is statically dead (eval.isNull is the constant false), so this skips it and emits a single values[i] = value;. Nullable children are unchanged.

This mirrors GenArrayData.genCodeToCreateArrayData in the same file, which already gates the per-element null write on !expr.nullable (so CreateArray / CreateMap already do this) — CreateNamedStruct was the lone holdout.

Why are the changes needed?

It removes a dead branch and the unreachable values[i] = null write per non-nullable field, shrinking the generated method. The win scales with struct width (e.g. struct(*) over a non-null schema), which helps with the JVM 64KB method / constant-pool limits, Janino compile time, and JIT work. Part of SPARK-56908.

Does this PR introduce any user-facing change?

No.

How was this patch tested?

Existing ComplexTypeSuite (CreateStruct, CreateNamedStruct, SPARK-22693 codegen) and CodeGenerationSuite wide-struct tests already exercise both the non-nullable child (the new branch) and nullable children via checkEvaluation, which runs the interpreted and codegen paths. Behavior is unchanged — the skipped write targets a slot that new Object[] already zero-initializes to null — so no new test is added, matching the in-file array (GenArrayData) convention.

Was this patch authored or co-authored using generative AI tooling?

Generated-by: Claude Code (Claude Opus 4.8)

[dongjoon-hyun]

+1, LGTM. Thank you, @LuciferYang .

[gengliangwang]

Flagging against the scope guidance added to the umbrella SPARK-56908.

This is a source-only dead-branch elimination: Janino constant-folds if (true)/if (false), so there is no bytecode change. Measured on TPC-DS whole-stage codegen, even the most frequent such patterns (×358–445, ~0.4% of generated source) saved compile time below the run-to-run noise floor (~0.7%). The changes that were worth it removed more than source — e.g. SPARK-57198 dropped an unreachable references[]/constant-pool entry, which Janino cannot recover.

Before merging, please confirm this removes more than source text (a references[]/constant-pool entry, or keeps a method under the 64KB limit) — raw occurrence count alone is not sufficient. If not, it adds codegen-logic complexity for negligible benefit and should be dropped.

[LuciferYang]

Close this PR after checking against the preceding criteria. thank you @gengliangwang @dongjoon-hyun