GH-53615: [C++][Parquet] Add ALP to file-level fuzz corpus

[wgtmac]

Rationale for this change

Follow up on #48345 by including ALP in the generated Parquet file-level fuzz corpus.

What changes are included in this PR?

Add Encoding::ALP to kFloatEncodings.

Are these changes tested?

N/A.

Are there any user-facing changes?

No. This only changes fuzzing seed generation, not production reader or writer behavior.

Was AI used for this PR?

AI prepared and reviewed the one-line change and the PR description.

PR code and description written by:

• Human
• AI

Reviewed before submission by:

• Human
• AI
• Not reviewed

Closes #53615.

• GitHub Issue: [C++][Parquet] Expand ALP coverage in fuzzing seed corpuses #53615

[github-actions]

⚠️ GitHub issue #53615 has been automatically assigned in GitHub to PR creator.

[Copilot]

🟢 Approval recommended

The focused change correctly integrates ALP into existing FLOAT and DOUBLE corpus generation.

0 open findings

What changed in this PR

Adds ALP-encoded FLOAT and DOUBLE samples to the Parquet file-level fuzz corpus.

Changes:

• Adds Encoding::ALP to supported floating-point corpus encodings.
• Reuses existing datasets and null-probability variants.

File	Description
cpp/​src/​parquet/​arrow/​generate_fuzz_corpus.cc	Enables ALP for generated floating-point columns.

🧠 Review effort: Balanced

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[wgtmac]

I didn't change the data samples since I don't have a good answer yet. AI has suggested me these cases for both FLOAT and DOUBLE:

Case	Suggested inputs	Why
Constants	0.0, 1.25, -1.25	Exercise zero-bit-width packing with zero, positive and negative frame-of-reference values.
Decimal-like data	Sequences at several decimal scales, with mixed signs	Exercise ALP’s typical workload and exponent/factor selection.
Vector boundaries	1, 1023, 1024, 1025 non-null values	Cover short inputs, a full default vector and a partial final vector.
Sparse exceptions	Decimal-like values with a few NaNs, infinities and -0.0, including exceptions at vector starts/ends	Exercise exception placeholders, positions and patching alongside normally encoded values.
All exceptions	Only NaNs, infinities and -0.0	Exercise vectors with no normally encoded values and no packed-value bytes.
Null/page combinations	Reuse the existing null patterns in both DataPage V1 and V2 files	Exercise non-null value counting and integration with definition levels and page layouts.

However, please correct me if I was wrong, these well-crafted cases will be altered in later fuzz runs at random so I don't think we need all the cases above. Do you have any suggestion? @pitrou

[pitrou]

However, please correct me if I was wrong, these well-crafted cases will be altered in later fuzz runs at random so I don't think we need all the cases above.

Are these cases covered by Parquet data files in parquet-testing?

Random mutation doesn't automatically produce all interesting cases, because of... randomness, so it's better to have a comprehensive seed corpus, but of course it's more tedious to develop and we might want to cut corners away.

[wgtmac]

Yes, most of them are already covered by alp_extended.zstd.parquet: constants, different decimal scales, sparse and all-exception vectors, NaN payloads, infinities, negative zero, and partial vectors with nulls, for both FLOAT and DOUBLE.

[alamb]

Makes sense to me -- thank you @wgtmac

[pitrou]

@github-actions crossbow submit fuzz

[pitrou]

@wgtmac Do you also attend to tackle the encoding-level Parquet fuzzer?

[github-actions]

Revision: 46f01a3

Submitted crossbow builds: ursacomputing/crossbow @ actions-57a3445bd3

Task	Status
test-build-cpp-fuzz

[wgtmac]

@wgtmac Do you also attend to tackle the encoding-level Parquet fuzzer?

The encoding-level corpus generator already includes ALP through SupportedEncodings():

arrow/cpp/src/parquet/decoder.cc

Lines 2759 to 2762 in 46f01a3

	case Type::FLOAT:
	case Type::DOUBLE:
	return ConstVectorRef<Type::FLOAT>(
	{Encoding::PLAIN, Encoding::BYTE_STREAM_SPLIT, Encoding::ALP});

.

[pitrou]

The encoding-level corpus generator already includes ALP through SupportedEncodings():

Oh, well, you're right (and I wrote this code 🙃).

[pitrou]

Thanks @wgtmac !