Parquet Variant Extract Benchmark

[abigalekim]

Description

(DRAFT)

Checklist

• I am familiar with the Contributing Guidelines.
• New or existing tests cover these changes.
• The documentation is up to date with these changes.

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[coderabbitai]

📝 Walkthrough

Summary by CodeRabbit

• New Features
  • Added a benchmark for extracting fields from experimental Parquet VARIANT data.
  • Supports testing primitive and array values across configurable row counts, nesting depths, data types, and match rates.
  • Enables performance comparisons between direct VARIANT conversion and path-based field extraction.

Walkthrough

Changes

VARIANT benchmark

Layer / File(s)	Summary
VARIANT value encoding cpp/benchmarks/io/parquet/experimental/variant/extract.cpp	Adds encoders for metadata, primitive and array leaves, and nested object values.
Column and extraction path construction cpp/benchmarks/io/parquet/experimental/variant/extract.cpp	Builds STRUCT VARIANT columns and maps benchmark parameters to extraction paths and target types.
Benchmark execution and registration cpp/benchmarks/io/parquet/experimental/variant/extract.cpp, cpp/benchmarks/CMakeLists.txt	Runs casting or path extraction across configurable rows, types, nesting depths, and hit rates. Registers the NVBench target.

Estimated code review effort: 3 (Moderate) | ~25 minutes

Possibly related PRs

• rapidsai/cudf#23036: Covers related cast_variant and extract_variant_field APIs and VARIANT encoding paths.
• rapidsai/cudf#23547: Optimizes metadata key lookup used by extract_variant_field.
• rapidsai/cudf#23614: Adds related Parquet VARIANT extraction and casting coverage.

Suggested labels: 2 - In Progress

Suggested reviewers: bdice, simoneves, vuule

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 inconclusive)

Check name	Status	Explanation	Resolution
Description check	❓ Inconclusive	The description contains only a draft marker and checklist, so it does not meaningfully describe the benchmark changes.	Add a concise summary of the Parquet Variant extraction benchmark and its scope.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly identifies the main change: a Parquet Variant extraction benchmark.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@cpp/benchmarks/io/parquet/experimental/variant/extract.cpp`:
- Around line 19-22: Add the direct declaration headers <cstdint>,
<cuda_runtime_api.h>, and <tuple> to the includes in extract.cpp for its
fixed-width integer types, cudaStreamSynchronize, and std::ignore usage.

🪄 Autofix

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: c1a8c46b-67f1-4a01-bbe5-91ef1b75588e

📥 Commits

Reviewing files that changed from the base of the PR and between 98832f4 and 1531d85.

📒 Files selected for processing (2)

• cpp/benchmarks/CMakeLists.txt
• cpp/benchmarks/io/parquet/experimental/variant/extract.cpp

[coderabbitai]

📐 Maintainability & Code Quality | 🟡 Minor | ⚡ Quick win

🧩 Analysis chain

🏁 Script executed:

#!/bin/bash
set -euo pipefail

file="cpp/benchmarks/io/parquet/experimental/variant/extract.cpp"
for header in cuda_runtime_api.h cstdint tuple; do
  rg -qF "`#include` <${header}>" "$file" ||
    printf 'Missing direct include: <%s>\n' "$header"
done

Repository: rapidsai/cudf

Length of output: 262

---

🏁 Script executed:

#!/bin/bash
set -euo pipefail

file="cpp/benchmarks/io/parquet/experimental/variant/extract.cpp"
printf '%s\n' '--- include block and symbol context ---'
sed -n '1,45p' "$file"
sed -n '205,240p' "$file"

printf '%s\n' '--- direct declarations and usages ---'
rg -n 'cudaStreamSynchronize|std::ignore|int(8|16|32|64)_t|uint(8|16|32|64)_t|size_t|ptrdiff_t|make_tuple|tie|tuple' "$file"

Repository: rapidsai/cudf

Length of output: 5883

---

🏁 Script executed:

#!/bin/bash
set -euo pipefail

printf '%s\n' '--- project usage of the required direct headers ---'
rg -n -g '*.{cpp,cu,h,hpp,cuh}' '`#include` <cuda_runtime_api\.h>|cudaStreamSynchronize|`#include` <cstdint>|`#include` <tuple>' cpp/benchmarks/io/parquet/experimental cpp/benchmarks | head -200

Repository: rapidsai/cudf

Length of output: 841

---

🏁 Script executed:

#!/bin/bash
set -euo pipefail

printf '%s\n' '--- CUDA runtime include conventions ---'
rg -n -g '*.{cpp,cu,h,hpp,cuh}' '`#include` <cuda_runtime(_api)?\.h>|cudaStreamSynchronize' cpp | head -200

Repository: rapidsai/cudf

Length of output: 3365

---

Add direct declaration headers.

This file uses fixed-width integer types, cudaStreamSynchronize, and std::ignore without their declaration headers. Add <cstdint>, <cuda_runtime_api.h>, and <tuple>.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@cpp/benchmarks/io/parquet/experimental/variant/extract.cpp` around lines 19 -
22, Add the direct declaration headers <cstdint>, <cuda_runtime_api.h>, and
<tuple> to the includes in extract.cpp for its fixed-width integer types,
cudaStreamSynchronize, and std::ignore usage.

Source: Coding guidelines