[dexhunter]

TL;DR

• Reduce execute_field overhead during synchronous query execution.
• Skip empty argument processing and pass async error-handler state as parameters.
• Large-schema introspection benchmark median: 256.18 ms → 246.13 ms, 3.9% less time.
• PYTHONPATH=src python -m pytest tests -q: 3,613 passed, 18 skipped.

Refs #298

What & Why

Each field currently calls argument coercion, even when the field and query have no arguments.
An inner async function also captures local values, which adds work to synchronous calls.

Use an empty dictionary directly when both argument collections are empty.
Give the async function explicit parameters so ordinary field execution does not need those captured values.

Results

Situation	Before	After	Change
GitHub schema introspection	256.18 ms	246.13 ms	3.9% less time
Async list with 1,000 items	1.191 ms	1.201 ms	Within variation: 0.9% more time
Large query pickle decoding	6.938 ms	7.165 ms	Within variation: 3.3% more time

These are median times from at least 50 rounds on CPython 3.12.13, Linux x86-64, and an AMD EPYC 9R14.
Both versions ran sequentially on the same CPU.
The async-list and pickle-decoding slowdowns are smaller than the larger interquartile spread from their respective runs.
The async benchmark resolves one field and spends most of its time processing list items.
Pickle decoding does not call the changed method.

Validation

Run on the base commit and this change:

PYTHONPATH=src python -m pytest tests/benchmarks/test_introspection_from_schema.py tests/benchmarks/test_async_iterable.py tests/benchmarks/test_serialization.py::test_pickle_large_query_decode --benchmark-enable --benchmark-min-rounds=50
python -m ruff check src tests
python -m ruff format --check src tests
python -m mypy src tests

Ruff, formatting, and mypy passed for all 355 source and test files.

Why this is safe and easy to maintain

The patch changes one method and adds no helpers or dependencies.
Resolver arguments, completion, returned values, and error handling stay unchanged.
Existing execution tests cover resolver information, subclass hooks, middleware, argument defaults, and nullable errors.

Scope

This changes Python execution overhead without adding features or changing public signatures.
It keeps the existing GraphQL execution rules and resolver-information constructor.

AI assistance

Codex gpt-6.1-sol generated candidates tracked with Weco Observe.
Codex also helped run automated checks and prepare this description.

Weco share link: https://dashboard.weco.ai/share/NKlrkTVvI9HA-J73ZOdnjmMXw0g4xOYq

[codspeed]

Merging this PR will improve performance by 18.7%

1 warning

⚠️ 1 benchmark with a significant change ran in a different software environment

Its base and head results were measured with a different OS or toolchain, such as compiler, runtime or library versions, so the change might come from this difference rather than from the code. Open the benchmark to see what changed.

See affected benchmarks

⚡ 1 improved benchmark
✅ 19 untouched benchmarks

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
⚡	test_execute_introspection_query	2.3 s	1.9 s	+18.7%

Tip

Curious why performance improved? Comment @codspeedbot explain why performance improved on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing dexhunter:perf/field-arguments (fd1d4a9) with main (f9bdf7c)}

[Cito]

Thanks! Results match main for queries with and without arguments, including async errors, and with this and #297 applied, introspection execution takes about 13% less time. This will be part of 3.3.2.