aio-libs / multidict / 36395529450

87%

push

github

Benchmark re-initializing a populated multidict (#1620)

<!-- Thank you for your contribution! -->

## What do these changes do?

Adds `reinit_items` (`d.__init__(items)`) and `reinit_clone`
(`d.__init__(other)` with `other` of the same kind, the
`md_clone_from_ht()` path) to `benchmarks/operations.py`, each at 200
and 20 items, plus four matching CodSpeed benchmarks. The existing
`ctor_*` rows always build a fresh object whose table is empty, so
replacing a live table (`_md_install_keys()`, #1619) was never measured;
#1619 needed a one-off script.

The target stays populated after each call, so the setup hands back the
same object and nothing is rebuilt inside the timed region. The rows are
multidict-only: `dict.__init__` on a live dict updates rather than
replaces. Multidict-only rows are not rendered into
`docs/benchmark.rst`, so the tables keep their shape and are not
regenerated; the exclusion paragraph gained one sentence saying why
re-initializing is left out. The CodSpeed tests carry no shard mark and
run in shard 1 next to the `create_*` benchmarks; recent runs show the
two shards within a few percent of each other on both builds.

## Are there changes in behavior for the user?

No.

## Is it a substantial burden for the maintainers to support this?

No; four rows in the registry and four CodSpeed tests following the
existing patterns.

## Related issue number

Follow-up to #1619.

## Checklist

- [x] I think the code is well written
- [x] Unit tests for the changes exist (`callgrind_driver.py
--self-check` asserts the re-initialized size)
- [x] Documentation reflects the changes
- [ ] If you provide code modification, please add yourself to
`CONTRIBUTORS.txt` N/A
- [x] Add a new news fragment into the `CHANGES/` folder
- [x] `make doc-spelling` passes and any new technical words are added
to `docs/spelling_wordlist.txt`

Drafted with Claude Code (Claude Opus 5.5); reviewed by @asvetlov.

<details>
<summary>Agent run details (optional, for re... (continued)

780 of 1560 branches covered (50.0%)

Branch coverage included in aggregate %.

20 of 24 new or added lines in 1 file covered. (83.33%)

8553 of 9146 relevant lines covered (93.52%)

1.87 hits per line