main

Warming up the performance wizard...

Rust crate with an existing criterion benchmark (benches/parse_render.rs)

Analyze repository

Swapped criterion for codspeed-criterion-compat; added benchmarks for feature-rich input, analyze/render phases and multi-file merging

Integrate CodSpeed criterion compat layer

7 benchmarks measured in simulation mode and uploaded

Validate benchmarks locally

Added .github/workflows/codspeed.yml (CodSpeedHQ/action@v5, simulation, OIDC) and README badge

Add CodSpeed CI workflow and badge

PR #512: CodSpeed workflow passed and CodSpeed posted its installation comment; marked ready for review

Open PR and monitor CI

Benchmarks have been successfully run and processed. The results are available in Codspeed

Processing benchmark results

CodSpeed is set up for pymermaider in PR #512 (https://github.com/diceroll123/pymermaider/pull/512). The CodSpeed workflow passed on the PR and CodSpeed posted its installation comment, which found 7 benchmarks. The PR is ready for review and @diceroll123 is requested as reviewer.

What's in the PR:
- **Benchmark library:** the existing criterion benchmarks now run through `codspeed-criterion-compat`. Your benchmark code doesn't change, and `cargo bench` works locally as before.
- **New benchmarks:** `benches/parse_render.rs` keeps the two existing benchmarks and adds five scenarios:
  - an input that uses many Python features (enums, ABCs, dataclasses, protocols, generics, overloads, private members), run with private members both shown and hidden
  - parsing and analysis only
  - rendering only
  - merging 20 files into one diagram
- **CI:** a new `.github/workflows/codspeed.yml` runs the benchmarks with `CodSpeedHQ/action@v5` in CPU simulation mode on pushes to `main`, on pull requests and on manual runs. It logs in to CodSpeed through GitHub's built-in identity tokens, so you don't need to add a token secret.
- **README:** a CodSpeed badge is added.

Other checks on the PR (CI, Run Tests, Lint and Format, Test WASM Build) hadn't finished when I last looked, so check them before merging.

Next step: merge the PR so `main` gets a baseline. After that, every pull request gets a performance report.

Performance History

Latest Results

Latest Branches