Cargo-criterion tool support

.gitignore

@@ -10,5 +10,6 @@
 /coverage
 /dist
 /.idea
+.release
 target/
 jmh-result.json

README.md

@@ -16,7 +16,7 @@ and monitor the results on GitHub Actions workflow.
 
 This action currently supports the following tools:
 
-- [`cargo bench`][cargo-bench] for Rust projects
+- [`cargo bench`][cargo-bench] and [`cargo criterion`][cargo-criterion] for Rust projects
 - `go test -bench` for Go projects
 - [benchmark.js][benchmarkjs] for JavaScript/TypeScript projects
 - [pytest-benchmark][] for Python projects with [pytest][]
@@ -83,6 +83,10 @@ context) properties. Like this:
 ]
 ```
 
+Furthermore, option `nativeBenchmarkDataDirPath` will also store your tool's native reports under github pages, i.e:
+
+https://umccr.github.io/htsget-rs/dev/bench/native/criterion/reports/
+
 ## Screenshots
 
 ### Charts on GitHub Pages
@@ -639,6 +643,7 @@ Every release will appear on your GitHub notifications page.
 [java-badge]: https://github.com/benchmark-action/github-action-benchmark/workflows/JMH%20Example/badge.svg
 [github-action]: https://github.com/features/actions
 [cargo-bench]: https://doc.rust-lang.org/cargo/commands/cargo-bench.html
+[cargo-criterion]: https://crates.io/crates/cargo-criterion
 [benchmarkjs]: https://benchmarkjs.com/
 [gh-pages]: https://pages.github.com/
 [examples-page]: https://benchmark-action.github.io/github-action-benchmark/dev/bench/

examples/criterion-rs/README.md

@@ -21,10 +21,22 @@ e.g.
 
 Note that you should run the benchmarks using the bencher output format.
 
+If one wishes to use JSON as output format, you can install `cargo-criterion` crate with the accompanying `tool:` YAML definition in the action:
+
+```yaml
+- name: Install cargo-criterion
+  uses: baptiste0928/cargo-install@v1
+  with:
+    crate: cargo-criterion
+- name: Run benchmarks
+  run: cargo criterion 1> output.json
+```
+
+Further details are described in [the cargo-criterion Rust example](../rust/README.md).
 
 ## Process benchmark results
 
-Store the benchmark results with step using the action. Please set `cargo` to `tool` input.
+Store the benchmark results with step using the action. Please set `cargo` to `tool` input:
 
 ```yaml
 - name: Store benchmark result
@@ -34,4 +46,14 @@ Store the benchmark results with step using the action. Please set `cargo` to `t
       output-file-path: output.txt
 ```
 
+Or `cargo-criterion` if JSON output is preferred instead.
+
+```yaml
+- name: Store benchmark result
+  uses: benchmark-action/github-action-benchmark@v1
+  with:
+      tool: 'cargo-criterion'
+      output-file-path: output.json
+```
+
 Please read ['How to use' section](https://github.com/benchmark-action/github-action-benchmark#how-to-use) for common usage.

examples/rust/README.md

@@ -1,5 +1,4 @@
-Rust example for benchmarking with `cargo bench`
-================================================
+# Rust example for benchmarking with `cargo bench`
 
 - [Workflow for this example](../../.github/workflows/rust.yml)
 - [Action log of this example](https://github.com/benchmark-action/github-action-benchmark/actions?query=workflow%3A%22Rust+Example%22)
@@ -46,4 +45,48 @@ Store the benchmark results with step using the action. Please set `cargo` to `t
 
 Please read ['How to use' section](https://github.com/benchmark-action/github-action-benchmark#how-to-use) for common usage.
 
+# Rust example for benchmarking with `cargo-criterion`
 
+In the previous section, both regular and criterion-rs can be used through the regular `cargo bench` facility, but there's an additional crate and cargo extension named [`cargo-criterion`][cargo-criterion].
+
+ The improvements in [cargo-criterion][cargo-criterion] do [match the goals of github-action-benchmark](https://crates.io/crates/cargo-criterion#goals), so it makes sense to include support for it.
+
+
+## Run benchmarks
+
+Official documentation for usage of `cargo criterion`:
+
+https://bheisler.github.io/criterion.rs/book/cargo_criterion/cargo_criterion.html
+
+.e.g: 
+
+```yaml
+- name: Run benchmarks 
+  run: cargo criterion 1> output.json
+```
+
+If you have a group of benchmarks, cargo criterion will output a [ndJSON][ndjson].
+
+## Process benchmarks results
+
+There are two notable differences in cargo-criterion:
+
+  1. Since the output is machine-readable JSON, the extract process only parses the result file and maps the required data into github-action-benchmark plotting system. In fact, [cargo-criterion][cargo-criterion] only supports JSON as `message-format` (output format).
+  2. cargo-criterion incorporates [its own HTML benchmark reports system][criterion-rs-own-html], which can be stored alongside if desired through the `native-benchmark-data-dir-path`.
+
+```yaml
+- name: Store benchmark result
+  uses: benchmark-action/github-action-benchmark@v1
+  with:
+    tool: 'cargo-criterion'
+    output-file-path: output.json
+    native-benchmark-data-dir-path: target/criterion
+```
+
+The native benchmark reports is simply copied from `target/criterion/reports` and pushed to the github results repo so that they are available under:
+
+https://YOUR_ORG.github.io/YOUR_REPO/dev/bench/native/criterion/reports/
+
+[cargo-criterion]: https://crates.io/crates/cargo-criterion
+[criterion-rs-own-html]: https://bheisler.github.io/criterion.rs/book/user_guide/plots_and_graphs.html
+[ndjson]: http://ndjson.org/