cargo-bench2md

Formatting cargo bench output into .md table.

Installation

The process is pretty straightforward:

git clone [email protected]:niki4smirn/cargo-bench2md.git
cd cargo-bench2md
cargo build --release
cd target/release

Here you can find the cargo-bench2md binary.

Usage

The app reads JSON-formatted info about benchmarks from stdin and prints results into stdout.

Example:

cargo +nightly bench -- --format json -Z unstable-options | ./cargo-bench2md > table.md

(Don't forget to use +nightly and -Z unstable-options to run benchmarks and format output as JSON)

Output

Results are separated into different tables if corresponding benchmarks are defined in different modules.

Example:

Human-readable output:

test magic::benchmarks::string::ahash       ... bench:         893 ns/iter (+/- 29)
test magic::benchmarks::string::fastmurmur3 ... bench:      25,247 ns/iter (+/- 2,352)
test magic::benchmarks::u64::ahash          ... bench:         774 ns/iter (+/- 39)
test magic::benchmarks::u64::fastmurmur3    ... bench:       5,268 ns/iter (+/- 221)

.md output:

### magic::benchmarks::string
**Note:** All measurements are in ns/iter.
| Name | Median | Deviation |
|------|-------:|----------:|
| ahash | 898 | 76 |
| fastmurmur3 | 25036 | 819 |

### magic::benchmarks::u64
**Note:** All measurements are in ns/iter.
| Name | Median | Deviation |
|------|-------:|----------:|
| ahash | 768 | 81 |
| fastmurmur3 | 5178 | 582 |