https://github.com/no-defun-allowed/re-benchmarks
Some regular expression benchmarks
https://github.com/no-defun-allowed/re-benchmarks
Last synced: 3 months ago
JSON representation
Some regular expression benchmarks
- Host: GitHub
- URL: https://github.com/no-defun-allowed/re-benchmarks
- Owner: no-defun-allowed
- Created: 2022-01-07T11:04:28.000Z (over 3 years ago)
- Default Branch: master
- Last Pushed: 2022-01-07T11:10:38.000Z (over 3 years ago)
- Last Synced: 2025-01-31T05:47:20.479Z (5 months ago)
- Language: Rust
- Size: 14.6 KB
- Stars: 1
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Some regular expression benchmarks
Okay, just one benchmark with a few variations. We play with literal
prefixes and submatching to get different behaviours out of regex
engines.
We test the regex `([0-9]+)x([0-9]+)` with and without
submatches/captures, and also with and without literal double quotes
around them (i.e. `"([0-9]+)x([0-9]+)"`), which provides more of a
prefix to optimise for.
The following results were produced on a Ryzen 5 1600 processor, with
the `simd-loops` one-more-re-nightmare branch, SBCL at commit
`2e6c6fac3b8caf618c0e4263244f7c3bed7f85ce`, and Rust 1.55.0.
| | one-more-re-nightmare | rust-lang/regex | Hyperscan |
|------------|----------------------:|----------------:|-------------:|
| neither | 1540 Mchar/s | 603 Mchar/s | 362 Mchar/s |
| submatches | 1510 Mchar/s | 172 Mchar/s | |
| quotes | 7260 Mchar/s | 2550 Mchar/s | 1040 Mchar/s |
| both | 6030 Mchar/s | 236 Mchar/s | |