https://github.com/genivia/ugrep-testing
ugrep pattern match testing and validation
https://github.com/genivia/ugrep-testing
Last synced: 12 months ago
JSON representation
ugrep pattern match testing and validation
- Host: GitHub
- URL: https://github.com/genivia/ugrep-testing
- Owner: Genivia
- License: bsd-3-clause
- Created: 2025-06-26T19:41:21.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2025-06-26T20:46:46.000Z (about 1 year ago)
- Last Synced: 2025-06-26T22:02:46.029Z (about 1 year ago)
- Language: Shell
- Homepage: https://ugrep.com
- Size: 35.1 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE.txt
Awesome Lists containing this project
README
# ugrep pattern match testing and validation
## Setup
Compile two small utilities `pick` and `trickle` to support the tests:
```console
$ ./setup.sh
```
## Run
Run the extensive barrage of tests, which can take hours to complete:
```console
$ ./run.sh
...
OK
```
Reports `OK` or halts when an error is detected, where `temp_words.txt` is the set of words seaarched as a pattern and `temp_results.txt` is the output with a problem.
The bulk of the tests are designed go through all possible pattern match methods and optimizations with randomized patterns. This is independent of the ugrep command line options. The SIMD optimizations with which ugrep was compiled are tested, when SIMD is enabled, which is one of SSE2, AVX2, AVX512BW, NEON, or AArch64.
## Data
- `words` extracted from `enwik8` with `ugrep -iwo '[a-z]+' enwik8 | sort -u > words`
- `enwik8` 100MB Wikipedia file
Note: we pick words of a specific byte length or within a byte length range to test with. Therefore, we test with ASCII words only to match byte lengths. This has no impact on validation of the internal byte-based pattern match methods that don't care what bytes represent, i.e. ASCII or UTF-8 or raw binary.