https://github.com/bbrtj/perl-validator-benchmark

Benchmark of Perl validation frameworks. Example results -> https://bbrtj.eu/blog/article/validation-frameworks-benchmark
https://github.com/bbrtj/perl-validator-benchmark

benchmark perl perl5 validation

Last synced: 15 days ago
JSON representation

Benchmark of Perl validation frameworks. Example results -> https://bbrtj.eu/blog/article/validation-frameworks-benchmark

• Host: GitHub
• URL: https://github.com/bbrtj/perl-validator-benchmark
• Owner: bbrtj
• Created: 2022-01-28T19:23:09.000Z (almost 3 years ago)
• Default Branch: master
• Last Pushed: 2023-08-24T16:06:29.000Z (over 1 year ago)
• Last Synced: 2025-01-07T01:58:21.021Z (20 days ago)
• Topics: benchmark, perl, perl5, validation
• Language: Perl
• Homepage:
• Size: 222 KB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Benchmarks for form validators

These benchmarks measure:

- how fast input hashes matching the specification can be validated at runtime

These benchmarks don't (currently?) measure:

- validation failure speed

- compilation time overhead

Object construction is part of the benchmark, but can be skipped if certain validator implementation does not require it.

## Example results

(https://bbrtj.eu/blog/article/validation-frameworks-benchmark)

## Running the benchmark

First, you have to install Carton for your perl and use it to install dependencies locally:

```

cpan Carton

carton install

```

The benchmark is run one case at a time:

```

carton exec ./benchmark.pl 

```

Results can be optionally filtered through `format.pl` script, which reduces horizontal size of the output:

```

carton exec ./benchmark.pl  | ./format.pl

```

## Cases

Here is the list of current benchmark cases with rationales behind them:

### single_field

We pass a single field `a`, which is required and can be any scalar value.

*Rationale*: we measure the runtime overhead each system has. The validation rule here is as easy as it can be on purpose. Results of this benchmark can be used to decide whether a validator should be used for simple data validation cases that will run very frequently.

### multiple_fields

We pass five fields `a`, `b`, `c`, `d`, `e`, all of which are required and strings.

*Rationale*: we can compare results of the previous single_field validator and see how having multiple rules affects validator's performance. A big drop in validation speed can indicate a poorly optimized system.

### array_of_objects

We pass a single field `a`, which is an array reference of hash references. There are 100 nested hashes in total. Each hash contains keys `b` - a number, and `c` - a string.

*Rationale*: We measure how efficiently validators can crawl a nested structure and whether they can deal with large amount of data. The amount of data can be configured by hand in `BenchmarkSetup` to see whether the performance decreases linearly or exponentially. This data can be used to decide whether a validator is fit for larger volume of data.

## Contributing

Contributions are welcome! If you wish to contribute a new validator, follow these steps:

1. Add your validator module into `cpanfile`. Run `carton install` to update `cpanfile.snapshot`

2. Add a new benchmark runner in `lib/Utils.pm`. Name your benchmark `Bench`, `` being module name without `::` separators

3. Choose which benchmark case you would like to implement. Add your benchmark name into `BenchmarkSetup.pm` file in that directory (in `sub participants`)

4. Create a file in that directory named `Bench.pm`. Implement validation code in that file

5. Ensure your validator works, then create a pull request