fix(pluginutils): improve regex performance

[bluwy]

Rollup Plugin Name: pluginutils

This PR contains:

• bugfix
• feature
• refactor
• documentation
• other

Are tests included?

• yes (bugfixes and features will not be merged without tests)
• no

Breaking Changes?

• yes (breaking changes will not be merged unless absolutely necessary)
• no

If yes, then include "BREAKING CHANGES:" in the first commit message body, followed by a description of what is breaking.

List any relevant issue numbers:
n/a

Description

Refactors some regex to use non-capturing groups (as captured groups are a little slower), and use regexp.prototype.exec instead of string.prototype.match for perf.

For the former change, it's hard to use an online tool to measure as the difference is quite small, but here's a local file you can copy using tinybench that you can run:

File

import { Bench } from 'tinybench'

const b = new Bench({ iterations: 1000 })

const choices = ['FunctionDeclaration', 'ClassDeclaration', 'VariableDeclaration', 'Identifier']
const strs = Array.from({ length: 10 }).map(_ => choices[Math.floor(Math.random() * choices.length)])

b.add('non-capture', () => {
  strs.forEach(s => { /(?:Function|Class)Declaration/.test(s) })
})

b.add('capture', () => {
  strs.forEach(s => { /(Function|Class)Declaration/.test(s) })
})

await b.warmup();
await b.run();

console.table(b.table());

node18.20.3 results:

┌─────────┬───────────────┬─────────────┬────────────────────┬──────────┬─────────┐
│ (index) │   Task Name   │   ops/sec   │ Average Time (ns)  │  Margin  │ Samples │
├─────────┼───────────────┼─────────────┼────────────────────┼──────────┼─────────┤
│    0    │ 'non-capture' │ '5,088,404' │ 196.5252384844669  │ '±0.63%' │ 2544203 │
│    1    │   'capture'   │ '4,831,673' │ 206.96762775341406 │ '±0.67%' │ 2415837 │
└─────────┴───────────────┴─────────────┴────────────────────┴──────────┴─────────┘

For the latter change, I've benchmarked this at lukeed/polka#210 which you can run the perf.link test for the results too.

[benmccann]

I would expect here and in the other that perhaps it'd be better to move the regex declaration outside the method declaration so that it doesn't have to be recompiled each method invocation and each iteration of this loop

The same would be true of your other PR as well

[bluwy]

Something I've found over time is that it doesn't necessarily help with perf by hoisting it out. Engines these days are able to optimize it, and in most cases only yield very small gains. This was also discussed at lukeed/polka#210

[benmccann]

the discussion there suggests the one place it matters is in loops, so it might still be worth changing here, but not in the other locations?

[bluwy]

Yeah perhaps we can try that here, but I'd prefer to have a benchmark on that first before making the change though

[shellscape]

Going to merge this. I think this is a micro optimization to begin with, so not going to ask for additional optimizations on top of that. Please keep in mind that these tools typically run in environments with enough computing power that micro perf improvements barely have an impact.