# Regex Builder & Parser

> Construct ReDoS-safe regular expressions and string parser logic for complex input string extraction.

- Skill: `rmazrim/regex-builder-parser` (Agent Skill)
- Install (CLI): `npx skillmds@latest add rmazrim/regex-builder-parser`
- Raw SKILL.md: https://api.skillmd.com/api/skills/rmazrim/regex-builder-parser/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: RMAzrim (https://skillmd.com/u/rmazrim)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/rmazrim/regex-builder-parser

---


# Regex Builder & Parser

## Prerequisites & Dependencies
- Node.js 18+ or Python 3.10+ with regex engine knowledge
- `npm i redoS` (for safety scoring) or Python's `re` module + `rstr` for generation
- Comfortable with character classes, quantifiers, grouping, and anchors

## Execution Steps
1. Clearly define the input format and extraction goal: what substrings, delimiters, or patterns must be captured
2. Sketch the regex on paper first: anchors (`^`, `$`), delimiters, optional groups, alternation (`|`)
3. Build the regex incrementally, testing each component against sample inputs before compositing
4. Use atomic groups `(?>...)` or possessive quantifiers `++`/`*`+` (if supported) to prevent backtracking exploits
5. Score the regex for ReDoS risk: `npm i redoS` → `checkRegex(regex)` should report low catastrophic backtracking risk
6. Favor simpler alternatives: string methods (`split`, `match`), `String.prototype.replace`, or parser combinators if the pattern exceeds ~20 characters or has nested quantifiers
7. Document the final regex with inline comments `/** @type {RegExp} */` and a brief explanation of each section

```javascript
// Safe regex: extract filename without extension from a path
// Breakdown: ^ anchors start, [^/]+ matches one or more non-slash chars, \. matches literal dot, $ anchors end
// Atomic group prevents backtracking on malicious inputs
const safeFilenameRegex = /^(?>[^/]+)\.([^.]+)$/;

// Test cases
const tests = [
  { input: '/path/to/document.pdf', expected: 'document', desc: 'pdf extension' },
  { input: 'archive.tar.gz', expected: 'archive', desc: 'double extension (should match first)' },
  { input: 'noext', expected: 'noext', desc: 'no dot, return basename' },
];

tests.forEach(({ input, expected, desc }) => {
  const match = input.match(safeFilenameRegex);
  const ok = match && match[1] === expected;
  console.log(`${desc}: "${input}" → ${match ? match[1] : 'null'} ${ok ? '✅' : '❌'}`);
});
```

```bash
# ReDoS safety check (Node)
const { checkRegex } = require('redoS');
const regex = /^(?>[^/]+)\.([^.]+)$/;
const result = checkRegex(regex, 'aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa.');
console.log(result); // { safe: true, maxIterations: ... } or warns about backtracking
```
```