nested-regex-groups

A lightweight TypeScript library that extends JavaScript's regular expressions with dot notation in capture group names to produce nested object structures.

Why?

Standard regex named capture groups return flat objects:

const match = '[email protected]'.match(/^(?<name>\w+)@(?<domain>\w+\.\w+)$/);
console.log(match.groups);
// { name: 'john', domain: 'example.com' }

With nested-regex-groups, use dot notation to create nested structures:

import { rx } from 'nested-regex-groups/template';

const parser = rx`^(?<user.name>\w+)@(?<user.domain>\w+\.\w+)$`;
const result = parser('[email protected]');

console.log(result.value);
// { user: { name: 'john', domain: 'example.com' } }

Installation

npm install nested-regex-groups

Features

• ✨ Dot notation in capture group names: (?<user.name>...) → { user: { name: ... } }
• 🎯 Multiple pattern matching: Try patterns in priority order
• 📦 Zero dependencies: Tiny footprint
• 🔒 Type-safe: Full TypeScript support with type inference
• 🚀 Fast: Minimal overhead over native regex
• 🧩 Composable: Build complex parsers from simple patterns
• 🌳 Tree-shakeable: Import only what you need with modular exports

Modular Imports

The library supports selective imports for optimal bundle size, especially useful in buildless environments:

// Import only what you need
import { splitStatements } from 'nested-regex-groups/split-statements';
import { parsePatterns } from 'nested-regex-groups/parse-patterns';

// Or import everything (backward compatible)
import { splitStatements, parsePatterns } from 'nested-regex-groups';

Available Subpath Exports

Core Utilities:

• nested-regex-groups/flat-to-nested - Convert dot-notation keys to nested objects
• nested-regex-groups/merge-results - Merge multiple parse results
• nested-regex-groups/split-statements - Split paragraphs into statements

Core Parsing:

• nested-regex-groups/nested-regex - Core parser with dot notation support
• nested-regex-groups/try-patterns - Try multiple patterns + create parser factory

Runtime Pattern Parsing:

• nested-regex-groups/parse-patterns - Parse patterns from strings/JSON config (includes parsePattern and parsePatterns)

Flat Group Parsing:

• nested-regex-groups/parse-grouped-captures - Parse with flat groups (no nesting)
• nested-regex-groups/parse-grouped-capture-statements - Parse multiple statements with flat groups

Nested Statement Parsing:

• nested-regex-groups/parse-pattern-statements - Parse multiple statements with nested groups

Template Tags:

• nested-regex-groups/template - Template tag API (rx, rxPattern, rxParser)

Buildless Environment Example

<script type="module">
  // Only loads split-statements.js (~1KB) instead of full library (~15KB)
  import { splitStatements } from './node_modules/nested-regex-groups/split-statements.js';
  
  const statements = splitStatements('First. Second. Third.');
  console.log(statements); // ['First', 'Second', 'Third']
</script>

Quick Start

JSON Config (Recommended for Runtime)

Perfect for configuration-driven applications:

import { parsePatterns } from 'nested-regex-groups';

// Load from JSON file
const config = {
  patterns: [
    {
      name: 'email',
      pattern: '^(?<user.name>\\w+)@(?<user.domain>\\w+)$'
    }
  ]
};

const parser = parsePatterns(config.patterns);
const result = parser('[email protected]');
// { user: { name: 'john', domain: 'example.com' } }

Template Tag (Recommended for Code)

The cleanest way for code-based patterns:

import { rx } from 'nested-regex-groups/template';

const parser = rx`^(?<person.name.first>\w+)\s+(?<person.name.last>\w+)$`;
const result = parser('John Doe');

if (result.success) {
  console.log(result.value);
  // { person: { name: { first: 'John', last: 'Doe' } } }
}

Multiple Patterns

import { rxParser, rxPattern } from 'nested-regex-groups/template';

const parser = rxParser([
  rxPattern('email')`^(?<user.name>\w+)@(?<user.domain>\w+\.\w+)$`,
  rxPattern('username')`^(?<user.name>\w+)$`
]);

const result1 = parser('[email protected]');
// { success: true, pattern: 'email', value: { user: { name: 'john', domain: 'example.com' } } }

const result2 = parser('john');
// { success: true, pattern: 'username', value: { user: { name: 'john' } } }

API

Statements Parsing (Multiple Statements in Paragraphs)

Parse paragraphs containing multiple statements separated by periods, with support for nested object structures.

splitStatements(input)

Splits a paragraph into individual statements based on period delimiters.

import { splitStatements } from 'nested-regex-groups';

const statements = splitStatements('First. Second. Third.');
// ['First', 'Second', 'Third']

// Handles optional chaining and escaped periods
splitStatements('Check #obj?.prop. Use file\\.txt.');
// ['Check #obj?.prop', 'Use file.txt']

Rules:

• Splits on . followed by whitespace or end of string
• Ignores ?. (optional chaining)
• Ignores \. (escaped period - becomes . in output)
• Trailing period on last statement is optional

parseGroupedCaptures(input, patternConfigs, options?)

Parses a single statement using flat group patterns (no dots in group names).

import { parseGroupedCaptures } from 'nested-regex-groups';

const patterns = [
  { name: 'comparison', pattern: '^(?<trigger>on|off)\\s+when\\s+(?<lhs>#\\w+)\\s+eq\\s+(?<rhs>#\\w+)$' }
];

const result = parseGroupedCaptures('on when #foo eq #bar', patterns);
// { success: true, pattern: 'comparison', value: { trigger: 'on', lhs: '#foo', rhs: '#bar' } }

Use when: You need flat objects without nesting.

parseGroupedCaptureStatements(input, patternConfigs, options?)

Parses multiple statements with flat group patterns.

import { parseGroupedCaptureStatements } from 'nested-regex-groups';

const patterns = [
  { name: 'comparison', pattern: '^(?<trigger>on|off)\\s+when\\s+(?<lhs>#\\w+)\\s+eq\\s+(?<rhs>#\\w+)$' },
  { name: 'boolean', pattern: '^(?<trigger>on|off)\\s+when\\s+(?<lhs>#\\w+)$' }
];

const result = parseGroupedCaptureStatements('on when #foo eq #bar. off when #baz.', patterns);
// {
//   success: true,
//   statements: [
//     { pattern: 'comparison', value: { trigger: 'on', lhs: '#foo', rhs: '#bar' } },
//     { pattern: 'boolean', value: { trigger: 'off', lhs: '#baz' } }
//   ]
// }

parsePatternStatements(input, patternConfigs, options?)

Parses multiple statements with nested group patterns (dots in group names).

import { parsePatternStatements } from 'nested-regex-groups';

const patterns = [
  { name: 'comparison', pattern: '^(?<trigger>on|off)\\s+when\\s+(?<lhs.id>#\\w+)\\s+eq\\s+(?<rhs.id>#\\w+)$' }
];

const result = parsePatternStatements('on when #foo eq #bar. off when #baz.', patterns);
// {
//   success: true,
//   statements: [
//     { pattern: 'comparison', value: { trigger: 'on', lhs: { id: '#foo' }, rhs: { id: '#bar' } } },
//     { pattern: 'comparison', value: { trigger: 'off', lhs: { id: '#baz' }, rhs: { id: '#baz' } } }
//   ]
// }

Use when: You need nested objects from multiple statements.

Default Values

Both parsePatternStatements and parseGroupedCaptureStatements support default values for optional capture groups. This is useful when you want to provide fallback values for groups that might not match.

import { parsePatternStatements } from 'nested-regex-groups';

const patterns = [
  { 
    name: 'comparison', 
    pattern: '^(?<trigger>on|off)?\\s*when\\s+(?<lhs.id>#\\w+)\\s+eq\\s+(?<rhs.id>#\\w+)$',
    defaultVals: { 
      trigger: 'on',           // Default for optional trigger
      'lhs.id': '#default'     // Dot notation works in defaults too
    }
  }
];

// When trigger is not specified, defaults to 'on'
const result = parsePatternStatements('when #foo eq #bar', patterns);
// {
//   success: true,
//   statements: [{
//     pattern: 'comparison',
//     value: { trigger: 'on', lhs: { id: '#foo' }, rhs: { id: '#bar' } }
//   }]
// }

// When trigger IS specified, parsed value overrides default
const result2 = parsePatternStatements('off when #foo eq #bar', patterns);
// {
//   success: true,
//   statements: [{
//     pattern: 'comparison',
//     value: { trigger: 'off', lhs: { id: '#foo' }, rhs: { id: '#bar' } }
//   }]
// }

Key behaviors:

• Default values are applied when optional groups don't match (return undefined)
• Parsed values always override defaults
• For nested patterns, use dot notation in default keys (e.g., 'lhs.id': '#default')
• For flat patterns, use simple keys (e.g., trigger: 'on')

parseParagraph(input, patternConfigs, options?)

Convenience alias for parsePatternStatements (the most common use case).

import { parseParagraph } from 'nested-regex-groups';

const result = parseParagraph('on when #foo eq #bar. off when #baz.', patterns);

StatementsResult<T>

Return type for statements parsing functions:

interface StatementsResult<T = any> {
  success: boolean;  // true if all statements parsed successfully
  statements: Array<{
    pattern?: string;   // Name of matched pattern
    value?: T;          // Parsed value (if successful)
    error?: string;     // Error message (if failed)
    matched?: string;   // Matched text
  }>;
}

Example: Real-world be-switched paragraph

import { parsePatternStatements } from 'nested-regex-groups';

const patterns = [
  {
    name: 'fullComparison',
    pattern: '^(?<trigger>on|off)\\s+when\\s+(?<lhs.id>#\\w+)(?:::(?<lhs.event>\\w+))?(?:\\?\\.(?<lhs.prop>\\w+))?\\s+(?<op>eq|gt|lt)\\s+(?<rhs.id>#\\w+)(?:::(?<rhs.event>\\w+))?(?:\\?\\.(?<rhs.prop>\\w+))?$'
  }
];

const paragraph = 'on when #lhs::change?.weight gt #rhs?.weight. off when #brother::change?.height lt #sister::input?.height.';
const result = parsePatternStatements(paragraph, patterns);

// {
//   success: true,
//   statements: [
//     {
//       pattern: 'fullComparison',
//       value: {
//         trigger: 'on',
//         lhs: { id: '#lhs', event: 'change', prop: 'weight' },
//         op: 'gt',
//         rhs: { id: '#rhs', prop: 'weight' }
//       }
//     },
//     {
//       pattern: 'fullComparison',
//       value: {
//         trigger: 'off',
//         lhs: { id: '#brother', event: 'change', prop: 'height' },
//         op: 'lt',
//         rhs: { id: '#sister', event: 'input', prop: 'height' }
//       }
//     }
//   ]
// }

See examples/statements-usage.ts for more examples.

Runtime Parsing (Recommended for JSON Config)

parsePattern(patternString, name?)

Parses a regex pattern string with dot notation and creates a parser. Perfect for loading patterns from JSON config files.

import { parsePattern } from 'nested-regex-groups';

// From JSON config
const config = { pattern: "^(?<user.name>\\w+)@(?<user.domain>\\w+)$" };
const parser = parsePattern(config.pattern);
const result = parser('[email protected]');

Parameters:

• patternString: string - Regex pattern with dots in group names
• name?: string - Optional name for error messages

Use when:

• Loading patterns from JSON files
• Patterns stored as strings in config
• Dynamic pattern loading at runtime

parsePatterns(patternConfigs, options?)

Creates a multi-pattern parser from JSON-like configuration objects.

import { parsePatterns } from 'nested-regex-groups';

const config = {
  patterns: [
    { name: 'email', pattern: "^(?<user.name>\\w+)@(?<user.domain>\\w+)$" },
    { name: 'username', pattern: "^(?<user.name>\\w+)$" }
  ]
};

const parser = parsePatterns(config.patterns);

Use when:

• Loading multiple patterns from JSON
• Configuration-driven applications
• Patterns need to be updated without code changes

See JSON_CONFIG.md for complete guide.

Template Tag API (Recommended for Code)

rx

Template tag for creating single-pattern parsers with dot notation.

import { rx } from 'nested-regex-groups/template';

const parser = rx`^(?<user.name>\w+)@(?<user.domain>\w+)$`;
const result = parser('[email protected]');

Benefits:

• Clean syntax with dots directly in the pattern
• Automatically handles String.raw behavior
• No manual groupMap needed

Use when:

• Patterns are part of your code
• You want type inference and IDE support
• Patterns rarely change

rxPattern

Creates a pattern definition for use with rxParser.

import { rxPattern } from 'nested-regex-groups/template';

const emailPattern = rxPattern('email', 'Email address')`^(?<user.name>\w+)@(?<user.domain>\w+)$`;

rxParser

Creates a multi-pattern parser from rxPattern definitions.

import { rxParser, rxPattern } from 'nested-regex-groups/template';

const parser = rxParser([
  rxPattern('email')`^(?<user.name>\w+)@(?<user.domain>\w+)$`,
  rxPattern('username')`^(?<user.name>\w+)$`
]);

See TEMPLATE_TAG.md for complete guide.

Core API (Advanced Use)

nestedRegex(pattern, options?)

Creates a parser from a single regex pattern with manual groupMap.

Parameters:

• pattern: RegExp - Regular expression with named capture groups (use underscores, not dots)
• options?: NestedRegexOptions - Options including name and groupMap

Returns: (input: string) => ParseResult

Example:

import { nestedRegex } from 'nested-regex-groups';

// Use underscores in regex, map to dots via groupMap
const parser = nestedRegex(/^(?<lhs_id>#\w+)\s+eq\s+(?<rhs_id>#\w+)$/, {
  name: 'comparison',
  groupMap: { lhs_id: 'lhs.id', rhs_id: 'rhs.id' }
});

const result = parser('#foo eq #bar');
// { lhs: { id: '#foo' }, rhs: { id: '#bar' } }

Note: For cleaner syntax, use rx template tag or parsePattern() instead.

createParser(patterns, options?)

Creates a parser that tries multiple patterns in order.

Parameters:

• patterns: ParsePattern[] - Array of pattern definitions
• options?: ParserOptions - Parser options

Returns: (input: string) => ParseResult & { pattern?: string }

Example:

import { createParser } from 'nested-regex-groups';

const parser = createParser([
  { 
    name: 'full', 
    regex: /^(?<a>\w+)\s+(?<b>\w+)$/
  },
  { 
    name: 'simple', 
    regex: /^(?<a>\w+)$/
  }
]);

tryPatterns(input, patterns, options?)

Tries multiple patterns against input (used internally by createParser).

Parameters:

• input: string - String to parse
• patterns: ParsePattern[] - Array of patterns to try
• options?: ParserOptions - Parser options

Returns: ParseResult & { pattern?: string }

flatToNested(groups)

Utility function to convert flat object with dot-notation keys to nested structure.

Parameters:

• groups: Record<string, string | undefined> - Flat object with dot-notation keys

Returns: Nested object

Example:

flatToNested({ 'user.name': 'John', 'user.age': '30' });
// { user: { name: 'John', age: '30' } }

mergeResults(results)

Merges multiple parse results into a single object.

Parameters:

• results: ParseResult[] - Array of parse results

Returns: T | null - Merged object or null if any parse failed

Types

All TypeScript type definitions are available in a dedicated types module:

// Import types from the types module
import type {
  ParseResult,
  ParseSuccess,
  ParseFailure,
  ParsePattern,
  PatternConfig,
  NestedRegexOptions,
  ParserOptions,
  StatementsResult
} from 'nested-regex-groups/types';

// Or import from main module (backward compatible)
import type { ParseResult, ParsePattern, PatternConfig } from 'nested-regex-groups';

Type Organization

Types are organized in separate files for better tree-shaking and discoverability:

• types/nested-regex-groups/parse-result.ts - ParseResult, ParseSuccess, ParseFailure
• types/nested-regex-groups/pattern.ts - ParsePattern, NestedRegexOptions
• types/nested-regex-groups/options.ts - ParserOptions
• types/nested-regex-groups/statements.ts - StatementsResult

Types Reference

ParseResult<T>

type ParseResult<T> = ParseSuccess<T> | ParseFailure;

interface ParseSuccess<T> {
  success: true;
  value: T;
  matched: string;  // The portion of input that matched
  rest: string;     // Remaining unparsed input
}

interface ParseFailure {
  success: false;
  error: string;
  position?: number;
}

ParsePattern

interface ParsePattern {
  name: string;
  regex: RegExp;
  description?: string;
}

PatternConfig

Configuration object for defining patterns from strings (used with parsePatterns, parsePatternStatements, etc.):

interface PatternConfig {
  name: string;
  pattern: string;              // Regex pattern as string
  description?: string;
  defaultVals?: Record<string, string>;  // Default values for optional groups
}

Example:

const config: PatternConfig = {
  name: 'comparison',
  pattern: '^(?<trigger>on|off)?\\s*when\\s+(?<lhs.id>#\\w+)\\s+eq\\s+(?<rhs.id>#\\w+)$',
  defaultVals: {
    trigger: 'on',
    'lhs.id': '#default'
  }
};

ParserOptions

interface ParserOptions {
  verbose?: boolean;  // Include detailed error messages
}

Real-World Example: be-switched Parser

This library was created to support parsing for be-switched, a template behavior for conditional content loading.

import { parsePatterns } from 'nested-regex-groups';

// Load from JSON config
const config = {
  patterns: [
    {
      name: 'fullComparison',
      pattern: '^(?<trigger>on|off)\\s+when\\s+(?<lhs.id>#\\w+)(?:::(?<lhs.event>\\w+))?(?:\\?\\.(?<lhs.prop>\\w+))?\\s+(?<op>equals|eq|lt|gt)\\s+(?<rhs.id>#\\w+)(?:::(?<rhs.event>\\w+))?(?:\\?\\.(?<rhs.prop>\\w+))?$',
      description: 'Comparison with events and properties'
    },
    {
      name: 'simpleComparison',
      pattern: '^(?<trigger>on|off)\\s+when\\s+(?<lhs.id>#\\w+)\\s+(?<op>equals|eq)\\s+(?<rhs.id>#\\w+)$',
      description: 'Simple comparison'
    },
    {
      name: 'booleanCondition',
      pattern: '^(?<trigger>on|off)\\s+when\\s+(?<lhs.id>#\\w+)$',
      description: 'Boolean condition'
    }
  ]
};

const beSwitchedParser = parsePatterns(config.patterns);

// Parse: "on when #lhs::change?.weight gt #rhs?.weight"
const result = beSwitchedParser('on when #lhs::change?.weight gt #rhs?.weight');

console.log(result);
// {
//   success: true,
//   pattern: 'fullComparison',
//   value: {
//     trigger: 'on',
//     lhs: { id: '#lhs', event: 'change', prop: 'weight' },
//     op: 'gt',
//     rhs: { id: '#rhs', prop: 'weight' }
//   }
// }

Design Philosophy

Why Array of Patterns?

For complex grammars, a single monolithic regex becomes:

• ❌ Unreadable (500+ characters)
• ❌ Unmaintainable (one change breaks everything)
• ❌ Poor error messages
• ❌ Difficult to extend

An array of patterns provides:

• ✅ Readable (each pattern ~100 chars)
• ✅ Maintainable (modify one pattern at a time)
• ✅ Better errors (know which pattern failed)
• ✅ Extensible (add new patterns easily)
• ✅ Testable (test each pattern independently)

Think of it like HTTP routing: Express.js doesn't use one giant regex for all routes—it uses an array of route patterns. Same principle applies here.

Migration Guide

Upgrading to Modular Exports

The library now supports selective imports for better tree-shaking and smaller bundle sizes in buildless environments. Your existing code continues to work without changes.

Before (Still Supported)

import { 
  splitStatements, 
  parsePatterns, 
  parsePatternStatements 
} from 'nested-regex-groups';

After (Selective Imports)

// Import only what you need
import { splitStatements } from 'nested-regex-groups/split-statements';
import { parsePatterns } from 'nested-regex-groups/parse-patterns';
import { parsePatternStatements } from 'nested-regex-groups/parse-pattern-statements';

Benefits of Selective Imports

• Smaller bundles: Load only the code you use (~1KB for splitStatements vs ~15KB for full library)
• Faster load times: Especially important in buildless environments
• Better tree-shaking: Even with bundlers, explicit imports help optimization
• Clearer dependencies: See exactly what your code uses

Module Grouping

Functions are grouped by their dependencies and common usage patterns:

• Zero-dependency utilities: split-statements, flat-to-nested, merge-results
• Core parsing: nested-regex, try-patterns, create-parser
• Runtime parsing: parse-pattern, parse-patterns
• Statement parsing: parse-grouped-captures, parse-grouped-capture-statements, parse-pattern-statements

When to Use Selective Imports

• Buildless environments: Always use selective imports to minimize loaded code
• Large applications: Reduce bundle size by importing only needed functions
• Library authors: Minimize dependencies for your consumers

When to Use Barrel Import

• Small scripts: When bundle size isn't a concern
• Using many functions: When you need most of the library anyway
• Rapid prototyping: When convenience matters more than optimization

Inspiration

This library was inspired by:

• Raku (Perl 6) Grammars - The gold standard for nested regex captures
• Parser Combinators - Composable parsing approach
• be-switched - Real-world need for declarative string parsing

License

MIT © Bruce B. Anderson

Contributing

Contributions welcome! Please open an issue or PR on GitHub.