SemReg: Semantic Regular Expressions

Table of Contents

• Installation
• Key Features
• Basic Usage
• API Reference
  • Core Function
  • Position Operators
  • Character Generators
  • Quantifiers
  • Groups
  • Compositors
  • Logical Operators
• Examples
  • Email Validation
  • URL Validation
• Custom Patterns
• TODO
  • Additional Character Classes
  • Lookahead and Lookbehind
  • Flags and Options
• License

SemReg is a TypeScript library for building regular expressions in a readable, maintainable way. It uses a functional, pipe-based approach that allows developers to compose regex patterns with a clear and expressive syntax.

Installation

npm install semreg
# or
yarn add semreg
# or
pnpm add semreg

Key Features

• 🔍 Readable Syntax: Replace cryptic regex patterns with a clear, expressive API
• 🧩 Composable: Build complex patterns by combining simple, reusable components
• 🛠️ Fully Typed: Complete TypeScript support with helpful type definitions
• 🧪 Well Tested: Comprehensive test suite ensures reliability

Basic Usage

import {
  regex,
  startOfLine,
  endOfLine,
  letters,
  digits,
  oneOrMore,
  literal,
} from "semreg";

// Create a simple regex for validating usernames (letters, digits, and underscores)
const usernameRegex = regex(
  startOfLine,
  oneOrMore(anyOf(letters, digits, literal("_"))),
  endOfLine
);

// Test the regex
usernameRegex.test("john_doe123"); // true
usernameRegex.test("invalid-username"); // false

API Reference

Core Function

• regex(...components): Combines multiple components to produce a RegExp object

Position Operators

• startOfLine: Matches the start of a line (^)
• endOfLine: Matches the end of a line ($)
• wordBoundary: Matches a word boundary (\b)
• nonWordBoundary: Matches a non-word boundary (\B)
• startOfInput: Matches the start of the input (\A)
• endOfInput: Matches the end of the input (\Z)

Character Generators

• letters: Matches any alphabetic character ([a-zA-Z])
• lowerLetters: Matches lowercase letters ([a-z])
• upperLetters: Matches uppercase letters ([A-Z])
• digits: Matches any digit ([0-9])
• literal(str): Matches the literal string provided, with special characters escaped
• whitespace(): Matches any whitespace character (\s)
• nonWhitespace(): Matches any non-whitespace character (\S)
• word(): Matches any word character (alphanumeric + underscore) (\w)
• nonWord(): Matches any non-word character (\W)
• any(): Matches any character except newline (.)
• range(from, to): Matches any character within the specified range ([from-to])

Quantifiers

• oneOrMore(component): Matches one or more occurrences (+)
• zeroOrMore(component): Matches zero or more occurrences (*)
• optional(component): Matches zero or one occurrence (?)
• repeat(component, options): Generic quantification. Use with exactly(n), atLeast(n), or between(min, max) to specify repetitions.
• exactly(n): Helper for repeat. Specifies exactly n occurrences ({n}).
• atLeast(n): Helper for repeat. Specifies at least n occurrences ({n,}).
• between(min, max): Helper for repeat. Specifies between min and max occurrences ({min,max}).

Groups

• group(...components): Creates a capturing group ((...))
• nonCapturingGroup(...components): Creates a non-capturing group ((?:...))
• namedGroup(name, ...components): Creates a named capturing group ((?<name>...))
• numberedBackreference(n): Backreference to the nth capturing group (\n).
• namedBackreference(name): Backreference to a named capturing group (\k<name>).

Compositors

• anyOf(...components): Matches any of the specified patterns ([...])
• sequence(...components): Defines an explicit sequence of patterns

Logical Operators

• or(...components): Creates an alternation between patterns (|)
• not(component): Creates a negated character set for the given component ([^...])

Examples

Email Validation

import {
  regex,
  startOfLine,
  endOfLine,
  letters,
  digits,
  literal,
  anyOf,
  oneOrMore,
  repeat,
  exactly,
  atLeast,
  between,
} from "semreg";

const emailRegex = regex(
  startOfLine,
  oneOrMore(anyOf(letters, digits, literal("._%+-"))),
  literal("@"),
  oneOrMore(anyOf(letters, digits, literal(".-"))),
  literal("."),
  repeat(letters, atLeast(2)),
  endOfLine
);

// Testing valid emails
emailRegex.test("[email protected]"); // true
emailRegex.test("[email protected]"); // true
emailRegex.test("[email protected]"); // true

// Testing invalid emails
emailRegex.test("invalid-email"); // false
emailRegex.test("@missingusername.com"); // false
emailRegex.test("user@domain"); // false

URL Validation

import {
  regex,
  startOfLine,
  endOfLine,
  letters,
  digits,
  literal,
  anyOf,
  oneOrMore,
  optional,
  zeroOrMore,
  repeat,
  exactly,
  atLeast,
  between,
  nonCapturingGroup,
  or,
} from "semreg";

const urlRegex = regex(
  startOfLine,
  or(literal("http"), literal("https")),
  literal("://"),
  optional(nonCapturingGroup(literal("www."))),
  oneOrMore(anyOf(letters, digits, literal(".-"))),
  literal("."),
  repeat(letters, between(2, 6)),
  optional(
    nonCapturingGroup(
      literal("/"),
      zeroOrMore(anyOf(letters, digits, literal("/._-")))
    )
  ),
  endOfLine
);

// Testing valid URLs
urlRegex.test("http://example.com"); // true
urlRegex.test("https://www.google.com"); // true
urlRegex.test("https://dev.to/path/to/resource"); // true

// Testing invalid URLs
urlRegex.test("ftp://example.com"); // false
urlRegex.test("http:/example.com"); // false
urlRegex.test("example.com"); // false

Custom Patterns

You can create your own reusable patterns:

import { regex, oneOrMore, letters, digits, literal, anyOf } from "semreg";

// Create a reusable pattern for alphanumeric strings
const alphanumeric = () => oneOrMore(anyOf(letters, digits));

// Use it in different contexts
const usernameRegex = regex(startOfLine, alphanumeric(), endOfLine);

const productCodeRegex = regex(
  startOfLine,
  literal("PROD-"),
  alphanumeric(),
  endOfLine
);

TODO

Operators that could be implemented soon

1. Lookahead and Lookbehind

• positiveLookahead(...): Positive lookahead ((?=...))
• negativeLookahead(...): Negative lookahead ((?!...))
• positiveLookbehind(...): Positive lookbehind ((?<=...))
• negativeLookbehind(...): Negative lookbehind ((?<!...))

2. Flags and Options

• caseInsensitive: Enable case-insensitive matching (i)
• global: Enable global matching (g)
• multiline: Enable multiline matching (m)

License

MIT