Parserator
⚠️ Warning: AI-Generated Documentation
This documentation was generated by AI and may contain inaccuracies or errors. While I hope it provides a useful introduction to the library's API, please verify important details through the source code or by testing. I'm working on comprehensive, human-written/verified documentation.
An elegant parser combinators library for TypeScript that makes building parsers as simple as combining LEGO blocks.
What are Parser Combinators?
Think of parser combinators as small, reusable parsing functions that you can combine to build complex parsers. Instead of writing one giant parsing function, you build many small ones and compose them together.
// Instead of this mess:
function parsePhoneNumber(input: string) {
// 20 lines of regex and string validation...
}
// You write this:
const phoneNumber = parser(function* () {
yield* char("(");
const areaCode = yield* many1(digit).expect("area code");
yield* char(")");
yield* char(" ");
const exchange = yield* many1(digit).expect("exchange");
yield* char("-");
const number = yield* many1(digit).expect("number");
return `(${areaCode.join("")}) ${exchange.join("")}-${number.join("")}`;
});
phoneNumber.parseOrThrow("(555) 123-4567"); // → "(555) 123-4567"
Quick Start
npm install parserator
# or
bun add parserator
import { Parser, char, string, many1, digit, parser } from "parserator";
// Parse a simple number
const number = many1(digit).map(digits => parseInt(digits.join("")));
number.parseOrThrow("123"); // → 123
// Parse coordinates like "(10, 20)"
const coordinate = parser(function* () {
yield* char("(").expect("opening parenthesis '('");
const x = yield* number;
yield* string(", ").expect("comma between coordinates");
const y = yield* number;
yield* char(")").expect("closing parenthesis ')'");
return { x, y };
});
coordinate.parseOrThrow("(10, 20)"); // → { x: 10, y: 20 }
Building Blocks
Parserator provides simple building blocks that you combine to parse complex formats:
Basic Parsers
import { char, string, digit, alphabet, regex } from "parserator";
char("a"); // Matches single character 'a'
string("hello"); // Matches exact string "hello"
digit; // Matches any digit 0-9
alphabet; // Matches any letter a-z, A-Z
regex(/[a-z]+/); // Matches using regular expressions
Combinators
import { many, many1, optional, or, sepBy } from "parserator";
many(digit); // Zero or more digits
many1(alphabet); // One or more letters
optional(char("-")); // Optional minus sign
or(string("yes"), string("no")); // Either "yes" or "no"
sepBy(digit, char(",")); // Digits separated by commas: "1,2,3"
Sequencing
// Traditional method chaining
const parser1 = string("Hello").then(char(" ")).then(string("World"));
// Generator syntax (recommended for complex parsers)
const parser2 = parser(function* () {
yield* string("Hello");
yield* char(" ");
yield* string("World");
return "Greeting parsed!";
});
Generator Syntax
The generator syntax (parser
) makes complex parsers readable and maintainable:
const jsonObject = parser(function* () {
yield* char("{");
yield* whitespace;
const properties = yield* sepBy(
parser(function* () {
const key = yield* stringLiteral;
yield* char(":");
yield* whitespace;
const value = yield* jsonValue;
return { key, value };
}),
char(",")
);
yield* whitespace;
yield* char("}");
return Object.fromEntries(properties.map(p => [p.key, p.value]));
});
Real-World Example: INI File Parser
import {
Parser,
parser,
char,
regex,
many,
optional,
or,
string,
atomic,
commit,
skipMany0
} from "parserator";
// Handle whitespace and comments
const whitespace = regex(/[ \t]+/);
const comment = regex(/[;#][^\n\r]*/);
const space = or(whitespace, comment);
const spaces = skipMany0(space);
function token<T>(parser: Parser<T>): Parser<T> {
return parser.trimLeft(spaces);
}
// Parse INI sections like [database]
const section = atomic(
parser(function* () {
yield* spaces;
yield* token(char("["));
yield* commit(); // Better error messages after this point
const name = yield* regex(/[^\]]+/)
.map(s => s.trim())
.expect("section name");
yield* char("]").expect("closing bracket ']'");
const properties = yield* many(
parser(function* () {
const key = yield* token(regex(/[a-zA-Z0-9_\-\.]+/));
yield* token(char("="));
yield* commit();
const value = yield* regex(/[^\r\n]*/).map(s => s.trim());
return { key, value };
})
);
return { type: "section", name, properties };
})
);
const iniFile = parser(function* () {
yield* spaces;
const sections = yield* many(section);
yield* eof;
return sections;
});
// Usage
const config = iniFile.parseOrThrow(`
[database]
host = localhost
port = 5432
[cache]
enabled = true
`);
Error Handling
Parserator provides excellent error messages with precise locations:
const parser = parser(function* () {
yield* string("function");
yield* commit(); // Commits to this parsing path
yield* char("(").expect("opening parenthesis after 'function'");
// ... rest of function parser
});
// Input: "function {"
// Error: Expected opening parenthesis after 'function' at line 1, column 9
Error Recovery
import { Either, ErrorFormatter } from "parserator";
const result = parser.parse(input);
if (Either.isLeft(result.result)) {
const formatter = new ErrorFormatter("ansi"); // or "html"
console.error(formatter.format(result.result.left));
} else {
console.log("Parsed:", result.result.right);
}
Advanced Features
Atomic Parsing
Prevents partial consumption on failure:
const complexExpression = parser(function* () {
yield* identifier;
yield* char("(");
yield* many(expression);
yield* char(")");
}).atomic(); // All-or-nothing parsing
Lookahead
Peek without consuming input:
const nextIsOperator = lookahead(or(string("++"), string("--"), string("+=")));
Custom Transformations
const hexColor = regex(/#[0-9a-fA-F]{6}/).map(hex => ({
r: parseInt(hex.slice(1, 3), 16),
g: parseInt(hex.slice(3, 5), 16),
b: parseInt(hex.slice(5, 7), 16)
}));
hexColor.parseOrThrow("#ff6b35");
// → { r: 255, g: 107, b: 53 }
Why Parser Combinators?
- Composable: Small parsers combine into larger ones
- Type-Safe: Full TypeScript support with precise types
- Readable: Code reads like the grammar you're parsing
- Testable: Each parser can be tested independently
- Maintainable: Easy to modify and extend grammars
API Reference
Core Classes
Parser<T>
- The main parser classParseErrorBundle
- Contains parsing errors with locationsErrorFormatter
- Formats errors for display
Basic Parsers
char(c)
- Match single characterstring(s)
- Match exact stringregex(r)
- Match regular expressiondigit
,alphabet
- Built-in character classes
Combinators
many(p)
,many1(p)
- Repetitionoptional(p)
- Optional parsingor(...parsers)
- AlternativessepBy(p, sep)
- Separated listsbetween(open, close, p)
- Delimited content
Generator Utilities
parser(fn)
- Generator-based parsingcommit()
- Prevent backtrackingatomic(p)
- All-or-nothing parsing
Examples
Check out the /examples
directory for complete parsers:
ini-parser.ts
- Configuration file parserjs-parser.ts
- Simplified JavaScript parser
License
MIT