multila-lexer

Lexical Analyzer for the Web and Node.js written in TypeScript.

Copyright 2022 by Andreas Schwenk

Licensed by GPLv3

Multila Website: https://www.multila.org

Personal Website: https://www.arts-and-sciences.com

Mail: [email protected]

Installation

npm install @multila/multila-lexer

Example

The following example program implements an LL(1) top-down parser for simple language with the following grammar, specified in EBNF. It uses multila-lexer to fetch tokens.

program = { assignment };
assignment = ID ":=" add ";";
add = mul { "+" mul };
mul = unary { "*" unary };
unary = ID | INT | "(" add ")";

A valid example program is for example:

# comment
x := 3 * (4+5);

Example code:

// import multila-lexer
const lex = require('@multila/multila-lexer');

function parse(src) {
  // create a new lexer instance
  const lexer = new lex.Lexer();

  // configuration
  lexer.configureSingleLineComments('#');

  // must add operators with two or more chars
  lexer.setTerminals([':=']);

  // source code to be parsed
  lexer.pushSource('mySource', src);
  parseProgram(lexer);
}

//G program = { assignment };
function parseProgram(lexer) {
  while (lexer.isNotEND()) {
    parseAssignment(lexer);
  }
}

//G assignment = ID ":=" add ";";
function parseAssignment(lexer) {
  const id = lexer.ID();
  console.log(id);
  lexer.TER(':=');
  parseAdd(lexer);
  lexer.TER(';');
  console.log('assign');
}

//G add = mul { "+" mul };
function parseAdd(lexer) {
  parseMul(lexer);
  while (lexer.isTER('+')) {
    lexer.next();
    parseMul(lexer);
    console.log('add');
  }
}

//G mul = unary { "*" unary };
function parseMul(lexer) {
  parseUnary(lexer);
  while (lexer.isTER('*')) {
    lexer.next();
    parseUnary(lexer);
    console.log('mul');
  }
}

//G unary = ID | INT | "(" add ")";
function parseUnary(lexer) {
  if (lexer.isID()) {
    const id = lexer.ID();
    console.log(id);
  } else if (lexer.isINT()) {
    const value = lexer.INT();
    console.log(value);
  } else if (lexer.isTER('(')) {
    lexer.next();
    parseAdd(lexer);
    lexer.TER(')');
  } else {
    lexer.error('expected ID or INT');
  }
}

// run
const src = `# comment
x := 3 * (4+5);`;
parse(src);

// the output is:
// x 3 4 5 add mul assign

Methods

Configuration

• enableEmitNewlines(value: boolean)

  Enables to emit newline (\n) tokens that can be tested by lexer.isNEWLINE() and consumed by lexer.NEWLINE(). Otherwise, newline characters are considered as white spaces.

• enableEmitHex(value: boolean)

  Enables to emit hexadecimal tokens that can be tested by lexer.isHEX() and consumed by lexer.HEX().

• enableEmitInt(value: boolean)

  Enables to emit integer tokens that can be tested by lexer.isINT() and consumed by lexer.INT().

• enableEmitReal(value: boolean)

  Enables to emit real valued tokens that can be tested by lexer.isREAL() and consumed by lexer.REAL().

• enableEmitBigint(value: boolean)

  Enables to emit big integer tokens that can be tested by lexer.isBIGINT() and consumed by lexer.BIGINT().

• enableEmitSingleQuotes(value: boolean)

  Enables to emit single quote tokens that can be tested by lexer.isSTR() and consumed by lexer.STR().

• enableEmitDoubleQuotes(value: boolean)

  Enables to emit double quote tokens that can be tested by lexer.isSTR() and consumed by lexer.STR().

• enableEmitIndentation(value: boolean)

  Enables to emit indentation tokens that can be tested by lexer.isINDENT() and consumed by lexer.INDENT(), as well as lexer.isOUTDENT() and consumed by lexer.OUTDENT(), respectively.

• enableBackslashLineBreaks(value: boolean)

  If enabled, a backslash (\) right before a newline concatenates the next line, ignoring indentation.

Input Files

TODO

Parsing

TODO

Error Handling

TODO

Tokens

• ID identifier
• INT integer constant
• REAL real valued constant
• EOS end of statement (usually ;)
• STR string constant in double quotes (") or single quotes (')
• TER terminal
• INDENT indentation begin
• OUTDENT indentation end
• NEWLINE newline (\n)
• EOS end of statement (; or \n)
• ENDend of input