You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

7.6 KiB

JavaScript API

In Node.js, require the PEG.js module:

// ES2015+ Module System
import peg from "pegjs";

// CommonJS
var peg = require("pegjs");

In browsers, include the PEG.js library in your web page or application using the <script> tag. If PEG.js detects an AMD loader, it will define itself as a module, otherwise the API will be available in the peg global object.

After this there are 3 methods and 1 class that you will mainly use:

var grammar = "start = ('a' / 'b')+";
var passes = [ /* check, transform and generate passes */ ];
var parser, session, ast;

ast = peg.parser.parse(grammar);
session = new peg.Session( { grammar, passes } );
parser = peg.compiler.compile(ast, session);
parser = peg.generate(grammar);

The most common method you will use is peg.generate(), which is the API alternative to the PEG.js CLI. More about each method is explained below, along with their respective options.

peg.parser.parse(input[, options])

You can simply parse a grammar and get it's AST by using the parser:

var ast = peg.parser.parse("start = ('a' / 'b')+");

The following option's are used by the PEG.js parser:

  • extractComments - If true, the parser will collect all comments in the grammar (default: false).
  • reservedWords - An array of words that the parser wont allow as labels for rules (default: ES5).

When extractComments is set to true, the parser will collect all comments in the grammar and return them on the comments property (as an object map) on the grammar AST node (the AST directly returned by the parser). Each comment property has it's offset as the property key, while the property value has the following structure:

  text: 'all text between /* or */, or // and end of line',
  multiline: true|false,
  location: location()

When set to false, the comments property will be set to null.

For more information about location(), as well more methods you can use in JavaScript actions within your grammar, see the helper's description in Action Execution Environment.

new peg.compiler.Session([options])

An instance of this class holds helpers (methods, objects with methods, constants, etc) that are shared between multiple calls to peg.compiler.compile().

This approach is used rather then a traditional compiler class because it provides the plugin developer with an easier way to replace certain components rather then creating their own version of PEG.js to override one or more internal methods.

The following option's are used by the Session API, but are currently all optional:

  • opcodes - An enum like hashmap (plain object) that is used by the bytecode and parser generators.
  • parser - A pre-generated PEG.js grammar parser that should return an instance of the PEG.js AST's Grammar class. Can be replaced to add additional syntax features, or allow an alternative syntax for the grammar.
  • passes - An object with each property being an array of methods that will check or alter the AST passed to them.
  • visitor - An object that should contain the ASTVisitor class for the PEG.js AST, as well as the build() helper.
  • vm - An object that should contain runInContext(), a wrapper for eval based on Node's vm.runInContext() method.
  • warn - A method called only when PEG.js encounters an error that doesn't stop the parser from being generated.
  • error - A method that will be called when PEG.js encounters an error that will stop the parser from being generated.

This class will also return some helper methods:

  • parse( input, options ) - an alias for this.parser.parse( input, options )
  • buildVisitor( functions ) - an alias for functions )
  • fatal( message, location ) - Will always throw an error (used by default for this.error() if not overridden).

peg.compiler.compile(ast, session[, options])

This method takes a parsed grammar (the PEG.js AST), sets default options, executes each pass currently within the session (passing the ast, session and options arguments to each one), then returns the result, which depends on the value of the output option.

var grammar = "start = ('a' / 'b')+";
var ast = peg.parser.parse(grammar);
var passes = peg.util.convertPasses( peg.compiler.passes );
var session = new peg.Session( { grammar, passes } );
var parser = peg.compiler.compile(ast, session);

You can tweak the generated parser by passing a third parameter with an options object. The following options are supported:

Option default value description
allowedStartRules first rule rules the generated parser is allowed to start parsing from
cache false makes the generated parser cache results, avoiding exponential parsing time in pathological cases but making the parser slower
context {} contains a map of variables used by peg.util.vm.runInContext() when the output option is set to "parser"
dependencies {} parser dependencies, the value is an object which maps variables used to access the dependencies to module IDs used to load them; valid only when format is set to "amd", "commonjs", "es", or "umd"
exportVar null name of an optional global variable into which the generated parser object is assigned to when no module loader is detected; valid only when format is set to "globals" or "umd"
features null map of optional features that are set to true by default: "text", "offset", "range", "location", "expected", "error", "filename" and "DefaultTracer"
format "bare" format of the generated parser ("amd", "bare", "commonjs", "es", "globals", or "umd"); valid only when output is set to "source"
header null adds additional comments or content after the Generated by ... comment; this option is only handled if it's an array or a string:
optimize "speed" selects between optimizing the generated parser for parsing speed ("speed") or code size ("size")
output "parser" if set to "parser", the method will return generated parser object; if set to "source", it will return generated parser source code as a string
trace false makes the generated parser trace its progress

The header options behavior will change depending on the option type:

  • [ string1, string2, ... ] will add each element (all expected to be strings) as a separate line comment
  • string will simply append the string (e.g. "/* eslint-disable */") after the Generated by ... comment

peg.generate(grammar[, options])

Will generate a parser from the given grammar (the input passed to peg.parser.parse()):

var parser = peg.generate("start = ('a' / 'b')+");

This method will return a generated parser object or its source code as a string (depending on the value of the output option - see above). It will throw an exception if the grammar is invalid. The exception will contain the usual message property with more details about the error, along with a location property to track the location of the error. You can easily tell if the exception was thrown by PEG.js because it will always have the name property set to GrammarError.

You can tweak the generated parser by passing a second parameter with an options object to peg.generate(). The following options are supported:

  • All options that you pass to peg.compiler.compile()
  • parser - an optional object with option's passed to the PEG.js parser (peg.parser.parse())
  • plugins - plugins to use (their use() method is executed by peg.generate())