luaparse

A Lua parser written in JavaScript, for my bachelor's thesis at Arcada.

Installation

Install through bower install luaparse or npm install luaparse.

Usage

CommonJS

var parser = require('luaparse');
var ast = parser.parse('i = 0');
console.log(JSON.stringify(ast));

AMD

require(['luaparse'], function(parser) {
  var ast = parser.parse('i = 0');
  console.log(JSON.stringify(ast));
});

Browser

<script src="luaparse.js"></script>
<script>
var ast = luaparse.parse('i = 0');
console.log(JSON.stringify(ast));
</script>

Parser Interface

Basic usage:

luaparse.parse(code, options);

The output of the parser is an Abstract Syntax Tree (AST) formatted in JSON.

The available options are:

• wait: false Explicitly tell the parser when the input ends.
• comments: true Store comments as an array in the chunk object.
• scope: false Track identifier scopes.
• locations: false Store location information on each syntax node.
• ranges: false Store the start and end character locations on each syntax node.
• onCreateNode: null A callback which will be invoked when a syntax node has been completed. The node which has been created will be passed as the only parameter.
• onCreateScope: null A callback which will be invoked when a new scope is created.
• onDestroyScope: null A callback which will be invoked when the current scope is destroyed.
• onLocalDeclaration: null A callback which will be invoked when a local variable is declared. The identifier will be passed as the only parameter.
• luaVersion: '5.1' The version of Lua the parser will target; supported values are '5.1', '5.2', '5.3' and 'LuaJIT'.
• extendedIdentifiers: false Whether to allow code points ≥ U+0080 in identifiers, like LuaJIT does. See 'Note on character encodings' below if you wish to use this option. Note: setting luaVersion: 'LuaJIT' currently does not enable this option; this may change in the future.

The default options are also exposed through luaparse.defaultOptions where they can be overriden globally.

There is a second interface which might be preferable when using the wait option.

var parser = luaparse.parse({ wait: true });
parser.write('foo = "');
parser.write('bar');
var ast = parser.end('"');

This would be identical to:

var ast = luaparse.parse('foo = "bar"');

AST format

If the following code is executed:

luaparse.parse('foo = "bar"');

then the returned value will be:

{
  "type": "Chunk",
  "body": [
    {
      "type": "AssignmentStatement",
      "variables": [
        {
          "type": "Identifier",
          "name": "foo"
        }
      ],
      "init": [
        {
          "type": "StringLiteral",
          "value": "bar",
          "raw": "\"bar\""
        }
      ]
    }
  ],
  "comments": []
}

Note on character encodings

Unlike strings in JavaScript, Lua strings are not Unicode strings, but bytestrings (sequences of 8-bit values); likewise, implementations of Lua parse the source code as a sequence of octets. However, the input to this parser is a JavaScript string, i.e. a sequence of 16-bit code units (not necessarily well-formed UTF-16). This poses a problem of how those code units should be interpreted, particularly if they are outside the Basic Latin block ('ASCII').

Currently, this parser handles Unicode input by encoding it in WTF-8, and reinterpreting the resulting code units as Unicode code points. This applies to string literals and (if extendedIdentifiers is enabled) to identifiers as well. Lua byte escapes inside string literals are interpreted directly as code points, while Lua 5.3 \u{} escapes are similarly decoded as UTF-8 code units reinterpreted as code points. It is as if the parser input was being interpreted as ISO-8859-1, while actually being encoded in UTF-8.

This ensures that no otherwise-valid input will be rejected due to encoding errors. Assuming the input was originally encoded in UTF-8 (which includes the case of only containing ASCII characters), it also preserves the following properties:

• String literal nodes representing the same string value in Lua (and identifier nodes, if extendedIdentifiers is enabled) will have the same interpretation in the AST: e.g. the Lua literals '💩', '\u{1f4a9}' and '\240\159\146\169' will all have "\u00f0\u009f\u0092\u00a9" in their .value property, and likewise local 💩 will have the same string in its .name property.
• The .length property of decoded string values in the AST is equal to the value that the # operator would return in Lua.

Maintaining those properties makes the logic of static analysers and code transformation tools simpler. However, it poses a problem when displaying strings to the user and serialising AST back into a string; to recover the original bytestrings, values transformed in this way will have to be encoded in ISO-8859-1.

Other solutions to this problem may be considered in the future. Some of them have been listed below, with their drawbacks:

• A mode that instead treats the input as if it were decoded according to ISO-8859-1 (or the x-user-defined encoding) and rejects code points that cannot appear in that encoding; may be useful for source code in encodings other than UTF-8
  • Still tricky to get semantics correctly
• Using an ArrayBuffer or Uint8Array for source code and/or string literals
  • May fail to be portable to older JavaScript engines
  • Cannot be (directly) serialised as JSON
  • Values of those types are fixed-length, which makes manipulation cumbersome; they cannot be incrementally built by appending.
  • They cannot be used as keys in objects; one has to use Map and WeakMap instead
• Using a plain Array of numbers in the range [0, 256)
  • Memory-inefficient
  • May bloat the JSON serialisation considerably
  • Cannot be used as keys in objects either
• Storing string literal values as ordinary String values, and requiring that escape sequences in literals constitute well-formed UTF-8; an exception is thrown if they do not
  • UTF-8 chauvinism; imposes semantics that may be unwanted
  • Reduced compatibility with other Lua implementations
• Like above, but instead of throwing an exception, ill-formed escapes are transformed to unpaired surrogates, just like Python's surrogateescape encoding error handler
  • Destroys the property that ("\xc4" .. "\x99") == "\xc4\x99"
  • If the AST is encoded in JSON, some JSON libraries may refuse to parse it

Custom AST

The default AST structure is somewhat inspired by the Mozilla Parser API but can easily be overriden to customize the structure or to inject custom logic.

luaparse.ast is an object containing all functions used to create the AST, if you for example wanted to trigger an event on node creations you could use the following:

var luaparse = require('luaparse'),
    events = new (require('events').EventEmitter);

Object.keys(luaparse.ast).forEach(function(type) {
  var original = luaparse.ast[type];
  luaparse.ast[type] = function() {
    var node = original.apply(null, arguments);
    events.emit(node.type, node);
    return node;
  };
});
events.on('Identifier', function(node) { console.log(node); });
luaparse.parse('i = "foo"');

this is only an example to illustrate what is possible and this particular example might not suit your needs as the end location of the node has not been determined yet. If you desire events you should use the onCreateNode callback instead).

Lexer

The lexer used by luaparse can be used independently of the recursive descent parser. The lex function is exposed as luaparse.lex() and it will return the next token up until EOF is reached.

Each token consists of:

• type expressed as an enum flag which can be matched with luaparse.tokenTypes.
• value
• line, lineStart
• range can be used to slice out raw values, eg. foo = "bar" will return a StringLiteral token with the value bar. Slicing out the range on the other hand will return "bar".

var parser = luaparse.parse('foo = "bar"', { wait: true });
parser.lex(); // { type: 8, value: "foo", line: 1, lineStart: 0, range: [0, 3] }
parser.lex(); // { type: 32, value: "=", line: 1, lineStart: 0, range: [4, 5]}
parser.lex(); // { type: 2, value: "bar", line: 1, lineStart: 0, range: [6, 11] }
parser.lex(); // { type: 1, value: "<eof>", line: 1, lineStart: 0, range: [11 11] }
parser.lex(); // { type: 1, value: "<eof>", line: 1, lineStart: 0, range: [11 11] }

Examples

Have a look in the examples directory of the repository for some code examples or check them out live.

luaparse(1)

The luaparse executable can be used in your shell by installing luaparse globally using npm:

$ npm install -g luaparse
$ luaparse --help

Usage: luaparse [option]... [file|code]...

Options:
  -c|--code [code]   parse code snippet
  -f|--file [file]   parse from file
  -b|--beautify      output an indenteted AST
  --[no]-comments    store comments. defaults to true
  --[no]-scope       store variable scope. defaults to false
  --[no]-locations   store location data on syntax nodes. defaults to false
  --[no]-ranges      store start and end character locations. defaults to false
  -q|--quiet         suppress output
  -h|--help
  -v|--version
  --verbose

Examples:
  luaparse --no-comments -c "locale foo = \"bar\""
  luaparse foo.lua bar.lua

Example usage

$ luaparse "i = 0"

{"type":"Chunk","body":[{"type":"AssignmentStatement","variables":[{"type":"Identifier","name":"i"}],"init":[{"type":"NumericLiteral","value":0,"raw":"0"}]}],"comments":[]}

Support

Has been tested in at least IE6+, Firefox 3+, Safari 4+, Chrome 10+, Opera 10+, Node 0.4.0+, RingoJS 0.8-0.9, Narwhal 0.3.2, Rhino 1.7R4-1.7R5, Nashorn 1.8.0.

Quality Assurance

TL;DR simply run make qa. This will run all quality assurance scripts but assumes you have it set up correctly.

Begin by cloning the repository and installing the development dependencies with npm install. To test AMD loading for browsers you should run bower install which will download RequireJS.

The luaparse test suite uses testem as a test runner, and because of this it's very easy to run the tests using different javascript engines or even on locally installed browsers. Currently the default runner uses PhantomJS and node so when using make test or npm test you should have PhantomJS installed.

Test runners

• make test uses PhantomJS and node.
• make testem-engines uses PhantomJS, node, narwhal, ringo, rhino and rhino 1.7R5. This requires that you have the engines installed.
• make test-node uses a custom command line reporter to make the output easier on the eyes while practicing TDD.
• By installing testem globally you can also run the tests in a locally installed browser.

Other quality assurance measures

• You can check the function complexity using complexity-report using make complexity-analysis
• Running make coverage will generate the coverage report. To simply check that all code has coverage you can run make coverage-analysis.
• make lint, make benchmark, make profile.

Documentation

By running make docs all documentation will be generated.

Projects using/extending luaparse

• luamin, a Lua minifier written by Mathias Bynens.
• Ace, an online code editor.

Acknowledgements

• Initial tests are scaffolded from yueliang and then manually checked for error.
• Much of the code is based on LuaMinify, the Lua source and Esprima. All awesome projects.

License

MIT