Plural Rule Parser

This is a PEG.js grammar for parsing plural rules as defined by CLDR. In addition to providing the grammar, a very lightweight js interface is available for parsing the grammar via PEG.js, running the parsed grammar over a passed in plural rule, and generating javascript code from the resulting AST.

The AST, which is the direct output of the grammar, is compatible with Mozilla's Spidermonkey Abstract Syntax Tree definition. It should be possible to consume the AST in other languages and/or generate code for other languages. However, I've only tested generation to javascript via Escodegen.

Installation

$> npm install plural-rule-parser --save

Example usage

var parser = require('plural-rule-parser'),
    rule = 'n is 1',
    result =  parser.parse(rule); // will throw if rule is invalid

console.log(result.toAst()); // {
                             //    "type": "Program",
                             //    "body": [
                             //      {
                             //        "type": "ExpressionStatement",
                             //        "expression": {
                             //          "type": "BinaryExpression",
                             //          "operator": "===",
                             //          "left": {
                             //            "type": "Identifier",
                             //            "name": "n"
                             //          },
                             //          "right": {
                             //            "type": "Literal",
                             //            "value": 1,
                             //            "raw": "1"
                             //          }
                             //        }
                             //      }
                             //    ]
                             //  }
console.log(result.toJsSource()); // 'n === 1;'
console.log(result.check(1)); // false
console.log(result.check(2)); // true

Test

npm test or node test/test.js

To consume the grammar directly

The recommended approach is to create a git submodule via this: git submodule add [email protected]:mattpowell/plural-rule-parser.git

Please open an issue if you need an alternate way to consume the grammar (e.g., a .tar.gz).

License

MIT