Espree started out as a fork of Esprima v1.2.2, the last stable published released of Esprima before work on ECMAScript 6 began. Espree is now built on top of Acorn, which has a modular architecture that allows extension of core functionality. The goal of Espree is to produce output that is similar to Esprima with a similar API so that it can be used in place of Esprima.
Install:
npm i espree --save
And in your Node.js code:
var espree = require("espree");
var ast = espree.parse(code);There is a second argument to parse() that allows you to specify various options:
var espree = require("espree");
var ast = espree.parse(code, {
// attach range information to each node
range: true,
// attach line/column location information to each node
loc: true,
// create a top-level comments array containing all comments
comments: true,
// attach comments to the closest relevant node as leadingComments and
// trailingComments
attachComment: true,
// create a top-level tokens array containing all tokens
tokens: true,
// try to continue parsing if an error is encountered, store errors in a
// top-level errors array
tolerant: true,
// specify the language version (3, 5, or 6, default is 5)
ecmaVersion: 5,
// specify which type of script you're parsing (script or module, default is script)
sourceType: "script",
// specify additional language features
ecmaFeatures: {
// enable JSX parsing
jsx: true,
// enable return in global scope
globalReturn: true,
// allow experimental object rest/spread
experimentalObjectRestSpread: true
}
});The primary goal is to produce the exact same AST structure and tokens as Esprima, and that takes precedence over anything else. (The AST structure being the ESTree API with JSX extensions.) Separate from that, Espree may deviate from what Esprima outputs in terms of where and how comments are attached, as well as what additional information is available on AST nodes. That is to say, Espree may add more things to the AST nodes than Esprima does but the overall AST structure produced will be the same.
Espree may also deviate from Esprima in the interface it exposes.
Issues and pull requests will be triaged and responded to as quickly as possible. We operate under the ESLint Contributor Guidelines, so please be sure to read them before contributing. If you're not sure where to dig in, check out the issues.
Espree is licensed under a permissive BSD 2-clause license.
npm test- run all linting and testsnpm run lint- run all lintingnpm run browserify- creates a version of Espree that is usable in a browser
- The
tokenize()method does not useecmaFeatures. Any string will be tokenized completely based on ECMAScript 6 semantics. - Trailing whitespace no longer is counted as part of a node.
letandconstdeclarations are no longer parsed by default. You must opt-in usingecmaFeatures.blockBindings.
In an effort to help those wanting to transition from other parsers to Espree, the following is a list of noteworthy incompatibilities with other parsers. These are known differences that we do not intend to change.
- Esprima counts trailing whitespace as part of each AST node while Espree does not. In Espree, the end of a node is where the last token occurs.
- Espree does not parse
letandconstdeclarations by default. - Error messages returned for parsing errors are different.
- There are two addition properties on every node and token:
startandend. These represent the same data asrangeand are used internally by Acorn.
- Esprima 2.x uses a different comment attachment algorithm that results in some comments being added in different places than Espree. The algorithm Espree uses is the same one used in Esprima 1.2.2.
ESLint had been relying on Esprima as its parser from the beginning. While that was fine when the JavaScript language was evolving slowly, the pace of development increased dramatically and Esprima had fallen behind. ESLint, like many other tools reliant on Esprima, has been stuck in using new JavaScript language features until Esprima updates, and that caused our users frustration.
We decided the only way for us to move forward was to create our own parser, bringing us inline with JSHint and JSLint, and allowing us to keep implementing new features as we need them. We chose to fork Esprima instead of starting from scratch in order to move as quickly as possible with a compatible API.
With Espree 2.0.0, we are no longer a fork of Esprima but rather a translation layer between Acorn and Esprima syntax. This allows us to put work back into a community-supported parser (Acorn) that is continuing to grow and evolve while maintaining an Esprima-compatible parser for those utilities still built on Esprima.
Yes. Since the start of ESLint, we've regularly filed bugs and feature requests with Esprima and will continue to do so. However, there are some different philosophies around how the projects work that need to be worked through. The initial goal was to have Espree track Esprima and eventually merge the two back together, but we ultimately decided that building on top of Acorn was a better choice due to Acorn's plugin support.
Acorn is a great JavaScript parser that produces an AST that is compatible with Esprima. Unfortunately, ESLint relies on more than just the AST to do its job. It relies on Esprima's tokens and comment attachment features to get a complete picture of the source code. We investigated switching to Acorn, but the inconsistencies between Esprima and Acorn created too much work for a project like ESLint.
We are building on top of Acorn, however, so that we can contribute back and help make Acorn even better.
All of them.
In general, we do not support experimental JavaScript features. We may make exceptions from time to time depending on the maturity of the features.