PyEsprima

A Python port of Esprima, the JavaScript parser.

Why shouldn't I use it?

It's pretty slow -- about two orders of magnitude slower to parse a 116k JS file. The code was semi-automatically translated from JavaScript, and it does a lot of string appending, which is fast in JS but slow in Python.

Why should I use it?

Shelling out to a NodeJS process is likely to be a better way to use Esprima from Python. However, writing the interprocess data-marshaling code is a bit annoying, so if you want to do a quick hack on small amounts of data, PyEsprima is a good way to get down to business. The API is almost exactly the same, so you can easily swap in an interface to the external Node process later on.

Also, since the code is semi-automatically translated using js2py, it's easier to keep up-to-date, and thus bit-rot is less of a problem.

Finally, you'll be glad to know that all the relevant tests from the original Esprima pass.

API

Aside from our handling of RegExps, we expose the same interface as Esprima / SpiderMonkey:

>>> import pyesprima

>>> print pyesprima.tokenize("1 + 1")
[{'type': 'Numeric', 'value': '1'},
 {'type': 'Punctuator', 'value': '+'},
 {'type': 'Numeric', 'value': '1'}]

>>> pyesprima.parse("1 + 1", loc=True)
{'body': [{'type': 'ExpressionStatement', 'expression': {'operator': '+',
'loc': {'start': {'column': 0, 'line': 1}, 'end': ...

Esprima creates RegExp objects for regular expression literals in the JS source. However, Python regular expressions do not have a global flag (it has re.findall instead), so we cannot map JS RegExps perfectly; instead, pyesprima.RegExp objects have a flags attribute that records what flags were used in the JS source.

Note that the returned objects are instances of pyesprima.jsdict, which simulates JavaScript dicts in having both a.x and a['x'] mean the same thing.

Installing

pip install pyesprima