How to use

[mrdeveloperdude]

Brother, your project is legitimately lit!

I was hoping to (ab)-use it together with esprima to create a simple javascript formatter/minifier in my pure python project (no dependency on nodejs or similar). Unfortunately I am unable to combine the two. So my question is, how do you parse your js to produce input for jscodegen?

If you are curious, here is my proof of concept code;

import esprima
from .jscodegen import CodeGenerator

generator = CodeGenerator(indent = 2)

def js2ast(js: str):
    return esprima.parseScript(js)

def ast2js(ast: dict):
    return generator.generate(ast)

raw = "var lol=123; bob='mystr'; function arne(){console.log(lol, bob);} arne(); alert('hello');"

output = ast2js( js2ast(raw) )

And it fails like this:

output = ast2js( js2ast(raw) )
../__init__.py:26: in ast2js
    return generator.generate(ast)
../jscodegen/__init__.py:430: in generate
    if self.is_statement(node):
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

self = <.jscodegen.CodeGenerator object at 0x7f8037a65c90>
node = {
    type: "Program",
    sourceType: "script",
    body: [
        {
            type: "VariableDeclaration",
      ...,
                    generator: False
                },
                arguments: []
            }
        }
    ]
}

    def is_statement(self, node):
>       return Syntax(node["type"]) in Statements
E       TypeError: 'Script' object is not subscriptable

../jscodegen/__init__.py:392: TypeError

[mrdeveloperdude]

I figured it out. Turns out esprima does not return a dictionary, but rather a hierarchy of Node objects. To turn this tree into a dictionary that can be read by jscodegen, simply call .toDict() on it.

NOTE: In the documentation for esprima js version the function is called .dict() for some reason (probably to avoid keyword overloading) it is .toDict() in the python version. This threw me off for a while.