Clean up JSON-like data before decoding as JSON

cleanup_json.py

import os, re, sys

comments_re = re.compile(
    r'//.*?$|/\*.*?\*/|\'(?:\\.|[^\\\'])*\'|"(?:\\.|[^\\"])*"',
    re.DOTALL | re.MULTILINE
)
trailing_commas_re = re.compile(
    r'(,)\s*}(?=([^"\\]*(\\.|"([^"\\]*\\.)*[^"\\]*"))*[^"]*$)')
    
def remove_comments(json_like):
    """
    Removes C-style comments from *json_like* and returns the result.  Example::
    
        >>> test_json = '''\
        {
            "foo": "bar", // This is a single-line comment
            "baz": "blah" /* Multi-line
            Comment */
        }'''
        >>> remove_comments('{"foo":"bar","baz":"blah",}')
        '{\n    "foo":"bar",\n    "baz":"blah"\n}'
    """
    def replacer(match):
        s = match.group(0)
        if s[0] == '/': return ""
        return s
    return comments_re.sub(replacer, json_like)

def remove_trailing_commas(json_like):
    """
    Removes trailing commas from *json_like* and returns the result.  Example::
    
        >>> remove_trailing_commas('{"foo":"bar","baz":"blah",}')
        '{"foo":"bar","baz":"blah"}'
    """
    return trailing_commas_re.sub("}", json_like)

# These two functions together can be used like so:
with open('some_file.json') as f:
    almost_json = remove_comments(f.read()) # Remove comments
    proper_json = remove_trailing_commas(almost_json) # Remove trailing commas