Determines the proportion of python code that is comments. To use, type `python comment_proportions.py file1.py [file2.py] [file3.py] [...]` in your terminal.

comment_proportions.py

from pygments import highlight
from pygments.lexers import get_lexer_by_name
from pygments.formatters import BBCodeFormatter

import re
import sys


def get_comments(code):
    """ Extracts comments and docstrings from python code.
        Uses pygments to mark comments and docstrings and
        then regular expressions to extract them.
        Args:
            - code (str) : the code to extract comments from
        Return:
            - comments (list<str>) : the comments and docstrings
                extracted from the code
    """
    comment_regex = re.compile(
        r"(?:\[color=#408080\]\[i\])(.*)(?:\[/i\]\[/color\])")
    docstring_regex = re.compile(
        r"(?:\[color=#BA2121\]\[i\])(.*?)(?:\[/i\]\[/color\])", re.DOTALL)

    lexer = get_lexer_by_name("python", stripall=True)
    formatter = BBCodeFormatter(linenos=False)
    result = highlight(code, lexer, formatter)

    comments = comment_regex.findall(result)
    docstrings = docstring_regex.findall(result)
    return comments + docstrings


def compute_comment_stats(file_list):
    """ Finds the total number of bytes, the number of bytes of comments,
        and computes the ratio of the two, for each file and in total.
        Args:
            - file_list (list<str>): the list of files to compute stats for
        Return:
            - stats (dict): a dict with filenames as keys and dicts of
                statistics as values.
    """
    stats = {}
    total_len = 0
    total_comment_len = 0

    for file in file_list:
        with open(file, 'r') as f:
            source_code = f.read()
            total_len += len(source_code)

            comments = get_comments(source_code)
            comment_len = sum([len(comment) for comment in comments])
            total_comment_len += comment_len

            stats[file] = {
                "comment bytes": comment_len,
                "total bytes": len(source_code),
                "ratio": comment_len / len(source_code)}

    stats["total"] = {
        "comment bytes": total_comment_len,
        "total bytes": total_len,
        "ratio": total_comment_len / total_len}

    return stats


if __name__ == '__main__':
    stats = compute_comment_stats(sys.argv[1:])
    print(stats)

This code is released under the .

If you liked this, check out cloc, it's a fantastic counter for lines of code.