Line Length

linelength.py

#! /usr/bin/env python
from collections import defaultdict
from os import walk
from os.path import join
from subprocess import check_output


args = ['git', 'blame', '-p']
authors = defaultdict(list)

for root, dirs, files in walk('.'):
    python_files = (join(root, f) for f in files if f[-3:] == '.py')

    for p in python_files:
        author = None
        record = False
        blame = check_output(args + [p]) 
        lines = (line for line in blame.splitlines())

        for line in lines:

            if line[:7] == 'author ':
                author = line[7:]
                record = False
            if line[:9] == 'filename ':
                record = True
                try:
                    line = lines.next()
                except StopIteration:
                    break

            if record:
                try:
                    if line.strip() != '':
                        authors[author].append(len(line))
                    line = lines.next()
                except StopIteration:
                    break

fields = (
    'Author',
    'lines > 80',
    '% lines > 80 / lines written',
    '% lines > 80 / total lines',
    'longest line'
    )

print ','.join(fields)

total_lines = sum(len(nums) for nums in authors.itervalues())

for author, nums in authors.iteritems():
    count = 0.0

    for n in nums:
        if n > 80:
            count += 1

    print '%s, %s, %s, %s, %s' % (
            author,
            count,
            count / len(nums),
            count / total_lines,
            max(nums)
            )