Python function for retrieving all git commits (from "git log" in the current directory) as a useful data structure

example_output.json

[
  {
    "date": "Mon Jan 23 13:03:00 2017 -0800", 
    "message": "Recommended by Django https://docs.djangoproject.com/en/1.10/ref/databases/#setting-sql-mode", 
    "title": "SET sql_mode='STRICT_TRANS_TABLES';", 
    "hash": "a395f401c9a6d3e8602b80ecc729d736da871f46", 
    "author": "Simon Willison <[email protected]>"
  }, 
  {
    "date": "Mon Jan 23 11:37:38 2017 -0800", 
    "message": "Turns out django.utils.log.NullHandler was removed in Django 1.9\n\nhttp://stackoverflow.com/a/34349623", 
    "title": "Fixed logging NullHandler", 
    "hash": "0a1da86196cb72ead43cd8ac2d256a79d4f3bd4a", 
    "author": "Simon Willison <[email protected]>"
  }
]

get_commits.py

import subprocess
import re

leading_4_spaces = re.compile('^    ')


def get_commits():
    lines = subprocess.check_output(
        ['git', 'log'], stderr=subprocess.STDOUT
    ).split('\n')
    commits = []
    current_commit = {}

    def save_current_commit():
        title = current_commit['message'][0]
        message = current_commit['message'][1:]
        if message and message[0] == '':
            del message[0]
        current_commit['title'] = title
        current_commit['message'] = '\n'.join(message)
        commits.append(current_commit)

    for line in lines:
        if not line.startswith(' '):
            if line.startswith('commit '):
                if current_commit:
                    save_current_commit()
                    current_commit = {}
                current_commit['hash'] = line.split('commit ')[1]
            else:
                try:
                    key, value = line.split(':', 1)
                    current_commit[key.lower()] = value.strip()
                except ValueError:
                    pass
        else:
            current_commit.setdefault(
                'message', []
            ).append(leading_4_spaces.sub('', line))
    if current_commit:
        save_current_commit()
    return commits

# Here's how I used this to find the longest commits in a repo
commits = get_commits()
commits.sort(key = lambda c: len(c['message']), reverse=True)
for commit in commits[:30]:
    print commit['title']
    print commit['author']
    print commit['hash']
    print 'https://github.com/simonw/repo_name/commit/' + commit['hash']
    print

# Added decode to get this to work for me 

lines = subprocess.check_output(
       ['git', 'log'], stderr=subprocess.STDOUT
       ).decode("utf-8").split("\n")