Git hook to insert word count into commit message.

prepare-commit-msg.py

#!/usr/bin/python3
#
# A hook script to prepare the commit log message.
#
# Called by git-commit with the name of the file that has the commit
# message, followed by the description of the commit message's source.
# The hook's purpose is to edit the commit message file. If the hook
# fails with a non-zero status, the commit is aborted.

import glob
import os.path
import sys

# Count the number of words in an English language text.
def count_words(chapter_path):
   word_count = 0
   chapters = glob.glob(os.path.join(chapter_path, '*.txt'))

   for chapter in chapters:
      with open(chapter, mode='r', encoding='utf-8') as chapter_file:
         chapter_text = chapter_file.read()
         word_count += len(chapter_text.split())

   return word_count

# Get the top-level repository path.
def get_project_path():
   # First argv entry is path to this hook file.
   hook_file = sys.argv[0]
   from os.path import dirname
   project_path = dirname(dirname(dirname(hook_file)))
   return project_path

# Read the commit message.
def get_commit_message(commit_filename):
   with open(commit_filename, mode='r', encoding='utf-8') as commit_file:
      lines = commit_file.readlines()
   return lines

# Write the new commit message contents.
def set_commit_message(commit_filename, lines):
   with open(commit_filename, mode='w', encoding='utf-8') as commit_file:
      commit_file.writelines(lines)

def inject_stats_message(commit_filename, stats):
   # Read the lines of the raw commit message.
   lines = get_commit_message(commit_filename)
   # NB: you can skip quotes around a dictionary key name within a
   # replacement field. Cf. PEP 3101: Advanced String Formatting.
   stat_string = "words: {stats[word_count]}".format(stats=stats)
   # The text to prepend to the commit message.
   message = ["\n", "\n", stat_string]
   message.reverse()
   # Inject our message text into the commit message.
   for line in message:
      lines.insert(0, line)
   # Store the updated commit message.
   set_commit_message(commit_filename, lines)

def get_statistics(chapter_path):
   # Add the utils/ directory to the python path so we can import the
   # word_count utilities.
   python_path = os.path.join(get_project_path(), 'utils')
   sys.path.append(python_path)
   import word_count
   # Data to be injected into the commit message.
   stats = {}
   # Get total word count for all files in chapters directory.
   stats['word_count'] = word_count.count_words(chapter_path)
   return stats

def main():
   # The only script argument is the name of the file where the commit
   # message is located.
   commit_filename = sys.argv[1]
   # Get the path to the chapters/ directory.
   chapter_path = os.path.realpath(os.path.join(get_project_path(), 'chapters'))
   # Collect some statistics about the text corpus.
   stats = get_statistics(chapter_path)
   # Inject the statistics into the commit message.
   inject_stats_message(commit_filename, stats)

if __name__ == '__main__':
   main()