git-repo-iter

#!/usr/bin/env python3
#
# git-repo-iter is a command line tool for collecting data from a git repository
# over time.
#
#
# Usage
#
#   git-repo-iter cmd [FILE]
#
#
# Input
# - CSV (without header) through stdin where the first column is a commit hash
# - Two positional arguments:
#   - (cmd) a shell command to run for each commit
#   - (FILE) the path to a file which contains previous results (optional)
#     - the contents of this file must be a CSV where the first column is a
#       commit hash
#
# Output
# - CSV with columns from stdin along with a new column containing the new
#   results from running cmd
#
#
# Example usage (Number of Java files in codebase by day)
#
#   git log --pretty='%H,%aI' --no-merges | awk -F, '!a[substr($2, 1, 10)]++' | git-repo-iter 'fd -e java | wc -l'

import csv
import os
import subprocess
import sys


cmd = sys.argv[1]
results_file = sys.argv[2] if len(sys.argv) > 2 else None

if results_file is not None and os.path.isfile(results_file):
    with open(results_file, newline='') as file:
        completed = { row[0] for row in csv.reader(file) }
else:
    completed = {}

writer = csv.writer(sys.stdout)

for row in csv.reader(sys.stdin):
    hash = row[0]
    if hash not in completed:
        subprocess.run(['git', 'checkout', hash, '--quiet'])

        output = subprocess.check_output(cmd, shell=True, text=True)

        row.append(output.strip())

        writer.writerow(row)
        sys.stdout.flush()