Back-dating Git commits based on file modification dates#

1
-rw-r--r--   1 simon  wheel   1.5K Sep  4  1991 Anchor.h
2
-rw-r--r--   1 simon  wheel   8.1K Sep  4  1991 Anchor.m
3
-rw-r--r--   1 simon  wheel   3.6K Dec  1  1993 Bugs.html
4
-rw-r--r--   1 simon  wheel   7.7K Jun  2  1994 Features.html

1
import os
2
import sys
3
import argparse
4
from datetime import datetime
5
from collections import defaultdict
6
import subprocess
7

8

9
def get_file_date(file_path):
10
    return datetime.fromtimestamp(os.path.getmtime(file_path)).date()
11

12

13
def group_files_by_date(directory):
14
    # I decided to group files edited on the same day in a single commit:
15
    grouped_files = defaultdict(list)
16
    for root, _, files in os.walk(directory):
17
        for file in files:
18
            if file.startswith("."):
19
                continue
20
            file_path = os.path.join(root, file)
21
            # I added this hack to skip the .git directory
22
            if ".git" in file_path:
23
                continue
24
            file_date = get_file_date(file_path)
25
            grouped_files[file_date].append(file_path)
26
    return grouped_files
27

28

29
def commit_files(files, date, author, author_email):
30
    print(files)
31
    for file in files:
32
        subprocess.run(["git", "add", file], check=True)
33

34
    commit_date = date.strftime("%Y-%m-%d 00:00:00")
35
    commit_message = f"Adding files from {date}"
36

37
    env = os.environ.copy()
38
    # Here'This is the most important bit: these environment variables are used
39
    # by Git to set the author and committer dates and names
40
    env["GIT_AUTHOR_DATE"] = commit_date
41
    env["GIT_COMMITTER_DATE"] = commit_date
42
    env["GIT_AUTHOR_NAME"] = author
43
    env["GIT_AUTHOR_EMAIL"] = author_email
44
    env["GIT_COMMITTER_NAME"] = author
45
    env["GIT_COMMITTER_EMAIL"] = author_email
46

47
    result = subprocess.run(
48
        ["git", "commit", "-m", commit_message],
49
        capture_output=True,
50
        text=True,
51
        check=True,
52
        env=env,
53
    )
54

55
    commit_hash = result.stdout.split()[1]
56
    return commit_hash
57

58

59
def main():
60
    parser = argparse.ArgumentParser(
61
        description="Populate Git repo with historical commits"
62
    )
63
    parser.add_argument("directory", help="Directory containing the files")
64
    parser.add_argument("--author", required=True, help="Author of the commits")
65
    # I added this option by hand:
66
    parser.add_argument("--email", required=True, help="Email of author")
67
    args = parser.parse_args()
68

69
    os.chdir(args.directory)
70

71
    if not os.path.exists(".git"):
72
        print(
73
            "Error: No .git directory found. Please initialize a Git repository first."
74
        )
75
        sys.exit(1)
76

77
    grouped_files = group_files_by_date(args.directory)
78

79
    for date, files in sorted(grouped_files.items()):
80
        commit_hash = commit_files(files, date, args.author, args.email)
81
        print(f"Commit: {commit_hash}, Date: {date}")
82

83

84
if __name__ == "__main__":
85
    main()