Retrieve a file or a directory from git history at specific reversion

git-history-extract.py

#!/usr/bin/env python3
#
# Usage:
#   git-history-extract.py [-o <output_path>] <rev> [<path>]
#
# Retrieve a file or a directory from git history at specific reversion (tag or commit).
# The restored file(s) are saved into destination path.
#
# You must run it in a directory which is inside a git repository.
#
# Limitations:
# - Not support LFS
# - Not support symbolic link

import subprocess
import argparse
import os


def parse_arguments():
    parser = argparse.ArgumentParser(
        description='Retrieve a file or a directory from git history at specific reversion')
    parser.add_argument('-o', '--output-path',
                        help='Path to save restored file. Print file content at console if not specify.')
    parser.add_argument('rev',
                        help='Revision of the commit to retrieve, can be SHA1 or tag. Default to current directory.')
    parser.add_argument('path', nargs='?', default='./',
                        help='Path to the file or directory to retrieve')
    return parser.parse_args()


def save_file(output_path, file_path, content, executable):
    full_file_path = os.path.join(output_path, file_path)
    dir_name = os.path.dirname(full_file_path)
    print('Save {}'.format(full_file_path))
    os.makedirs(dir_name, exist_ok=True)
    with open(full_file_path, 'wb') as file:
        file.write(content)
    if executable:
        os.chmod(full_file_path, 0o755)


if __name__ == '__main__':

    args = parse_arguments()

    ls_tree = subprocess.check_output(
        ['git', 'ls-tree', '-rz', args.rev, args.path]).decode().split('\0')

    for line in ls_tree:
        if line:
            meta, file = line.split('\t')
            mode, _, object = meta.split(' ')
            content = subprocess.check_output(
                ['git', 'cat-file', 'blob', object])
            if args.output_path:
                save_file(args.output_path, file, content,
                        executable=(mode == '100755'))
            else:
                print('\n>>>>> {}'.format(file))
                try:
                    print(content.decode())
                except:
                    print('[ ... binary content ... ]')

output-example

# Example: output to console

$ git-history-extract.py v3

>>>>> 1658350664.jpg
[ ... binary content ... ]

>>>>> dir1.1/dir with space/name with space.x.y.txt
This file has LF ending.
LF

>>>>> dir1.1/script.sh
echo Runnable

>>>>> dir1.1/中文文件名.txt
This file has CRLF ending.
中文
hello



# Example: output to folder

$ git-history-extract.py v3 -o ../
Save ../1658350664.jpg
Save ../dir1.1/dir with space/name with space.x.y.txt
Save ../dir1.1/script.sh
Save ../dir1.1/中文文件名.txt

$ ls -l dir1.1
total 16
drwxr-xr-x  3 myid  staff  96 Jun 12 14:05 dir with space
-rwxr-xr-x  1 myid  staff  13 Jun 11 16:03 script.sh
-rw-r--r--  1 myid  staff  43 Jun 12 18:24 中文文件名.txt