targz2git.py

import glob
import os
import tarfile
from datetime import datetime
from shutil import copytree, rmtree
from tarfile import ReadError

from git import *

repo = Repo(os.path.dirname(__file__))


def copy_and_overwrite(from_path, to_path):
    if os.path.exists(to_path):
        rmtree(to_path)
    copytree(from_path, to_path)


data_dir = os.path.join(os.path.dirname(__file__), 'data')
xml_dir = os.path.join(os.path.dirname(__file__), 'xml')

services = [
    'allegro',
]

for service in services:

    os.chdir(os.path.join(data_dir, service))

    archive_names = glob.glob("*.tar.bz2")
    archive_names.sort()

    for archive_name in archive_names:
        service_xml_dir = os.path.join(xml_dir, service)
        unpacked_data_dir = archive_name.replace('.tar.bz2', '')
        commit_date = datetime.strptime(unpacked_data_dir, '%Y%m%d%H%M%S')
        commit_date_str = commit_date.strftime("%Y-%m-%d %H:%M:%S")

        try:
            tar = tarfile.open(archive_name)
            tar.extractall()
            tar.close()
        except ReadError, e:
            with open(os.path.join(service_xml_dir, 'errors.log'), 'a') as f:
                msg = "File %s: %s" % (archive_name, str(e))
                print >> f, msg,
            commit_msg = msg
        else:
            copy_and_overwrite(unpacked_data_dir, service_xml_dir)
            rmtree(unpacked_data_dir)
            commit_msg = "Service: %s\nDate: %s" % (service, commit_date_str)

        repo.index.add([service_xml_dir])

        os.environ["GIT_AUTHOR_DATE"] = os.environ["GIT_COMMITTER_DATE"] = commit_date_str


        repo.index.commit(commit_msg)