python实现完全备份和增量备份

import hashlib
import tarfile
from time import strftime
import os

import pickle


def check_md5(fname):
    m = hashlib.md5()

    with open(fname, 'rb') as fobj:
        while True:
            data = fobj.read(4096)
            if not data:
                break
            m.update(data)

    return m.hexdigest()

def full_backup(src, dst, md5file):
    fname = '%s_full_%s.tar.gz' % (os.path.basename(src), strftime('%Y%m%d'))
    fname = os.path.join(dst, fname)

    # 打压缩包
    tar = tarfile.open(fname, 'w:gz')
    tar.add(src)
    tar.close()

    # 计算每个文件的md5值
    md5dict = {}
    for path, folders, files in os.walk(src):
        for file in files:
            key = os.path.join(path, file)
            md5dict[key] = check_md5(key)

    # 把md5值字典保存到文件
    with open(md5file, 'wb') as fobj:
        pickle.dump(md5dict, fobj)

def incr_backup(src, dst, md5file):
    fname = '%s_incr_%s.tar.gz' % (os.path.basename(src), strftime('%Y%m%d'))
    fname = os.path.join(dst, fname)

    # 计算当前目录下文件的md5值
    md5dict = {}
    for path, folders, files in os.walk(src):
        for file in files:
            key = os.path.join(path, file)
            md5dict[key] = check_md5(key)

    # 取出前一天的文件md5值
    with open(md5file, 'rb') as fobj:
        old_md5 = pickle.load(fobj)

    # 找出变化的文件和新增的文件，把它们压缩
    tar = tarfile.open(fname, 'w:gz')
    for key in md5dict:
        if old_md5.get(key) != md5dict[key]:
            tar.add(key)
    tar.close()

    # 把当前的md5字典写到文件中，以便下一次比较使用
    with open(md5file, 'wb') as fobj:
        pickle.dump(md5dict, fobj)

if __name__ == '__main__':
    src = '/tmp/demo/security'
    dst = '/tmp/demo/backup'
    md5file = '/tmp/demo/backup/md5.data'
    if strftime('%a') == 'Mon':
        full_backup(src, dst, md5file)
    else:
        incr_backup(src, dst, md5file)