문서의 전체 백업 및 증가분 백업

import hashlib                        #      ，       
import tarfile                           #      
from time import strftime      #      
import os                  
import pickle                        #pickle                 

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

    with open(fname, 'rb') as fobj:    #        ，      
        while True:
            data = fobj.read(4096)        #    4k  
            if not data:
                break
            m.update(data)                    #       ，   md5 

    return m.hexdigest()                  #     ，  md5   
  
def full_backup(src, dst, md5file):                #        
    fname = '%s_full_%s.tar.gz' % (os.path.basename(src), strftime('%Y%m%d'))              #     ：xxx_full_20190710.tar.gz
#         :/tmp/demp/backup/xxx_full_20190710.tar.gz
    fname = os.path.join(dst, fname)      

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

    #        md5 
    md5dict = {}                    #       md5 
#os.walk      [  ][    ][   ]
    for path, folders, files in os.walk(src):      
        for file in files:          
#                ，             
            key = os.path.join(path, file)      #           
            md5dict[key] = check_md5(key)    #  md5 

    #  md5        
#wb    ，       ，     
    with open(md5file, 'wb') as fobj:          
        pickle.dump(md5dict, fobj)      #  pickle.dump  

#        
#  ：        md5 ，       md5      ，       

def incr_backup(src, dst, md5file):            
    fname = '%s_incr_%s.tar.gz' % (os.path.basename(src), strftime('%Y%m%d'))      #xxx_incr_20190710.tar.gz
#/tmp/demo/backup/xxx_incr_20190710.tar.gz
    fname = os.path.join(dst, fname)  

    #           md5 
    md5dict = {}
    for path, folders, files in os.walk(src):
        for file in files:
            key = os.path.join(path, file)
            md5dict[key] = check_md5(key)

    #         md5 
    with open(md5file, 'rb') as fobj:
        old_md5 = pickle.load(fobj)

    #              ，     
    tar = tarfile.open(fname, 'w:gz')
    for key in md5dict:      #        md5 
#     ，old_md5.get(key)  ，     ，old_md5.get(key)   md5    ，       
        if old_md5.get(key) != md5dict[key]:      
            tar.add(key)
    tar.close()

    #     md5       ，         
    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'        #md5   
    if strftime('%a') == 'Mon':                #        
        full_backup(src, dst, md5file)
    else:
        incr_backup(src, dst, md5file)