end0tknr's kipple - 新web写経開発

http://d.hatena.ne.jp/end0tknr/ から移転しました

python で smb 接続し、dir以下を再帰探索

以下のように書くと、よさそう

#!/usr/local/bin/python
# -*- coding: utf-8 -*-

import datetime
import os
import sys
import platform
from smb.SMBConnection import SMBConnection
import time

conf = {"user_id":"ないしょ",
        "user_pw":"ないしょ",
        "remote_netbios":'CBBないしょ4252',
        "remote_ip":"ないしょ",
#        "remote_port":139,
        "remote_port":445,      # tera station?に対しては
        "is_direct_tcp": True,  #ここがTrue & port445がよさそう
        "remote_path":"ファイルサーバー",
        "sleep": 0.2}

def main():
    smb_conn = SMBConnection(conf["user_id"],
                             conf["user_pw"],
                             platform.uname().node, # client netbios
                             conf["remote_netbios"],
                             is_direct_tcp=conf["is_direct_tcp"])
    try:
        smb_conn.connect(conf["remote_ip"],conf["remote_port"])
    except Exception as e:
        print('Connect failed. Reason: %s', e)
        return False

    start_dir = ""
    if len(sys.argv) == 2:
        start_dir = sys.argv[1]
    else:
        print("USAGE:",sys.argv[0],"start_dir")
        return None

    files_summary = {}
    aggregate_dir( smb_conn, start_dir, files_summary)
    
    smb_conn.close()

    sorted_summary = sorted(files_summary.items(), key=lambda x:x[0])
    for summary in sorted_summary:
        yyyy,file_type = summary[0].split("\t")
        disp_str = "\t".join([str(yyyy),
                              file_type,
                              str(summary[1]["byte"]),
                              str(summary[1]["count"]) ])
        print( disp_str)
    

    

def aggregate_dir( smb_conn, org_dir, files_summary ):
    print("aggregating ...", org_dir)
    time.sleep(conf["sleep"])
    
    dir_or_files = smb_conn.listPath(conf["remote_path"],org_dir)

    for dir_or_file in dir_or_files:
        if dir_or_file.filename=="." or dir_or_file.filename=="..":
            continue
        
        if dir_or_file.isDirectory:
            next_dir = org_dir +"/"+ dir_or_file.filename
            aggregate_dir( smb_conn, next_dir, files_summary)
            continue

        last_unix_time = dir_or_file.last_write_time
        if last_unix_time < dir_or_file.last_access_time:
            last_unix_time = dir_or_file.last_access_time
            
        atime = date_str_from_unixtime(last_unix_time)
        file_ext = dir_or_file.filename.split(".")[-1].upper()
        summary_key = atime +"\t"+file_ext
        
        if (summary_key in files_summary) == False:
            files_summary[summary_key] = {"byte":0, "count":0}

        files_summary[summary_key]["count"] += 1
        files_summary[summary_key]["byte"]  += dir_or_file.file_size

        
def date_str_from_unixtime(org_val):
    dt = datetime.datetime.fromtimestamp(org_val)
#    print(dt.strftime('%Y-%m-%d %H:%M:%S'))
    return dt.strftime('%Y')
    
if __name__ == '__main__':
    main()