历经几次换域名,以及从GAE-wordpress造成不少历史遗留的图片分布在不同域名或gae目录下。

主要分布
http://oldres.fengsage.com
http://www.fengsage.com
http://fensageblog.appspot.com

原来使用代码实现图片代理访问。但自己越看越不爽。索性写个脚本一次性把原来的图片下载到wordpress本地。并且更新数据库。主要是圣诞节闲的蛋疼。

代码有不少问题：

1. 不用php,因为懒得看wordpress插件制作。
2. 因为是python脚本,所以必须是有执行python的权限
3. 没有生成wordpress media记录。不能后台编辑多媒体

wordpress插件实现：

http://wordpress.org/extend/plugins/velvet-blues-update-urls/  等发现的时候脚本已经写好了。悲哀～～

脚本笔记简单。分享下。

#!/usr/bin/env python
# --*-- encoding:utf-8 --*--
'''
Created on 2011-12-25
@author: fred 
'''

import MySQLdb
import re
import urllib
import time
import os

WP_POST_CONTENT = u"wp_posts"
SIMPLE_IMG_URL = r'' 
NEW_IMG_URL = u''
DOWNLOAD_PATH = u'wp-content/uploads/%s'%(time.strftime('%m/%d', time.localtime(time.time())))

REPLACE_LIST = ['http://oldres.fengsage.com','../media','http://www.fengsage.com']

conn = MySQLdb.connect(host="localhost",
                         user="root",
                         passwd="zhufeng",
                         db="wordpress",
                         charset='utf8')


def get_post_list():
    cursor = conn.cursor()
    cursor.execute("SET NAMES utf8")
    cursor.execute("select * from %s"%WP_POST_CONTENT);
    return cursor.fetchall()

def find_img_urls(str):
    pattern = re.compile(SIMPLE_IMG_URL) 
    match = pattern.findall(str)
    if match:
        return match
    return None

def process_imgs(img_urls,replace):
    result = {}
    for img in img_urls:
        for rp in replace:
            if img.find(rp)>=0:
                print "\tdownload picture"
                new_img_ath = download_img(img)
                result.update({img:new_img_ath})
    return result
                
def download_img(img_url):
    IMG_NAME_REG = r'([^./]*).(png|jpg|gif|jpeg|bmp)'
    m = re.findall(IMG_NAME_REG, img_url)
    if m:
        name,suf = m[0]
        if os.path.isdir(DOWNLOAD_PATH) == False:
            os.makedirs(DOWNLOAD_PATH)
        file_path = u'%s/%s.%s'%(DOWNLOAD_PATH,name,suf)
        downloaded_image = file(file_path.encode('utf-8'), "wb")
        try:
            image_on_web = urllib.urlopen(img_url.encode('utf-8'))
            while True:
                buf = image_on_web.read(65536)
                if len(buf) == 0:
                    break
                downloaded_image.write(buf)
            downloaded_image.close()
            image_on_web.close()
            return file_path
        except:
            print '\tdownload img failture:%s'%img_url

def refresh_content(post_content,result):
    for k in result.keys():
        post_content = post_content.replace(k,'%s/%s'%(NEW_IMG_URL,result[k]))
    return post_content

def update_content(post_id,post_content):
    sql = "update "+WP_POST_CONTENT+" set post_content=%s where ID=%s"
    cursor = conn.cursor()
    cursor.execute("SET NAMES utf8")
    cursor.execute(sql,(post_content,post_id))
    
    
def go():
    post_list = get_post_list()
    for post in post_list:
        post_id = post[0]
        post_content = post[4]
        post_title = post[5]
        if post_id and post_content and post_title:
            print "process article ID:%s"%(post_id)
            img_urls = find_img_urls(post_content)
            if img_urls:
                result = process_imgs(img_urls,REPLACE_LIST)
                if result and len(result)>0:
                    post_content = refresh_content(post_content,result)
                    try:
                        update_content(post_id,post_content)
                    except Exception, e:
                        print 'update database failture:%s'%e
                        

if __name__ == '__main__': 
    print go()