使用pyspider爬取巨量淘宝MM图片

具体搭建步骤不再赘述，这里主要使用到了fakeagent，phantomjs和proxy

pyspider的爬取相当智能，在不能获取图片的时候会适当的暂停一段时间再试探性的爬取，配合fakeagent，proxypool和phantomjs，爬取成功率在90%以上。

代码是扒的别人的然后修改提高速度和成功率的，数据总量在百G左右，磁盘大的可以扒一扒。

代码如下：

#!/usr/bin/env python
# -*- encoding: utf-8 -*-
# Created on 2016-03-25 00:59:45
# Project: taobaomm

from pyspider.libs.base_handler import *
from fake_useragent import UserAgent
import base64
import requests
import random
import sys
reload(sys)
sys.setdefaultencoding('UTF-8') 

PAGE_START = 1
PAGE_END = 4301
DIR_PATH = '/root/images/tbmm'

class Handler(BaseHandler):

    r = requests.get(u'http://127.0.0.1:5010/get_all/')
    proxy = random.choice(eval(r.text))
    ua = UserAgent()

    crawl_config = {
        "proxy": proxy,
        "headers":{
            "User-Agent": ua.random
        },
    }

    def __init__(self):
        self.base_url = 'https://mm.taobao.com/json/request_top_list.htm?page='
        self.page_num = PAGE_START
        self.total_num = PAGE_END
        self.deal = Deal()

    def on_start(self):
        while self.page_num <= self.total_num:
            url = self.base_url + str(self.page_num)
            self.crawl(url, callback=self.index_page)
            self.page_num += 1

    def index_page(self, response):
        for each in response.doc('.lady-name').items():
            self.crawl(each.attr.href, callback=self.detail_page, fetch_type='js')

    def detail_page(self, response):
        domain = response.doc('.mm-p-domain-info li > span').text()
        if domain:
            page_url = 'https:' + domain
            self.crawl(page_url, callback=self.domain_page)

    def domain_page(self, response):
        name = base64.b64encode(response.doc('.mm-p-model-info-left-top dd > a').text())
        dir_path = self.deal.mkDir(name)
        brief = response.doc('.mm-aixiu-content').text()
        if dir_path:
            imgs = response.doc('.mm-aixiu-content img').items()
            count = 1
            self.deal.saveBrief(brief, dir_path, name)
            for img in imgs:
                url = img.attr.src
                if url:
                    extension = self.deal.getExtension(url)
                    file_name = name + str(count) + '.' + extension
                    count += 1
                    self.crawl(img.attr.src, callback=self.save_img,
                               save={'dir_path': dir_path, 'file_name': file_name})

    def save_img(self, response):
        content = response.content
        dir_path = response.save['dir_path']
        file_name = response.save['file_name']
        file_path = dir_path + '/' + file_name
        self.deal.saveImg(content, file_path)

import os

class Deal:
    def __init__(self):
        self.path = DIR_PATH
        if not self.path.endswith('/'):
            self.path = self.path + '/'
        if not os.path.exists(self.path):
            os.makedirs(self.path)

    def mkDir(self, path):
        path = path.strip()
        dir_path = self.path + path
        exists = os.path.exists(dir_path)
        if not exists:
            os.makedirs(dir_path)
            return dir_path
        else:
            return dir_path

    def saveImg(self, content, path):
        f = open(path, 'wb')
        f.write(content)
        f.close()

    def saveBrief(self, content, dir_path, name):
        file_name = dir_path + "/" + name + ".txt"
        f = open(file_name, "w+")
        f.write(content.encode('utf-8'))

    def getExtension(self, url):
        extension = url.split('.')[-1]
        return extension