用Python 爬虫爬取贴吧图片

版权声明:本文为博主原创文章,未经博主允许不得转载。 https://blog.csdn.net/u012705410/article/details/47685417

之前一直在看机器学习,遇到了一些需要爬取数据的内容,于是稍微看了看Python爬虫,在此适当做一个记录。我也没有深入研究爬虫,大部分均是参考了网上的资源。

先推荐两个Python爬虫的教程,网址分别是http://cuiqingcai.com/1052.htmlhttp://ddswhu.com/2015/03/25/python-downloadhelper-premium/ ,我就是看这两个当做基本的入门。

有兴趣才有动力,既然学了爬虫,那就先爬取美女照片吧。当然,这里先并不是爬取好友的照片了,比如QQ空间、人人之类的,因为这些还涉及模拟登录之类的操作。为了方便,还是爬取贴吧里的图片。主要是参考了Github上这个项目:https://github.com/Yixiaohan/show-me-the-code (Python 练习册,每天一个小项目),感觉还是很不错的。

先来看上面那个项目里的第13题,要求是爬取这个网址中的图片:http://tieba.baidu.com/p/2166231880,参考了网友renzongxian 的代码(使用版本为Python 3.4),如下:

# 爬取贴吧图片,网址:http://tieba.baidu.com/p/2166231880

import urllib.request
import re
import os


def fetch_pictures(url):
    html_content = urllib.request.urlopen(url).read()
    r = re.compile('<img pic_type="0" class="BDE_Image" src="(.*?)"')
    picture_url_list = r.findall(html_content.decode('utf-8'))

    os.mkdir('pictures')
    os.chdir(os.path.join(os.getcwd(), 'pictures'))
    for i in range(len(picture_url_list)):
        picture_name = str(i) + '.jpg'
        try:
            urllib.request.urlretrieve(picture_url_list[i], picture_name)
            print("Success to download " + picture_url_list[i])
        except:
            print("Fail to download " + picture_url_list[i])

if __name__ == '__main__':
    fetch_pictures("http://tieba.baidu.com/p/2166231880")

运行之后,会生成一个名为pictures的文件夹,自动下载图片并存放进去(图片就不截图了)。

为了测试该程序,我看了另一个Python爬虫的代码,准备把网址替换一下,即爬取这个网址中的图片:http://tieba.baidu.com/p/2460150866 ,只需要将上述代码稍作修改(把网址换一下,文件夹改一下)即可,如下

# 爬取贴吧图片,网址:http://tieba.baidu.com/p/2460150866

import urllib.request
import re
import os


def fetch_pictures(url):
    html_content = urllib.request.urlopen(url).read()
    r = re.compile('<img pic_type="0" class="BDE_Image" src="(.*?)"')
    picture_url_list = r.findall(html_content.decode('utf-8'))

    os.mkdir('photos')
    os.chdir(os.path.join(os.getcwd(), 'photos'))
    for i in range(len(picture_url_list)):
        picture_name = str(i) + '.jpg'
        try:
            urllib.request.urlretrieve(picture_url_list[i], picture_name)
            print("Success to download " + picture_url_list[i])
        except:
            print("Fail to download " + picture_url_list[i])

if __name__ == '__main__':
    fetch_pictures("http://tieba.baidu.com/p/2460150866")

爬取的图片会存放在photos文件夹中,截图如下:
p2

好的,那下面就来爬取一组明星的图片吧,比如刘诗诗,网址如下:http://tieba.baidu.com/p/2854146750 ,如果仍然直接把代码中的网址和文件夹名改一下,会发现并没有下载成功。所以,编程还是得老老实实看代码啊,光想着直接换个网址实在太偷懒(当然,想法可以有,这样才可以写出更智能、更进步的程序嘛)。

代码在抓取网页之后,利用正则表达式找到了图片的网址(URL),即关键在代码的这一行

r = re.compile('<img pic_type="0" class="BDE_Image" src="(.*?)"')

上面那两个例子之所以会成功,是因为网页源代码中(可以用谷歌浏览器查看)图片的网址确实是这个形式的,所以匹配到了。而刘诗诗照片的网址呢,查看一下,是这样的:
p2

也就是里面img标签里的代码形式是不一样的,不是

<img pic_type="0" class="BDE_Image" src="(.*?)"

而是

<img class="BDE_Image" src="(.*?)" pic_ext="jpeg"  pic_type="0"

所以要把正则表达式的内容修改一下,整体代码如下:

# 爬取贴吧刘诗诗的图片,网址:http://tieba.baidu.com/p/2854146750

import urllib.request
import re
import os


def fetch_pictures(url):
    html_content = urllib.request.urlopen(url).read()
    r = re.compile('<img class="BDE_Image" src="(.*?)" pic_ext="jpeg"  pic_type="0"')
    picture_url_list = r.findall(html_content.decode('utf-8'))

    os.mkdir('liushishi')
    os.chdir(os.path.join(os.getcwd(), 'liushishi'))
    for i in range(len(picture_url_list)):
        picture_name = str(i) + '.jpg'
        try:
            urllib.request.urlretrieve(picture_url_list[i], picture_name)
            print("Success to download " + picture_url_list[i])
        except:
            print("Fail to download " + picture_url_list[i])

if __name__ == '__main__':
    fetch_pictures("http://tieba.baidu.com/p/2854146750")

运行后,爬取成功,截图如下:
p3

嗯,这种用代码自动下载的感觉就是爽!(哈哈,逃)

所以下载图片的关键在于正则表达式的匹配。下面再举两个例子。

先来一个简单的。网址是:http://tieba.baidu.com/p/3884688092 ,代码如下:

# 爬取贴吧图片,网址:http://tieba.baidu.com/p/3884688092

import urllib.request
import re
import os


def fetch_pictures(url):
    html_content = urllib.request.urlopen(url).read()
    r = re.compile('<img class="BDE_Image" pic_type="\d" width="\d\d\d" height="\d\d\d" src="(.*?)"')
    picture_url_list = r.findall(html_content.decode('utf-8'))

    os.mkdir('test')
    os.chdir(os.path.join(os.getcwd(), 'test'))
    for i in range(len(picture_url_list)):
        picture_name = str(i) + '.jpg'
        try:
            urllib.request.urlretrieve(picture_url_list[i], picture_name)
            print("Success to download " + picture_url_list[i])
        except:
            print("Fail to download " + picture_url_list[i])

if __name__ == '__main__':
    fetch_pictures("http://tieba.baidu.com/p/3884688092")

关键仍在于正则表达式那一行:

r = re.compile('<img class="BDE_Image" pic_type="\d" width="\d\d\d" height="\d\d\d" src="(.*?)"')

不过我写的匹配比较拙劣,也就爬取了前4张图片而已,有待完善。

再来一个自己喜欢的。之前看到了一个人手绘龙珠的图片,就想爬取下来,网址在:http://tieba.baidu.com/p/3720487356 ,代码如下:

# 爬取贴吧龙珠图片,网址:http://tieba.baidu.com/p/3720487356

import urllib.request
import re
import os


def fetch_pictures(url):
    html_content = urllib.request.urlopen(url).read()
    r = re.compile('<img class="BDE_Image" pic_type="0" width="\d\d\d" height="\d\d\d" src="(.*?)"')
    picture_url_list = r.findall(html_content.decode('utf-8'))

    os.mkdir('longzhu')
    os.chdir(os.path.join(os.getcwd(), 'longzhu'))
    for i in range(len(picture_url_list)):
        picture_name = str(i) + '.jpg'
        try:
            urllib.request.urlretrieve(picture_url_list[i], picture_name)
            print("Success to download " + picture_url_list[i])
        except:
            print("Fail to download " + picture_url_list[i])

if __name__ == '__main__':
    fetch_pictures("http://tieba.baidu.com/p/3720487356")

效果也不是很好,只是爬取了首页的5张图而已:
p4

以上代码均是用Python 3.4写的。不过,用Python 2爬虫也很方便,于是也参考了网上Python 2爬虫的代码。

先看一个网上的示例(不好意思,又是爬取美女图片,网上爬美女程序太多),是用Python 2.7写的:

#!/bin/env python
# -*- coding: utf-8 -*- 
# 爬取网上图片,网址:http://jandan.net/ooxx

#导入模块
import urllib2
import re
import os
import glob

#设定抓取页数
page_amount = 2

#抓取首页的html代码   
def get_page(url):
    req = urllib2.Request(url)
    req.add_header('User-Agent','Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/40.0.2214.111 Safari/537.36')      #缺省部分填上浏览器字符串
    response = urllib2.urlopen(req)
    html = response.read().decode('utf-8')
    return html

#抓取图片
def read_image(url):
    req = urllib2.Request(url)
    req.add_header('User-Agent','Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/40.0.2214.111 Safari/537.36')      #缺省部分填上浏览器字符串
    response = urllib2.urlopen(url)
    html = response.read()
    return html

#得到当前的最新页面数,从这个页面开始倒着爬,因为用了这个脚本以后以前的图可能已经看过了
def get_current_page_number(html):
    match = re.search(r'<span class="current-comment-page">\[(.*)\]</span>',html)
    return match.group(1)

#得到图片列表
def get_picturs_url_list(url):
    html = get_page(url)
    l = re.findall(r'<p><img src="http://.*.sinaimg.cn/mw600/.*jpg" /></p>',html)
    result = []
    for string in l:
        src = re.search(r'"(.*)"',string)
        result.append(str(src.group(1)))    #解决Unicode编码开头问题,有空好好补下编码和字符规范
    return result

#下载图片并存储到本地文件夹
def image_save(url,number):
    number = str(number)
    print '正在抓取第',number,'张'
    filename = number + '.jpg'
    with open(filename,'wb') as fp:
        img = read_image(url)
        fp.write(img)

#准备存放图片的文件夹,并进入到指定路径
def floder_prepare(floder):
    a = glob.glob('*')
    if floder not in a:
        os.mkdir(floder)
    os.chdir(floder)

#主函数
def main():
    html = get_page('http://jandan.net/ooxx')
    number = int(get_current_page_number(html))
    l = []
    amount = 0
    for n in range(0,page_amount):
        url = 'http://jandan.net/ooxx/page-' + str(number-n) + '#comments'
        l += get_picturs_url_list(url) 
    floder_prepare('picture')
    for url in l:
        amount += 1
        image_save(url,amount)

if __name__ == '__main__':
    main()
    print '全部抓完啦'

上面的程序也是自动下载图片,并自动建立了一个pciture文件夹存放(图片就不截图了)。于是我就把它适当修改用来重新爬取上面的龙珠图片,代码如下:

#!/bin/env python
# -*- coding: utf-8 -*- 
# 爬取网上图片,网址:http://tieba.baidu.com/p/3720487356

#导入模块
import urllib2
import re
import os
import glob

#设定抓取页数
page_amount = 5

#抓取首页的html代码   
def get_page(url):
    req = urllib2.Request(url)
    req.add_header('User-Agent','Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/40.0.2214.111 Safari/537.36')      #缺省部分填上浏览器字符串
    response = urllib2.urlopen(req)
    html = response.read().decode('utf-8')
    return html

#抓取图片
def read_image(url):
    req = urllib2.Request(url)
    req.add_header('User-Agent','Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/40.0.2214.111 Safari/537.36')      #缺省部分填上浏览器字符串
    response = urllib2.urlopen(url)
    html = response.read()
    return html


#得到图片列表
def get_picturs_url_list(url):
    html = get_page(url)
    l = re.findall(r'<img class="BDE_Image" pic_type="0" width="\d\d\d" height="\d\d\d" src="(.*?)"',html)
    return l

#下载图片并存储到本地文件夹
def image_save(url,number):
    number = str(number)
    print '正在抓取第',number,'张'
    filename = number + '.jpg'
    with open(filename,'wb') as fp:
        img = read_image(url)
        fp.write(img)

#准备存放图片的文件夹,并进入到指定路径
def floder_prepare(floder):
    a = glob.glob('*')
    if floder not in a:
        os.mkdir(floder)
    os.chdir(floder)

#主函数
def main():
    number = 5
    l = []
    amount = 0
    for n in range(0,page_amount):
        url = 'http://tieba.baidu.com/p/3720487356?pn=' + str(number-n)
        l += get_picturs_url_list(url) 
    floder_prepare('pics_longzhu')
    for url in l:
        amount += 1
        image_save(url,amount)

if __name__ == '__main__':
    main()
    print '全部抓完啦'

由于这个帖子目前总共有5页,所以就设定了爬取页数,然后从最后一页倒着逐页全部爬取下来了,截图如下:
p5

嗯,果然效果好了不少。当然,用Python 2爬取图片的代码还有很多,以Github上那个项目里的13题为例,最后再贴两个代码以供交流学习。

代码(1)

# -*- coding: utf-8 -*-
# 爬取贴吧图片,网址:http://tieba.baidu.com/p/2166231880

import urllib2
from HTMLParser import HTMLParser
from traceback import print_exc
from sys import stderr

class _DeHTMLParser(HTMLParser):
    '''
    利用HTMLParse来解析网页元素
    '''
    def __init__(self):
        HTMLParser.__init__(self)
        self.img_links = []

    def handle_starttag(self, tag, attrs):
        if tag == 'img':
            # print(attrs)
            try:
                if ('pic_type','0') in attrs:
                    for name, value in attrs:
                        if name == 'src':
                            self.img_links.append(value)
            except Exception as e:
                print(e)


        return self.img_links


def dehtml(text):
    try:
        parser = _DeHTMLParser()
        parser.feed(text)
        parser.close()
        return parser.img_links
    except:
        print_exc(file=stderr)
        return text


def main():
    html = urllib2.urlopen('http://tieba.baidu.com/p/2166231880')
    content = html.read()
    print(dehtml(content))
    i = 0
    for img_list in dehtml(content):
        img_content = urllib2.urlopen(img_list).read()
        path_name = str(i)+'.jpg'
        with open(path_name,'wb') as f:
            f.write(img_content)
        i+=1

if __name__ == '__main__':
    main()

代码(2)

# coding:utf-8
# 爬取贴吧图片,网址:http://tieba.baidu.com/p/2166231880

import requests, re, os

url = 'http://tieba.baidu.com/p/2166231880'

header = {
        'Accept': '*/*',
        'Accept-Encoding':'gzip,deflate,sdch',
        'Accept-Language':'zh-CN,zh;q=0.8',
        'Connection':'keep-alive'
        }
html = requests.get(url,headers = header)

data = html.content.decode('utf-8')
find = re.compile(r'<img pic_type="0" class="BDE_Image" src="(.*?).jpg" bdwater')
result = find.findall(data)

for img_url in result:
    name = img_url.split('/')[-1]
    img_url = img_url+'.jpg'
    html = requests.get(img_url,headers = header)
    im = html.content
    with open(name+'.jpg','wb')as f:
        f.write(im)

猜你喜欢

转载自blog.csdn.net/u012705410/article/details/47685417