零python基础--爬虫实践总结

# encoding=utf8 
#设置编辑源py文件的编码格式为utf8
import requests, sys, chardet, os, time, random, time
from bs4 import BeautifulSoup

reload(sys)  #必须要重新加载
sys.setdefaultencoding("utf8")

print sys.getdefaultencoding(), sys.getfilesystemencoding()  # utf8 mbcs:MBCS(Multi-ByteChactacterSystem,即多字节字符系统)它是编码的一种类型,而不是某个特定编码的名称
path = os.getcwd() #获取当前文件所在目录
newPath = os.path.join(path, "Computer")
if not os.path.isdir(newPath):
    os.mkdir(newPath) #新建文件夹
destFile = unicode(newPath + "/题目.docx","utf-8) #存为word也可以，不过后续用office编辑后，保存的时候总需要另存为；用unicode()后，文件名取中文名不会变成乱码

#最常见的模拟浏览器，伪装headers
headers = {
    ‘User-Agent‘: ‘Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_2) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/47.0.2526.80 Safari/537.36‘
}


def downLoadHtml(url):
    html = requests.get(url, headers=headers)
    content = html.content
    contentEn = chardet.detect(content).get("encoding", "utf-8")
    # print contentEn  #GB2312
    try:
        tranCon = content.decode(contentEn).encode(sys.getdefaultencoding())#转换网页内容编码格式；消除中文乱码
    except Exception:
        return content #用了编码转换，为什么还是存在少量页面异常？
    # print tranCon
    else:
        return tranCon


def parseHtml(url):
    # print url, "now"
    content = downLoadHtml(url)
    contentEn = chardet.detect(content).get("encoding", "utf-8")
    soup = BeautifulSoup(content, "html.parser")  # soup.name  [document] BeautifulSoup 对象表示的是一个文档的全部内容
    # 查找下一页url
    theUL = soup.find("ul", {"class": "con_updown"})
    theLi = theUL.find("li")
    href = theLi.find("a").get("href")
    preUrl = None
    if href:
        print href, "next"
        preUrl = href

    # 查找所需内容
    topics = []
    try:
        divCon = soup.find("div", attrs={"class": "con_nr"})
        if divCon:
            subjects = divCon.find_all("p")  # __len__属性不是整数，而是：method-wrapper ‘__len__‘ of ResultSet object
            index = 0 #借助index标识查找第几个，还有别的方式？
            for res in subjects:
                #跳过不想要的导读行内容
                if index == 0 and res.string == "【导读】":
                    index = 1  # 跳出循环也要加1
                    continue  # 跳过 导读
                topic = res.string  # res有子标签及文本，就会返回None
                if topic:
                    #按需要，只留下纯文本，保存到文件
                    try:
                        parsed = topic.decode(contentEn).encode("utf8")
                    except Exception:
                        topics.append("本页面解码有误，请自行查看: " + url + "\n")  # ‘%d‘ %index str(index) 数字转字符串
                        break
                    else:
                        topics.append(parsed + "\n")
                index = index + 1
            topics.append("\n")
        else:
            topics.append("本页面查找试题有误，请自行查看: " + url + "\n")
    except Exception:
        topics.append("本页面解析有误，请自行查看: " + url + "\n")

    fp = open(destFile, ‘a‘)  # a追加写
    fp.writelines(topics)
    fp.close()
    return preUrl


#执行.py文件的入口
if __name__ == ‘__main__‘:
    i = 0 #记录处理了多少页面
    next = "http://xxxxx/1.html" #起始页面
    print "start time:", time.strftime(‘%Y-%m-%d %H:%M:%S‘, time.localtime(time.time())) #打印时间，看跑了多久
    print next, "start"
    while next and i < 1000:
        next = parseHtml(next)
        i = i + 1
        #sTime = random.randint(3, 8) #随机整数 [3,8)
        #time.sleep(sTime)  # 休息：防反爬
    print "end time:", time.strftime(‘%Y-%m-%d %H:%M:%S‘, time.localtime(time.time()))
    print "i =", i, "url:", next
    fp = open(destFile, ‘a‘)  # a追加写
    fp.writelines(["lastPage：" + str(next) + "\n", "total:" + str(i) + "\n"])  # None及数字：无法和字符串用 + 拼接
    fp.close()