[python]书籍信息爬虫示例

背景说明

需要收集一些书籍信息，以豆瓣书籍条目作为源，得到一些有效书籍信息，并保存到本地数据库。

获取书籍分类标签

具体可参考这个链接：
https://book.douban.com/tag/?view=type
然后将这些分类标签链接存到本地某个文件，存储内容如下

  https://book.douban.com/tag/小说
https://book.douban.com/tag/外国文学
https://book.douban.com/tag/文学
https://book.douban.com/tag/随笔
https://book.douban.com/tag/中国文学
https://book.douban.com/tag/经典
https://book.douban.com/tag/日本文学
https://book.douban.com/tag/散文
https://book.douban.com/tag/村上春树
https://book.douban.com/tag/诗歌
https://book.douban.com/tag/童话
......

获取书籍信息，并保存本地数据库

假设已经建好mysql表，如下：

  CREATE TABLE `book_info` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `bookid` varchar(64) NOT NULL COMMENT 'book ID',
  `tag` varchar(32) DEFAULT '' COMMENT '分类目录',
  `bookname` varchar(256) NOT NULL COMMENT '书名',
  `subname` varchar(256) NOT NULL COMMENT '二级书名',
  `author` varchar(256) DEFAULT '' COMMENT '作者',
  `translator` varchar(256) DEFAULT '' COMMENT '译者',
  `press` varchar(128) DEFAULT '' COMMENT '出版社',
  `publishAt` date DEFAULT '0000-00-00' COMMENT '出版日期',
  `stars` float DEFAULT '0' COMMENT '评分',
  `price_str` varchar(32) DEFAULT '' COMMENT '价格string',
  `hotcnt` int(11) DEFAULT '0' COMMENT '评论人数',
  `bookdesc` varchar(8192) DEFAULT NULL COMMENT '简介',
  `updateAt` timestamp NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP COMMENT '修改日期',
  PRIMARY KEY (`id`),
  UNIQUE KEY `idx_bookid` (`bookid`),
  KEY `idx_bookname` (`bookname`),
  KEY `hotcnt` (`hotcnt`),
  KEY `stars` (`stars`),
  KEY `idx_tag` (`tag`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8 COMMENT='书籍信息';

并已实现相关爬虫逻辑，主要用到了BeautifulSoup包，如下：

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

import re
import logging
import requests
import pymysql
import random
import time
import datetime
from hashlib import md5
from bs4 import  BeautifulSoup

logging.basicConfig(level=logging.INFO,
                    format='[%(levelname)s][%(name)s][%(asctime)s]%(message)s',
                    datefmt='%Y-%m-%d %H:%M:%S')

class DestDB:
    Host = "192.168.1.10"
    DB = "spider"
    Table = "book_info"
    User = "test"
    Pwd = "123456"

def connect_db(host, db, user, pwd):
    conn = pymysql.connect(
        host=host,
        user=user,
        passwd=pwd,
        db=db,
        charset='utf8',
        connect_timeout=3600) #,
#        cursorclass=pymysql.cursors.DictCursor)
    conn.autocommit(True)
    return conn

def disconnect_db(conn, cursor):
    cursor.close()
    conn.close()

#提取评价人数，如果评价人数少于10人，按10人处理
def hotratings(person):
    try:
        ptext = person.get_text().split()[0]
        pc = int(ptext[1:len(ptext)-4])
    except ValueError:
        pc = int(10)
    return pc

# 持久化到数据库
def save_to_db(tag, book_reslist):
    dest_conn = connect_db(DestDB.Host, DestDB.DB, DestDB.User, DestDB.Pwd)
    dest_cursor = dest_conn.cursor()

    isql = "insert ignore into book_info "
    isql += "(`bookid`,`tag`,`author`,`translator`,`bookname`,`subname`,`press`,"
    isql += "`publishAt`,`price_str`,`stars`,`hotcnt`,`bookdesc`) values "
    isql += ",".join(["(%s)" % ",".join(['%s']*12)]*len(book_reslist))

    values = []
    for row in book_reslist:
        # 暂时将md5(bookname+author)作为bookid唯一指
        bookid = md5(("%s_%s"%(row[0],row[2])).encode('utf-8')).hexdigest()
        values.extend([bookid, tag]+row[:10])

    dest_cursor.execute(isql, tuple(values))
    disconnect_db(dest_conn, dest_cursor)

# 处理每一次访问的页面
def do_parse(tag, url):
    page_data = requests.get(url)
    soup = BeautifulSoup(page_data.text.encode("utf-8"), "lxml")
    # 提取标签信息
    tag = url.split("?")[0].split("/")[-1]
    # 抓取作者，出版社信息
    details = soup.select("#subject_list > ul > li > div.info > div.pub")
    # 抓取评分
    scores = soup.select("#subject_list > ul > li > div.info > div.star.clearfix > span.rating_nums")
    # 抓取评价人数
    persons = soup.select("#subject_list > ul > li > div.info > div.star.clearfix > span.pl")
    # 抓取书名
    booknames = soup.select("#subject_list > ul > li > div.info > h2 > a")
    # 抓取简介 
    descs = soup.select("#subject_list > ul > li > div.info > p")
    # 从标签信息中分离内容
    book_reslist = []
    for detail, score, personCnt, bookname, desc in zip(details, scores, persons, booknames, descs):
        try:
            subtitle = ""
            title_strs = [s.replace('\n', '').strip() for s in bookname.strings]
            title_strs = [s for s in title_strs if s]
            # 部分书籍有二级书名
            if not title_strs:
                continue
            elif len(title_strs) >= 2:
                bookname, subtitle = title_strs[:2]
            else:
                bookname = title_strs[0]

            # 评分人数
            hotcnt = hotratings(personCnt)
            desc = desc.get_text()
            stars = float('%.1f' % float(score.get_text() if score.get_text() else "-1"))

            author, translator, press, publishAt, price = [""]*5
            detail_texts = detail.get_text().replace('\n', '').split("/")
            detail_texts = [s.strip() for s in detail_texts]

            # 部分书籍无译者信息
            if len(detail_texts) == 4:
                author, press, publishAt, price = detail_texts[:4]
            elif len(detail_texts) >= 5:
                author, translator, press, publishAt, price = detail_texts[:5]
            else:
                continue

            # 转换出版日期为date类型
            if re.match('^[\d]{4}-[\d]{1,2}', publishAt):
                dts = publishAt.split('-')
                publishAt = datetime.date(int(dts[0]), int(dts[1]), 1)
            else:
                publishAt = datetime.date(1000, 1, 1)

            book_reslist.append([author, translator, bookname, subtitle, press, 
                                 publishAt, price, stars, hotcnt, desc])
        except Exception as e:
            logging.error(e)

    logging.info("insert count: %d" % len(book_reslist))
    if len(book_reslist) > 0:
        save_to_db(tag, book_reslist)
        book_reslist = []
    return len(details)

def main():
    with open("book_tags.txt") as fd:
        tags = fd.readlines()
        for tag in tags:
            tag = tag.strip()
            logging.info("current tag url: %s" % tag)
            for idx in range(0, 1000000, 20):
                try:
                    url = "%s?start=%d&type=T" % (tag.strip(), idx)
                    cnt = do_parse(tag.split('/')[-1], url)
                    if cnt < 10:
                        break
                    # 睡眠若干秒，降低访问频率
                    time.sleep(random.randint(10, 15))
                except Exception as e:
                    logging.warn("outer_err: %s" % e)
            time.sleep(300)

if __name__ == "__main__":
    main()

小结

以上代码基于python3环境来运行；
需要首先安装BeautifulSoup: pip install bs4
爬取过程中需要控制好访问频率；
需要对一些信息进行异常处理，比如译者信息、评论人数等。