scrapy爬虫成长日记之创建工程-抽取数据-保存为json格式的数据

转自 http://www.tuicool.com/articles/y6fErea

环境：CentOS 6.0 虚拟机

scrapy（如未安装可参考 安装python爬虫scrapy踩过的那些坑和编程外的思考 ）

1、创建工程cnblogs

[root@bogon share]# scrapy startproject cnblogs
2015-06-10 15:45:03 [scrapy] INFO: Scrapy 1.0.0rc2 started (bot: scrapybot)
2015-06-10 15:45:03 [scrapy] INFO: Optional features available: ssl, http11
2015-06-10 15:45:03 [scrapy] INFO: Overridden settings: {}
New Scrapy project 'cnblogs' created in:
    /mnt/hgfs/share/cnblogs

You can start your first spider with:
    cd cnblogs
    scrapy genspider example example.com

2、查看下工程的结构

[root@bogon share]# tree cnblogs/
cnblogs/
├── cnblogs
│   ├── __init__.py
│   ├── items.py #用于定义抽取网页结构
│   ├── pipelines.py #将抽取的数据进行处理
│   ├── settings.py #爬虫配置文件
│   └── spiders
│       └── __init__.py
└── scrapy.cfg #项目配置文件

3、定义抽取cnblogs的网页结构，修改items.py

这里我们抽取四个内容：

• 文章标题
• 文章链接
• 文在所在的列表页URL
• 摘要

[root@bogon cnblogs]# vi cnblogs/items.py
# -*- coding: utf-8 -*-
# Define here the models for your scraped items
#
# See documentation in:
# http://doc.scrapy.org/en/latest/topics/items.html

import scrapy
class CnblogsItem(scrapy.Item):
  # define the fields for your item here like:
  # name = scrapy.Field()
  title = scrapy.Field()
  link = scrapy.Field()
  desc = scrapy.Field()
  listUrl = scrapy.Field()
  pass

4、创建spider

[root@bogon cnblogs]# vi cnblogs/spiders/cnblogs_spider.py
#coding=utf-8
import re
import json
from scrapy.selector import Selector
try:
  from scrapy.spider import Spider
except:
  from scrapy.spider import BaseSpider as Spider
from scrapy.utils.response import get_base_url
from scrapy.utils.url import urljoin_rfc
from scrapy.contrib.spiders import CrawlSpider, Rule
from scrapy.contrib.linkextractors.sgml import SgmlLinkExtractor as sle
from cnblogs.items import *
class CnblogsSpider(CrawlSpider):
  #定义爬虫的名称
  name = "CnblogsSpider"
  #定义允许抓取的域名,如果不是在此列表的域名则放弃抓取
  allowed_domains = ["cnblogs.com"]
  #定义抓取的入口url
  start_urls = [
    "http://www.cnblogs.com/rwxwsblog/default.html?page=1"
  ]
  # 定义爬取URL的规则，并指定回调函数为parse_item
  rules = [
    Rule(sle(allow=("/rwxwsblog/default.html\?page=\d{1,}")), #此处要注意?号的转换，复制过来需要对?号进行转义。
             follow=True,
             callback='parse_item')
  ]
  #print "**********CnblogsSpider**********"
  #定义回调函数
  #提取数据到Items里面，主要用到XPath和CSS选择器提取网页数据
  def parse_item(self, response):
    #print "-----------------"
    items = []
    sel = Selector(response)
    base_url = get_base_url(response)
    postTitle = sel.css('div.day div.postTitle')
    #print "=============length======="
    postCon = sel.css('div.postCon div.c_b_p_desc')
    #标题、url和描述的结构是一个松散的结构，后期可以改进
    for index in range(len(postTitle)):
      item = CnblogsItem()
      item['title'] = postTitle[index].css("a").xpath('text()').extract()[0]
      #print item['title'] + "***************\r\n"
      item['link'] = postTitle[index].css('a').xpath('@href').extract()[0]
      item['listUrl'] = base_url
      item['desc'] = postCon[index].xpath('text()').extract()[0]
      #print base_url + "********\n"
      items.append(item)
      #print repr(item).decode("unicode-escape") + '\n'
    return items

注意：

首行要设置为： #coding=utf-8 或  # -*- coding: utf-8 -*- 哦！否则会报错。

SyntaxError: Non-ASCII character '\xe5' in file /mnt/hgfs/share/cnblogs/cnblogs/spiders/cnblogs_spider.py on line 15, but no encoding declared; see http://python.org/dev/peps/pep-0263/ for details

spider的名称为： CnblogsSpider ，后面会用到。

5、修改pipelines.py文件

[root@bogon cnblogs]# vi cnblogs/pipelines.py
# -*- coding: utf-8 -*-
# Define your item pipelines here
#
# Don't forget to add your pipeline to the ITEM_PIPELINES setting
# See: http://doc.scrapy.org/en/latest/topics/item-pipeline.html

from scrapy import signals
import json
import codecs
class JsonWithEncodingCnblogsPipeline(object):
  def __init__(self):
    self.file = codecs.open('cnblogs.json', 'w', encoding='utf-8')
  def process_item(self, item, spider):
    line = json.dumps(dict(item), ensure_ascii=False) + "\n"
    self.file.write(line)
    return item
  def spider_closed(self, spider):
    self.file.close()

注意类名为 JsonWithEncodingCnblogsPipeline 哦！settings.py中会用到

6、修改settings.py，添加以下两个配置项

ITEM_PIPELINES = {
    'cnblogs.pipelines.JsonWithEncodingCnblogsPipeline': 300,
}
LOG_LEVEL = 'INFO'

7、运行spider，scrapy crawl 爬虫名称（cnblogs_spider.py中定义的name）

[root@bogon cnblogs]# scrapy crawl CnblogsSpider

8、查看结果more cnblogs.json（pipelines.py中定义的名称 ）

more cnblogs.json

9、如果有需要可以将结果转成txt文本格式，可参考另外一篇文章 python将json格式的数据转换成文本格式的数据或sql文件

源码可在此下载： https://github.com/jackgitgz/cnblogs

10、相信大家还会有疑问，我们能不能将数据直接保存在数据库呢？答案是可以的，接下来的文章会逐一介绍，敬请期待。

参考资料：

http://doc.scrapy.org/en/master/

http://blog.csdn.net/HanTangSongMing/article/details/24454453