python中类的继承，初始化,super()__init__()方法

每日星语：

“你知道年轻的时候穷最大的问题是什么吗？”

“什么”

### “以为所有得不到的一切都是穷的错。眼睛盯着遥远的地方，却看不到身边一朵花的美丽”

2018-09-07 20:30:30 星期五

今天遇到一个问题，爬虫在定义网页规则时，因为有十几个站点来自于统一网站，刚开始考虑去复制一份，只需要修改web_site变量即可。但考虑到两点，其一，站点过多，复制麻烦，其二，也是最重要的一点，当网站结构变化时，修改维护起来过于冗余复杂，所以就使用了模板继承的办法！

刚开时circHeNan.py代码如下

# encoding: utf-8
from datetime import datetime
from .base_stock import BaseStock

class CircHeNan(BaseStock):

def is_detail_url(self, dom):
return dom("#ess_mailrightpane")

def parse_detail_url(self, dom, params):

web_site = "河南委员会"
store_json = {
"info:title": dom("#tab_content)").text(),
"info:publish_time": dom("#tab_content > tbody:nth-child(1)").text(),
"info:source": web_site,
"info:author": "",
"info:content": dom("#tab_content > tbody").outerHtml(),
"info:tag": dom("head meta[name=\"keywords\"]").attr("content") or "",
"info:dese": dom("head meta[name=\"description\"]").attr("content") or "",
"info:crawl_time": datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
"info:url": params["info:url"],
"info:channel": "",
"info:laiyuan": web_site,
"info:contain_image": "True" if dom("#tab_content > tbody")("img") else "False"
}
return store_json

由于该类继承与BaseStock

class BaseStock(object):

def __init__(self):
self.html_clean = html_clean.HTMLClean()
self.html = ""
self.web_site=""

考虑到BsaeStock为父类，当子类重写init时，子类就不会继承父类的init，所以需要对父类中init重写完整

将circ.py定义成模板

# encoding: utf-8
import re
from datetime import datetime
from .base_stock import BaseStock
from components import html_clean

class Csrc(BaseStock):
def __init__(self):
self.html_clean = html_clean.HTMLClean()
self.html = ""
self.web_site = ""

def is_detail_url(self, dom):
return dom(".content")

def parse_detail_url(self, dom, params):
store_json = {

"info:title": dom("#tab_content)").text(),
"info:publish_time": dom("#tab_content > tbody:nth-child(1)").text(),
"info:source":self.web_site,
"info:author": "",
"info:content": dom("#tab_content > tbody").outerHtml(),
"info:tag": dom("head meta[name=\"keywords\"]").attr("content") or "",
"info:dese": dom("head meta[name=\"description\"]").attr("content") or "",
"info:crawl_time": datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
"info:url": params["info:url"],
"info:channel": "",
"info:laiyuan": self.web_site,
"info:contain_image": "True" if dom("#tab_content > tbody")("img") else "False"
}
return store_json

修改circHeNan.py文件里的内容如下：

from .circ import Circ
from components import html_clean

class CircHeNan(Circ):
def __init__(self):
self.html_clean = html_clean.HTMLClean()
self.html = ""
self.web_site = "河南委员会"

这时候的circHeNan.py虽然只有很少的几行，但这时候的代码效果和刚开始的一样了

更新文章

2018-09-11 20:43:54 星期二

在多重类的继承过程中，其实存在这super().__init__()的方法，这个方法可以很好的解决冲突问题，上面的文掌可以不需要这么复杂，只需要用super()完美解决
将模板文件circ.py修改如下

import re
from datetime import datetime
from .base_stock import BaseStock
class Csrc(BaseStock):
def __init__(self):
super(Csrc,self).__init__()
self.web_site = ""
def is_detail_url(self, dom):
return dom(".content")
def parse_detail_url(self, dom, params):
store_json = {
"info:title": dom("#tab_content)").text(),
"info:publish_time": dom("#tab_content > tbody:nth-child(1)").text(),
"info:source":self.web_site,
"info:author": "",
"info:content": dom("#tab_content > tbody").outerHtml(),
"info:tag": dom("head meta[name=\"keywords\"]").attr("content") or "",
"info:dese": dom("head meta[name=\"description\"]").attr("content") or "",
"info:crawl_time": datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
"info:url": params["info:url"],
"info:channel": "",
"info:laiyuan": self.web_site,
"info:contain_image": "True" if dom("#tab_content > tbody")("img") else "False"
}
return store_json

而circHeNan.py的文件可以修改为

from .circ import Circ
class CircHeNan(Circ):
def __init__(self):
super()__init__()
self.web_site = "河南委员会"

简单案例可以参考这个这个页面
的实例部分，看懂了就理解了