Selenimu做爬虫续

前面曾写过一篇Selenimu做爬虫的文章，大概只说了些如何爬取信息，这篇文章主要接着来说下比较具体的用途。

假设你是某个财经网站的质量人员，上头很想知道自己网站的数据是否正确，（这里插一句，做财经的应该知道，例如国内股票的原始数据都是从上证或深证来的，之后有很多的数据提供商从2大交易所购买数据，这个不是一般公司可以做到的，看看level-2和超赢等卖得那么红火就知道了，国外的blomberg也类似。），那么就需要提供一个dashboard来展示自己的数据和其他同行（多数是竞争对手啦）的数据，这样直观且方便跟踪，同时对其他人也有说服力。

我了解下来，很多都是通过人工采样的方式来对比，比如建个excel表把自己和其他网站数据填充，这个做法费时又费力，其实selenium可以帮助我们提高效率。

大致的思路如下：

1. 获取交易所信息及股票代码

在Selenimu做爬虫已经介绍过，找某个网站股票综合页面，例如新浪就能获取到某交易所的股票代码，并抓取下来，当然，如果自己后台数据库有这个数据，那就更好了。

 

2. 创建一个CrawlSettings.py文件，用来存取比较网站的信息，如网站名称、URL、财经指标的xpath,示例如下，

 

 1 #!/usr/bin/python2.5.2
 2 #-*- coding: utf8 -*-
 3 
 4 import codecs
 5 
 6 PageProfiles = (
 7   {
 8     "Name" : "sina",
 9     "Link" : {
10       "sz" : "http://finance.sina.com.cn/realstock/company/sz%s/nc.shtml",
11       "sh" : "http://finance.sina.com.cn/realstock/company/sh%s/nc.shtml",
12     },
13     "Pattern" : {
14       "PriceClose" : "//span[@id='itemPrevious2']",
15       "QuoteLast" : "//h3[@id='itemCurrent']",
16       "BookValuePerShareYear":"//td[@id='cwjk_mgjzc']",
17       "EPS" : "//td[@id='cwjk_mgsy']",
18       "MarketCap" :"//span[@id='totalMart2']",
19       "PE" :"//td[@id='pe_ratio_lyr']",
20       "Float":"//div[@id='con04-0']/table/tbody/tr[4]/td[2]",
21       "CashPerShareYear" : "//div[@id='con04-0']/table/tbody/tr[2]/td[1]",
22       "Volume":"//span[@id='itemVolume2']",
23     },
24     "PostProcess" : {
25       "all" : {
26         "RemoveString" : [r",", r"%", r"元"]
27       },
28     },
29    ...
30 )
31 
32 SiteProfiles = {
33   "sina" : {
34     "BaseLink" : "http://finance.sina.com.cn",
35     "PageProfileNames" : ["sina","sina_DPSRecentYear","sina_other"]
36   },
37   ...
38 }
39 
40 CrawlProfiles = {
41   "test" : {
42       "Sites" : ["baidu","sina","yahoo"],
43       "Criteria" : ["Float", "PE", "MarketCap", "EPS", "PriceClose", "QuoteLast", "BookValuePerShareYear", 
44                     "CurrentRatioYear", "NetProfitMarginPercent", "GrossMargin", "CashPerShareYear", 
45                     "DPSRecentYear", "High52Week", "Low52Week","Volume"],  
46       "TickerListFile" : "tickerList.txt",
47       "ResultFiles" : "result_%s.txt",
48       "NumTickerPerSite" : "20"
49     },
50   ...
51 }
52 
53 ExchangesProfiles={
54     "HongKong":"HKG",
55     "UK":"LON",
56     "Canada":["CVE","TSE"],
57     "France":"EPA",
58     "Italy":["ISE","BIT"],
59     "TaiWan":"TPE",
60     # "China Mainland":["SHA","SHE"],
61         ...
62 }
63 
64 class CrawlSettings:
65   
66   def GetCrawlProfile(self, name):
67     if name in CrawlProfiles:
68       return CrawlProfiles[name]
69     return None
70   
71   def GetSiteProfile(self, name):
72     if name in SiteProfiles:
73       return SiteProfiles[name]
74     return None
75   
76   def GetPageProfile(self, name):
77     for prof in PageProfiles:
78       if prof["Name"] == name:
79         return prof
80     return None
81   
82   def LoadTickerList(self, filename):
83     f = codecs.open(filename, 'r', 'utf-8')
84     list = []
85     for line in f:
86       list.append(line.strip())
87 
88     f.close()
89     return list
90 
91   def GetStockType(self, ticker):
92       if str(ticker)[0]=='0' or  str(ticker)[0]=='2':
93           return "sz"
94       elif str(ticker)[0]=='6' or str(ticker)[0]=='9':
95           return "sh"
96       else:
97           return "all"

 

 

3. 爬取页面数据，FinanceAutomationCrawler.py；参考Selenimu做爬虫

 

4. 比较和保存数据，FinanceAutomationResult.py

 

 

 1 #!/usr/bin/python2.5.2
 2 #-*- coding: utf8 -*-
 3 
 4 import sys
 5 import codecs
 6 
 7 
 8 class FinanceAutomationResult:
 9   ###  Save / Load / Generate the result comparison ###
10 
11   def SaveAsSiteResults(self, results, siteList, criteriaList, tickerList, fileName):
12     ### results[site][criteria][ticker] 
13     ###  Save as "result_site.txt": 
14     ### ticker   criteria1 criteria2 .. 
15     ### ticker1    1        2
16     ### ticker2    1        2
17     ###
18     for site in siteList:
19     ...
20 
21   def SaveAsCriteriaResults(self, results, siteList, criteriaList, tickerList, fileName):
22     ### results[site][criteria][ticker] 
23     ### Save as "result_criteria.txt": 
24     ### ticker   site1     site2  .. 
25     ### ticker1    1        2
26     ### ticker2    1        2
27     ###
28     for criteria in criteriaList:
29     ...
30 
31   def CompareResult(self, results, siteList,criteriaList, tickerList):
32     
33     for criteria in criteriaList:
34     ...

 

好了，废话不多说，大概的思路如下图： 

 

 

 

 最后祝大家清明节快乐！