1.背景

項目需求，要求獲得github的repo的api，以便可以提取repo的數據進行分析。研究了一天，終于解決了這個問題，雖然效率還是比較低下。

因為github的那個顯示repo的api，列出了每個repo的詳細信息，而且是json格式的?，F在貌似還沒有找到可以分析多個json格式數據的方法，所以用的是比較蠢得splite加re的方法。如果大家有更好的方法，不發留言討論！

2.代碼

            
import re
import os

def GetUrl(num):
 str = os.popen("curl -G https://api.github.com/repositories?since=%d"%(num)).read()
 pattern = '"url"'
 pattern1='repos'
 urls=str.split(',\n')  
 for i in urls:
  if pattern in i and pattern1 in i:   
#  text1=i.splite(':')
  text=re.compile('"(.*?)"').findall(i)[1]
  print text
if __name__=='__main__':
 GetUrl(1000)
          

??? 其中num的值指的是頁面的id，我們可以做一個循環，不斷增大num的值，就可以無限提取repo。因為github的api對于流量是有限制的，所以這么做是一個可行的方法。

效果如下（提取下來的repo的api地址）：

https://api.github.com/repos/wycats/merb-core

https://api.github.com/repos/rubinius/rubinius

https://api.github.com/repos/mojombo/god

https://api.github.com/repos/vanpelt/jsawesome

https://api.github.com/repos/wycats/jspec

https://api.github.com/repos/defunkt/exception_logger

https://api.github.com/repos/defunkt/ambition

https://api.github.com/repos/technoweenie/restful-authentication

https://api.github.com/repos/technoweenie/attachment_fu

https://api.github.com/repos/topfunky/bong

https://api.github.com/repos/Caged/microsis

https://api.github.com/repos/anotherjesse/s3

https://api.github.com/repos/anotherjesse/taboo

https://api.github.com/repos/anotherjesse/foxtracs

https://api.github.com/repos/anotherjesse/fotomatic

https://api.github.com/repos/mojombo/glowstick

https://api.github.com/repos/defunkt/starling

https://api.github.com/repos/wycats/merb-more

https://api.github.com/repos/macournoyer/thin

https://api.github.com/repos/jamesgolick/resource_controller

https://api.github.com/repos/jamesgolick/markaby

https://api.github.com/repos/jamesgolick/enum_field

https://api.github.com/repos/defunkt/subtlety

https://api.github.com/repos/defunkt/zippy

https://api.github.com/repos/defunkt/cache_fu

https://api.github.com/repos/KirinDave/phosphor