首先導入模塊，用help查看相關文檔
>>> from urlparse import urljoin >>> help(urljoin) Help on function urljoin in module urlparse: urljoin(base, url, allow_fragments=True) Join a base URL and a possibly relative URL to form an absolute interpretation of the latter.
意思就是將基地址與一個相對地址形成一個絕對地址，然而講的太過抽象
接下來，看幾個例子，從例子中發現規律。
>>> urljoin("http://www.google.com/1/aaa.html","bbbb.html") 'http://www.google.com/1/bbbb.html' >>> urljoin("http://www.google.com/1/aaa.html","2/bbbb.html") 'http://www.google.com/1/2/bbbb.html' >>> urljoin("http://www.google.com/1/aaa.html","/2/bbbb.html") 'http://www.google.com/2/bbbb.html' >>> urljoin("http://www.google.com/1/aaa.html","http://www.google.com/3/ccc.html") 'http://www.google.com/3/ccc.html' >>> urljoin("http://www.google.com/1/aaa.html","http://www.google.com/ccc.html") 'http://www.google.com/ccc.html' >>> urljoin("http://www.google.com/1/aaa.html","javascript:void(0)") 'javascript:void(0)'
規律不難發現，但是並不是萬事大吉了，還需要處理特殊情況，如鏈接是其本身，鏈接中包含無效字符等
url = urljoin("****","****")
### find()查找字符串函數，如果查到：返回查找到的第一個出現的位置。否則，返回-1
if url.find("'")!=-1:
continue
### 只取井號前部分
url = url.split('#')[0]
### 這個isindexed()是我自己定義的函數，判斷該鏈接不在保存鏈接的數據庫中
if url[0:4]=='http' and not self.isindexed(url):
###newpages = set(),無序不重復元素集
newpages.add(url)