在url中使用中文其实是一个坏习惯，会带来一系列的转码问题， 我更喜欢英文译名或者id来标识某个uri。但是现实往往是残酷的， 特别是在我们调用别人服务时候，有时候被逼无奈使用中文URL。

Python中unicode转码一向是让人头疼的问题。数次碰壁之后，我也摸出了一些门道， 研读完Python字符串的encode与decode 之后，就自认为找到了万金油，谁知道这次又碰上这个老冤家。

Traceback (most recent call last):
02	  File "<stdin>", line 1, in <module>
03	  File "/usr/lib/python2.6/urllib2.py", line 126, in urlopen
04	    return _opener.open(url, data, timeout)
05	  File "/usr/lib/python2.6/urllib2.py", line 391, in open
06	    response = self._open(req, data)
07	  File "/usr/lib/python2.6/urllib2.py", line 409, in _open
08	    '_open', req)
09	  File "/usr/lib/python2.6/urllib2.py", line 369, in _call_chain
10	    result = func(*args)
11	  File "/usr/lib/python2.6/urllib2.py", line 1170, in http_open
12	    return self.do_open(httplib.HTTPConnection, req)
13	  File "/usr/lib/python2.6/urllib2.py", line 1142, in do_open
14	    h.request(req.get_method(), req.get_selector(), req.data, headers)
15	  File "/usr/lib/python2.6/httplib.py", line 914, in request
16	    self._send_request(method, url, body, headers)
17	  File "/usr/lib/python2.6/httplib.py", line 951, in _send_request
18	    self.endheaders()
19	  File "/usr/lib/python2.6/httplib.py", line 908, in endheaders
20	    self._send_output()
21	  File "/usr/lib/python2.6/httplib.py", line 780, in _send_output
22	    self.send(msg)
23	  File "/usr/lib/python2.6/httplib.py", line 759, in send
24	    self.sock.sendall(str)
25	  File "<string>", line 1, in sendall
26	UnicodeEncodeError: 'ascii' codec can't encode characters in position 7-8: ordinal not in range(128)

这次错误引发是在 urlopen() 引起的，很有特色，开始使用 url.encode('utf-8')

1. ascii + unicode 测试

	>>> 'a' + u'b'
02	>>> '你' + u'好'
03	Traceback (most recent call last):
04	  File "<stdin>", line 1, in <module>
05	UnicodeDecodeError: 'ascii' codec can't decode byte 0xe4 in position 0: ordinal not in range(128)
06	>>> u'你' + u'好'
07	u'\u4f60\u597d'
08	>>> u'a' + '你' + u'好'
09	Traceback (most recent call last):
10	  File "<stdin>", line 1, in <module>
11	UnicodeDecodeError: 'ascii' codec can't decode byte 0xe4 in position 0: ordinal not in range(128)
上面的测试说明ascii码和unicode码相连操作，结论是有中文记得带上u，就不会有问题。 Python默认解码器是ascii，无法解码unicode中的中文。

2. urllib2的测试
01	>>> import urllib2
02	>>> h1 = 'http://baidu.com'
03	>>> urllib2.urlopen(h1)
04	<addinfourl at 153439532 whose fp = <socket._fileobject object at 0xb74e51ac>>
05	>>> h2 = u'http://baidu.com'
06	>>> urllib2.urlopen(h2)
07	<addinfourl at 153440236 whose fp = <socket._fileobject object at 0x925912c>>
08	>>> h3 = 'http://baidu.com?w=测试'
09	>>> urllib2.urlopen(h3)
10	<addinfourl at 153482348 whose fp = <socket._fileobject object at 0x92593ac>>
11	>>> h4 = u'http://baidu.com?w=测试'
12	>>> urllib2.urlopen(h4)
13	Traceback (most recent call last):
14	  File "<stdin>", line 1, in <module>
15	  File "/usr/lib/python2.6/urllib2.py", line 126, in urlopen
16	    return _opener.open(url, data, timeout)
17	  File "/usr/lib/python2.6/urllib2.py", line 391, in open
18	    response = self._open(req, data)
19	  File "/usr/lib/python2.6/urllib2.py", line 409, in _open
20	    '_open', req)
21	  File "/usr/lib/python2.6/urllib2.py", line 369, in _call_chain
22	    result = func(*args)
23	  File "/usr/lib/python2.6/urllib2.py", line 1170, in http_open
24	    return self.do_open(httplib.HTTPConnection, req)
25	  File "/usr/lib/python2.6/urllib2.py", line 1142, in do_open
26	    h.request(req.get_method(), req.get_selector(), req.data, headers)
27	  File "/usr/lib/python2.6/httplib.py", line 914, in request
28	    self._send_request(method, url, body, headers)
29	  File "/usr/lib/python2.6/httplib.py", line 951, in _send_request
30	    self.endheaders()
31	  File "/usr/lib/python2.6/httplib.py", line 908, in endheaders
32	    self._send_output()
33	  File "/usr/lib/python2.6/httplib.py", line 780, in _send_output
34	    self.send(msg)
35	  File "/usr/lib/python2.6/httplib.py", line 759, in send
36	    self.sock.sendall(str)
37	  File "<string>", line 1, in sendall
38	UnicodeEncodeError: 'ascii' codec can't encode characters in position 7-8: ordinal not in range(128)

这个测试说明， urllib2.urlopen()

so，请尽量英文url，非要用中文，请记得转码。