python中的urlencode错误

Question

我有这个：

a = {'album': u'Metamorphine', 'group': 'monoku', 'name': u'Son Of Venus (Danny\xb4s Song)', 'artist': u'Leandra', 'checksum': '2836e33d42baf947e8c8adef48921f2f76fcb37eea9c50b0b59d7651', 'track_number': 8, 'year': '2008', 'genre': 'Darkwave', 'path': u'/media/data/musik/Leandra/2008. Metamorphine/08. Son Of Venus (Danny\xb4s Song).mp3', 'user_email': 'diegueus9@gmail.com', 'size': 6624104}
data = urllib.urlencode(mp3_data)

然后这引发了一个异常：

Traceback (most recent call last):
  File "playkud.py", line 44, in <module>
    main()
  File "playkud.py", line 29, in main
    craw(args, options.user_email, options.group)
  File "/home/diegueus9/workspace/playku/src/client/playkud/crawler/crawler.py", line 76, in craw
    index(root, file, data, user_email, group)
  File "/home/diegueus9/workspace/playku/src/client/playkud/crawler/crawler.py", line 58, in index
    done = add_song(data[mp3file])
  File "/home/diegueus9/workspace/playku/src/client/playkud/service.py", line 32, in add_song
    return make_request(URL+'add_song/', data)
  File "/home/diegueus9/workspace/playku/src/client/playkud/service.py", line 14, in make_request
    data = urllib.urlencode(dict([k.encode('utf-8'),v] for k,v in mp3_data.items()))
  File "/usr/lib/python2.5/urllib.py", line 1250, in urlencode
    v = quote_plus(str(v))
UnicodeEncodeError: 'ascii' codec can't encode character u'\xb4' in position 19: ordinal not in range(128)

还有在ipython（2.5）中：

In [7]: urllib.urlencode(a)
UnicodeEncodeError                        Traceback (most recent call last)

/home/diegueus9/<ipython console> in <module>()

/usr/lib/python2.5/urllib.pyc in urlencode(query, doseq)
   1248         for k, v in query:
   1249             k = quote_plus(str(k))
-> 1250             v = quote_plus(str(v))
   1251             l.append(k + '=' + v)
   1252     else:

UnicodeEncodeError: 'ascii' codec can't encode character u'\xb4' in position 19: ordinal not in range(128)

我该怎么解决这个问题呢？

Answer 1

问题在于，你的mp3_data字典里有些值是无法用默认编码表示的unicode字符串（而其他的则是ASCII字符，还有一些是整数）。你可以通过在把这些值传给urlencode()之前先进行编码来解决这个问题。在/home/diegueus9/workspace/playku/src/client/playkud/service.py文件的make_request()函数的第14行，试着把这段代码：

data = urllib.urlencode(dict([k.encode('utf-8'),v] for k,v in mp3_data.items()))

改成这段代码：

data = urllib.urlencode(dict([k.encode('utf-8'),unicode(v).encode('utf-8')] for k,v in mp3_data.items()))

Answer 2

如果你在使用Django框架，可以看看Django的QueryDict类，它里面有一个叫做urlencode()的方法。

另外，你也可以直接使用这个帮助函数，urlencode。这个函数基本上就是把其他回答中提到的功能封装了一下，实际上是对原来的urllib.encode的一个包装。

Answer 3

urlencode库需要的数据格式是str，而且它对Unicode数据处理得不好，因为它没有提供指定编码的方式。所以你可以试试下面的方法：

mp3_data = {'album': u'Metamorphine',
     'group': 'monoku',
     'name': u'Son Of Venus (Danny\xb4s Song)',
     'artist': u'Leandra',
     'checksum': '2836e33d42baf947e8c8adef48921f2f76fcb37eea9c50b0b59d7651',
     'track_number': 8,
     'year': '2008', 'genre': 'Darkwave',
     'path': u'/media/data/musik/Leandra/2008. Metamorphine/08. Son Of Venus (Danny\xb4s Song).mp3',
     'user_email': 'diegueus9@gmail.com',
     'size': 6624104}

str_mp3_data = {}
for k, v in mp3_data.iteritems():
    str_mp3_data[k] = unicode(v).encode('utf-8')
data = urllib.urlencode(str_mp3_data)

我做的就是在把字典传入urlencode函数之前，确保所有数据都用UTF-8编码成str格式。