Django html页面 'ascii' codec can't encode characters in position 8-10: ordinal not

用Django开发的页面，之前用的是python3.X，后来又换成python2.X后各种报错，编码问题，于是在所有python文件开头加了编码：#coding=utf-8

但是后来发现，有些文件加了#coding=utf-8还是不起作用，如现在在一个网页报错：

于是在Django项目的views.py下加了如下四行代码：

# coding=utf-8
import sys
reload(sys)
sys.setdefaultencoding('utf-8')

后来发现有网友和我的问题一样，但是他的解决方法和我不一样：https://stackoverflow.com/questions/27435622/python-ascii-codec-cant-encode-character-u-xe9-in-position-5-ordinal-not

我也贡献了我的答案my  solution:

Hi,your problem is same with mine,just encode question, you can add bellow code in your views.py of django project:

# coding=utf-8
import sys
reload(sys)
sys.setdefaultencoding('utf-8')

my problem:
Error during template rendering
In template D:PythonProjectsDjangoProjectguestsign emplatessignguest_manage.html, error at line 72

Python - 'ascii' codec can't encode character u'xe9' in position 5: ordinal not in range(128)

Ask Question

up vote0down votefavorite

1

I'm trying to run a quick Django application that pulls data from Google AdWords and exposes the names of accounts that are managed by an agency. When doing so, I get the following error:

UnicodeEncodeError at /account-hierarchy/
'ascii' codec can't encode character u'xe9' in position 5: ordinal not in range(128)

Here's the snippet:

<table class="pretty-table">
  <thead>
    <tr>
      <td>Customer ID</td>
      <td>Client Name</td>
      <td>Can Manage Clients</td>
      <td>Account Currency</td>
    </tr>
  </thead>
  {% for account in managed_accounts %}
    <tr>
      {% for field in account %}
        <td>{{ field }}</td>
      {% endfor %}
    </tr>
  {% endfor %}
</table>

where the call to {{ field }} is the problematic line.

I have already added

<meta http-equiv="content-type" content="text/html; charset=utf-8">

to the template I am rendering, but it still fails, so I believe the problem is not on the HTML template, but rather on the Python/Django engine.

Any ideas how I can fix it?

Here's the View code that renders the template:

def account_hierarchy(request):
manager_ids = settings.MANAGER_IDS
managed_accounts = []
for manager_id in manager_ids:
    managed_accounts.extend(adwords_utils.getManagedAccounts(manager_id))
return render_to_response('simple-table.html', {"managed_accounts": managed_accounts})

UPDATED Question

• Python Version: 2.7.6
• Models.py is currently empty

What's also curious is that if I remove this:

{% for field in account %}
    <td>{{ field }}</td>
  {% endfor %}

and I just print out the main array:

{{ managed_accounts }}

it works just fine. Not sure what's going on.

Curious fact #2: As I managed to output the full array, I checked for character 'é' and I didn't find it on the final output. Not sure where it was coming from.

python django python-unicode

shareimprove this question

edited Dec 12 '14 at 2:30

asked Dec 12 '14 at 1:52 daniel_c05 6,951115274

Perhaps define the __unicode__ method in your models.py would work – dazedconfused Dec 12 '14 at 1:57 

Could you also include some related codes in your models.py? – dazedconfused Dec 12 '14 at 2:13

I am not using any models yet. So models.py is empty. – daniel_c05 Dec 12 '14 at 2:15

Then you would probably need to encode some of the data in managed_accounts – dazedconfused Dec 12 '14 at 2:18

add a comment

1 Answer

activeoldestvotes

up vote4down vote

The problem is likely that, in some place in your code, you accidentally defined a data structure to be a Python byte string, when you should have made it a Python Unicode string. This leads Django and Python to convert from Django's Unicode string to your byte string, in a default way, using the ASCII codec.

The error message gives some clues:

• The error is perhaps occurring when you call account_hierarchy()
• The phrase character u'xe9' in position 5: ordinal not in range(128) means that Python is trying to convert Unicode character 'é' (U+00E9) into an ASCII value from 0..127. Look in your data for an 'é'. Which field is it in?
• The phrase 'ascii' codec means the conversion is likely inadvertent. If you intend to use UTF-8, you wouldn't have called the ASCII codec intentionally. But when you cause a conversion but don't specify a codec, Python uses ASCII.
• The word encode means conversion from Unicode to byte-oriented encoding. You have read Python's Unicode HOWTO article a couple of times, haven't you?

Function render_to_response() uses settings DEFAULT_CHARSET and DEFAULT_CONTENT_TYPE](https://docs.djangoproject.com/en/dev/ref/settings/#std:setting-DEFAULT_CONTENT_TYPE). They should default to 'utf-8' and 'text/html', which should be appropriate for generating a UTF-8 encoded HTML page, but check.

I like the suggestion that you check your models.py to be sure that your models are defined in terms of Unicode data types, not byte string data types. Update: you say models.py is empty, so that won't be much help.

Character handling, and Unicode vs byte strings, are handled differently in Python 2 and Python 3. Which version of Python are you using? Update: Python 2.7.6, thanks. The Unicode HOWTO I linked to above is for Python 2.7.x.

If you make sure your code handles strings as Unicode throughout, unless you really want a byte string, that will likely fix this problem.

Update: Consider modifying your template to give you debugging information. Try something like these expressions, to see what's really in managed_accounts:

</thead>
<tr><td>managed_accounts is: {{ repr(managed_accounts) }}</tr></td>
{% for account in managed_accounts %}
  <tr>
  {% for field in account %}
    <td>{{ type(field) }}, {{ repr(field) }}</td>
  {% endfor %}
  </tr>
{% endfor %}

[Updated in response to multiple updates from original poster.]

shareimprove this answer

edited Dec 12 '14 at 7:21

answered Dec 12 '14 at 2:21 Jim DeLaHunt 6,50812948

Nice and detailed answer! – dazedconfused Dec 12 '14 at 2:29

add a comment

Your Answer