UnicodeDecodeError

[honi]

I think I've found another bug. I have a TextField in a model, and in the admin site, when I input a latin character, for example: á, after I submit the change_form and try to edit that object I get this error.

Here's the traceback:

Django Version: 1.3
Python Version: 2.6.1

Traceback:
File "/Users/jonito/mingus/proyectos/galerias-belgrano/bootstrap/lib/python2.6/site-packages/django/core/handlers/base.py" in get_response

1.             response = middleware_method(request, response)
   

   File "/Users/jonito/mingus/proyectos/galerias-belgrano/bootstrap/src/django-htmlmin/htmlmin/middleware.py" in process_response

2.         response.content = html_minify(response.content)
   

   File "/Users/jonito/mingus/proyectos/galerias-belgrano/bootstrap/src/django-htmlmin/htmlmin/minify.py" in html_minify

3.         html_code = html_code.replace(script, TAGS_PATTERN % (tag, index))
   

Exception Type: UnicodeDecodeError at /admin/chunks/chunk/2/
Exception Value: ('ascii', '<textarea id="id_content" rows="10" cols="40" name="content" class="vLargeTextField">Lunes a S\xc3\xa1bados de 9 a 20hs.</textarea>', 94, 95, 'ordinal not in range(128)')

[gilsondev]

Used in smart_unicode __unicode__ ? Will probably be it.

The data can come in ASCII, and can not convert to unicode.

[honi]

I'm sorry, I didn't understand you completely.

This also happens for any <textarea> content. For example, you have a form with a field rendered as a textarea. If you submit the form with a latin character in the textarea, but the form doesn't validate, when the form is rendered again showing the errors, and where normally bounded fields show their values, this error happens.

[fsouza]

Sorry for the delay, this week is a little hard, I'll take a deep look in the weekend :)

Thanks in advance for reporting.

[marcusalmeida]

Like said, the data come in ASCII and when try to replace "script" in html_code not convert to unicode.
Could we use "string.unicode('utf-8')" in line 20 in minify.py? Like this?
html_code = html_code.replace(script.decode('utf-8'), TAGS_PATTERN %(tag, index))

Is there any problem with it ?

[fsouza]

Hi @marcusalmeida, I'll fix it today or this weekend (in the worst case).

Sorry for this long delay.

[fsouza]

Just fixed it and released 0.5 version, please upgrade.

Thank you for reporting.

[marcusalmeida]

Nice. I'll upgrade.

[honi]

Excellent! Thanks