NameError: name 'unicode' is not defined

[leocaza]

Hi,

Perhaps you will be able to help me.

I downloaded the "tldextract" in order to run "newspaper" (https://github.com/codelucas/newspaper), but I'm getting an error message "NameError: name 'unicode' is not defined". Why is unicode not defined, I can see "unicode = str" in the remote.py file.

After running this :

  import newspaper

  cnn_paper = newspaper.build('http://www.fox13now.com')
  

This error message appear :

File "C:\anaconda3\lib\tldextract\remote.py", line 56, in _decode_utf8
   if not isinstance(text, unicode):
**NameError: name 'unicode' is not defined**
In remote.py (from tldextract) :

This part seem to be the problem :

d```
ef _decode_utf8(text):
""" Decode from utf8 to Python unicode string.
The suffix list, wherever its origin, should be UTF-8 encoded.
"""
if not isinstance(text, unicode):
return unicode(text, 'utf-8')
else:
return text

Do you understand where the error came from?

[john-kurkowski]

It should be defined for Python 3 on this line. What version of Python are you using? Maybe we need a more robust version check.

[leocaza]

Python 3.6.0 |Continuum Analytics, Inc.| (default, Dec 23 2016, 11:57:41) [MSC v.1900 64 bit (AMD64)] on win32

[john-kurkowski]

I can't reproduce on

Python 3.6.0 (default, Apr 28 2017, 17:32:58)
[GCC 4.2.1 Compatible Apple LLVM 8.1.0 (clang-802.0.42)] on darwin

Maybe you have a copy of Python 2's urlparse installed in your Python 3.6.0? That would trick this version check. I think this warrants a more direct sys.version_info check, instead of try/except ImportError. Pythonistas love EAFP but yeesh it's indirect and unintuitive in this case.

[leocaza]

You are a genius. Thanks mate

[john-kurkowski]

Glad that worked around the issue for you! I'ma keep this open. The clever EAFP Python version check has overstayed it's welcome. Don't want it to bite someone again.