Dealing With a TypeError in Python 3

I recently upgraded to Python 3 (I know, took me a while!) and realized that one of my scripts that writes JSON to a file no longer works!

This is a simplified version of what I’m doing:

>>> import json
>>> x = {"mark": {"name": "Mark"}, "michael": {"name": "Michael"}  } 
>>> json.dumps(x.values())
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/local/Cellar/python3/3.6.0/Frameworks/Python.framework/Versions/3.6/lib/python3.6/json/__init__.py", line 231, in dumps
    return _default_encoder.encode(obj)
  File "/usr/local/Cellar/python3/3.6.0/Frameworks/Python.framework/Versions/3.6/lib/python3.6/json/encoder.py", line 199, in encode
    chunks = self.iterencode(o, _one_shot=True)
  File "/usr/local/Cellar/python3/3.6.0/Frameworks/Python.framework/Versions/3.6/lib/python3.6/json/encoder.py", line 257, in iterencode
    return _iterencode(o, 0)
  File "/usr/local/Cellar/python3/3.6.0/Frameworks/Python.framework/Versions/3.6/lib/python3.6/json/encoder.py", line 180, in default
    o.__class__.__name__)
TypeError: Object of type 'dict_values' is not JSON serializable

Python 2.7 would be perfectly happy:

>>> json.dumps(x.values())
'[{"name": "Michael"}, {"name": "Mark"}]'

The difference is in the results returned by the values method:

# Python 2.7.10
>>> x.values()
[{'name': 'Michael'}, {'name': 'Mark'}]

# Python 3.6.0
>>> x.values()
dict_values([{'name': 'Mark'}, {'name': 'Michael'}])
>>>

Python 3 no longer returns an array. Instead, we have a dict_values wrapper around the data.

Luckily this is easy to resolve; we just need to wrap the call to values with a call to list:

>>> json.dumps(list(x.values()))
'[{"name": "Mark"}, {"name": "Michael"}]'

This versions works with Python 2.7, as well, so if I accidentally run the script with an old version, the world isn’t going to explode!