15

I'm working with the following data:

[{"title": null, "metric1": 361429, "metric2": 36,},{"title": null, "metric1": 253798, "metric2": 48}]

When I attempt to assign this data to a variable in Python (with the aim of parsing it out), I receive the following error message:

Traceback (most recent call last):
File "<stdin>", line 1, in <module>
NameError: name 'null' is not defined

From my research, it appears that the None is Python's null. What I'm wondering is, is it possible to change the null's in my data to None's using Python?

I've tried creating a string out of the data, assigning it to data, and replacing the null's that way:

data = data.replace('null','None')

but that results in a string of the data itself:

data = '[{"title": None, "metric1": 361429, "metric2": 36,},{"title": None, "metric1": 253798, "metric2": 48}]'

and I can't figure out how to turn it from a string back into JSON.

EDIT: I am copying and pasting this data into the Python interpreter from a separate source.

Daniel
  • 2,345
  • 4
  • 19
  • 36
  • 3
    The correct way to convert a JSON string into a python object is [json.loads](https://docs.python.org/2/library/json.html#json.loads) -- no string munging necessary. – kevingessner Mar 03 '15 at 23:19
  • That's not a json string. The comma at the end of `"metric2": 36,` shouldn't be there. – tdelaney Mar 03 '15 at 23:59
  • 3
    It seems like you are pasting this into the python interpreter (which means its python syntax, not JSON). Is that the case, and if so, why not mention it? – tdelaney Mar 04 '15 at 00:02
  • @tdelaney you are correct! I am copying and pasting into the interpreter. Apologies for not mentioning; it wasn't on purpose, just a detail that slipped my mind. Good call. – Daniel Mar 09 '15 at 21:37

2 Answers2

46

Much simpler!

Just assign None to null before assigning that list to a variable:

null = None
var = [{"title": null, "metric1": 361429, "metric2": 36,},{"title": null, "metric1": 253798, "metric2": 48}]

Then you won't need to do the rather unnecessary conversion to a string (and back to a Python object with json.loads) only to replace null by None.

But that is only really necessary if you're copy-pasting that code from some source. Otherwise, the canonical answer is to use json.loads (or json.load).

Oliver W.
  • 13,169
  • 3
  • 37
  • 50
13

As is mentioned above, you don't need to replace "null" for "None"

Just

import json
parsed_data = json.loads(data)
levi
  • 22,001
  • 7
  • 73
  • 74
  • Thanks @levi! This makes perfect sense. Will do for future json and python endeavors. +1 for your answer. – Daniel Mar 09 '15 at 21:40