JSON is a popular way of storing data in a key/value type arrangement so that the data can be parsed easily later. For instance, here is a very simple JSON snippet:

{
"name":"tux",
"health":"23",
"level":"4"
}

If you are like me, three questions probably spring to your mind:

1. That looks an awful lot like a Python dictionary.

   Yes, it looks exactly like a Python dictionary. They are shockingly similar. If you are comfortable with Python lists and dictionaries, you will feel right at home with JSON.

2. I don't feel comfortable with dictionaries, can't I just use a delimited text file?

   You can, but you will have to write parsers for it yourself. If your data gets very complex, the parsing can get pretty ugly.

   That is not to say that you should not use a simple delimited text file if that is all that your programme needs. For example, I would not want to open a config file as a user and find that I have to format all my options as valid JSON.

   Just know that JSON is out there and available, and that the JSON Python module has some little features that make your life easier when dealing with sets of data.

3. Why not use XML instead?

   You can. Mostly one should use the most appropriate format for one's project. I'm a big fan of XML, but sometimes JSON makes more sense.

I am not going to make this post about teaching the JSON format. If you need clarification on how to structure data into JSON, go through a tutorial on it somewhere; there are several good ones online. Honestly, it's not that complex; you can think of JSON as nested dictionaries.

Starting from scratch, let's say that you write a programme that by nature gathers data as it runs. When the user quits, you want to save the data to a file so that when the user resumes the app later, they can load the file back in and pick up where they left off.

Storing Data as JSON

At its most basic, the JSON data structure is basically the same as a Python dictionary, and in fact the nice thing about JSON is that it can be directly imported into a Python dictionary. Usually, however, you are resorting to JSON because you have somewhat complex data, so in the sample code we will use a dictionary-within-a-dictionary:

#!/usr/bin/env python

game = {'tux': {'health': 23, 'level': 4}, 'beastie': {'health': 13, 'level': 6}}
# you can always add more to your dictionary

game['konqi'] = {'health': 18, 'level': 7}

That code creates a ditionary called game which stores the player name and a corresponding dictionary of attributes about how the player is doing in the progress of the game. As you can see after the comment, adding new players is simple.

Now let's see how to save that data to a save file.

## continued...
import json

with open('dosiero.json', 'w') as outfile:
    json.dump(game, outfile)

That would be your save command. Simple as that, all the structured content of your game dictionary is committed to a file on your hard drive.

Reading Data from a JSON File

If you are saving data to JSON, you probably will evenually want to read the data back into Python. For this, Python features the function json.load

import json

dosiero = open('dosiero.json')
game = json.load(dosiero)

print game['tux']     # prints {'health': 23, 'level': 4}
print game['tux']['health']    # prints 23
print game['tux']['level']     # prints 4

# when finished, close the file

json_data.close()

As you can see, JSON integrates surprisingly well with Python, so it's a great format when your data fits in with its model.

Have fun!

[EOF]

Made with Free Software.