Working with JSON Files#

Goals of this lecture#

We’ve already discussed the basics of reading and writing text (.txt) files.

But another common file type is a .json file, which allows you to store structured data.

This lecture will cover:

  • What is a .json file? What is JSON more generally and why is it useful?

  • How do we read in a .json file?

  • How do we write a .json file?

  • JSON files vs. JSON strings.

What is a .json file?#

A .json file is a file written in the JSON file format. It allows us to store structured data objects consisting of key-value pairs.

What is JSON?#

JSON = JavaScript Object Notation.

  • Standard format for representing and transmitting data.

    • “Standard” = different people/systems agree to use this format to send and receive information.

  • Represents data in key-value pairs.

Check-in#

What else have we seen that represents data in key-value pairs?

A Python dict is a collection of key-value pairs#

A dictionary (dict) stores key-value pairs.

my_class = {'Code': '1',
           'Department': 'CSS',
           'Instructor': 'Trott',
            'Prerequisite': True,
           'Enrollment': 120}
print(my_class)

{'Code': '1', 'Department': 'CSS', 'Instructor': 'Trott', 'Prerequisite': True, 'Enrollment': 120}

my_class['Department']

'CSS'

JSON and dict: an analogy#

Conceptually, JSON accomplishes the same goals as a Python dict.

  • In fact, Python programmers often convert a dict into a JSON str when they want to store it in a file.

  • Similarly, you can read in a .json file and convert the contents into a dict.

Bottom line: we’re not dealing with a fundamentally new data sturcture––it’s another standardized way to represent key-value pairs.

Reading in a .json file#

Reading in a .json file shares some similarities with reading .txt files.

  • Must specify a file path.

  • File path can be either absolute or relative.

But there are also some important differences:

  • To read in a .json file, we’ll need to import the json library.

  • json.load will read in a structured .json file as a dict, not a str.

Example: simple file#

Here, we will work with a simple .json file: data/restaurant.json.

  • The file contains a structured representation of a restaurant.

  • We use json.load(...) to load this representation as a dict.

## This imports the json library
import json

## As with normal .txt. files, we use "open" to open the target restaurant
with open("data/restaurant.json", "r") as fp:
    ## use json.load to load as dict
    info = json.load(fp)

info

{'Name': 'Plumeria', 'Location': 'University Heights', 'Cuisine': 'Thai'}

load creates a dict#

Now, we can work with the contents of this file as we would any dict.

info['Name']

'Plumeria'

info['Location']

'University Heights'

info['Cuisine']

'Thai'

Check-in#

Try reading in another file that’s stored in data: data/school.json.

What is the value of the Name key?

### Your code here

Solution#

## As before, we use "open" to open the target file
with open("data/school.json", "r") as fp:
    ## use json.load to load as dict
    school_info = json.load(fp)

## Get name of school
school_info['Name']

'UCSD'

Writing a .json file#

Often, you’ll want to write a structured dict to a file.

  • Useful for storing information, so you can access it later.

  • Useful for transmitting information between programs.

We can use json.dump(...) to write (or “dump”) a dict into a .json file.

Simple example: course#

To start out, let’s use the my_class dict we defined earlier.

my_class['Code']

'1'

To write this to a file, we:

  • open (create) a file with the name we want to call it.

  • Use json.dump(dict_name, filename).

with open("course.json", "w") as fp:
    json.dump(my_class, fp)

Checking that this worked#

with open("course.json", "r") as fp:
    course_info = json.load(fp)
print(course_info)

{'Code': '1', 'Department': 'CSS', 'Instructor': 'Trott', 'Prerequisite': True, 'Enrollment': 120}

Check-in#

Create a new dict called my_info. Add the following keys/values:

  • Name.

  • Major.

Then, use json.dump to write this dict to a .json file called my_info.json to your own computer (in whichever directory you prefer).

### Your code here

Solution#

my_info = {'Name': 'Sean', 'Major': 'Cognitive Science'}
with open("data/my_info.json", "w") as fp:
    json.dump(my_info, fp)

JSON files vs. JSON strings#

The load and dump methods can be used to read and write a dict from/to a .json file.

However, Python can also represent JSON as a str.

  • To read a dict from a JSON str, use loads (load + string).

  • To write a dict into a JSON str, use dumps (dump + string).

json.dumps#

  • Input: a dict.

  • Output: a JSON str.

json_str = json.dumps(my_class)
json_str

'{"Code": "1", "Department": "CSS", "Instructor": "Trott", "Prerequisite": true, "Enrollment": 120}'

type(json_str)

str

json.loads#

  • Input: a JSON str.

  • Output: a dict.

Other objects besides dicts#

  • Technically, you can use dumps/loads for other objects, such as str, list, and more.

  • Though in my experience, a dict is the most common format.

json.dumps([1, 2, 3])

'[1, 2, 3]'

json.loads('[1, 2, 3]')

[1, 2, 3]

Conclusion#

This was a brief introduction to working with .json files. Hopefully you have a better handle on:

  • What is a .json file?

  • How do I read and write .json files?