PythonのJSONファイルの読み取り–ファイルからJSONをロードしてダンプを解析する方法

ようこそ!PythonでJSONファイルを操作する方法を学びたい場合は、この記事が役に立ちます。

あなたは学びます:

  • JSON形式が非常に重要である理由。
  • その基本構造とデータ型。
  • JSONとPython辞書がPythonでどのように連携するか。
  • Python組み込み  jsonモジュールの操作方法。
  • JSON文字列をPythonオブジェクトに、またはその逆に変換する方法。
  • 使用方法loads()dumps()
  • JSON文字列を自動的にインデントする方法。
  • を使用してPythonでJSONファイルを読み取る方法 load()
  • を使用してPythonでJSONファイルに書き込む方法 dump()
  • もっと!

準備はできたか?さぁ、始めよう!✨

🔹はじめに:JSONとは何ですか?

JSON形式は、もともとJavaScript(Web開発に使用されるプログラミング言語)の構文に触発されました。しかしそれ以来、それは言語に依存しないデータ形式になり、今日使用しているプログラミング言語のほとんどはJSONを生成して読み取ることができます。

JSONの重要性とユースケース

JSONは基本的に、データを保存または表現するために使用される形式です。その一般的な使用例には、Web開発ファイルと構成ファイルが含まれます。

理由を見てみましょう:

  • Web開発: JSONは通常、サーバーからクライアントにデータを送信するために使用され、Webアプリケーションではその逆に使用されます。
  • 構成ファイル: JSONは、構成と設定の保存にも使用されます。たとえば、Google Chromeアプリを作成するにmanifest.jsonは、アプリの名前、説明、現在のバージョン、その他のプロパティと設定を指定するために呼び出されるJSONファイルを含める必要があります。  

🔸JSONの構造と形式

JSON形式が何に使用されるかがわかったので、ピザの注文のデータを表す例を使用して、その基本構造を見てみましょう。

{ "size": "medium", "price": 15.67, "toppings": ["mushrooms", "pepperoni", "basil"], "extra_cheese": false, "delivery": true, "client": { "name": "Jane Doe", "phone": null, "email": "[email protected]" } }

JSON形式の主な特徴は次のとおりです。

  • 中括弧で囲まれた一連のキーと値のペアがあります{}
  • 各キーは、次の形式を使用して特定の値にマップされます。
"key":  

💡ヒント:引用符が必要な値は、二重引用符で囲む必要があります。

  • キーと値のペアはコンマで区切ります。最後のペアのみの後にコンマはありません。
{ "size": "medium", # Comma! "price": 15.67 }

💡ヒント:通常、データを読みやすくするために、さまざまなレベルのインデントを使用してJSONをフォーマットします。この記事では、Pythonを使用してインデントを自動的に追加する方法を学習します。

JSONデータ型:キーと値

JSONファイルには、キーと値に有効なデータ型を決定する特定のルールがあります。

  • キーは文字列である必要があります。
  • は、文字列、数値、配列、ブール値(true/ falsenull、、またはJSONオブジェクトのいずれかです。

Pythonドキュメントによると:

JSONのキーと値のペアのキーは、常にタイプstrです。辞書がJSONに変換されると、辞書のすべてのキーが文字列に強制変換されます。

スタイルガイド

According to the Google JSON Style Guide:

  • Always choose meaningful names.
  • Array types should have plural key names. All other key names should be singular. For example: use "orders" instead of "order" if the corresponding value is an array.
  • There should be no comments in JSON objects.

🔹 JSON vs. Python Dictionaries

JSON and Dictionaries might look very similar at first (visually), but they are quite different. Let's see how they are "connected" and how they complement each other to make Python a powerful tool to work with JSON files.

JSON is a file format used to represent and store data whereas a Python Dictionary is the actual data structure (object) that is kept in memory while a Python program runs.

How JSON and Python Dictionaries Work Together

When we work with JSON files in Python, we can't just read them and use the data in our program directly. This is because the entire file would be represented as a single string and we would not be able to access the key-value pairs individually.

Unless...

We use the key-value pairs of the JSON file to create a Python dictionary that we can use in our program to read the data, use it, and modify it (if needed).

This is the main connection between JSON and Python Dictionaries. JSON is the string representation of the data and dictionaries are the actual data structures in memory that are created when the program runs.

Great. Now that you know more about JSON, let's start diving into the practical aspects of how you can work with JSON in Python.

🔸 The JSON Module

Luckily for us, Python comes with a built-in module called json. It is installed automatically when you install Python and it includes functions to help you work with JSON files and strings.

We will use this module in the coming examples.

How to Import the JSON Module

To use json in our program, we just need to write an import statement at the top of the file.

Like this:

With this line, you will have access to the functions defined in the module. We will use several of them in the examples.

💡 Tip: If you write this import statement, you will need to use this syntaxto call a function defined in the json module:

🔹 Python and JSON Strings

To illustrate how some of the most important functions of the json module work, we will use a multi-line string with JSON format.

JSON String

Particularly, we will use this string in the examples. It is just a regular multi-line Python string that follows the JSON format.

data_JSON = """ { "size": "Medium", "price": 15.67, "toppings": ["Mushrooms", "Extra Cheese", "Pepperoni", "Basil"], "client": { "name": "Jane Doe", "phone": "455-344-234", "email": "[email protected]" } } """
  • To define a multi-line string in Python, we use triple quotes.  
  • Then, we assign the string to the variable data_JSON.

💡 Tip: The Python Style Guide recommends using double quote characters for triple-quoted strings.  

JSON String to Python Dictionary

We will use the string with JSON format to create a Python dictionary that we can access, work with, and modify.

To do this, we will use the loads() function of the json module, passing the string as the argument.

This is the basic syntax:

Here is the code:

# Import the module import json # String with JSON format data_JSON = """ { "size": "Medium", "price": 15.67, "toppings": ["Mushrooms", "Extra Cheese", "Pepperoni", "Basil"], "client": { "name": "Jane Doe", "phone": "455-344-234", "email": "[email protected]" } } """ # Convert JSON string to dictionary data_dict = json.loads(data_JSON) 

Let's focus on this line:

data_dict = json.loads(data_JSON)
  • json.loads(data_JSON) creates a new dictionary with the key-value pairs of the JSON string and it returns this new dictionary.
  • Then, the dictionary returned is assigned to the variable data_dict.

Awesome! If we print this dictionary, we see this output:

{'size': 'Medium', 'price': 15.67, 'toppings': ['Mushrooms', 'Extra Cheese', 'Pepperoni', 'Basil'], 'client': {'name': 'Jane Doe', 'phone': '455-344-234', 'email': '[email protected]'}}

The dictionary has been populated with the data of the JSON string. Each key-value pair was added successfully.

Now let's see what happens when we try to access the values of the key-value pairs with the same syntax that we would use to access the values of a regular Python dictionary:

print(data_dict["size"]) print(data_dict["price"]) print(data_dict["toppings"]) print(data_dict["client"])

The output is:

Medium 15.67 ['Mushrooms', 'Extra Cheese', 'Pepperoni', 'Basil'] {'name': 'Jane Doe', 'phone': '455-344-234', 'email': '[email protected]'}

Exactly what we expected. Each key can be used to access its corresponding value.

💡 Tip: We can use this dictionary just like any other Python dictionary. For example, we can call dictionary methods, add, update, and remove key-value pairs, and more. We can even use it in a for loop.

JSON to Python: Type Conversion

When you use loads() to create a Python dictionary from a JSON string, you will notice that some values will be converted into their corresponding Python values and data types.

This table presented in the Python Documentation for the json module summarizes the correspondence from JSON data types and values to Python data types and values:

💡 Tip: The same conversion table applies when we work with JSON files.

Python Dictionary to JSON String

Now you know how to create a Python dictionary from a string with JSON format.

But sometimes we might need to do exactly the opposite, creating a string with JSON format from an object (for example, a dictionary) to print it, display it, store it, or work with it as a string.

To do that, we can use the dumps function of the json module, passing the object as argument:

💡 Tip: This function will return a string.

This is an example where we convert the Python dictionary client into a string with JSON format and store it in a variable:

# Python Dictionary client = { "name": "Nora", "age": 56, "id": "45355", "eye_color": "green", "wears_glasses": False } # Get a JSON formatted string client_JSON = json.dumps(client)

Let's focus on this line:

client_JSON = json.dumps(client)
  • json.dumps(client) creates and returns a string with all the key-value pairs of the dictionary in JSON format.
  • Then, this string is assigned to the client_JSON variable.

If we print this string, we see this output:

{"name": "Nora", "age": 56, "id": "45355", "eye_color": "green", "wears_glasses": false}

💡 Tip: Notice that the last value (false) was changed. In the Python dictionary, this value was False but in JSON, the equivalent value is false. This helps us confirm that, indeed, the original dictionary is now represented as a string with JSON format.

If we check the data type of this variable, we see:

So the return value of this function was definitely a string.

Python to JSON: Type Conversion

A process of type conversion occurs as well when we convert a dictionary into a JSON string. This table from the Python Documentation illustrates the corresponding values:

How to Print JSON With Indentation

If we use the dumps function and we print the string that we got in the previous example, we see:

{"name": "Nora", "age": 56, "id": "45355", "eye_color": "green", "wears_glasses": false}

But this is not very readable, right?

We can improve the readability of the JSON string by adding indentation.

To do this automatically, we just need to pass a second argument to specify the number of spaces that we want to use to indent the JSON string:

💡 Tip: the second argument has to be a non-negative integer (number of spaces) or a string. If indent is a string (such as "\t"), that string is used to indent each level (source).

Now, if we call dumps with this second argument:

client_JSON = json.dumps(client, indent=4)

The result of printing client_JSON is:

{ "name": "Nora", "age": 56, "id": "45355", "eye_color": "green", "wears_glasses": false }

That's great, right? Now our string is nicely formatted. This will be very helpful when we start working with files to store the data in a human-readable format.

How to Sort the Keys

You can also sort the keys in alphabetical order if you need to. To do this, you just need to write the name of the parameter sort_keys and pass the value True:

💡 Tip: The value of sort_keys is False by default if you don't pass a value.

For example:

client_JSON = json.dumps(client, sort_keys=True)

Returns this string with the keys sorted in alphabetical order:

{"age": 56, "eye_color": "green", "id": "45355", "name": "Nora", "wears_glasses": false}

How to Sort Alphabetically and Indent (at the same time)

To generate a JSON string that is sorted alphabetically and indented, you just need to pass the two arguments:

In this case, the output is:

{ "age": 56, "eye_color": "green", "id": "45355", "name": "Nora", "wears_glasses": false }

💡 Tip: You can pass these arguments in any order (relative to each other), but the object has to be the first argument in the list.

Great. Now you know how to work with JSON strings, so let's see how you can work with JSON files in your Python programs.

🔸 JSON and Files

Typically, JSON is used to store data in files, so Python gives us the tools we need to read these types of file in our program, work with their data, and write new data.

💡 Tip: a JSON file has a .json extension:

Let's see how we can work with .json files in Python.

How to Read a JSON File in Python

Let's say that we created an orders.json file with this data that represents two orders in a pizza shop:

{ "orders": [ { "size": "medium", "price": 15.67, "toppings": ["mushrooms", "pepperoni", "basil"], "extra_cheese": false, "delivery": true, "client": { "name": "Jane Doe", "phone": null, "email": "[email protected]" } }, { "size": "small", "price": 6.54, "toppings": null, "extra_cheese": true, "delivery": false, "client": { "name": "Foo Jones", "phone": "556-342-452", "email": null } } ] }

Please take a moment to analyze the structure of this JSON file.

Here are some quick tips:

  • Notice the data types of the values, the indentation, and the overall structure of the file.
  • The value of the main key "orders" is an array of JSON objects (this array will be represented as list in Python). Each JSON object holds the data of a pizza order.

If we want to read this file in Python, we just need to use a with statement:

💡 Tip: In the syntax above, we can assign any name to file (green box). This is a variable that we can use within the with statement to refer to the file object.

The key line of code in this syntax is:

data = json.load(file)
  • json.load(file) creates and returns a new Python dictionary with the key-value pairs in the JSON file.
  • Then, this dictionary is assigned to the data variable.

💡 Tip: Notice that we are using load() instead of loads(). This is a different function in the json module. You will learn more about their differences at the end of this article.

Once we have the content of the JSON file stored in the data variable as a dictionary, we can use it to do basically anything we want.

Examples

For example, if we write:

print(len(data["orders"]))

The output is 2 because the value of the main key "orders" is a list with two elements.

We can also use the keys to access their corresponding values. This is what we typically do when we work with JSON files.

For example, to access the toppings of the first order, we would write:

data["orders"][0]["toppings"]
  • First, we select the main key "orders"
  • Then, we select the first element in the list (index 0).
  • Finally, we select the value that corresponds to the key "toppings"

You can see this "path" graphically in the diagram:

If we print this value, the output is:

['mushrooms', 'pepperoni', 'basil']

Exactly what we expected. You just need to "dive deeper" into the structure of the dictionary by using the necessary keys and indices. You can use the original JSON file/string as a visual reference. This way, you can access, modify, or delete any value.

💡 Tip: Remember that we are working with the new dictionary. The changes made to this dictionary will not affect the JSON file. To update the content of the file, we need to write to the file.

How to Write to a JSON File

Let's see how you can write to a JSON file.

The first line of the with statement is very similar. The only change is that you need to open the file in 'w' (write) mode to be able to modify the file.

💡 Tip: If the file doesn't exist already in the current working directory (folder), it will be created automatically. By using the 'w' mode, we will be replacing the entire content of the file if it already exists.

There are two alternative ways to write to a JSON file in the body of the with statement:

  • dump
  • dumps

Let's see them in detail.

First Approach: dump

This is a function that takes two arguments:

  • The object that will be stored in JSON format (for example, a dictionary).
  • The file where it will be stored (a file object).

Let's say that the pizza shop wants to remove the clients' data from the JSON file and create a new JSON file called orders_new.json with this new version.

We can do this with this code:

# Open the orders.json file with open("orders.json") as file: # Load its content and make a new dictionary data = json.load(file) # Delete the "client" key-value pair from each order for order in data["orders"]: del order["client"] # Open (or create) an orders_new.json file # and store the new version of the data. with open("orders_new.json", 'w') as file: json.dump(data, file)

This was the original version of the data in the orders.json file. Notice that the "client" key-value pair exists.

{ "orders": [ { "size": "medium", "price": 15.67, "toppings": ["mushrooms", "pepperoni", "basil"], "extra_cheese": false, "delivery": true, "client": { "name": "Jane Doe", "phone": null, "email": "[email protected]" } }, { "size": "small", "price": 6.54, "toppings": null, "extra_cheese": true, "delivery": false, "client": { "name": "Foo Jones", "phone": "556-342-452", "email": null } } ] } 

This is the new version in the orders_new.json file:

{"orders": [{"size": "medium", "price": 15.67, "toppings": ["mushrooms", "pepperoni", "basil"], "extra_cheese": false, "delivery": true}, {"size": "small", "price": 6.54, "toppings": null, "extra_cheese": true, "delivery": false}]}

If you analyze this carefully, you will see that the "clients" key-value pair was removed from all the orders.

However, there is something missing in this file, right?

Please take a moment to think about this... What could it be?

Indentation, of course!

The file doesn't really look like a JSON file, but we can easily fix this by passing the argument indentation=4 to dump().

Now the content of the file looks like this:

{ "orders": [ { "size": "medium", "price": 15.67, "toppings": [ "mushrooms", "pepperoni", "basil" ], "extra_cheese": false, "delivery": true }, { "size": "small", "price": 6.54, "toppings": null, "extra_cheese": true, "delivery": false } ] }

What a difference! This is exactly what we would expect a JSON file to look like.

Now you know how to read and write to JSON files using load() and dump(). Let's see the differences between these functions and the functions that we used to work with JSON strings.  

🔹 load() vs. loads()

This table summarizes the key differences between these two functions:

💡 Tip: Think of loads() as "load string" and that will help you remember which function is used for which purpose.

🔸 dump() vs. dumps()

Here we have a table that summarizes the key differences between these two functions:

💡 Tip: Think of dumps() as a "dump string" and that will help you remember which function is used for which purpose.

🔹 Important Terminology in JSON

Finally, there are two important terms that you need to know to work with JSON:

  • Serialization: converting an object into a JSON string.
  • Deserialization: converting a JSON string into an object.

🔸 In Summary

  • JSON (JavaScript Object Notation) is a format used to represent and store data.
  • It is commonly used to transfer data on the web and to store configuration settings.
  • JSON files have a .json extension.
  • You can convert JSON strings into Python objects and vice versa.
  • You can read JSON files and create Python objects from their key-value pairs.
  • You can write to JSON files to store the content of Python objects in JSON format.

I really hope you liked my article and found it helpful. Now you know how to work with JSON in Python. Follow me on Twitter @EstefaniaCassN and check out my online courses.