Application Contexts

Describing the Application Context

• An application context tracks application-level data from:

  • A request
  • A CLI command
  • Other activity
• Flask avoids passing an application to each function
• It does this by creating objects g and current_app
• These are proxies that can be accessed at any point

Issue with Importing app Instances

• Every Flask application object has attributes
• The config attribute is one example
• Attributes are useful to access within views and CLI commands
• However, importing an app instance within a module is prone to circular imports
• An app instance can't be imported using the app factory pattern
• Flask solves this problem with the application context

The Purpose of the Application Context

• With application contexts, we don't refer to an app directly
• Instead, we use a current_app proxy
• This points to the application handling the current activity
• Flask automatically pushes an application context when handling a request

• The following functions have access to current_app:

  • View functions
  • Error handlers
  • Other functions

Defining the Structure of Application Contexts

• There is a stack that manages application contexts

• This stack is referred to as the application context stack

  • In the code, this is _app_ctx_stack

• Eac element in the stack is a single application context

  • In the code, this is an AppContext class

• Each application context is essentially a list of two objects:

  • A single g object
  • A single current_app object

Implementing an Application Context in Flask

• Flask needs a Flask object to create an application context
• Flask needs an environment dictionary to create a request context
• The environment dictionary has all the client data and is taken from the browser
• The applications and request context are both created simultaneously

• The steps include the following:

  1. A Flask object is created
  2. An application context is created when an app variable is assigned
  3. A request context is then created when getting an environment dictionary
  4. View functions are invoked
  5. Gain access to objects g, current_app, request, and session

Lifetieme of the Context

• The application context is created and destroyed as necessary
• A Flask application begins by handling a request
• Then, it pushes an application context and a request context

• Once a request ends, it does the following:

  • Pops the request context
  • Then, pops the application context
• Typically, an application context will have the same lifetime as a request

Storing Data with the Application Context

• The application context is used for storing common data
• Specifically, this data refers to a request or CLI command

• Flask provides the g object for this purpose

  • g stands for global
  • This refers to data being global within a context
• The data on g is lost after the context ends

• Implying, it is not used for storing data between requests

  • To store data across requests, we can use session or store data in a database
• It is a simple namespace object that has the same lifetime as an application context

References

• Basics of the Application Context
• Describing App Factory Patterns