Skip to content


Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?

Latest commit

Thanks to injector 0.20.0 invalid forward references in return type
annotation positions no longer break injection/detecting dependencies
(which is what GH-69 is actually about).

This PR fixes #69

This PR also effectively reverts [1] as [1] was breaking forward
reference support.

One more thing being fixed here is class-based view failures
that started happening for some reason, reordering some internal
logic handles that (we prioritize treating views as class-based views
and all is well).

[1] Fixes: 4d8bfe8 ("(wip) fix: name 'Response' is not defined")

Git stats


Failed to load latest commit information.
Latest commit message
Commit time


Adds Injector support to Flask, this way there's no need to use global Flask objects, which makes testing simpler.

Injector is a dependency-injection framework for Python, inspired by Guice. You can find Injector on PyPI and Injector documentation on Read the Docs.

Flask-Injector is compatible with CPython 3.7+. As of version post-0.13.0 it requires Injector version 0.20.0 or greater and Flask 2.1.2 or greater.

GitHub project page:

PyPI package page:



Flask-Injector lets you inject dependencies into:

  • views (functions and class-based)
  • before_request handlers
  • after_request handlers
  • teardown_request handlers
  • template context processors
  • error handlers
  • Jinja environment globals (functions in app.jinja_env.globals)
  • Flask-RESTFul Resource constructors
  • Flask-RestPlus Resource constructors
  • Flask-RESTX Resource constructors

Flask-Injector supports defining types using function annotations (Python 3), see below.


As Flask-Injector uses Injector under the hood you should find the Injector documentation, including the Injector API reference, helpful. The Injector README provides a tutorial-level introduction to using Injector.

The Flask-Injector public API consists of the following:

  • FlaskInjector class with the constructor taking the following parameters:
    • app, an instance of`flask.Flask` [mandatory] – the Flask application to be used
    • modules, an iterable of Injector modules [optional] – the Injector modules to be used.
    • injector, an instance of injector.Injector [optional] – an instance of Injector to be used if, for some reason, it's not desirable for FlaskInjector to create a new one. You're likely to not need to use this.
    • request_scope_class, an injector.Scope subclass [optional] – the scope to be used instead of RequestScope. You're likely to need to use this except for testing.
  • RequestScope class – an injector.Scope subclass to be used for storing and reusing request-scoped dependencies
  • request object – to be used as a class decorator or in explicit bind() calls in Injector modules.

Creating an instance of FlaskInjector performs side-effectful configuration of the Flask application passed to it. The following bindings are applied (if you want to modify them you need to do it in one of the modules passed to the FlaskInjector constructor):

  • flask.Flask is bound to the Flask application in the (scope: singleton)
  • flask.Config is bound to the configuration of the Flask application
  • flask.Request is bound to the current Flask request object, equivalent to the thread-local flask.request object (scope: request)

Example application using Flask-Injector

import sqlite3
from flask import Flask, Config
from flask.views import View
from flask_injector import FlaskInjector
from injector import inject

app = Flask(__name__)

# Configure your application by attaching views, handlers, context processors etc.:

def bar():
    return render("bar.html")

# Route with injection
def foo(db: sqlite3.Connection):
    users = db.execute('SELECT * FROM users').all()
    return render("foo.html")

# Class-based view with injected constructor
class Waz(View):
    def __init__(self, db: sqlite3.Connection):
        self.db = db

    def dispatch_request(self, key):
        users = self.db.execute('SELECT * FROM users WHERE name=?', (key,)).all()
        return 'waz'

app.add_url_rule('/waz/<key>', view_func=Waz.as_view('waz'))

# In the Injector world, all dependency configuration and initialization is
# performed in modules (
# The same is true with Flask-Injector. You can see some examples of configuring
# Flask extensions through modules below.

# Accordingly, the next step is to create modules for any objects we want made
# available to the application. Note that in this example we also use the
# Injector to gain access to the `flask.Config`:

def configure(binder):

# Initialize Flask-Injector. This needs to be run *after* you attached all
# views, handlers, context processors and template globals.

FlaskInjector(app=app, modules=[configure])

# All that remains is to run the application

See for a more complete example, including Flask-SQLAlchemy and Flask-Cache integration.

Supporting Flask Extensions

Typically, Flask extensions are initialized at the global scope using a pattern similar to the following.

app = Flask(__name__)
ext = ExtClass(app)

def view():
    # Use ext object here...

As we don't have these globals with Flask-Injector we have to configure the extension the Injector way - through modules. Modules can either be subclasses of injector.Module or a callable taking an injector.Binder instance.

from injector import Module

class MyModule(Module):
    def provide_ext(self, app: Flask) -> ExtClass:
        return ExtClass(app)

def main():
    app = Flask(__name__)

    # attach your views etc. here

    FlaskInjector(app=app, modules=[MyModule])

Make sure to bind extension objects as singletons.