Transport Vite apps to Flask (or Quart)
pip install flask-vite-transporter
Note (Flask/Quart): When including credentials in fetch requests in the vite app. You must visit the serve app first to set the credentials.
For example, if the serve app is running on http://127.0.0.1:5001
, you must visit this address first.
This won't be needed in production, as it's expected that the Vite app will be served from the same domain.
The pyproject.toml file is used to store what Vite apps are available.
Adding the following to the pyproject.toml file will transfer all the Vite
apps listed in the vite_app_dirs
list to the serving app listed in the serve_app
key.
pyproject.toml
:
[tool.flask_vite_transporter]
npm_exec = "npm"
npx_exec = "npx"
serve_app = "app_flask"
vite_apps = ["frontend"]
The compiling of the Vite apps requires the npx
and npm
be
available. You can use absolute paths here.
npm_exec
is used to run npm install
if your Vite app does not
have the node_modules
folder.
npx
is used to run the Vite app build command.
serve_app
is the app that will serve the Vite compiled files.
vite_app_dirs
is a list of directories that contain Vite apps.
You can send over many Vite apps to the serving app, they will be accessible within template files.
You can see what apps can be compiled by running:
vt list
vt pack
This will create a dist
folder in each Vite app directory with the compiled files.
vt transport
This will move the compiled files to the serving app.
You can also run the pack
and transport
commands together:
vt pack transport
The Vite apps are compiled into a dist
folder, the files contained
in this folder are then moved to a folder called vite
in the serving app.
Any js file that is compiled that contains an asset reference will
replace assets/
with /--vite--/{app_name}
.
This requires that all assets in the Vite app stay in the assets
folder, and are imported in the
frontend project in a way that the Vite compile stage can find them.
The Vite apps can be compiled in different modes:
vt pack -m development
# or
vt pack -m your-named-mode
An example of pack
and transport
together:
vt pack -m dev transport
# or
vt pack transport -m dev
These mode values are accessible via import.meta.env.MODE
in the Vite app.
flask-vite-transporter creates a couple of Flask / Quart context processors that match the Vite apps to a Flask / Quart template.
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
{{ vt_head('app_vite_demo') }}
<title>Test</title>
</head>
<body>
{{ vt_body() }}
</body>
</html>
vt_head(
vite_app_name: str # The name of the Vite app to load
)
vt_body(
root_id: str = "root", # The id of the root element
noscript_message: str = "You need to enable JavaScript to run this app.",
)
from flask import Flask, render_template
from flask_vite_transporter import ViteTransporter
def create_app():
app = Flask(__name__)
ViteTransporter(app)
@app.route("/")
def index():
return render_template("index.html")
return app
from quart import Quart, render_template
from flask_vite_transporter.quart import QuartViteTransporter
def create_app():
app = Quart(__name__)
QuartViteTransporter(app)
@app.route("/")
async def index():
return await render_template("index.html")
return app
Setting:
ViteTransporter(app, cors_allowed_hosts=["http://127.0.0.1:5003"])
This is to allow the Vite app to communicate with the app.
Note: It's recommended to remove this in production.
We will be using a package call pyqwe
to run commands from the pyproject file.
Installing the development requirements will install pyqwe
:
pip install -r requirements/tests.txt
Use pyqwe
to install the local version of flask-vite-transporter:
pyqwe install
The serve_app
under tool.flask_vite_transporter
is currently set to use
the Flask demo app.
pyqwe flask_plus_vite
You should be able to visit the Flask app and the Vite app from the link in the terminal. Change something in the Vite app, save, then in a separate terminal run:
vt pack transport
The Vite app will be compiled, and the files will be moved to the Flask app. Visiting the Flask app from the link in the terminal should show the changes.