Awesome
Inertia.js FastAPI Adapter
<!-- TOC -->Installation
You can install the package via pip:
pip install fastapi-inertia
Configuration
You can configure the adapter by passing a InertiaConfig
object to the Inertia
class.
The following options are available:
key | default | options | description |
---|---|---|---|
environment | development | development,production | The environment to use |
version | 1.0.0 | Any valid string | The version of your server |
json_encoder | InertiaJsonEncoder | Any class that extends json.JSONEncoder | The JSON encoder used to encode page data when HTML is returned |
manifest_json_path | "" | Any valid path | The path to the manifest.json file. Needed in production |
dev_url | http://localhost:5173 | Any valid url | The URL to the development server |
ssr_url | http://localhost:13714 | Any valid url | The URL to the SSR server |
ssr_enabled | False | True,False | Whether to enable SSR. You need to install the httpx package, to have set the manifest_json_path and started the SSR server |
root_directory | src | Any valid path | The directory in which is located the javascript code in your frontend. Will be used to find the relevant files in your manifest.json. |
entrypoint_filename | main.js | Any valid file | The entrypoint for you frontend. Will be used to find the relevant files in your manifest.json. |
assets_prefix | "" | Any valid string | An optional prefix for your assets. Will prefix the links generated from the assets mentioned in manifest.json. |
use_flash_messages | False | True,False | Whether to use flash messages. You need to use Starlette's SessionMiddleware to use this feature |
flash_message_key | messages | Any valid string | The key to use for flash errors |
use_flash_errors | False | True,False | Whether to use flash errors |
flash_error_key | errors | Any valid string | The key to use for flash errors |
templates | None | A Jinja2Templates instance | The templates instance in which Inertia will look for the root_template_filename template |
root_template_filename | index.html | Any valid jinja2 template file | The file which will be used to render your inertia application |
Examples
You can see different full examples in the examples
directory
Usage
Create a Jinja2Template
In order to use the Inertia.js adapter, you have to create a Jinja2Template that the library will use.
It must have both an inertia_head
and an inertia_body
tag in it.
inertia_head
is where the library will place the code that supposedly goes inside the HTMLhead
taginertia_body
is where the library will place the code that supposedly goes inside the HTMLbody
tag
You can find the simplest example in inertia/tests/templates/index.html
.
You should then register the folder in which you put this file as the directory of your Jinja2Templates
templates = Jinja2Templates(directory=template_dir)
This option should be passed to the InertiaConfig class presented below, under the templates
key.
If you choose a different template file name than index.html
, you can also pass the root_template_filename
key with, as value, your template file name.
Set up the dependency
This Inertia.js adapter has been developed to be used as a FastAPI dependency. To use it, you first need to set up the dependency, with your desired configuration.
inertia_dependency.py
from fastapi import Depends
from typing import Annotated
from inertia import InertiaConfig, inertia_dependency_factory, Inertia
inertia_config = InertiaConfig(
# Your desired configuration
)
inertia_dependency = inertia_dependency_factory(
inertia_config
)
InertiaDependency = Annotated[Inertia, Depends(inertia_dependency)]
You can then access the InertiaDependency
in your route functions, and use it to render your pages.
Rendering a page
To render a page, you can use the render
method of the Inertia
class. It takes two arguments:
- The name of the page
- The data to pass to the page
main.py
from fastapi import FastAPI, Depends
from inertia import InertiaResponse, InertiaVersionConflictException, inertia_version_conflict_exception_handler
from inertia_dependency import InertiaDependency
app = FastAPI()
app.add_exception_handler(InertiaVersionConflictException, inertia_version_conflict_exception_handler)
@app.get('/', response_model=None)
async def index(inertia: InertiaDependency) -> InertiaResponse:
return inertia.render('Index', {
'name': 'John Doe'
})
Rendering assets
As your front-end framework likely references assets that are not served by FastAPI, you need to mount a static directory to serve these assets.
main.py
import os
from fastapi import FastAPI
from fastapi.staticfiles import StaticFiles
from inertia_dependency import inertia_config
app = FastAPI()
webapp_dir = (
os.path.join(os.path.dirname(__file__), "..", "webapp", "dist")
if inertia_config.environment != "development"
else os.path.join(os.path.dirname(__file__), "..", "webapp", "src")
)
app.mount("/src", StaticFiles(directory=webapp_dir), name="src")
app.mount(
"/assets", StaticFiles(directory=os.path.join(webapp_dir, "assets")), name="assets"
)
Sharing data
To share data, in Inertia, is basically to add data before even entering your route. This is useful, for example, to add a user to all your pages that expects your user to be logged in.
main.py
from fastapi import FastAPI, Depends
from inertia import InertiaResponse, InertiaVersionConflictException, inertia_version_conflict_exception_handler
from inertia_dependency import InertiaDependency
app = FastAPI()
app.add_exception_handler(InertiaVersionConflictException, inertia_version_conflict_exception_handler)
def current_user(inertia: InertiaDependency):
inertia.share(user={
'name': 'John Doe'
})
@app.get('/', response_model=None, dependencies=[Depends(current_user)])
async def index(inertia: InertiaDependency) -> InertiaResponse:
"""
Because of the dependency, and as we are sharing the user data, the user data will be available in the page.
"""
return inertia.render('Index')
Flash messages
With the inertia dependency, you have access to a flash
helper method that allows you to add flash messages to your pages.
This is useful to display messages to the user after a form submission, for example.
Those messages are called flash
messages as they are only displayed once.
You need to have set use_flash_messages
to True
in your configuration to use this feature.
You need to have the SessionMiddleware
enabled in your application to use this feature.
main.py
from fastapi import FastAPI, Depends
from starlette.middleware.sessions import SessionMiddleware
from inertia import InertiaResponse, InertiaVersionConflictException, inertia_version_conflict_exception_handler
from inertia_dependency import InertiaDependency
app = FastAPI()
app.add_exception_handler(InertiaVersionConflictException, inertia_version_conflict_exception_handler)
app.add_middleware(SessionMiddleware, secret_key="secret")
@app.get('/', response_model=None)
async def index(inertia: InertiaDependency) -> InertiaResponse:
inertia.flash('Index was reached successfully', category='success')
return inertia.render('Index')
Flash errors
If you handle form submissions in your application, and if you do all validation at the pydantic level,
a malformed payload will raise a RequestValidationError
exception.
You can use the inertia_request_validation_exception_handler
to handle this exception and display the errors to the user.
It supports error bags, so you can display multiple errors at once.
If the request is not from Inertia, it will fallback to FastAPI's default error handling.
In order to use this feature, you need to have set use_flash_errors
to True
in your configuration.
You also need to have the SessionMiddleware
enabled in your application to use this feature.
main.py
from fastapi import FastAPI, Depends
from pydantic import BaseModel, model_validator
from typing import Any
from fastapi.exceptions import RequestValidationError
from starlette.middleware.sessions import SessionMiddleware
from inertia import InertiaResponse, InertiaVersionConflictException, inertia_version_conflict_exception_handler, inertia_request_validation_exception_handler
from inertia_dependency import InertiaDependency
app = FastAPI()
app.add_exception_handler(InertiaVersionConflictException, inertia_version_conflict_exception_handler)
app.add_exception_handler(RequestValidationError, inertia_request_validation_exception_handler)
app.add_middleware(SessionMiddleware, secret_key="secret")
class Form(BaseModel):
name: str
@model_validator(mode="before")
@classmethod
def name_must_contain_doe(cls, data: Any):
if 'Doe' not in data.name:
raise ValueError('Name must contain Doe')
@app.post('/', response_model=None)
async def index(data: Form, inertia: InertiaDependency) -> InertiaResponse:
return inertia.render('Index')
Redirect to an external URL
If you want to redirect the user to an external URL, you can use the location
method of the Inertia
class.
It takes one argument: the URL to redirect to.
main.py
from fastapi import FastAPI, Depends
from inertia import InertiaResponse, InertiaVersionConflictException, inertia_version_conflict_exception_handler
from inertia_dependency import InertiaDependency
app = FastAPI()
app.add_exception_handler(InertiaVersionConflictException, inertia_version_conflict_exception_handler)
@app.get('/', response_model=None)
async def index(inertia: InertiaDependency) -> InertiaResponse:
return inertia.location('https://google.fr')
Redirect back
If you want to redirect the user back (for example, after a form submission), you can use the back
method of the Inertia
class.
It will use the Referer
header to redirect the user back.
If you're on a GET
request, the status code will be 307
. Otherwise, it will be 303
.
That ways, it will trigger a new GET request to the referer URL.
main.py
from fastapi import FastAPI, Depends
from inertia import InertiaResponse, InertiaVersionConflictException, inertia_version_conflict_exception_handler
from inertia_dependency import InertiaDependency
app = FastAPI()
app.add_exception_handler(InertiaVersionConflictException, inertia_version_conflict_exception_handler)
@app.get('/', response_model=None)
async def index(inertia: InertiaDependency) -> InertiaResponse:
return inertia.back()
Enable SSR
To enable SSR, you need to set ssr_enabled
to True
in your configuration.
You also need to have set the manifest_json_path
to the path of your manifest.json
file.
You need to have the httpx
package installed to use this feature.
This can be done through the following command:
pip install httpx
Frontend documentation
There is no particular caveats to keep in mind when using this adapter. However, here's an example of how you would set up your frontend to work with this adapter.
For a classic build
[!NOTE]
To build the project, you can run thevite build
command
vite.config.js
import { fileURLToPath } from "node:url";
import { dirname } from "path";
import { defineConfig } from "vite";
import vue from "@vitejs/plugin-vue";
const projectRoot = dirname(fileURLToPath(import.meta.url));
// https://vitejs.dev/config/
export default defineConfig({
plugins: [vue()],
resolve: {
alias: {
"@": `${projectRoot}/src`,
},
},
build: {
manifest: "manifest.json",
outDir: "dist",
rollupOptions: {
input: "src/main.js",
},
},
});
main.js
import { createApp, h } from "vue";
import { createInertiaApp } from "@inertiajs/vue3";
createInertiaApp({
resolve: (name) => {
const pages = import.meta.glob("./Pages/**/*.vue", { eager: true });
return pages[`./Pages/${name}.vue`];
},
setup({ el, App, props, plugin }) {
createApp({ render: () => h(App, props) })
.use(plugin)
.mount(el);
},
});
For a SSR build
[!NOTE]
To build the project, you can run thevite build
andvite build --ssr
commands
To serve the Inertia SSR server, you can run thenode dist/ssr/ssr.js
command
vite.config.js
import { fileURLToPath } from "node:url";
import { dirname } from "path";
import { defineConfig } from "vite";
import vue from "@vitejs/plugin-vue";
const projectRoot = dirname(fileURLToPath(import.meta.url));
// https://vitejs.dev/config/
export default defineConfig(({ isSsrBuild }) => ({
plugins: [vue()],
resolve: {
alias: {
"@": `${projectRoot}/src`,
},
},
build: {
manifest: isSsrBuild ? false : "manifest.json",
outDir: isSsrBuild ? "dist/ssr" : "dist/client",
rollupOptions: {
input: isSsrBuild ? "src/ssr.js" : "src/main.js",
},
},
}));
main.js
import { createSSRApp, h } from "vue";
import { createInertiaApp } from "@inertiajs/vue3";
createInertiaApp({
resolve: (name) => {
const pages = import.meta.glob("./Pages/**/*.vue", { eager: true });
return pages[`./Pages/${name}.vue`];
},
setup({ el, App, props, plugin }) {
createSSRApp({ render: () => h(App, props) })
.use(plugin)
.mount(el);
},
});
ssr.js
import { createInertiaApp } from "@inertiajs/vue3";
import createServer from "@inertiajs/vue3/server";
import { renderToString } from "@vue/server-renderer";
import { createSSRApp, h } from "vue";
createServer((page) =>
createInertiaApp({
page,
render: renderToString,
resolve: (name) => {
const pages = import.meta.glob("./Pages/**/*.vue", { eager: true });
return pages[`./Pages/${name}.vue`];
},
setup({ App, props, plugin }) {
return createSSRApp({
render: () => h(App, props),
}).use(plugin);
},
})
);
Performance note
With the implementation proposed above, you'll be loading the whole page on the first load. This is because everything will be bundled in the same file. If you want to split your code, you can use the following implementation.
helper.js
(taken from laravel vite plugin inertia helpers)
export async function resolvePageComponent<T>(
path: string | string[],
pages: Record<string, Promise<T> | (() => Promise<T>)>
): Promise<T> {
for (const p of Array.isArray(path) ? path : [path]) {
const page = pages[p];
if (typeof page === "undefined") {
continue;
}
return typeof page === "function" ? page() : page;
}
throw new Error(`Page not found: ${path}`);
}
main.js
import { createApp, h } from "vue";
import { createInertiaApp } from "@inertiajs/vue3";
import { resolvePageComponent } from "@/helper.js";
createInertiaApp({
resolve: (name) => {
return resolvePageComponent(
`./Pages/${name}.vue`,
import.meta.glob("./Pages/**/*.vue")
);
},
setup({ el, App, props, plugin }) {
createApp({ render: () => h(App, props) })
.use(plugin)
.mount(el);
},
});