Execute create-next-app
with Yarn or npx to bootstrap the example:
npx create-next-app --example pass-server-data pass-server-data-app
# or
yarn create next-app --example pass-server-data pass-server-data-app
Download the example or clone the repo:
curl https://codeload.github.com/zeit/next.js/tar.gz/master | tar -xz --strip=2 next.js-master/examples/pass-server-data
cd pass-server-data
Install it and run:
npm install
npm run dev
# or
yarn
yarn dev
Deploy it to the cloud with now (download)
now
If you already have a custom server which has local data (for instance cached data from an API call, or data read
from a file at startup) that you wish to make available in the Next.js page, you can pass that data in the query
parameter of nextApp.render()
.
This is not the only way to pass data. You could also expose an endpoint and make a fetch()
call to localhost, or you could
import server-side code with eval
(necessary to prevent webpack from trying to package your server code). However both
solutions leave something to be desired in either performance or elegance.
This example shows the express server at server.js
reading in a file at load time with static data (this could also have been
data cached from an API call) in operations/get-item.js
. It has two routes: a home page, and an item page. The item page uses
data from the get-item operation, passed as a query parameter in routes/item.js
.
We use this data in pages/item.js
if rendered server-side, or make a fetch request if rendered client-side.
The server knows whether or not to use next.js to render the route based on the Accept header, which will be
application/json
when we fetch client-side.
Take a look at the following files:
- server.js
- routes/item.js
- pages/item.js
- operations/get-item.js