The URLs in our app right now are built on a hack: the hash. It's the default because it will always work, but there's a better way.
Modern browsers let JavaScript manipulate the URL without making an http
request, so we don't need to rely on the hash (#) portion of the url
to do routing, but there's a catch (we'll get to it later).
Open up index.js and import browserHistory instead of hashHistory.
// index.js
// ...
// bring in `browserHistory` instead of `hashHistory`
import { Router, Route, browserHistory, IndexRoute } from 'react-router'
render((
<Router history={browserHistory}>
{/* ... */}
</Router>
), document.getElementById('app'))Now go click around and admire your clean URLs.
Oh yeah, the catch. Click on a link and then refresh your browser. What happens?
Cannot GET /repos
Your server needs to deliver your app no matter what URL comes in, because your app, in the browser, is manipulating the URL. Our current server doesn't know how to handle the URL.
The Webpack Dev Server has an option to enable this. Open up
package.json and add --history-api-fallback.
"start": "webpack-dev-server --inline --content-base . --history-api-fallback"We also need to change our relative paths to absolute paths in
index.html since the URLs will be at deep paths and the app, if it
starts at a deep path, won't be able to find the files.
<!-- index.html -->
<!-- index.css -> /index.css -->
<link rel="stylesheet" href="/index.css">
<!-- bundle.js -> /bundle.js -->
<script src="/bundle.js"></script>Stop your server if it's running, then npm start again. Look at those
clean URLs :)