Tired of giant frameworks like Angular?
Trying to avoid complicated stuff like MVC, Typescript or JSX and just want to create a website in good old plain JS HTML CSS?
Want to make a small website feel 100x better because it’s a Single-Page Application?
Then this is the framework for you!
Navigating your website with a
tags like this
<a href="login">Login</a>
will, instead of fully refreshing the page, dynamically load the requested content (in this caselogin.html
), and switch out the old content with the new one.
Additionally, LDF will wait for external resources (like stylesheets or images) to load before displaying the page. Demonstration with a throttled connection and cpu:
The script will also change the browser’s navigation bar via the History-API, so that the page can be refreshed and browser-features like back and forward will work.
If you want to navigate your page via Javascript, all you have to do is call the nav
function of the global ldf
object like this, with the parameter being the page you want to navigate to:
ldf.nav("login");
You can include css and js in your html files just like normal, they will also be dynamically loaded.
For a small demonstration, check out the demo
directory, or check it out live.
Notes:
If you dynamically add a
tags via Javascript (not by changing the page with ldf
but with for example document.createElement
), you’ll have to invoke ldf.updatePageLinks()
in order for a
tags to not reload the page.
Just include the script in your index page like this:
<script src="ldf.js"></script>
You also have to add a div element with the id “ldf”(configurable) in your index.html
where you want your content to be like this:
<div id="ldf"></div>
LDF expects a folder structure like this (directory name can be configured):
index.html required
pages/ required
├── index/ required
│ ├── index.html required
│ ├── index.css
│ └── index.js
└── examplepage/
├── examplepage.html
└── styles.css
Each folder in the pages
directory represents a page (duh) and will be loaded if a user navigates to the route named after the folder. So, for example if a user clicks on an a
tag with href="examplepage"
, LDF will attempt to load pages/examplepage/examplepage.html
.
The page index
is required, and is the default content you’ll see without any navigation.
LDF currently allows for some configuration.
// The selector for which content should be switched out on page change
ldf.mainselector = "#ldf";
// The content displayed when LDF receives a 404 from the server while requesting a page
// Set this to undefined if you want LDF to display the error page it received from your server
ldf.notfound = "<div>Page not Found</div>";
// The directory your pages are stored on the server, default is /pages
ldf.pagedir = "/pages";
// A base url that is prepended for every navigation
// Used in my case for github pages since that has a required base url
ldf.baseurl = "";
// The css selector string LDF will wait on before displaying the page.
// By default it waits for all stylesheets and all elements with a src attribute (like images), but not scripts.
ldf.waitselector = "link[rel='stylesheet'],:not(script)[src]";
// Use hash urls, by default it's true. Hash urls look like this: domain.com/#login
// The advantage of these is that you don't need a web server to handle refresh events, but the url doesn't look that good.
ldf.hash = true;
You can also use the functions provided by ldf to load pages into other elements like demonstrated here:
Methods you can use:
ldf.load(selector, location, \<optional\>query);
ldf.load("#header", "/header", "?fancy=true");
// external content will be waited on before the content is changed just like switching pages
ldf.change(selector, htmlContent);
ldf.change("#footer", "<div>new footer</div><script src='...'/>");
If you want to have an example that shows the potential uses for this, you can find one here
If you don’t want to use hash urls (more info a few lines above), then here’s the instruction for you:
In order for the refresh function to work, you will have to include a special case in your web server.
You have to look at the request if it’s not /
but a request to a page, and send back your index.html
. But you can’t just always send back index.html
when the requested file isn’t found, unless you want your site to break if a user visits a missing page, since LDF relies on getting an error from a request to a missing page.
A small implementation of this I used in the demo/example project (with nodejs and express) looks like this:
// Static resources
app.use(express.static(__dirname));
// Special case, always sends back index.html if the request is only 1 level.
// This works because requests to resources are 3 levels deep (pages/resource/resource.html)
app.get("/:page", (req, res) => {
res.sendFile(require("path").join(__dirname, "index.html"));
});
I wrote a java library to quickstart development with ldf and the templating engine mustache. It also solves the problems with non-hash url as mentioned above. You can find information about the library here
This project was just thrown together in a few hours based on a way messier implementation, which I used in my diploma thesis. If you find a bug or want to add an improvement, please open an issue/pull request.