nano-hcx

v0.0.4

Published

3 years ago

HTML Compiler Extensions

Downloads

0High
0Medium
0Low

anywhichway

tlx react angular svelte vue hyperApp uhtml lighterhtml hypeHTML

Nano HCX

JavaScript Template Literals in HTML

Core Library:

1) ES 2017 - 6.1K raw, 3.4K minify, 1.6k gzip

What
Installation
Basic Examples
API
Special Attributes
Advanced Examples
Notes
Release History

What

HTML Compiler eXtensions (HCX) flips JSX on its head. Rather than make JavaScript handle HTML with something like JSX, HCX makes JavaScript template literal notation i.e. ${ ... javascript } available directly in HTML.

It's this simple:

<div>${firstName} ${lastName}</div>
<div>${phone} ${email}</div>

Nano HCX will be the core of v1 of 'hcx'. The rest of 'hcx' BETA has yet to be fully re-written, but visit it if you want to see the type of capabability (not API) that is coming.

Nano HCX allows designers to express a UI as HTML and CSS at whatever micro, macro, or monolithic scale they wish and then hand-off to programmers to layer in functionality. Designers can continue to adjust much of the HTML while programmers are at work. For designers that wish to code, nano-hcx makes the transition into bits and pieces of JavaScript easier than moving into a full build/code oriented environment.

HCX:

Eliminates the need for control flow attribute directives; however, they will be directly supported in hcx-plus and custom directives can be added.
Eliminates the need for content replacement directives like VUE's v-text. You just reference static or reactive data directly in your HTML, e.g. instead of <div v-text="message"></div> just use <div>${message}</div>. This also means that the VUE filter syntax is un-neccesary, e.g. instead of <span v-text="message | capitalize"></span> use <span>${message.toUpperCase()}</span> or even the new JavaScript pipe operator when it becomes available <span>${message |> capitalize}</span>.
Does not use a virtual DOM, it's reactive data dependency tracker laser targets just those DOM nodes that need updates, so diffing is not necessary.
Does not require a special interpolation function that complicates templates like most other template literal based approaches.
Does not require a build environment/pre-compilation process.
Allows direct use of the JavaScript debugger in templates.
Supports server side rendering via a pre-built Express rendering engine.
Sanitizes HTML using DOMPurify or a sanitizer of your choice.
Is far smaller, and we think simpler and more flexible, than many other options.

Installation

npm install nano-hcx

The file nano-hcx.js is isomorphic and ready for the browser, Node.js or Nano.

Browser

<script src="<path>/nano-hcx.js"></script>

Node.js

HCX curently runs in the most recent versions of Chrome, Firefox, and Edge.

npm install nano-hcx

If you don't want to copy files out of node_modules/nano-hcx into your prefered JavaScripr directory and are using Express, try modulastic to expose the hcx files directly.

Basic Examples

In the most simple case, a document body can be bound to a model and rendered as shown in ./examples/simple.html:

<html>
<head>
	<script src="../nano-hcx.js" type="text/javascript"></script>
	<script>
		var data = nHCX.reactor({name:"joe",age:27,children:["mary","jane","jack"]});
	</script>
</head>
	<body onload="nHCX.render(document.body,{data})">
	${name} has these children: ${data.children.join(",")}
	</body>
</html>

If you are new to template literals, see ...

So long as the data and code between ${ ... } delimiters does not contain any > characters required to specify HTML tags and the placement does not violate HTML layout rules (tables can be pretty finicky), you can put HCX templates just about anywhere. If you need to insert HTML or if the un-interpreted content might violate HTML layout rules, you can use a standard <script> tag with type text/hcx to wrap your code as shown in ./examples/table.html below.

<html>
<head>
	<script src="../nano-hcx.js" type="text/javascript"></script>
	<script>
		var data = {name:"joe",age:27,children:["mary","jane","jack"]};
	</script>
</head>
<body onload="nHCX.render(document.body,{data})">
	${name} has these children:
	<script type="text/hcx">
	<table>
	${
		children.reduce((accum,item) => accum +=`<tr><td>${item}</td></tr>`,"")
	}
	</table>
	</script>
</body>
</html>

Although it is possible to make individual elements reactive in nano-hcx, the default approach is to wrap data as a reactor. This way any element in which the data appears will re-render if the data changes as shown in ./examples/reactor.html.

It is not really necessary in this small example, but you may notice the removeAttribute call in the on-load. If you are using a DOM node as its own replacement, the node will render in an unresolved state first. For large nodes, this may create an odd UI containing may ${ ... } delimited expressions. By starting the node in a hidden state you can avoid this. You should also typically be more targetted in your use of render than providing an entire document body.

<html>
<head>
<script src="../nano-hcx.js" type="text/javascript"></script>
<script>
	var data = nHCX.reactor({name:"joe",age:27,children:["mary","jane","jack"]});
</script>
</head>
<body onload="nHCX.render(document.body,{data}).then(targets => targets[0].removeAttribute('hidden'))" hidden>
	${name} has these children:
	<ul>
	<script type="text/hcx">
	${
		children.reduce((accum,item) => accum +=`<li>${item}</li>`,"")
	}
	</script>
	</ul>
	<script>
		setTimeout(() => {
			alert("Ready for change?");
			data.children[0] = "jim";
		},1000);
	</script>
	</body>
</html>

API

render(target(s)[,{source,data,root,shadow,sanitize,scripts,...rest}]) returns a Promise for an array of target(s) in a resolved state.

Replaces any ${ ... } delimited content the source and inserts the result into the target(s). During processing, this is bound to the containing element.

target(s) Any of a DOM node, and array of DOM nodes, or a CSS selector. If an array or CSS selector each target will have the options applied.

All options are exactly that, optional.

source An optional DOM node or an HTML string containing ${ ... }. If no source is supplied, the initial content of the target is used.
data An object to use for replacing values in ${ ... } delimited areas. If it is a reactor (see below), then changes will automatically force the re-rendering of the sections of the HTML that reference the properties of the data changed.
root The root DOM node for initiating the CSS based selection of targets. By default, the current document.
shadow If provided, the source is rendered into a shadow DOM rooted in the target. The value of shadow can be 'open' or 'closed'. By using styles and scripts nested inside <template> tags as the source, the style and scripts can have their impact isolated to the shadow DOM. When scripts are processed by hcx in a shadow scope, they have shadowHost available as a variable set to the target in which they are rendered. Also see the special attribute :scoped.
santize The function to use to sanitize HTML to improve security. If DOMPurfiy is present in the scope of render invocation, it will automatically be used unless sanitize is set to something else. The default value for sanitize without DOMPurify is, (html) => html, essentially a no-op.
scripts A boolean indicating if standard JavaScript blocks should be evaluated when processing the source. This can be very powerful; however, it brings with it some security risks. If DOMPurify is used for sanitize, then scripts are completely removed unless scripts is set to true. If DOMPurify is not used, scripts still won't be executed unless scripts is set to true.
...rest Allows for the passing of custom options to support the larger hcx library.

There are also some undocumented options to provide felxibility when introducing the larger hcx library. These are documented with hcx.

reactor(data) - Takes data an object and returns a Proxy that tracks the impact of the data on the rendered HTML and responds to changes of the data to re-render the HTML. If primitive data is passed in, just returns the primitive data.

Special Attributes

:assign=>object> - Uses Object.assign to augment the current data with the parsed value of >object>. The new values apply at any time after the attribute has been evaluated. For a first rendering, this will mean the duration of the document. For subsequent renderings, the augmentation will be available earlier in the document.

:data=>object> - Sets the data for the scope of the element on which the attribute exists to the parsed value of >object>. Make sure to put quotes around theobject properties, e.g.

<div :data='{"name":"joe","age":27}'>Name: ${name} Age: ${age}</div>

data->property>=>value> - Sets the propertyon the current data to thevalue`.

:reactive - Makes the data associated with the element reactive to changes so that the element will be automatically re-rendered for changes. Since the Proxy generated by :reactive is not externally accessable, its use only makes sense in the context of child <script> contents within the scope of the element.

:remove - If true, unlike the attribute hidden or the style display:none, the node and its children are not fully processed and the node is actually removed from the DOM. This is typically facilitated via template resolution, e.g. <div :remove="${new Date().getHours()<12}">Afternoon message</div>. It is primarilly useful in lightening the payload when doing server side rendering, or lightening the processing for large nested DOM nodes.

:scoped - When hcx evaluates JavaScript scripts (not text/hcx scripts) the this value for the script will be set to the parent element of the script and the data variable will be available. Also see shadow above where the script automatically gets a shadowHost variable, even if not scoped. The example ./examples/scopesandshadows.html illustrates how this works.

<html>
<head>
	<script src="../nano-hcx.js" type="text/javascript"></script>
	<script>
		var data = {name:"joe",age:27,children:["mary","jane","jack"]};
		var children = [];
	</script>
	<template id="example">
		<style>
			p { font-size: 150% }
		</style>
		<p>Larger font</p>
		<script :scoped>
			children = data.children;
			var shadowVar = "secret";
			console.log(shadowHost);
			console.log(this);
			console.log(shadowVar);
		</script>
	</template>
</head>
<body onload="nHCX.render('#app',{data,scripts:true,shadow:'open',source:document.getElementById('example')})">
	<div id="app"></div>
	<p>Normal font. Open debugger to see output.</p>
	<script>
		setTimeout(() => { console.log(children);console.log('shadowVar type:',typeof(shadowVar))},1000)
	</script>
</body>
</html>

Advanced Examples

You can insert debugger statements in your templates as shown in ./examples/debugger.html.

<html>
<head>
<script src="../nano-hcx.js" type="text/javascript"></script>
<script>
	var data = nHCX.reactor({name:"joe",age:27,children:["mary","jane","jack"]});
</script>
</head>
<body onload="nHCX.render(document.body,{data,scripts:true})">
	${name} has these children:
	<ul>
	<script type="text/hcx">
	${
		children.reduce((accum,item) => {debugger; return accum +=`<li>${item}</li>`;},"")
	}
	</script>
	</ul>
	<div id="app">
	<script :scoped>
		alert(`My id is ${this.id}. The first child is ${data.children[0]}`);
	</script>
	</div>
	<script>
		setTimeout(() => {
			alert("Ready for change?");
			data.children[0] = "jim";
		},1000);
	</script>
	</body>
</html>

There is a pre-built Express rendering engine available in the file express.js. This can be added to Express as shown in ./examples/express/index.js.

const express = require("express")
const app = express()
const port = 3000

app.engine("html",require("../../express.js"));
app.set("views", "./views");
app.set("view engine", "html");
app.get("/*", (req, res) => {
	if(req.url==="/favicon.ico") {
		res.status = 404;
		res.end();
		return;
	}
	res.render(req.url.substring(1), {data:{message: "Hello there!" }}) 
 	// replace 'data' with any data you want, you can also pass additional HCX options as part of this object
	// if you do not pass data, it can still be added using element attributes in the browser
})

app.listen(port, () => console.log(`HCX example app listening at http://localhost:${port}`));

Per the API description, scripts can be scoped to their containing element and executed as shown in ./examples/script.html

<html>
<head>
<script src="../nano-hcx.js" type="text/javascript"></script>
<script>
	var data = nHCX.reactor({name:"joe",age:27,children:["mary","jane","jack"]});
</script>
</head>
<body onload="nHCX.render(document.body,{data,scripts:true})">
	${name} has these children:
	<ul>
	<script type="text/hcx">
	${
		children.reduce((accum,item) => accum +=`<li>${item}</li>`,"")
	}
	</script>
	</ul>
	<div id="app">
	<script :scoped>
		alert(`My id is ${this.id}. The first child is ${data.children[0]}`);
	</script>
	</div>
	<script>
		setTimeout(() => {
			alert("Ready for change?");
			data.children[0] = "jim";
		},1000);
	</script>
	</body>
</html>

Notes

nano-hcx was developed while hcx was in BETA and will serve as the core of the production release of hcx.

Release History (Reverse Chronological Order)

2020-06-30 v0.0.4 - Added unit tests.

2020-06-28 v0.0.3 - Added unit tests. Enhanced docs. Made render asynchronous.

2020-06-24 v0.0.2 - Fixed issue with undefined `source'. Enhanced documentation and examples.

2020-06-23 v0.0.1 - First public release.

Published

Vulnerabilities

Links

Maintainers

Keywords

Readme