µhtml

Social Media Photo by Andrii Ganzevych on Unsplash

micro html is a ~2.5K lighterhtml subset to build declarative and reactive UI via template literals tags.

📣 Community Announcement

Please ask questions in the dedicated discussions repository, to help the community around this project grow ♥

V2.8 Update

• added µhandlers foreign export to enable arbitrary attributes handling!

import {html, foreign} from 'uhtml';

const handler = (node, name, value) => {
  // P, any, {data: 123}
  console.log(node, name, value);
  // return null/undefined to remove it
  return value.data;
};

html`<p any=${foreign(handler, {data: 123})}>foreign</p>`;

V2.5 Update

• an interpolated value, within a DOM element, can now be a function, enabling a world of µhtml extending possibilities, including intents, hence aligning the behavior with both lighterhtml and hyperHTML. That is: <el>${callback}</el>! The callback will be invoked with the comment pin/placeholder as unique argument, where its parentNode would be the element containing such comment, if needed, and its returned value will be passed along the same mechanism that resolves already all other cases.

V2.4 Update

• a new ?attribute=${value} prefix, with a question mark, has landed, after this long debate, and based to the fact µhtml never wants to be ambiguous. However, the mighty lit-html put an end to the debate, disambiguating through a ? question mark prefix, which explicits developers intents, and works well across browsers. So that's it: whenever you expect an attribute to be in, or out, use ?name=${value} and win the day 🥳

V2.2 Update

• the new.js file has been renamed as es.js to align with other modules of mine that follow the same pattern.
• this module now exports its very same utilities via uhtml/async, in order to automatically resolve asynchronous values passed along the template. Please note this means that exported render, html, and svg tags, are all asynchronous, hence these all return a promise.
• the async.js file is now published too, compatible with ES2015+ browsers (no async / await used)

How To

Example

import {render, html, svg} from 'uhtml/async';

render(document.body, html`a${Promise.resolve('b')}c`);

How To Use µhtml

Install the module via npm i uhtml and consume it as such:

import {render, html, svg} from 'uhtml';
// const {render, html, svg} = require('uhtml');

render(document.body, html`<h1>Hello 👋 µhtml</h1>`);

Alternatively you can use a CDN such as unpkg, as shown in this demo.

<script src="https://unpkg.com/uhtml">/* global uhtml */</script>
<!-- or -->
<script type="module">
import {render, html, svg} from 'https://unpkg.com/uhtml?module';
</script>

API Documentation

Most information about µhtml are written in the documentation file, but following you can read most essential details.

The module exports the following functionalities:

• a render(where, what) function to populate the where DOM node with what content, which can be a DOM node, or the returning value of html and svg tags. The render function returns the where DOM node itself.
• a html template literal tag, to produce any sort of HTML content.
• a svg template literal tag, to produce any sort of SVG content.
• both html and svg implements a .for(reference[, id]) template tag function for keyed weak relationships within the node. Please don't overuse this feature, as 90% of the time is not necessary, and it could make the rendering slower than it should. Also, consider the ref attribute, in case a reference to the current node is needed at any time.
• both html and svg implements a .node template tag function for one-off HTML or SVG creation. Please don't use html.node one off nodes within render(...) calls, as this utility exists to help creating fragments or nodes that should be manually added to the DOM, and not through render calls.

Any element can have one or more attribute, either interpolated or not.

render(document.body, html`
  <div id="main"
        class=${`content ${extra}`}
        data-fancy=${fancy}>
    <p contenteditable=${editable}
        @click=${listener}
        onclick=${listener}
        class="${['container', 'user'].join(' ')}">
      Hello ${user.name}, feel free to edit this content.
    </p>
  </div>
`);

These are the rules to follow for attributes:

• interpolated attributes don't require the usage of quotes, but these work either ways. name=${value} is OK, and so is name="${value}" or even name='${value}'
• you cannot have sparse attribute interpolations: always use one interpolation to define each attribute that needs one, but never write things like style="top:${x};left${y}" as the parser will simply break with the error bad template. Use template literals within interpolations, if you want to obtain exact same result: style=${`top:${x};left${y}`}
• if the passed value is null or undefined, the attribute will be removed. If the value is something else, it will be set as is as value. If the attribute was previously removed, the same attribute will be placed back again. If the value is the same as it was before, nothing happens
• if the attribute name starts with on or @, as example, onclick=${...} or @click=${...}, it will be set as listener. If the listener changes, the previous one will be automatically removed. If the listener is an Array like [listener, {once:true}], the second entry of the array would be used as listener's options.
• if the attribute starts with a . dot, as in .setter=${value}, the value will be passed directly to the element per each update. If such value is a known setter, either native elements or defined via Custom Elements, the setter will be invoked per each update, even if the value is the same
• new: if the attribute starts with a ? question mark, as in ?hidden=${value}, the value will be toggled, accordingly with its truthy, or falsy, value.
• if the attribute name is ref, as in ref=${object}, the object.current property will be assigned to the node, once this is rendered, and per each update. If a callback is passed instead, the callback will receive the node right away, same way React ref does. Please note that conditional renders will not cleanup the reference, if this is not assigned to the new node.
• if the attribute name is aria, as in aria=${object}, aria attributes are applied to the node, including the role one.
• if the attribute name is .dataset, as in .dataset=${object}, the node.dataset gets populated with all values.

Following an example of both aria and data cases:

// the aria special case
html`<div aria=${{labelledBy: 'id', role: 'button'}} />`;
//=> <div aria-labelledby="id" role="button"></div>

// the data special case
html`<div .dataset=${{key: 'value', otherKey: 'otherValue'}} />`;
//=> <div data-key="value" data-other-key="otherValue"></div>

It is possible to place interpolations within any kind of node, and together with text or other nodes too.

render(document.body, html`
  <table>
    ${lines.map((text, i) => html`
      <tr><td>Row ${i} with text: ${text}</td></tr>
    `)}
  </table>
`);

There are only few exceptional nodes that do not allow sparse content within themselves:

• <plaintext>${content}</plaintext>, deprecated, yet it cannot contain comments
• <script>${content}</script>, it can contain comments, but only whole text can be replaced
• <style>${content}</style>, it cannot contain comments
• <textarea>${content}</textarea>, same as above
• <title>${content}</title>, same as above
• <xmp>${content}</xmp>, same as above

Following an example on how to populate these nodes (wrong + right way):

// DON'T DO THIS
render(document.body, html`
  <style>
    body { font-size: ${fontSize}; }
  </style>
  <textarea>
    Write here ${user.name}
  </textarea>
`);

// DO THIS INSTEAD
render(document.body, html`
  <style>
  ${`
    body { font-size: ${fontSize}; }
  `}
  </style>
  <textarea>
  ${`
    Write here ${user.name}
  `}
  </textarea>
`);

Beside nodes where the content will be inevitably just text, like it is for style or textarea, as example, every other interpolation can contain primitives, as strings, numbers, or even booleans, or the returned value of html or svg, plus regular DOM nodes.

The only special case are Array of either primitives, or returned values from html or svg.

render(document.body, html`
  <ul>
    <li>This is ${'primitive'}</li>
    <li>This is joined as primitives: ${[1, 2, 3]}</li>
    ${lines.map((text, i) => html`
      <li>Row ${i} with content: ${text}</li>
    `)}
  </ul>
`);

The second what argument of the render(where, what) signature can be either a function, which returning value will be used to populate the content, the result of html or svg tags, or a DOM node, so that it is possible to render within a render.

const Button = selector => {
  const button = document.querySelector(selector);
  return count => render(button, html`Clicks: ${count}`);
};

const Clicker = selector => {
  const button = Button(selector);
  return function update(count) {
    return render(document.body, html`
      <div onclick=${() => update(++count)}>
        Click again:
        ${button(count)}
      </div>
    `);
  };
}

const clicker = Clicker('#btn-clicker');
clicker(0);

µhtml html and svg tags implement exact same API offered by lighterhtml.

This means that both html.for(reference[, id]) and svg.for(reference[, id]) will weakly relate the node with the reference, and an optional unique id, instead of using its internal auto-referenced algorithm.

render(document.body, html`
  <ul>
    ${items.map(item => html.for(item)`
      <li>Keyed row with content: ${item.text}</li>
    `)}
  </ul>
`);

In more than one occasion developers got bitten by the fact µhtml can produce some cryptic error when null, or empty content, is provided as interpolation/hole.

The problem is pretty simple:

• don't loop / pass data that should not be render
• sanitize data upfront instead of expecting this library to magically understand where, and how, such data should not be rendered

A basic example would be the following:

const dirtyData = [
  {name: 'first'},
  null,
  {name: 'second'}
];

render(document.body, html`
  <ul>
    ${dirtyData.map(item => {
      // ⚠ this should not happen in the first place!
      if (!item)
        return null;
      return html`<li>${item.name}</li>`;
    })}
  </ul>
`);

There are at least two workarounds to consider:

• use data.filter(item => validate(item)).map(...)
• returns an actual node:

  if (!item)
    return html`<!--ignore-->`;

Between these two workarounds, I believe the cleanest one is the former: sanitize data upfront!

The benefits are:

• there are no unnecessary nodes in the DOM
• there are no weird cases to consider
• the mapping is pure, data in, node out

Custom Elements are either brought you, in a simplified manner, via µce (micro custom elements), which is a 3K library based on µhtml, or via vanilla JS, as demoed in WebComponents.dev.

This module works in IE11, Edge, and every other Desktop to Mobile browser, including KaiOS.

µhtml vs lighterhtml

You could read an exhaustive summary of features differences, but the first thing to keep in mind, is that lighterhtml is at pair with uhtml features, but not vice-versa, meaning if you need anything more, you can always switch to lighterhtml later on, and without changing a single line of code.

Following a list of other points to consider when choosing µhtml instead of lighterhtml (or vice-versa).

• there are no sparse attributes, each attribute must have a single interpolated value: attribute=${value} is OK, attribute="${a}${b}" is not, and attribute="some ${'partial'}" is not allowed neither.
• the interpolations are simple: primitive, or array of primitives, and nodes, or array of nodes.
• the style attribute is not special at all: if you want to pass objects there, please transform these as you prefer.
• the domdiff has been replaced with udomdiff, with a new blazing fast and super small diffing algorithm written from scratch
• the template argument is not normalized. If you target browsers with issue with such argument, please be sure you transpile your code with latest Babel before shipping to production
• no domtagger whatsoever, you can't change the current behavior of the library in any way

• uhtml should not suffer any of the IE11/Edge issues, or invalid SVG attributes warnings, as the parsing is done differently 🎉
• nested html and svg are allowed like in lighterhtml. The version 0 of this library didn't allow that, hence it was more "surprise prone". uhtml in that sense is more like a drop-in replacement for lighterhtml, and vice-versa
• keyed results via htmlfor(...) or svg.for(...), as well as one-off node creation, via html.node or svg.node are the same found in lighterhtml
• both aria=${object}, and ref=${...} special attributes work same as lighterhtml
• the .dataset=${object} helper work the same as lighterhtml
• the .property=${...} direct setter is still available
• self closing nodes are also supported, go wild with <custom-elements /> or even <span />
• the wire parsing logic has been simplified even more, resulting in slightly better bootstrap and update performance
• it's half of lighterhtml production size, mostly because ...
• there are no 3rd parts dependencies, except for µparser, µdomdiff, and for @ungap/create-content, needed only for IE11, but removable via @ungap/degap, same way I've done it here, or babel-plugin-remove-ungap. The compressed final size difference is just around ~0.2K though.

• be an essential/ideal companion for wickedElements, hookedElements, or any third part that would like a lighterhtml like API, without the extra weight
• keep it as simple as possible, but not simpler
• see if there is room for improvements in lighterhtml whenever uhtml simplifications allow to be ported there

V2 Breaking Change

The recently introduced data helper could conflict with some node such as <object>, hence it has been replaced by the .dataset utility. Since element.dataset = object is an invalid operation, the sugar to simplify data- attributes is now never ambiguous and future-proof: <element .dataset=${...} /> it is.