2vdom

2vdom

A node library for parsing HTML into vdom of your choice.

Differences from html-virtualize:

• 2vdom doesn't restrict you to virtual-dom. Pass in the jsx-compatible pragma function from your framework (or write one), and it just generates the right vdom for you.
• html-virtualize depends on vtree, which is an obsolete implementation of virtual-dom. 2vdom doesn't depend on any specific vdom implementation, so it doesn't get obsolete as long as jsx doesn't.
• As a result, 2vdom has much fewer dependencies. It only depends on htmlparser2.

Usage

Installation

> npm install 2vdom

Top-level API

parse(pragma, html | buffer)

• pragma: jsx-compatible pragma function, e.g. React.createElement
• html: html to parse. May be an ordinary js string or Node.js Buffer
• Returns: parsed tree produced by your pragma function

parse.stream(pragma, stream)

• stream: Node.js ReadStream containing the html to be parsed
• Returns: a Promise that resolves to the parsed tree.

Usage with React / Deku / ...

let parse = require('2vdom');

let html = "<html>...</html>";
let pragma = React.createElement || deku.element || <some jsx pragma fn>;

let vdom = parse(pragma, html);
// Buffer works too:
parse(pragma, fs.readFileSync('<filename>'))

Usage with virtual-dom (hyperscript)

let pragma = (tagname, attrs, ...children) =>
  h(tagname, attrs, children);

// carry on as usual ...
let vdom = parse(pragma, html);

Usage with node.js streams

let fileStream = fs.createReadStream('<filename>');

parse.stream(pragma, fileStream)
.then(function(vdom) {
  // your code here ...
});

Limitations

There's a number of limitations of 2vdom that should not affect normal usage. However, if you encounter them in any actual use case or if you have good ideas for overcoming them, please issue or PR.

• 2vdom doesn't currently preserve comments. So if you rely on IE conditional comments, you are in a bit of a trouble here. This is an inherent limitation of JSX, because there's no way of conveying comments to vdom pragma functions.
• 2vdom doesn't currently preserve HTML directives e.g. <!doctype> Again, because there's no way of conveying these to pragma functions.
• 2vdom expects a single element (i.e. <html>) at document root. If you have multiple ones, 2vdom throws away every one but the last element. This is not a technical limitation, so let me know if you need this feature.

Testing

> jasmine

License

2-Clause BSD