mellie

Package VersionHex Docs

A library for working with HTML that supports JavaScript and Erlang Targets

Installation

gleam add mellie

mellie uses htmgrrrl for parsing on Erlang targets and htmlparser2 for JavaScript targets

If targeting JavaScript, you also need to install htmlparser2 using the relevant package manager:

# using npm

npm install htmlparser2@12



# or using pnpm

pnpm add htmlparser2@12

If targeting Erlang you shouldn't need to install any additional dependencies

Usage

Full documentation can be found at https://hexdocs.pm/mellie

HTML Authoring

mellie provides functions to simplify authoring of HTML. These functions are generated from MDN and are provided in the mellie.html and mellie.attr namespaces

Parsing HTML can be done using the parse function. This will always return a full HTML document including the html, head, and body elements

let input =

  "

<html>

  <head><title>Page Title</title></head>

  <body>

    <h1>Hello World</h1>

  </body>

</html>

"



let assert Ok(parsed) =

  input

  |> mellie.parse



parsed

|> mellie.to_document_string

|> birdie.snap("basic html parsing")

Elements can also be created using the provided html and attr functions for HTML elements and attributes respectively:

let content =

  html.main([], [

    html.h1([], [html.text("My heading")]),

    html.p([attr.class("some-class")], [html.text("My body text")]),

  ])

Custom elements and attributes can also be defined using the millie.element and millie.attribute functions directly:

let content =

  html.main([], [

    mellie.element("my-custom-element", [], [

      html.p([], [html.text("My body text")]),

    ]),

    html.br([]),

  ])

Querying and Updating HTML

The package provides utilities for updating or querying the parsed HTML. Additionally, it has support for converting from mellie.element.ElementTree to other trees, for example a conversion to presentable_soup looks a bit like this:

import presentable_soup as soup



let soup_tree = transform_tree(my_el, soup.ElementNode, soup.TextNode)

HTML Parsing and Normalization

A primary of this package is to normalize behavior between JavaScript and Erlang environments for applications or libraries that target both. As platform deviation is considered a bug

Another important consideration of the library is to ensure that whitespace is retained to avoid subtle bugs when working with whitespace sensitive elements since this may be influenced by external CSS and as such can be context dependent

See the CSS white-space property doc or the HTML pre element doc on MDN for more information

Development

Commands needed for development are outlined in maskfile.md