Emmet markup abbreviation parser

Parses given Emmet markup abbreviation into AST. Parsing is performed in two steps: first it tokenizes given abbreviation (useful for syntax highlighting in editors) and then tokens are analyzed and converted into AST nodes as plain, JSON-serializable objects.

Note that AST tree in most cases cannot be used directly for output: for example, AST node produced from .foo.bar element misses element name and contains two class attributes with foo and bar values (not a single class with foo bar value).

Usage

You can install it via npm:

npm install @emmetio/abbreviation

Then add it into your project:

import parse from '@emmetio/abbreviation';

const tree = parse('div#foo>span.bar*3');
/* {
    type: 'Abbreviation',
    children: [{
        type: 'AbbreviationNode',
        name: 'div',
        attributes: [...],
        children: [...]
    }]
} */

The returned tree contains AbbreviationNode items: a node with name, attributes and/or text content. E.g. an element that can be represented somehow. Repeated and grouped nodes like a>(b+c)*3 are automatically converted and duplicated as distinct AbbreviationNode with distinct .repeat property which identifies node in repeating sequence.

Abbreviation syntax

Emmet abbreviation element has the following basic parts:

name.class#id[attributes?, ...]{text value}*repeater/

• name — element name, like div, span etc. Stored as node.name property.
• [attributes] — list of attributes. Each attribute is stored as AbbreviationAttribute instance and can be accessed by node.getAttribute(name). Each attribute can be written in different formats:
  • attr — attribute with empty value.
  • attr=value — attribute with value. The value may contain any character except space or ].
  • attr="value" or attr='value' — attribute with value in quotes. Quotes are automatically removed. Expression values like attr={value} are supported and can be identified by valueType: "expression" property.
  • attr. — boolean attribute, e.g. attribute without value, like required in <input>.
  • !attr – implicit attribute, will be outputted if its value is not empty. Used as a placeholder to preserve attribute order in output.
  • ./non/attr/value — value for default attribute. In other words, anything that doesn’t match a attribute name characters. Can be a single- or double-quotted as well. Default attribute is stored with null as name and should be used later, for example, to resolve predefined attributes.
• .class — shorthand for class attribute. Note that an element can have multiple classes, like .class1.class2.class3.
• #id — shorthand for id attribute.
• {text} — node’s text content
• *N — element repeater, tells parser to create N copies of given node.
• / — optional self-closing operator. Marks element with node.selfClosing = true.

Operators

Each element of abbreviation must be separated with any of these operators:

elem1+elem2>elem3

• + — sibling operator, adds next element as a next sibling of current element in tree.
• > — child operator, adds next element as a child of current element.
• ^ — climb-up operator, adds next element as a child of current element’s parent node. Multiple climb-up operators are allowed, each operator moves one level up by tree.

Groups

A set of elements could be grouped using (), mostly for repeating and for easier elements nesting:

a>(b>c+d)*4+(e+f)

Groups can be optionally concatenated with + operator.