HTML-adapted text

Convert semtext-pairs into html, treating labels as css selectors that specify elements to generate.

(section.content-section){{
    (h1.section-title){{The section title}}
    (p.text-content){{Some text content here}}
    (div.photo-grid){{
        (img[src='an/image1.png']){{description}}
        (img[src='an/image2.png']){{description}}
    }}
}}

It uses emmet to parse the selectors, with some limitations.

What it doesn't do

1. It will ignore any descendants, specified by the selector: div#address-container>span#address won't generate the span with it's attributes: it only will generate the div with it's attributes.
2. It will ignore the text content: {text content} (probably, it won't parse it)
3. It doesn't support @emmetio/abbreviaton's selector grouping with braces (see their readme): the semtext-pairs syntax doesn't support nested braces - this (a>(b>c+d)*4){{an/url}} will break the parsing.

img: the alt attribute

It will treat first-occurring text in the img entity as the alt attribute: (img){{image description}} will produce <img alt="image description">. It will ignore any other entities in the entity, e.g. (img){{(div){{illegal div}} image description}} will produce <img alt=" image description">.

XSS prevention

Generally, I suppose, the logic outlined in showdown article applies here. I.e., sanitize the generated html as opposed to the semtext-pairs formatted text.

Sanitizing encoded content doesn't work

By default, htmlat-rich escapes any html occurring in the text nodes. A consequence of that is that the escaped nodes become invisible to sanitizer and don't get erased.

You can prevent escaping by setting the spanTextNodes option to true. However, then the content of text nodes will be wrapped in spans, which might not be what we want in the end. spanToTextTextNodes converts the spans into Text nodes.

In conclusion, if you want to sanitize the output, what you should do is:

1. convert with spanTextNodes enabled
2. sanitize
3. convert spans to Text nodes with spanToTextTextNodes

If any html tags left in the text after sanitization, they will be escaped.

CLI

Here is a cli with this package under the hood and some additional features.