Tell the parser to skip or alias specific tokens?

[ispringle]

I am working on writing some extensions so I can take a markdown file and render it in a way that works with tufte-css. The problem I'm facing is that I'd like to process child tokens of my footer, but I do not want to process them if the child would be a paragraph, instead I'd like that to just render as text. So this is the extension:

const regex = /^(@footer ?(paragraph|[^\n]*)(?:\n|$))+/

const extension = {
    name: 'footer',
    level: 'block',
    start(src) {
        return src.match(/@footer/)?.index;
    },
    tokenizer(src, _tokens) {
        const cap = regex.exec(src);
        if (cap) {
            const text = cap[0].replace(/^@footer[ \t]?/gm, '');
            const tokens = this.lexer.blockTokens(text);
            return {
                type: 'footer',
                raw: cap[0],
                tokens,
                text
            };
        }
    },
    
    renderer(token) {
        return `<footer>${this.parser.parse(token.tokens)}</footer>`
    }
}

export default extension

It takes this:

@footer <cite>[Le Bourgeois gentilhomme](https://www.gutenberg.org/files/2992/2992-h/2992-h.htm)</cite>

And renders this:

<footer>
      <p>
       <cite><a href="https://www.gutenberg.org/files/2992/2992-h/2992-h.htm">Le Bourgeois gentilhomme</a></cite>
      </p>
</footer>

However, I want it to render as:

<footer>
       <cite><a href="https://www.gutenberg.org/files/2992/2992-h/2992-h.htm">Le Bourgeois gentilhomme</a></cite>
</footer>

Is there a way to omit the paragraph but allow to render all tokens?

[UziTech]

If there will only be inline tokens you can do this.parser.parseInline(token.tokens) in the renderer.

[ispringle]

Thanks that did the trick!