Markdown reader: 2-space-indented dd paragraph separated

[jtojnar]

Explain the problem.

apple
: pomaceous

  fruit

results in fruit in a separate paragraph outside the definition list:

<dl>
<dt>apple</dt>
<dd>
pomaceous
</dd>
</dl>
<p>fruit</p>

This is inconsistent with 4-space separated version

apple
:   pomaceous

    fruit

and from unordered list, where the second paragraph is correctly placed inside the list item:

- pomaceous

  fruit

<ul>
<li><p>pomaceous</p>
<p>fruit</p></li>
</ul>

I can open PR later.

Pandoc version?

pandoc version 3.7.0.2 on try.pandoc.org

Using the defaultpandoc --from markdown --to html5 --no-highlight

[jgm]

As documented:

A term may have multiple definitions, and each definition may consist of one or more block elements (paragraph, code block, list, etc.), each indented four spaces or one tab stop. The body of the definition (not including the first line) should be indented four spaces.

[jgm]

The old "four space rule" has been replaced for ordered and bulleted lists by the more flexible syntax used in commonmark, but for definition lists we retain the four space rule. I don't remember now whether there was any reason for this, or if it's just an omission.

I closed this because pandoc's behavior conforms to the documentation, but it might be worth changing this behavior nonetheless.

Note that commonmark_x (which includes a definition list extension) will do what you expect here:

% pandoc -f commonmark_x  
apple
: pomaceous

  fruit
^D
<dl>
<dt>apple</dt>
<dd>
pomaceous
fruit
</dd>
</dl>

[jgm]

Actually, the commonmark_x behavior doesn't look right either, because we have lost the distinction between paragraphs. That also needs a look.

[jgm]

OK, I remember now. Note:

% pandoc -f commonmark_x
apple

: pomaceous

  fruit
<dl>
<dt>apple</dt>
<dd>
<p>pomaceous</p>
<p>fruit</p>
</dd>
</dl>

For definition lists, the tight/loose distinction is triggered by the blank line, or lack of a blank line, between the term and the definitions. And the effect of a "tight" list is that content is rendered without paragraph tags. That explains the result, though it's arguably not optimal.