Enhancement: Visual support for Pandoc-style ::: Div blocks (matching, folding, navigation)

[fretwurst]

Enhancement: Visual pairing and navigation support for Pandoc ::: Div blocks in Quarto documents

Context

Quarto relies heavily on Pandoc-style Div blocks using the ::: syntax to structure content, style callouts, create columns, collapsibles, etc. These blocks are increasingly used for layout and interactivity, especially in Reveal.js presentations or Quarto books.

However, the ::: syntax has no native editor support for:

• Bracket matching (like {} or ()),
• Folding by nesting level,
• Visual pairing or indentation to show the block structure.

This makes it difficult to maintain clean documents, especially when using modular includes or multiple nested Divs with 4+ colons like ::::::. Mistakes (unclosed blocks) are only caught as Pandoc warnings such as:

[WARNING] Div at line 3629 unclosed at line 17148, closing implicitly.

Problem

As users, we are forced to:

• Manually count colons (::: vs :::::),
• Comment every block open/close manually,
• Search large documents line-by-line for mismatches.

This is error-prone and doesn't scale with complex or reusable content (e.g., via include:).

Proposal

Positron (or VS Code via the Quarto extension) could provide:

• Bracket matching for ::: blocks, similar to {} or [].
• Folding support based on nesting depth or ::: levels.
• Gutter icons or scope highlighting for nested Divs.
• Optional: Linting for unmatched ::: pairs, like unclosed HTML tags.

Even simple support such as scope matching or minimal indentation guides would drastically improve usability for Quarto authors.

Example Use Case

:::::: {.callout-note collapse="true"}
::::: {.columns}
:::: {.column width="60%"}
- Content…
::::
:::: {.column width="40%"}
- Image…
::::
:::::
::::::

Currently, it's hard to track visually whether all blocks are closed correctly, especially across includes.

Closing Thoughts

As Quarto becomes more widely used for complex documents, this is an increasingly pressing issue. A lightweight implementation (even optional) would provide great benefit to the growing base of scientific and technical users.

---

[juliasilge]

Thank you for this suggestion, @fretwurst! 🙌 I think improvements in how we handle ::: blocks would be quite helpful for folks.

Can you open this as a feature request for the Quarto VS Code extension? That is where this type of functionality would be implemented.