fix(asciidoc): preserve ordered lists and literal blocks

[yushuosun]

AsciiDocBackend treated a bare-dot ordered item such as ". Download" as a caption because the list matcher only accepted explicit markers such as "1.". That caption buffer was then reused by later blocks or left unflushed, which is why list items and following paragraphs disappeared or moved. The backend also had no delimiter state for four-dot literal blocks, so their contents fell into ordinary text buffering.

This change recognizes AsciiDoc auto-numbered markers (including nested dot markers), preserves ordered-list semantics, and keeps blank lines plus picture/literal + continuation blocks with the current list item. Literal blocks now become code items, pending block captions are emitted before them, and caption detection no longer accepts a dot followed by whitespace.

I regenerated the two affected AsciiDoc references. The existing procedure fixture now keeps all four numbered steps, their images, and the Procedure/Verification labels in source order.

Issue resolved by this Pull Request:
Resolves #4117

Validation:

• uv run --no-sync pytest -q tests/test_backend_asciidoc.py — 10 passed
• Changed-file prek hooks — Ruff, formatting, ty, Tach, license/header and repository checks passed

Checklist:

• Documentation has been updated, if necessary. (No API or configuration change.)
• Examples have been added, if necessary. (Existing reference fixtures cover the end-to-end output.)
• Tests have been added, if necessary.

[github-actions]

✅ DCO Check Passed

Thanks @yushuosun, all your commits are properly signed off. 🎉

[Copilot]

Pull request overview

Fixes Docling’s AsciiDoc backend parsing so ordered lists using auto-numbered dot markers (., .., etc.) and literal blocks (....) are preserved in the resulting DoclingDocument/Markdown output, addressing silent content loss and reordering reported in #4117.

Changes:

• Recognize auto-numbered ordered list markers (including nested dot markers) and keep ordered-list semantics intact.
• Add literal-block (....) delimiter handling so literal blocks become CodeItems instead of being absorbed into paragraph buffering.
• Regenerate affected AsciiDoc regression groundtruth outputs and add targeted backend tests for ordered lists and literal blocks.

Reviewed changes

Copilot reviewed 4 out of 4 changed files in this pull request and generated 2 comments.

File	Description
docling/backend/asciidoc_backend.py	Updates list-item detection, list/literal-block parsing flow, and caption recognition to preserve ordered lists and literal blocks.
tests/test_backend_asciidoc.py	Adds regression tests covering auto-numbered ordered lists and literal blocks plus following text retention.
tests/data/asciidoc/groundtruth/asciidoc_03.asciidoc.md	Updates golden output to reflect correct ordered-list and section ordering for the procedure fixture.
tests/data/asciidoc/groundtruth/asciidoc_02.asciidoc.md	Updates golden output for nested ordered lists to render as ordered items.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[codecov]

Codecov Report

❌ Patch coverage is 93.84615% with 4 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
docling/backend/asciidoc_backend.py	93.84%	2 Missing and 2 partials ⚠️

📢 Thoughts on this report? Let us know!

[mergify]

Merge Protections

🟢 All 2 merge protections satisfied — ready to merge.

Show 2 satisfied protections

🟢 Enforce conventional commit

Make sure that we follow https://www.conventionalcommits.org/en/v1.0.0/

• title ~= ^(fix|feat|docs|style|refactor|perf|test|build|ci|chore|revert)(?:\(.+\))?(!)?:

🟢 Require two reviewer for test updates

When test data is updated, we require two reviewers

• #approved-reviews-by >= 2

[PeterStaar-IBM]

lgtm!