fix: preserve code blocks in embedded markdown files (fixes #106)

[bingryan]

Summary

• Fixes issue Export of files with embedding of .md files containing mermaid doesn't work #106 where embedded mermaid code blocks were converted to rendered HTML instead of preserving original markdown code block syntax
• Changes getEmbedMap() to read raw markdown directly from source files instead of extracting from rendered DOM
• Preserves all code blocks (mermaid, math, etc.) in their original format

Changes

• Rewrote getEmbedMap() function to read raw markdown from source files
• Added getHeadingContent() helper to extract heading sections
• Added getEmbedContentFromSource() helper to unify embed content extraction
• Maintains support for:
  • ![[file.md]] - full file embeds
  • ![[file.md#heading]] - heading section embeds
  • ![[file.md#^blockid]] - block reference embeds
• Preserves removeYamlHeader setting behavior

Test plan

• Test embedding a file containing mermaid code blocks
• Verify mermaid code is preserved in exported markdown (not converted to HTML)
• Test heading-based embeds (![[file.md#heading]])
• Test block reference embeds (![[file.md#^blockid]])
• Verify removeYamlHeader setting still works correctly

🤖 Generated with Claude Code

Summary by CodeRabbit

• New Features

  • Support embedding specific blocks and headings, plus inline block embeds when needed.
  • Configurable option to remove YAML headers from embedded content.

• Bug Fixes

  • More reliable embed resolution by extracting content from source text instead of rendered output.
  • Improved error handling and safer fallback behavior for unresolved embeds.

• Refactor

  • Internal improvements to embed parsing and content normalization.

_{✏️ Tip: You can customize this high-level summary in your review settings.}

[coderabbitai]

Caution

Review failed

The pull request is closed.

📝 Walkthrough

Walkthrough

Refactors embed resolution in src/utils.ts to extract embed content directly from raw markdown (headings, blocks, whole files) using new helpers, with YAML-header removal support and safer error handling returning null on failures.

Changes

Cohort / File(s)	Summary
Embed resolution and helpers src/utils.ts	Replaces HTML-based embed parsing with source-based extraction; adds getHeadingContent, getEmbedContentFromSource, getBlockContent, and parseEmbedLink; reworks getEmbedMap to key by embed link and store quoted markdown content; supports block/heading/file embeds, optional YAML header stripping via plugin.settings.removeYamlHeader, and logs errors returning null instead of throwing.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Possibly related PRs

• fix: add inline block embeds support to fix #114 #119: Adds getBlockContent and parseEmbedLink and implements inline block-embed resolution—strongly overlaps with these helper and embed-resolution changes.
• fix: prevent undefined when embed content not loaded (#116) #118: Adjusts embed-replacement guards to avoid inserting undefined when embedMap entries are missing; closely related to the embedMap key/lookup changes here.

Poem

🐰 I nibble lines from markdown beds,
Pull out headings, blocks, and threads,
Wrap them in quotes, neat and bright,
Hopping through code in morning light —
A tiny rabbit, binding embeds 🥕

Pre-merge checks and finishing touches

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 50.00% which is insufficient. The required threshold is 80.00%.	You can run @coderabbitai generate docstrings to improve docstring coverage.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title 'fix: preserve code blocks in embedded markdown files (fixes #106)' directly describes the main change: switching from DOM-based embed extraction to raw markdown parsing to preserve code block syntax in embeds, which is the primary objective of the PR.

---

📜 Recent review details

Configuration used: defaults

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 7a000fb and 20597d1.

📒 Files selected for processing (1)

• src/utils.ts

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 1

Caution

Some comments are outside the diff and can’t be posted inline due to platform limitations.

⚠️ Outside diff range comments (1)

src/utils.ts (1)

4-4: Remove unused import.

The htmlToMarkdown import is no longer used after refactoring to read raw markdown directly instead of parsing HTML DOM.

🔎 Proposed fix

-import { TAbstractFile, TFile, TFolder, htmlToMarkdown, CachedMetadata } from "obsidian";
+import { TAbstractFile, TFile, TFolder, CachedMetadata } from "obsidian";

🧹 Nitpick comments (1)

src/utils.ts (1)

515-575: Consider extracting YAML header removal logic.

The YAML header removal logic is duplicated in three places (lines 546-548, 559-561, 566-568). While this works correctly, you could extract it to reduce duplication.

🔎 Optional refactor to deduplicate YAML header removal

 async function getEmbedContentFromSource(
     plugin: MarkdownExportPlugin,
     embedLink: string,
     currentPath: string
 ): Promise<string | null> {
     const parsed = parseEmbedLink(embedLink, currentPath);
 
     if (!parsed.filePath) {
         return null;
     }
 
     try {
         const file = plugin.app.vault.getAbstractFileByPath(parsed.filePath);
         if (!(file instanceof TFile)) {
             return null;
         }
 
         let content = await plugin.app.vault.cachedRead(file);
 
         // Extract specific section if needed
         if (parsed.blockId) {
             const blockContent = await getBlockContent(
                 plugin,
                 parsed.filePath,
                 parsed.blockId
             );
-            // Remove YAML header if configured (in case block is at start of file)
-            if (plugin.settings.removeYamlHeader && blockContent) {
-                return blockContent.replace(EMBED_METADATA_REGEXP, "");
-            }
-            return blockContent;
+            content = blockContent;
         }
-
-        if (parsed.heading) {
+        else if (parsed.heading) {
             const headingContent = await getHeadingContent(
                 plugin,
                 parsed.filePath,
                 parsed.heading
             );
-            // Remove YAML header if configured (in case heading is at start of file)
-            if (plugin.settings.removeYamlHeader && headingContent) {
-                return headingContent.replace(EMBED_METADATA_REGEXP, "");
-            }
-            return headingContent;
+            content = headingContent;
         }
 
         // Remove YAML header if configured for full file content
-        if (plugin.settings.removeYamlHeader) {
+        if (content && plugin.settings.removeYamlHeader) {
             content = content.replace(EMBED_METADATA_REGEXP, "");
         }
 
         return content;
     } catch (error) {
         console.error("Error getting embed content from source:", error);
         return null;
     }
 }

📜 Review details

Configuration used: defaults

Review profile: CHILL

Plan: Pro

📥 Commits

Reviewing files that changed from the base of the PR and between 375f99c and 7a000fb.

📒 Files selected for processing (1)

• src/utils.ts

🧰 Additional context used

🧬 Code graph analysis (1)

src/utils.ts (1)

src/config.ts (1)

• EMBED_METADATA_REGEXP (8-9)

🪛 GitHub Actions: Lint and build

src/utils.ts

[error] 4-4: ESLint: 'htmlToMarkdown' is defined but never used. (no-unused-vars) | Command: eslint src/ --ext .js,.jsx,.ts,.tsx

🪛 GitHub Check: build_and_test (18)

src/utils.ts

[failure] 597-597:
'embedValue' is never reassigned. Use 'const' instead

🪛 GitHub Check: build_and_test (20)

src/utils.ts

[failure] 597-597:
'embedValue' is never reassigned. Use 'const' instead

🔇 Additional comments (2)

src/utils.ts (2)

452-513: LGTM: Heading extraction logic is sound.

The function correctly identifies headings, extracts content up to the next heading of equal or higher level, and handles errors gracefully by returning null.

---

660-715: The path resolution for root-level files is already correctly handled. The code uses ternary operators that properly evaluate empty currentDir strings (which result from substring(0, -1) when files have no directory separator), falling back to use the link text directly. No issues found.