diff --git a/renderers/markdown_toc.go b/renderers/markdown_toc.go index 6026527..e64ad93 100644 --- a/renderers/markdown_toc.go +++ b/renderers/markdown_toc.go @@ -84,8 +84,10 @@ func processTOC(content []byte, opts *TOCOptions) ([]byte, error) { return content, nil } - // Generate the TOC HTML - toc, err := generateTOC(content, opts) + // Generate the TOC HTML from the same DOM tree + // This ensures that {:.no_toc} paragraphs removed by extractHeadings() + // are also removed from the final output + toc, err := generateTOCFromDoc(doc, opts) if err != nil { return nil, err } @@ -109,9 +111,8 @@ func processTOC(content []byte, opts *TOCOptions) ([]byte, error) { // Extract the body content (html.Render wraps in ...) result := extractBodyContent(buf.Bytes()) - // Note: We don't remove {:.no_toc} markers from the final output - // They remain as literal text when inside headings/paragraphs - // Only {:.no_toc} in sibling paragraphs (IAL markers) are removed during TOC generation + // Note: {:.no_toc} in sibling paragraphs (IAL markers) are removed by extractHeadings() + // during TOC generation. Literal {:.no_toc} text inside headings remains in the output. return result, nil } @@ -137,15 +138,11 @@ func shouldProcessTOC(content []byte) bool { return false } -// generateTOC parses HTML content and creates a table of contents -func generateTOC(content []byte, opts *TOCOptions) (string, error) { - // Parse the HTML document - doc, err := html.Parse(bytes.NewReader(content)) - if err != nil { - return "", err - } - - // Extract headings +// generateTOCFromDoc creates a table of contents from an already-parsed HTML document +// This allows the caller to reuse the DOM and benefit from any modifications made +// during heading extraction (e.g., removal of {:.no_toc} paragraphs) +func generateTOCFromDoc(doc *html.Node, opts *TOCOptions) (string, error) { + // Extract headings (this also removes {:.no_toc} sibling paragraphs from the DOM) headings := extractHeadings(doc) // Filter headings by level if opts is provided diff --git a/renderers/markdown_toc_test.go b/renderers/markdown_toc_test.go index d014ddf..fca4ef4 100644 --- a/renderers/markdown_toc_test.go +++ b/renderers/markdown_toc_test.go @@ -304,6 +304,62 @@ func TestNoTocInline(t *testing.T) { } } +// TestNoTocParagraphRemoval tests that {:.no_toc} sibling paragraphs are removed from output +// This is a regression test for issue #100 +func TestNoTocParagraphRemoval(t *testing.T) { + markdown := `* TOC +{:toc} + +## White/wheat loaf + +## Should not appear +{:.no_toc} + +## Focaccia + +## Ciabatta` + + html, err := renderMarkdown([]byte(markdown)) + if err != nil { + t.Fatalf("Error rendering markdown: %v", err) + } + + htmlStr := string(html) + + // The {:.no_toc} paragraph should be completely removed from output + if containsString(htmlStr, "{:.no_toc}") { + t.Error("{:.no_toc} marker should be removed from output, but found it") + } + + // The excluded heading should NOT appear in the TOC + tocStart := strings.Index(htmlStr, "") + if tocStart == -1 || tocEnd == -1 { + t.Fatal("Could not find TOC in output") + } + tocContent := htmlStr[tocStart : tocEnd+5] + + if containsString(tocContent, "Should not appear") { + t.Error("Heading marked with {:.no_toc} should not appear in TOC") + } + + // But the excluded heading should still appear in the document body + if !containsString(htmlStr, "