Document Metadata: Support nested sectPr in tables and text boxes

[JSv4]

Description

The GetDocumentMetadata() function currently only processes section properties (w:sectPr) at the document body's top level. Section breaks that occur inside tables or text boxes are not detected.

Current Behavior

The CollectSectionData method in WmlToHtmlConverter.cs (lines 850-910) only iterates over top-level elements:

var blockElements = body.Elements()
    .Where(e => e.Name == W.p || e.Name == W.tbl || e.Name == W.sectPr)
    .ToList();

This means:

• ✅ sectPr inside paragraph properties (w:p/w:pPr/w:sectPr) is handled
• ✅ Document-level sectPr at end of body (w:body/w:sectPr) is handled
• ❌ sectPr inside tables (w:tbl/.../w:sectPr) is NOT detected
• ❌ sectPr inside text boxes/shapes is NOT detected

Expected Behavior

The metadata extraction should scan the entire document tree for sectPr elements, not just top-level ones.

Impact

• Documents with section breaks inside tables may report incorrect section counts
• Paragraph/table indices per section may be inaccurate for complex documents
• This is an edge case - most documents don't have section breaks inside tables

Suggested Implementation

1. Use body.Descendants(W.sectPr) or a recursive scan to find all sectPr elements
2. Determine the proper ordering of sections based on document position
3. Associate content (paragraphs/tables) with their containing sections

Related

• PR feat(npm): add document metadata API for lazy loading (Issue #44 Phase 3) #50: Document Metadata API for Lazy Loading
• Issue WASM execution blocks UI thread, preventing loading states from rendering #44: Performance improvements and lazy loading

Labels

enhancement, metadata-api