Skip to content

fix: concise no-TOC titles and distinct same-page summaries#360

Open
MS170998 wants to merge 1 commit into
VectifyAI:mainfrom
MS170998:fix/toc-title-and-duplicate-summaries-340-341
Open

fix: concise no-TOC titles and distinct same-page summaries#360
MS170998 wants to merge 1 commit into
VectifyAI:mainfrom
MS170998:fix/toc-title-and-duplicate-summaries-340-341

Conversation

@MS170998

Copy link
Copy Markdown

Summary

  • Ask no-TOC prompts for short synthesized titles; clamp overlong titles
  • Partition same-page sibling text by title anchors
  • Deduplicate summary generation for identical text / container nodes

Closes #341
Closes #340

Test plan

  • python -m unittest tests.test_toc_title_summary -v
  • Index a no-TOC PDF with same-page numbered list items; titles stay short
  • Parent/children on one page no longer share identical summaries

No-TOC prompts now ask for short synthesized titles instead of verbatim
sentences, with a clamp safety net. Same-page siblings get title-anchored
text slices and summary generation dedupes identical text / container nodes.

Closes VectifyAI#341
Closes VectifyAI#340

Co-authored-by: Cursor <cursoragent@cursor.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

1 participant