PDF table extraction for RAG and LLM — convert PDF tables to clean HTML. Fast, local, no GPU. Handles merged cells, line-wrapped text, no serialization.
