"""LegacyHUB adapter for the shared Docling recognition engine."""

from __future__ import annotations

from pathlib import Path

from teamhub_document_recognition_engine import (
    ExtractedBlock,
    ExtractedFigure,
    ExtractedPage,
    ExtractedTable,
    ExtractionResult,
)
from teamhub_document_recognition_engine import extract as _engine_extract

from app.config import settings

# Re-exported engine models: chunker, table/figure processors and tests import
# them from this adapter instead of depending on the engine package directly.
__all__ = [
    "ExtractedBlock",
    "ExtractedFigure",
    "ExtractedPage",
    "ExtractedTable",
    "ExtractionResult",
    "extract",
]


def extract(pdf_path: Path) -> ExtractionResult:
    return _engine_extract(pdf_path, docling_ocr_enabled=settings.docling_ocr_enabled)