Spaces:
Sleeping
Sleeping
Implement OCR pipeline enhancements: Pydantic schemas, validation gate, retry, and evaluation harness (excluding images)
497c3ef | from __future__ import annotations | |
| import json | |
| from pathlib import Path | |
| import pytest | |
| from pydantic import ValidationError | |
| from bl_ocr.domain.models import DeliveryNote | |
| PROJECT_ROOT = Path(__file__).resolve().parents[1] | |
| def _example() -> dict: | |
| return json.loads((PROJECT_ROOT / "example.json").read_text(encoding="utf-8")) | |
| def test_models_parse_example(): | |
| note = DeliveryNote.model_validate(_example()) | |
| assert note.products[0].pharmacist_price_ttc.value == 5.35 | |
| assert note.totals.line_count.value == 3 | |
| def test_models_forbid_extra_keys(): | |
| annotation = _example() | |
| annotation["products"][0]["_repaired"] = "leak" | |
| with pytest.raises(ValidationError): | |
| DeliveryNote.model_validate(annotation) | |
| def test_models_enforce_confidence_bounds(): | |
| annotation = _example() | |
| annotation["products"][0]["quantity"]["confidence"] = 1.5 | |
| with pytest.raises(ValidationError): | |
| DeliveryNote.model_validate(annotation) | |
| def test_models_allow_null_values(): | |
| annotation = _example() | |
| annotation["products"][0]["expiry_date"]["value"] = None | |
| note = DeliveryNote.model_validate(annotation) | |
| assert note.products[0].expiry_date.value is None | |