tests/test_source_population.py

from __future__ import annotations

import importlib.util
from pathlib import Path


ROOT = Path(__file__).resolve().parents[1]
SPEC = importlib.util.spec_from_file_location(
    "convert_wisemoor", ROOT / "scripts/convert_wisemoor_to_training_candidates.py"
)
assert SPEC and SPEC.loader
MODULE = importlib.util.module_from_spec(SPEC)
SPEC.loader.exec_module(MODULE)


def test_wisemoor_description_survives_candidate_conversion():
    record = {
        "product_id": 7,
        "name": "Observed formula",
        "description": "A source-observed floral woody description.",
        "notes": ["source note"],
        "slug": "observed-formula",
        "source_url": "https://example.test/formula/7",
        "formula": {"formula_rows": [
            {"ingredient": "benzyl acetate", "amount": 40},
            {"ingredient": "linalool", "amount": 30},
            {"ingredient": "vanillin", "amount": 30},
        ]},
    }
    candidate, audit = MODULE.convert_record(
        record,
        mapping={
            "benzyl acetate": "140-11-4",
            "linalool": "78-70-6",
            "vanillin": "121-33-5",
        },
        min_row_coverage=1.0,
        min_amount_coverage=1.0,
        min_resolved_components=3,
    )
    assert audit["accepted"] is True
    assert candidate["description"] == record["description"]
    assert candidate["metadata"]["description_provenance"] == "source_observed_wisemoor_product_text"