"""Regression: dataset YAML files parse and have the expected shape.

Bug captured: schema_docs.yaml had values that began with a quote character
(e.g. `gender: 'M' or 'F'.`), which YAML treats as a quoted scalar — anything
after the closing quote is a parse error.
"""

from pathlib import Path

import pytest
import yaml

DATASETS_DIR = Path(__file__).resolve().parent.parent / "api" / "datasets"
DATASETS = [p.name for p in DATASETS_DIR.iterdir() if p.is_dir() and not p.name.startswith("_")]


@pytest.mark.parametrize("dataset", DATASETS)
def test_schema_docs_parses(dataset):
    docs = yaml.safe_load((DATASETS_DIR / dataset / "schema_docs.yaml").read_text())
    assert docs is not None
    assert "tables" in docs
    assert isinstance(docs["tables"], dict) and docs["tables"]


@pytest.mark.parametrize("dataset", DATASETS)
def test_metrics_parses(dataset):
    metrics = yaml.safe_load((DATASETS_DIR / dataset / "metrics.yaml").read_text())
    assert metrics is not None
    assert "metrics" in metrics
    assert isinstance(metrics["metrics"], dict) and metrics["metrics"]


def test_financial_columns_with_quotes_survive_parse():
    """Specific regression: values that contain quoted codes (M, F, OWNER,
    PRIJEM, etc.) must parse without truncation."""
    docs = yaml.safe_load(
        (DATASETS_DIR / "financial" / "schema_docs.yaml").read_text()
    )
    cols = docs["tables"]["client"]["columns"]
    assert '"M"' in cols["gender"] and '"F"' in cols["gender"]

    disp = docs["tables"]["disp"]["columns"]
    assert '"OWNER"' in disp["type"] and '"DISPONENT"' in disp["type"]

    trans = docs["tables"]["trans"]["columns"]
    assert '"PRIJEM"' in trans["type"] and '"VYDAJ"' in trans["type"]