beta_shape_pnpi/tests/integration/test_cli_flows.py

"""CLI flow tests: command-line options must reach the config in every
data-source flow (JSON file and paceENSDF), and degenerate exports must
exit non-zero without writing a file (regression D2/D4).

The CLI is invoked in-process via beta_spectrum.cli.main(argv) so the
paceENSDF dataset cache (loaded once per process) is shared with the
rest of the test session.
"""

from __future__ import annotations

import json
from typing import Any, Dict, List, Optional

import numpy as np
import pandas as pd
import pytest

from beta_spectrum.cli import main

pytestmark = pytest.mark.integration


def _base_json() -> Dict[str, Any]:
    return {
        "parent_symbol": "Tc99",
        "Z_parent": 43,
        "Z_daughter": 44,
        "A_number": 99,
        "endpoint_MeV": 0.294,
        "transition_type": "A",
    }


def write_json(tmp_path: Any, extra: Optional[Dict[str, Any]] = None) -> str:
    data = _base_json()
    if extra:
        data.update(extra)
    path = tmp_path / "input.json"
    path.write_text(json.dumps(data))
    return str(path)


def header_lines_of(path: str) -> List[str]:
    lines: List[str] = []
    with open(path, "r") as f:
        for line in f:
            if line.startswith("#"):
                lines.append(line.rstrip("\n"))
            else:
                break
    return lines


def read_export(path: str) -> pd.DataFrame:
    return pd.read_csv(path, sep=None, engine="python", comment="#")


def n_points_of(header: List[str]) -> int:
    for h in header:
        if h.startswith("# n_points:"):
            return int(h.split(":")[1])
    raise AssertionError(f"n_points header missing: {header}")


class TestCLIJsonFlows:
    """Options must propagate through the JSON config flow (D2)."""

    def test_json_honors_normalize_to_one_and_tab_separator(
        self, tmp_path: Any
    ) -> None:
        src = write_json(tmp_path)
        out = str(tmp_path / "to_one.csv")
        main([src, "--output", out, "--normalize-to-one", "--csv-separator", "tab"])

        with open(out) as f:
            body = "".join(ln for ln in f if not ln.startswith("#"))
        assert "\t" in body, "CSV must be tab-separated"

        df = read_export(out)
        assert "total_spectrum" not in df.columns
        area = np.trapezoid(df["branch_1_spectrum"], df["energy_MeV"])
        assert np.isclose(area, 1.0, rtol=1e-3)

    def test_json_defaults_keep_total_column(self, tmp_path: Any) -> None:
        src = write_json(tmp_path)
        out = str(tmp_path / "default.csv")
        main([src, "--output", out])

        df = read_export(out)
        assert "total_spectrum" in df.columns
        area = np.trapezoid(df["total_spectrum"], df["energy_MeV"])
        assert np.isclose(area, 1.0, rtol=1e-3)

    def test_json_cli_overrides_file_physics_values(self, tmp_path: Any) -> None:
        src = write_json(tmp_path, {"e_step_MeV": 0.001, "neutrino_mass_MeV": 0.0})
        out = str(tmp_path / "overrides.csv")
        main([src, "--output", out, "--e-step", "0.02", "--neutrino-mass", "0.15"])

        header = header_lines_of(out)
        assert any("neutrino_mass_MeV: 0.150000" in h for h in header)
        expected_points = int(np.arange(0.02, 0.294, 0.02).size)
        assert n_points_of(header) == expected_points

    def test_json_file_values_used_when_flags_absent(self, tmp_path: Any) -> None:
        src = write_json(tmp_path, {"e_step_MeV": 0.005})
        out = str(tmp_path / "file_values.csv")
        main([src, "--output", out])

        header = header_lines_of(out)
        expected_points = int(np.arange(0.005, 0.294, 0.005).size)
        assert n_points_of(header) == expected_points


class TestCLIDegenerateFlows:
    """Degenerate exports (m_nu above every endpoint) must fail cleanly."""

    def test_degenerate_intensity_mode_exits_nonzero_no_file(
        self, tmp_path: Any
    ) -> None:
        src = write_json(tmp_path)
        out = str(tmp_path / "deg.csv")
        with pytest.raises(SystemExit) as exc:
            main([src, "--output", out, "--neutrino-mass", "0.35"])
        assert exc.value.code == 1
        assert not (tmp_path / "deg.csv").exists()

    def test_degenerate_to_one_mode_exits_nonzero_no_file(self, tmp_path: Any) -> None:
        src = write_json(tmp_path)
        out = str(tmp_path / "deg_to_one.csv")
        with pytest.raises(SystemExit) as exc:
            main(
                [
                    src,
                    "--output",
                    out,
                    "--neutrino-mass",
                    "0.35",
                    "--normalize-to-one",
                ]
            )
        assert exc.value.code == 1
        assert not (tmp_path / "deg_to_one.csv").exists()


@pytest.mark.requires_ensdf
class TestCLIPaceENSDF:
    """paceENSDF flow honors the same export options."""

    def test_paceensdf_honors_normalize_to_one(self, tmp_path: Any) -> None:
        out = str(tmp_path / "p32.csv")
        main(["P32", "--output", out, "--normalize-to-one"])

        df = read_export(out)
        assert "total_spectrum" not in df.columns
        area = np.trapezoid(df["branch_1_spectrum"], df["energy_MeV"])
        assert np.isclose(area, 1.0, rtol=1e-3)

    def test_paceensdf_degenerate_exits_nonzero_no_file(self, tmp_path: Any) -> None:
        out = str(tmp_path / "tc99deg.csv")
        with pytest.raises(SystemExit) as exc:
            main(["Tc99", "--output", out, "--neutrino-mass", "0.35"])
        assert exc.value.code == 1
        assert not (tmp_path / "tc99deg.csv").exists()