"""Tests for the JSON output-formatting helpers."""

import math

import numpy as np

from mcltspice.output_format import compact_list, downsample_indices, round_sig


class TestRoundSig:
    def test_significant_figures(self):
        assert round_sig(123456.789, 3) == 123000.0
        assert round_sig(0.00123456, 3) == 0.00123

    def test_zero_and_nonfinite_pass_through(self):
        assert round_sig(0.0) == 0.0
        assert math.isinf(round_sig(float("inf")))
        assert math.isnan(round_sig(float("nan")))


class TestCompactList:
    def test_decimal_places(self):
        assert compact_list([1.23456, 2.98765], decimal_places=2) == [1.23, 2.99]

    def test_sig_figs(self):
        assert compact_list([123456.0], sig_figs=3) == [123000.0]

    def test_no_rounding_returns_floats(self):
        out = compact_list([1, 2, 3])
        assert out == [1.0, 2.0, 3.0]
        assert all(isinstance(v, float) for v in out)


class TestDownsampleIndices:
    def test_no_downsample_when_under_limit(self):
        idx = downsample_indices(50, 100)
        assert np.array_equal(idx, np.arange(50))

    def test_stride_for_transient(self):
        idx = downsample_indices(1000, 100)
        assert len(idx) <= 100
        # plain stride: evenly spaced
        assert idx[0] == 0
        assert idx[1] - idx[0] == idx[2] - idx[1]

    def test_peak_preserving_keeps_a_narrow_spike(self):
        # A flat magnitude with one sharp spike: peak-preserving must keep it.
        n = 1000
        mag = np.ones(n, dtype=complex)
        mag[497] = 1000.0  # narrow resonance peak
        idx = downsample_indices(n, 50, signals=[mag])
        assert 497 in idx  # spike survived downsampling