import re
import kalpaa.read_bin_csv
import pathlib
import dataclasses
import logging
import typing
import numpy

_logger = logging.getLogger(__name__)


TEST_DATA_DIR = pathlib.Path(__file__).resolve().parent / "test_files"


def test_regex_matches():
	apsd_v_1 = "APSD_V_dot1_mean"

	actual_match1 = re.match(kalpaa.read_bin_csv.BINNED_HEADER_REGEX, apsd_v_1)

	# For reference, REGEX is currently: APSD_(?P<measurement_type>\w+)_(?P<dot_name>\w+)_(?P<summary_stat>mean|stdev)\s*
	assert actual_match1 is not None
	groups = actual_match1.groupdict()
	assert groups["measurement_type"] == "V"
	assert groups["dot_name"] == "dot1"
	assert groups["summary_stat"] == "mean"


def test_parse_headers(snapshot):
	example_headers = [
		# using these headers from recent run: APSD_V_dot1_mean, APSD_V_dot1_stdev, APSD_V_dot2_mean, APSD_V_dot2_stdev, APSD_V_line_mean, APSD_V_line_stdev, APSD_V_triangle1_mean, APSD_V_triangle1_stdev, APSD_V_triangle2_mean, APSD_V_triangle2_stdev, APSD_V_uprise1_mean, APSD_V_uprise1_stdev, APSD_V_uprise2_mean, APSD_V_uprise2_stdev
		"APSD_V_dot1_mean",
		"APSD_V_dot1_stdev",
		"APSD_V_dot2_mean",
		"APSD_V_dot2_stdev",
		"APSD_V_line_mean",
		"APSD_V_line_stdev",
		"APSD_V_triangle1_mean",
		"APSD_V_triangle1_stdev",
		"APSD_V_triangle2_mean",
		"APSD_V_triangle2_stdev",
		"APSD_V_uprise1_mean",
		"This is not a valid header",
		"CPSD_correlation_V_dot1_dot2_mean",
		"CPSD_correlation_V_dot1_dot2_stdev",
		"CPSD_phase_V_dot1_dot2_mean",
		"CPSD_phase_V_dot1_dot2_stdev",
	]

	# force logger to be used for now
	_logger.debug("parsing headers for test")

	def null_asdict(dataclz) -> typing.Optional[dict]:
		if dataclz is None:
			return None
		return dataclasses.asdict(dataclz)

	actual_parsed = [
		null_asdict(kalpaa.read_bin_csv._parse_bin_header(h)) for h in example_headers
	]
	assert actual_parsed == snapshot


def test_binned_data_dot_measurement(snapshot):

	dots_json = TEST_DATA_DIR / "dots.json"
	csv_file = TEST_DATA_DIR / "test_binned_apsd_V.csv"

	actual_read = kalpaa.read_bin_csv.read_dots_and_binned(dots_json, csv_file)

	assert dataclasses.asdict(actual_read) == snapshot


def test_binned_data_dot_measurement_costs(snapshot):

	dots_json = TEST_DATA_DIR / "dots.json"
	v_csv_file = TEST_DATA_DIR / "test_binned_apsd_V.csv"
	ex_csv_file = TEST_DATA_DIR / "test_binned_apsd_Ex.csv"

	# it's overkill but while we have the mental model of what the input form is of these numpy arrays we should record it!
	test_dipole_1 = [10, 20, 30, 0.4, 0.5, 0.6, 0.7]
	test_dipole_2 = [15, 25, 35, -4, 0, 0, 0.11]
	test_dipoles_configuration_1 = [test_dipole_1]
	test_one_dipole_config2 = [test_dipole_2]
	all_one_dipole_configs = numpy.array(
		[test_dipoles_configuration_1, test_one_dipole_config2]
	)

	test_two_dipole_config1 = [test_dipole_1, test_dipole_2]
	all_two_dipole_configs = numpy.array([test_two_dipole_config1])

	binned_v = kalpaa.read_bin_csv.read_dots_and_binned(dots_json, v_csv_file)
	measurements_v = binned_v.measurements(["dot1"])

	binned_ex = kalpaa.read_bin_csv.read_dots_and_binned(dots_json, ex_csv_file)
	measurements_ex = binned_ex.measurements(["dot1"])
	_logger.warning(measurements_ex)

	v_log_noise_stdev_cost_func = measurements_v.stdev_cost_function(use_log_noise=True)
	ex_log_noise_stdev_cost_func = measurements_ex.stdev_cost_function(
		use_log_noise=True
	)

	v_linear_noise_stdev_cost_func = measurements_v.stdev_cost_function(
		use_log_noise=False
	)
	ex_linear_noise_stdev_cost_func = measurements_ex.stdev_cost_function(
		use_log_noise=False
	)

	result_dict = {
		"log": {
			"v": {
				"one": v_log_noise_stdev_cost_func(all_one_dipole_configs),
				"two": v_log_noise_stdev_cost_func(all_two_dipole_configs),
			},
			"ex": {
				"one": ex_log_noise_stdev_cost_func(all_one_dipole_configs),
				"two": ex_log_noise_stdev_cost_func(all_two_dipole_configs),
			},
		},
		"linear": {
			"v": {
				"one": v_linear_noise_stdev_cost_func(all_one_dipole_configs),
				"two": v_linear_noise_stdev_cost_func(all_two_dipole_configs),
			},
			"ex": {
				"one": ex_linear_noise_stdev_cost_func(all_one_dipole_configs),
				"two": ex_linear_noise_stdev_cost_func(all_two_dipole_configs),
			},
		},
	}

	assert result_dict == snapshot


def test_read_csv_with_pairs(snapshot):

	# dots_json = TEST_DATA_DIR / "dots.json"
	# v_csv_file = TEST_DATA_DIR / "test_binned_apsd_V.csv"
	# ex_csv_file = TEST_DATA_DIR / "test_binned_apsd_Ex.csv"
	pair_data_csv = TEST_DATA_DIR / "test_simple_pair_V.csv"

	actual_read = kalpaa.read_bin_csv.read_bin_csv(pair_data_csv)

	assert dataclasses.asdict(actual_read) == snapshot