import json

import platform

import pytest

from loop_to_python_api.api import (

initialize_exception_handlers,

generate_prediction,

get_prediction_dates,

get_prediction_values_and_dates,

get_dose_recommendations,

get_glucose_effect_velocity,

get_glucose_effect_velocity_dates,

get_glucose_effect_velocity_and_dates,

get_active_carbs,

get_active_insulin,

get_loop_recommendations,

percent_absorption_at_percent_time,

piecewise_linear_percent_rate_at_percent_time,

linear_percent_rate_at_percent_time,

get_dynamic_carbs_on_board,

insulin_percent_effect_remaining,

)

def get_generate_prediction_input():

with open('python_tests/test_files/generate_prediction_input.json', 'r') as f:

return json.load(f)

def get_loop_algorithm_input():

with open('python_tests/test_files/loop_algorithm_input.json', 'r') as f:

return json.load(f)

def get_dynamic_carbs_input():

with open('python_tests/test_files/dynamic_carbs_input.json', 'r') as f:

return json.load(f)

def test_initialize_exception_handlers():

result = initialize_exception_handlers()

assert result is None

def test_generate_prediction():

prediction_input = get_generate_prediction_input()

prediction_values = generate_prediction(prediction_input)

assert isinstance(prediction_values, list)

def test_get_prediction_dates():

prediction_input = get_generate_prediction_input()

prediction_dates = get_prediction_dates(prediction_input)

assert isinstance(prediction_dates, list)

def test_get_prediction_values_and_dates():

prediction_input = get_generate_prediction_input()

values, dates = get_prediction_values_and_dates(prediction_input)

# Assertions to check the types

assert isinstance(values, list), "The prediction values should be a list."

assert isinstance(dates, list), "The prediction dates should be a list."

assert all(isinstance(value, (int, float)) for value in values), "All prediction values should be integers or floats."

assert all(isinstance(date, str) for date in dates), "All prediction dates should be strings."

@pytest.mark.skipif(platform.system() == "Windows", reason="Windows compatibility issue - test disabled for Windows builds")

def test_get_dose_recommendations():

loop_algorithm_input = get_loop_algorithm_input()

dose_recommendations = get_dose_recommendations(loop_algorithm_input)

assert isinstance(dose_recommendations, dict)

def test_get_glucose_effect_velocity():

prediction_input = get_generate_prediction_input()

glucose_effect_velocity = get_glucose_effect_velocity(prediction_input)

assert isinstance(glucose_effect_velocity, list)

def test_get_glucose_effect_velocity_dates():

prediction_input = get_generate_prediction_input()

glucose_effect_velocity_dates = get_glucose_effect_velocity_dates(prediction_input)

assert isinstance(glucose_effect_velocity_dates, list)

@pytest.mark.skipif(platform.system() == "Windows", reason="Windows compatibility issue - test disabled for Windows builds")

def test_get_glucose_effect_velocity_values_and_dates():

loop_algorithm_input = get_loop_algorithm_input()

values, dates = get_glucose_effect_velocity_and_dates(loop_algorithm_input)

# Assertions to check the types

assert isinstance(values, list), "The prediction values should be a list."

assert isinstance(dates, list), "The prediction dates should be a list."

assert all(isinstance(value, (int, float)) for value in values), "All prediction values should be integers or floats."

def test_get_active_carbs():

loop_algorithm_input = get_loop_algorithm_input()

active_carbs = get_active_carbs(loop_algorithm_input)

assert isinstance(active_carbs, float)

def test_get_active_insulin():

loop_algorithm_input = get_loop_algorithm_input()

active_insulin = get_active_insulin(loop_algorithm_input)

assert isinstance(active_insulin, float)

def test_percent_absorption_at_percent_time():

result = percent_absorption_at_percent_time(0.2)

assert isinstance(result, float)

def test_piecewise_linear_percent_rate_at_percent_time():

result = piecewise_linear_percent_rate_at_percent_time(0.2)

assert isinstance(result, float)

def test_linear_percent_rate_at_percent_time():

result = linear_percent_rate_at_percent_time(0.2)

assert isinstance(result, float)

@pytest.mark.skip(reason="Known unit conversion issue: 'g is not compatible with mg/dL·s' - see README.md Known Issues section")

def test_get_dynamic_carbs_on_board():

dynamic_carbs_input = get_dynamic_carbs_input()

dynamic_carbs_on_board = get_dynamic_carbs_on_board(dynamic_carbs_input)

assert isinstance(dynamic_carbs_on_board, float)

def test_get_loop_recommendations():

loop_algorithm_input = get_loop_algorithm_input()

loop_recommendations = get_loop_recommendations(loop_algorithm_input)

assert isinstance(loop_recommendations, str)

@pytest.mark.skipif(platform.system() == "Windows", reason="Windows compatibility issue - test disabled for Windows builds")

def test_insulin_percent_effect_remaining():

# Test with typical rapid-acting insulin parameters

result = insulin_percent_effect_remaining(

minutes=60, # 60 minutes after injection

action_duration=360, # 6 hours total duration

peak_activity_time=75, # Peak at 75 minutes

delay=10 # 10 minute delay

)

# Basic assertions

assert isinstance(result, float), "Result should be a float"

assert 0.0 <= result <= 1.0, f"Result should be between 0.0 and 1.0, got {result}"

# Test edge cases

# At time 0 (before delay), should have close to 100% effect remaining

result_start = insulin_percent_effect_remaining(0, 360, 75, 10)

assert result_start > 0.9, f"At start, should have >90% remaining, got {result_start}"

# At very end of action duration, should have very little effect remaining

result_end = insulin_percent_effect_remaining(360, 360, 75, 10)

assert result_end < 0.1, f"At end, should have <10% remaining, got {result_end}"

# At peak time, should have less than 100% but more than end

result_peak = insulin_percent_effect_remaining(75, 360, 75, 10)

assert result_end < result_peak < result_start, "Effect should decrease over time"