import sys
import numpy as np
import scipy.optimize
from logging import getLogger
from cam_server.pipeline.data_processing import functions, processor

_logger = getLogger(__name__)

def gauss_function_op(x, H, A, x0, sigma):
    return H + A * np.exp(-(x - x0)**2 / (2 * sigma**2))

def gauss_fit_op(x, y):
    H = min(y)
    A = max(y) - min(y)
    x0 = x[y.argmax()]
    surface = np.trapz((y - H), x=x)
    sigma = surface / (A * np.sqrt(2 * np.pi)) if A != 0 else 1
    try:
        opt_par, cov_par = scipy.optimize.curve_fit(gauss_function_op, x, y, p0=[H, A, x0, sigma])
    except BaseException as e:
        opt_par = np.array([H, A, x0, sigma]).astype("float64")
    return opt_par
    
def gauss_op(x, y):
    H, A, x0, sigma = gauss_fit_op(x, y)
    sigma = abs(sigma)
    gauss = gauss_function_op(x, H, A, x0, sigma)
    y = y - H     # remove offset
    y = y / sum(y) if sum(y) != 0 else y     # normalise to 1
    condition = np.abs(x - x0) > 5 * sigma
    y[condition] = 0.0   # remove outliers
    com = sum(x * y)
    x = x - com
    rms = np.sqrt(np.abs(sum(x**2 * y)))     # y can have negative values
    return gauss, H, A, x0, sigma, com, rms

def get_fw(x, y, level=0.5):
    try:
        ymax, ymin =  np.amax(y),  np.amin(y)
        thr =  (ymax - ymin) * level
        imax, i1, i2 = np.argmax(y), 0, len(x)-1
        for i in range(imax - 1, 0, -1):
            if (y[i] - ymin) <= thr:
                i1 = i
                break
        for i in range(imax + 1, len(x), 1):
            if (y[i] - ymin) <= thr:
                i2 = i
                break
        slope1 = (y[i1 + 1] - y[i1]) / (x[i1 + 1] - x[i1])
        slope2 = (y[i2] - y[i2 - 1]) / (x[i2] - x[i2 - 1])
        dx1 = (ymin + thr - y[i1]) / slope1
        dx2 = (ymin + thr - y[i2]) / slope2
        fw = abs((x[i2] + dx2) - (x[i1] + dx1))
        return fw
    except:
        return 0.0

def process_image(image, pulse_id, timestamp, x_axis, y_axis, parameters, bsdata=None):
  try:
    ret = processor.process_image(image, pulse_id, timestamp, x_axis, y_axis, parameters, bsdata)
    replace = parameters.get("replace", True)
    level = parameters.get("fw_threshold", 0.5)
    sufix = "" if replace else "_op"
    x_axis = ret["x_axis"]
    y_axis = ret["y_axis"]
    x_profile = ret["x_profile"]
    y_profile = ret["y_profile"]
    gauss, H, A, x0, sigma, com, rms = gauss_op(x_axis, x_profile)
    ret["x_fit_offset" + sufix] = H
    ret["x_fit_amplitude" + sufix] = A
    ret["x_fit_mean" + sufix] = x0
    ret["x_fit_standard_deviation" + sufix] = sigma
    ret["x_center_of_mass" + sufix] = com
    ret["x_rms" + sufix] = rms
    fw = get_fw(x_axis, x_profile, level)
    ret["x_fw" + sufix] = fw
    gauss, H, A, x0, sigma, com, rms = gauss_op(y_axis, y_profile)
    ret["y_fit_offset" + sufix] = H
    ret["y_fit_amplitude" + sufix] = A
    ret["y_fit_mean" + sufix] = x0
    ret["y_fit_standard_deviation" + sufix] = sigma
    ret["y_center_of_mass" + sufix] = com
    ret["y_rms" + sufix] = rms
    fw = get_fw(y_axis, y_profile, level)
    ret["y_fw" + sufix] = fw
    return ret
  except Exception as e:
    _logger.warning(sys.exc_info())
    return None