import glob 
import sys

sys.path.append('../..')
from natsort import natsort
import SRFlow.code.options.options as option

import torch
from SRFlow.code.utils.util import opt_get
from SRFlow.code.models.SRFlow_model import SRFlowModel
import numpy as np
import os
import cv2


def fiFindByWildcard(wildcard):
    return natsort.natsorted(glob.glob(wildcard, recursive=True))


def load_model(conf_path):
    opt = option.parse(conf_path, is_train=False)
    opt['gpu_ids'] = None
    opt = option.dict_to_nonedict(opt)
    model = SRFlowModel(opt, 0)

    model_path = opt_get(opt, ['model_path'], None)
    model.load_network(load_path='models/SRFlow/35000_G.pth', network=model.netG)
    return model, opt


def predict(model, lr):
    model.feed_data({"LQ": t(lr)}, need_GT=False)
    model.test()
    visuals = model.get_current_visuals(need_GT=False)
    return visuals.get('rlt', visuals.get("SR"))


def t(array): return torch.Tensor(np.expand_dims(array.transpose([2, 0, 1]), axis=0).astype(np.float32)) / 255


def rgb(t): return (
        np.clip((t[0] if len(t.shape) == 4 else t).detach().cpu().numpy().transpose([1, 2, 0]), 0, 1) * 255).astype(
    np.uint8)


def imread(path):
    return cv2.imread(path)[:, :, [2, 1, 0]]


def imwrite(path, img):
    os.makedirs(os.path.dirname(path), exist_ok=True)
    cv2.imwrite(path, img[:, :, [2, 1, 0]])


def imCropCenter(img, size):
    h, w, c = img.shape

    h_start = max(h // 2 - size // 2, 0)
    h_end = min(h_start + size, h)

    w_start = max(w // 2 - size // 2, 0)
    w_end = min(w_start + size, w)

    return img[h_start:h_end, w_start:w_end]


def impad(img, top=0, bottom=0, left=0, right=0, color=255):
    return np.pad(img, [(top, bottom), (left, right), (0, 0)], 'reflect')