import sys, os
sys.path.insert(0, os.path.join(os.path.dirname(__file__), "..", "services", "recolor"))
import numpy as np
import cv2
from PIL import Image
from fabric_recolor.masks import _silhouette_bands
from fabric_recolor.labcolor import rgb_to_lab

for name, mp, kp in [
    ("detail", "data/media/assets/20/20282cb1bd8a58596050b99b9f29e079145a7bcc45ea46b8523ae2389801fbdc.png", "data/media/assets/20/20282cb1bd8a58596050b99b9f29e079145a7bcc45ea46b8523ae2389801fbdc.mask.png"),
    ("packshot_front", "data/media/assets/79/7987376a67865864247263e5028a83950a2fd900740bbe339fab1ff8e6b3fbb2.png", "data/media/assets/79/7987376a67865864247263e5028a83950a2fd900740bbe339fab1ff8e6b3fbb2.mask.png"),
    ("hero", "data/media/assets/f0/f06282d9ba621d137d7c978cf931792ffbcb6b5c9764cf94e964f83ded13ddcf.png", "data/media/assets/f0/f06282d9ba621d137d7c978cf931792ffbcb6b5c9764cf94e964f83ded13ddcf.mask.png"),
]:
    rgb = np.asarray(Image.open(mp).convert("RGB"), dtype=np.uint8)
    mask = np.asarray(Image.open(kp).convert("L"), dtype=np.uint8)
    if mask.shape != rgb.shape[:2]:
        mask = cv2.resize(mask, (rgb.shape[1], rgb.shape[0]), interpolation=cv2.INTER_NEAREST)
    mb = (mask > 128).astype(np.uint8)
    band, defect = _silhouette_bands(mb, 6)
    lab = rgb_to_lab(rgb)
    med = np.median(lab[mb > 0], axis=0)
    h, w = mb.shape
    # hull i jego otoczenie
    cnts, _ = cv2.findContours(mb, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
    hull = np.zeros_like(mb)
    for c in cnts:
        cv2.fillPoly(hull, [cv2.convexHull(c)], 1)
    outside = ((1 - mb) & (1 - hull)).astype(np.uint8)  # tlo poza otoczka
    n, lab_c, stats, _ = cv2.connectedComponentsWithStats(defect, connectivity=8)
    print(f"== {name}")
    k3 = cv2.getStructuringElement(cv2.MORPH_ELLIPSE, (3, 3))
    for i in range(1, n):
        area = stats[i, cv2.CC_STAT_AREA]
        if area < 5000:
            continue
        comp = (lab_c == i).astype(np.uint8)
        rim = cv2.dilate(comp, k3) & (1 - comp)  # piksele sasiadujace
        to_mask = int((rim & mb).sum())
        to_outside = int((rim & outside).sum())
        to_defect = int(rim.sum()) - to_mask - to_outside  # inne defekty
        dl = np.linalg.norm(lab[comp > 0].mean(axis=0) - med)
        ys, xs = np.nonzero(comp)
        edge = xs.min() == 0 or xs.max() == w - 1 or ys.min() == 0 or ys.max() == h - 1
        print(f"  defekt {i}: area={area} ({100*area/mb.size:.2f}%) dE={dl:.1f} krawedz={edge} rim: maska={to_mask} tlo={to_outside} ({100*to_mask/max(1,to_mask+to_outside):.0f}% przylega do maski)")