import easyocr
import numpy as np
import cv2
import re

reader = easyocr.Reader(['en'], gpu=False)

def enhance_image(img):
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    blur = cv2.GaussianBlur(gray, (5, 5), 0)
    _, thresh = cv2.threshold(blur, 120, 255, cv2.THRESH_BINARY_INV)

    # Find contours to crop region with max text
    contours, _ = cv2.findContours(thresh, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
    height, width = gray.shape
    digit_region = gray  # default

    if contours:
        largest = max(contours, key=cv2.contourArea)
        x, y, w, h = cv2.boundingRect(largest)
        pad = 10
        x = max(x - pad, 0)
        y = max(y - pad, 0)
        w = min(w + 2 * pad, width - x)
        h = min(h + 2 * pad, height - y)
        digit_region = gray[y:y+h, x:x+w]

    # Resize and sharpen
    digit_region = cv2.resize(digit_region, None, fx=2, fy=2, interpolation=cv2.INTER_CUBIC)
    kernel = np.array([[0, -1, 0], [-1, 5, -1], [0, -1, 0]])
    digit_region = cv2.filter2D(digit_region, -1, kernel)

    return digit_region

def extract_weight_from_image(pil_img):
    try:
        img = np.array(pil_img)
        img = cv2.cvtColor(img, cv2.COLOR_RGB2BGR)
        processed = enhance_image(img)

        results = reader.readtext(processed)

        best_weight = None
        best_conf = 0.0

        for (_, text, conf) in results:
            text = text.lower().strip()
            text = text.replace(",", ".").replace("o", "0").replace("s", "5")
            text = text.replace("kgs", "").replace("kg", "")
            text = re.sub(r"[^\d\.]", "", text)

            if len(text) >= 2 and re.fullmatch(r"\d{1,4}(\.\d{1,3})?", text):
                if conf > best_conf:
                    best_weight = text
                    best_conf = conf

        if not best_weight:
            return "Not detected", 0.0

        if "." in best_weight:
            int_part, dec_part = best_weight.split(".")
            best_weight = f"{int_part.lstrip('0') or '0'}.{dec_part}"
        else:
            best_weight = best_weight.lstrip("0") or "0"

        return best_weight, round(best_conf * 100, 2)

    except Exception as e:
        return f"Error: {str(e)}", 0.0