"""Small collection of image helper functions factored out of sfp_module.
These are intentionally lightweight wrappers that avoid depending on the
SfpConnectorModule instance state; callers should pass necessary parameters
(e.g. normalize method) when required.
"""
from typing import Optional
import os
import logging

try:
    import numpy as np
except Exception:  # pragma: no cover - best-effort
    np = None

try:
    from PIL import Image, ImageOps
except Exception:
    Image = None
    ImageOps = None

# Try to import the local normalize_image if available
try:
    from .image_processing import normalize_image
except Exception:
    normalize_image = None

logger = logging.getLogger("VideoReceiverSFP.image_helpers")


def looks_like_encoded_image(payload: bytes) -> bool:
    if not payload:
        return False
    if payload.startswith(b"\x89PNG\r\n\x1a\n"):
        return True
    if payload.startswith(b"\xff\xd8\xff"):
        return True
    if payload.startswith(b"BM"):
        return True
    return False


def extract_index_array(width: int, height: int, row_stride: int, pixel_data: bytes):
    if np is None:
        return None
    total_bytes = row_stride * height
    if len(pixel_data) < total_bytes:
        logger.warning("extract_index_array: pixel_data shorter than expected (got=%d need=%d)", len(pixel_data), total_bytes)
        return None
    try:
        arr = np.frombuffer(pixel_data[:total_bytes], dtype=np.uint8)
        row_pixels = width
        if row_stride == row_pixels:
            arr = arr.reshape((height, width))
        else:
            arr = arr.reshape((height, row_stride))[:, :width]
        return arr.copy()
    except Exception:
        logger.exception("extract_index_array: failed to reshape pixel data")
        return None


def build_grayscale_frame(width, height, row_stride, row_pixels, pixel_data, normalize_method: Optional[str] = None):
    if Image is None:
        return None
    try:
        if row_stride == row_pixels:
            img = Image.frombytes("L", (width, height), pixel_data)
        else:
            rows = []
            for r in range(height):
                start = r * row_stride
                rows.append(pixel_data[start:start + row_pixels])
            img = Image.frombytes("L", (width, height), b"".join(rows))

        normalized_img = None
        if normalize_method == 'cp' and normalize_image is not None:
            try:
                if np is not None:
                    arr = np.frombuffer(img.tobytes(), dtype=np.uint8).reshape((height, width))
                    norm = normalize_image(arr, target_type=np.uint8)
                    if norm is not None:
                        normalized_img = Image.fromarray(norm, mode='L')
            except Exception:
                logger.exception("build_grayscale_frame: controlpanel normalize_image failed")
        elif normalize_method == 'autocontrast':
            try:
                if ImageOps is not None:
                    normalized_img = ImageOps.autocontrast(img)
            except Exception:
                normalized_img = None

        if normalized_img is not None:
            frame = normalized_img.convert("RGB")
            try:
                dumps_dir = os.path.join(os.getcwd(), "dumps")
                os.makedirs(dumps_dir, exist_ok=True)
                dbg_path = os.path.join(dumps_dir, "VideoReceiverSFP_last_normalized.png")
                frame.save(dbg_path)
                logger.info("build_grayscale_frame: saved normalized debug frame to %s", dbg_path)
            except Exception:
                pass
            return frame
        return img.convert("RGB")
    except Exception:
        logger.exception("build_grayscale_frame: failed to build grayscale frame")
        return None