# Copyright 2026 SARMAE Authors and The HuggingFace Inc. team.
"""Image processor for SARMAE models (self-contained for trust_remote_code)."""

from typing import Optional, Union

import numpy as np

from transformers.image_processing_utils import BaseImageProcessor, BatchFeature, get_size_dict
from transformers.image_transforms import resize, to_channel_dimension_format
from transformers.image_utils import (
    ChannelDimension,
    ImageInput,
    PILImageResampling,
    infer_channel_dimension_format,
    to_numpy_array,
    valid_images,
    validate_preprocess_arguments,
)
from transformers.utils import TensorType, filter_out_non_signature_kwargs, logging


logger = logging.get_logger(__name__)


def _repeat_grayscale_channels(image: np.ndarray, target_channels: int, input_data_format: ChannelDimension) -> np.ndarray:
    if input_data_format == ChannelDimension.FIRST:
        num_channels = image.shape[0]
        if num_channels == target_channels:
            return image
        if num_channels == 1:
            return np.repeat(image, target_channels, axis=0)
        return image[:target_channels]
    num_channels = image.shape[-1]
    if num_channels == target_channels:
        return image
    if num_channels == 1:
        return np.repeat(image, target_channels, axis=-1)
    return image[..., :target_channels]


def _prepare_image_batch(images: ImageInput) -> list:
    if isinstance(images, np.ndarray):
        images = [images]
    elif not isinstance(images, (list, tuple)):
        images = [images]

    prepared = []
    for image in images:
        array = to_numpy_array(image)
        if array.ndim == 2:
            array = np.expand_dims(array, axis=-1)
        prepared.append(array)
    return prepared


class SarmaeImageProcessor(BaseImageProcessor):
    model_input_names = ["pixel_values"]

    def __init__(
        self,
        do_resize: bool = True,
        size: Optional[dict[str, int]] = None,
        resample: PILImageResampling = PILImageResampling.BILINEAR,
        do_rescale: bool = True,
        rescale_factor: float = 1 / 255.0,
        do_normalize: bool = True,
        image_mean: Optional[Union[float, list[float]]] = None,
        image_std: Optional[Union[float, list[float]]] = None,
        do_convert_rgb: bool = False,
        repeat_grayscale_channels: bool = True,
        **kwargs,
    ):
        super().__init__(**kwargs)
        size = size if size is not None else {"height": 224, "width": 224}
        self.do_resize = do_resize
        self.size = size
        self.resample = resample
        self.do_rescale = do_rescale
        self.rescale_factor = rescale_factor
        self.do_normalize = do_normalize
        self.image_mean = image_mean
        self.image_std = image_std
        self.do_convert_rgb = do_convert_rgb
        self.repeat_grayscale_channels = repeat_grayscale_channels

    @filter_out_non_signature_kwargs()
    def preprocess(
        self,
        images: ImageInput,
        do_resize: Optional[bool] = None,
        size: Optional[dict[str, int]] = None,
        resample: Optional[PILImageResampling] = None,
        do_rescale: Optional[bool] = None,
        rescale_factor: Optional[float] = None,
        do_normalize: Optional[bool] = None,
        image_mean: Optional[Union[float, list[float]]] = None,
        image_std: Optional[Union[float, list[float]]] = None,
        return_tensors: Optional[Union[str, TensorType]] = None,
        data_format: Union[str, ChannelDimension] = ChannelDimension.FIRST,
        input_data_format: Optional[Union[str, ChannelDimension]] = None,
        do_convert_rgb: Optional[bool] = None,
        repeat_grayscale_channels: Optional[bool] = None,
    ):
        do_resize = do_resize if do_resize is not None else self.do_resize
        size = get_size_dict(size if size is not None else self.size, default_to_square=True)
        resample = resample if resample is not None else self.resample
        do_rescale = do_rescale if do_rescale is not None else self.do_rescale
        rescale_factor = rescale_factor if rescale_factor is not None else self.rescale_factor
        do_normalize = do_normalize if do_normalize is not None else self.do_normalize
        image_mean = image_mean if image_mean is not None else self.image_mean
        image_std = image_std if image_std is not None else self.image_std
        do_convert_rgb = do_convert_rgb if do_convert_rgb is not None else self.do_convert_rgb
        repeat_grayscale_channels = (
            repeat_grayscale_channels if repeat_grayscale_channels is not None else self.repeat_grayscale_channels
        )

        if do_normalize and (image_mean is None or image_std is None):
            raise ValueError("Normalization requires `image_mean` and `image_std` with one value per channel.")

        images = _prepare_image_batch(images)
        if not valid_images(images):
            raise ValueError("Invalid image type. Must be PIL, numpy, or torch tensor.")

        validate_preprocess_arguments(
            do_rescale=do_rescale,
            rescale_factor=rescale_factor,
            do_normalize=do_normalize,
            image_mean=image_mean,
            image_std=image_std,
            do_resize=do_resize,
            size=size,
            resample=resample,
        )

        processed_images = []
        for image in images:
            image = to_numpy_array(image)
            if do_convert_rgb:
                image = self._convert_image_to_rgb(image)
            if input_data_format is None:
                try:
                    input_data_format = infer_channel_dimension_format(image)
                except ValueError:
                    input_data_format = ChannelDimension.LAST
            if repeat_grayscale_channels:
                image = _repeat_grayscale_channels(image, target_channels=3, input_data_format=input_data_format)
            if do_resize:
                image = resize(
                    image,
                    size=(size["height"], size["width"]),
                    resample=resample,
                    input_data_format=input_data_format,
                )
            if do_rescale:
                image = image * rescale_factor
            if do_normalize:
                image = self.normalize(image=image, mean=image_mean, std=image_std, input_data_format=input_data_format)
            image = to_channel_dimension_format(image, data_format, input_channel_dim=input_data_format)
            processed_images.append(image)

        return BatchFeature(data={"pixel_values": processed_images}, tensor_type=return_tensors)


__all__ = ["SarmaeImageProcessor"]