2magiDialogCut.py

import requests
import json
from PIL import Image
import os


def getImgCoordinatesByMagi(image_path:"包含后缀的文件路径") -> "全部对话坐标list,失败返回none":
    headers = {
        'authority': 'rogerxavier-fastapi-t5-magi.hf.space',
        'method': 'GET',
        'scheme': 'https',
        'Accept': '*/*',
        'Accept-Encoding': 'gzip, deflate, br, zstd',
        'Accept-Language': 'zh-CN,zh;q=0.9',
        'Cookie': 'spaces-jwt=eyJhbGciOiJFZERTQSJ9.eyJyZWFkIjp0cnVlLCJwZXJtaXNzaW9ucyI6eyJyZXBvLmNvbnRlbnQucmVhZCI6dHJ1ZX0sIm9uQmVoYWxmT2YiOnsia2luZCI6InVzZXIiLCJfaWQiOiI2NDJhNTNiNTE2ZDRkODI5M2M5YjdiNzgiLCJ1c2VyIjoicm9nZXJ4YXZpZXIifSwiaWF0IjoxNzE2NjExMTE3LCJzdWIiOiIvc3BhY2VzL3JvZ2VyeGF2aWVyL2Zhc3RhcGlfdDVfbWFnaSIsImV4cCI6MTcxNjY5NzUxNywiaXNzIjoiaHR0cHM6Ly9odWdnaW5nZmFjZS5jbyJ9.W00jo8kiRgwCpq5aaGhaPE2RP2jLOyvfimjyIfHVhP1gs7NHkBkRzVTFqYv3TRtZoHNPTiFiI5Ehu12KP06sDQ',
        'Dnt': '1',
        'Priority': 'u=1, i',
        'Sec-Ch-Ua': '"Chromium";v="124", "Google Chrome";v="124", "Not-A.Brand";v="99"',
        'Sec-Ch-Ua-Mobile': '?0',
        'Sec-Ch-Ua-Platform': '"Windows"',
        'Sec-Fetch-Dest': 'empty',
        'Sec-Fetch-Mode': 'cors',
        'Sec-Fetch-Site': 'same-origin',
        'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36'
    }
    with open(image_path, 'rb') as file:
        image_bytes = file.read()
    files = {
        "image": image_bytes,
    }
    magi_fastapi_base = 'https://rogerxavier-fastapi-t5-magi.hf.space/getCoordinates'
    try:
        resp = requests.post(magi_fastapi_base, files=files, headers=headers) ##是仓库public后好像就api可用了
        print(resp.json())
        return resp.json()["texts"]
    except Exception as e:
        print(e)
        return None #意外情况返回


def path_to_format(old_path)->"比如/path/to/your/image/1.jpg  ->/path/to/your/image/100.jpg 经过03d":
    # 获取文件名和文件扩展名
    file_path, file_name = os.path.split(old_path)
    file_name, file_extension = os.path.splitext(file_name)
    # 提取文件名中的数字部分（假设数字部分在文件名的末尾）
    number_str = ''.join(filter(str.isdigit, file_name))
    number = int(number_str)
    # 格式化数字部分为03d格式
    formatted_number = "{:03d}".format(number)
    # 生成新的文件名
    new_file_name = f"{file_name.replace(number_str, formatted_number)}{file_extension}"
    new_absolute_path = os.path.join(file_path, new_file_name)
    return new_absolute_path


def save_img(new_save_path:"新文件的保存路径(包含后缀)",old_img_path:"旧文件路径(包含后缀)")->"void生成新的文件保存 ,传入旧文件路径是为了删除有问题的旧文件":
    print(new_save_path)
    # 原始照片文件名
    original_image = old_img_path
    #打开原始照片
    # img = Image.open(original_image)
    # text_bboxes_for_all_images = getImgCoordinatesByMagi(original_image)
    # if text_bboxes_for_all_images is not None:
    #     for index, box in enumerate(text_bboxes_for_all_images):
    #         cropped_img = img.crop(tuple(box))
    #         # 保存裁剪后的图片，并按照数字大小依次排序命名(包括本身的也要3d,比如 0.jpg - >000.000_cropped.jpg)
    #
    #         cropped_img.save(f"{os.path.splitext(new_save_path)[0]}_{index:03d}_cropped.jpg")
    # else:
    #     print("图片识别有问题,准备删除")
    #     os.remove(original_image)


    ##防止文件打开无法删除

    text_bboxes_for_all_images = getImgCoordinatesByMagi(original_image)
    if text_bboxes_for_all_images is not None:
        with Image.open(original_image) as img:
            for index, box in enumerate(text_bboxes_for_all_images):
                cropped_img = img.crop(tuple(box))
                # 保存裁剪后的图片，并按照数字大小依次排序命名(包括本身的也要3d,比如 0.jpg - >000.000_cropped.jpg)

                cropped_img.save(f"{os.path.splitext(new_save_path)[0]}_{index:03d}_cropped.jpg")
    else:
        print("图片识别有问题,准备删除")
        os.remove(original_image)


if __name__ == '__main__':
    # 获取需要裁剪的无水印漫画位置
    img_path = 'manga1'
    subdir_path = os.path.join(os.getcwd(), img_path)

    # 图片素材获取（包含子目录下所有图片）
    image_files = []
    for root, dirs, files in os.walk(subdir_path):
        for file in files:
            if file.endswith(".jpg") or file.endswith(".png"):
                image_files.append(os.path.relpath(os.path.join(root, file)))

    # 创建处理后的子目录在与image_files同级目录下
    processed_subdir_path = os.path.join(os.path.dirname(subdir_path), f"{img_path}2")#加2表示经过2阶段处理
    os.makedirs(processed_subdir_path, exist_ok=True)

    # 对image_files进行某种处理，生成新图片，并保存在处理后的子目录中
    for img_file in image_files:
        # 处理图片的代码（这里仅作示例）
        # 假设处理后的图片为new_img
        img_dir = os.path.dirname(img_file)
        new_img_dir = os.path.join(processed_subdir_path, img_dir)
        os.makedirs(new_img_dir, exist_ok=True)

        new_img_path = os.path.join(new_img_dir, os.path.basename(img_file))
        new_img_path = path_to_format(new_img_path)

        if not os.path.exists(f"{os.path.splitext(new_img_path)[0]}_{1:03d}_cropped.jpg"):
            # 如果已经处理过那么跳过 ->只看第一个数字片段就行
            # 处理图片并保存
            save_img(new_save_path=new_img_path, old_img_path=img_file)
        else:
            print(f"Skipping {new_img_path} as it already exists.")