dsProject/dsLightRag/Util/DocxUtil.py

import logging
import os
import subprocess
import uuid

from PIL import Image
import os


# 或者如果你想更详细地控制日志输出
logger = logging.getLogger('DocxUtil')
logger.setLevel(logging.INFO)
handler = logging.StreamHandler()
handler.setFormatter(logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s'))
logger.addHandler(handler)
logging.basicConfig(format="%(levelname)s:%(message)s", level=logging.INFO)


def resize_images_in_directory(directory_path, max_width=640, max_height=480):
    """
    遍历目录下所有图片并缩放到指定尺寸
    :param directory_path: 图片目录路径
    :param max_width: 最大宽度
    :param max_height: 最大高度
    """
    # 支持的图片格式
    valid_extensions = ('.jpg', '.jpeg', '.png', '.bmp', '.gif')

    for root, _, files in os.walk(directory_path):
        for filename in files:
            if filename.lower().endswith(valid_extensions):
                file_path = os.path.join(root, filename)
                try:
                    with Image.open(file_path) as img:
                        # 计算缩放比例
                        width, height = img.size
                        ratio = min(max_width / width, max_height / height)
                        # 如果图片已经小于目标尺寸，则跳过
                        if ratio >= 1:
                            continue
                        # 计算新尺寸并缩放
                        new_size = (int(width * ratio), int(height * ratio))
                        resized_img = img.resize(new_size, Image.Resampling.LANCZOS)

                        # 保存图片（覆盖原文件）
                        resized_img.save(file_path)
                        logger.info(f"已缩放: {file_path} -> {new_size}")
                except Exception as e:
                    logger.error(f"处理 {file_path} 时出错: {str(e)}")


import hashlib


def calculate_docx_md5(docx_file_path):
    """
    计算docx文件的MD5哈希值
    :param docx_file_path: docx文件路径
    :return: MD5哈希字符串
    """
    # 以二进制模式读取文件
    with open(docx_file_path, 'rb') as f:
        file_content = f.read()

    # 创建MD5哈希对象
    md5_hash = hashlib.md5()

    # 更新哈希值
    md5_hash.update(file_content)

    # 返回16进制格式的哈希值
    return md5_hash.hexdigest()


def get_docx_content_by_pandoc(docx_file):
    # 最后拼接的内容
    content = ""
    # 计算 docx_file 的字符串md5值
    md5_value = calculate_docx_md5(docx_file)

    # 将docx_file去掉扩展名
    prefix = docx_file.split(".")[0].split("/")[-1]
    temp_markdown = os.path.join('./static/markdown/', prefix + '.md')
    # 调用pandoc将docx文件转换成markdown
    path = "./static/Images/" + md5_value
    if not os.path.exists(path):
        os.mkdir("./static/Images/" + md5_value)
    subprocess.run(['pandoc', docx_file, '-f', 'docx', '-t', 'markdown', '-o', temp_markdown,
                    '--extract-media=./static/Images/' + md5_value])
    # 遍历目录 './static/Images/'+file_name 下所有的图片，缩小于640*480的尺寸上

    resize_images_in_directory('./static/Images/' + md5_value + '/media')
    # 读取然后修改内容，输出到新的文件
    img_idx = 0  # 图片索引
    with open(temp_markdown, 'r', encoding='utf-8') as f:
        for line in f:
            line = line.strip()
            if not line:
                continue
            # 跳过图片高度描述行
            if line.startswith('height=') and (line.endswith('in"}') or line.endswith('in"')):
                continue
            # height="1.91044072615923in"
            # 使用find()方法安全地检查图片模式
            is_img = line.find("![](") >= 0 and (
                    line.find(".png") > 0 or
                    line.find(".jpg") > 0 or
                    line.find(".jpeg") > 0
            )
            if is_img:
                # ![](media/image3.png){width="3.1251607611548557in"
                # height="3.694634733158355in"}
                # ![](../static/Images/01b20e04085e406ea5375791da58a60f/media/image3.png){width="3.1251607611548557in"
                pos = line.find(")")
                q = line[:pos + 1]
                q = q.replace("./static", ".")
                # q = q[4:-1]
                # q='<img src="'+q+'" alt="我是图片">'
                img_idx += 1
                content += q + "\n"
            else:
                content += line.strip().replace("**", "") + "\n"
                content = content.replace("\phantom", "")
    # 将content回写到markdown文件
    with open(temp_markdown, 'w', encoding='utf-8') as f:
        f.write(content)
    # 删除临时文件 output_file
    # os.remove(temp_markdown)
    return content.replace("\n\n", "\n").replace("\\", "")