dsProject/dsLightRag/Util/DocxUtil.py

import os
import subprocess
import uuid


def get_docx_content_by_pandoc(docx_file):
    # 最后拼接的内容
    content = ""
    # output_file 设置为临时目录下的uuid.md
    file_name = uuid.uuid4().hex
    temp_markdown = os.path.join('./static/markdown/', file_name + '.md')
    # 调用pandoc将docx文件转换成markdown
    os.mkdir("./static/Images/" + file_name)
    subprocess.run(['pandoc', docx_file, '-f', 'docx', '-t', 'markdown', '-o', temp_markdown,
                    '--extract-media=./static/Images/' + file_name])
    # 读取然后修改内容，输出到新的文件
    wmf_idx = 0  # wmf索引
    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"}'):
                continue
            # 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]
                img_idx += 1
                content += q + "\n"
            else:
                content += line.strip().replace("**", "") + "\n"

    # 删除临时文件 output_file
    os.remove(temp_markdown)
    return content.replace("\n\n", "\n").replace("\\","")