# 安装所有格式的支持（推荐新手）
pip install "markitdown[all]"

# 或者只装你需要的格式
pip install "markitdown[pdf, docx, pptx]"
            

# 基本用法：转换后输出到屏幕
markitdown report.pdf

# 保存到文件
markitdown report.pdf -o report.md

# 管道操作：从其他命令读取输入
cat report.pdf | markitdown

# 使用 Azure 文档智能服务
markitdown report.pdf -d -e "<your-endpoint>"
            

from markitdown import MarkItDown

md = MarkItDown()
result = md.convert("季度报告.xlsx")
print(result.text_content)
            

from markitdown import MarkItDown
from openai import OpenAI

client = OpenAI()
md = MarkItDown(
    llm_client=client,
    llm_model="gpt-4o",
)
result = md.convert("产品截图.jpg")
print(result.text_content)
            

class DocumentConverter:
    def accepts(self, file_stream,
                 stream_info, **kwargs) -> bool:
        "我能处理这个文件吗？"
        raise NotImplementedError

    def convert(self, file_stream,
                 stream_info, **kwargs):
        "转换它！"
        raise NotImplementedError
            

@dataclass(kw_only=True, frozen=True)
class StreamInfo:
    mimetype: Optional[str] = None
    extension: Optional[str] = None
    charset: Optional[str]   = None
    filename: Optional[str]  = None
    local_path: Optional[str]= None
    url: Optional[str]       = None
            

# 排序后逐个尝试
sorted_registrations = sorted(
    self._converters,
    key=lambda x: x.priority
)

for converter_registration in sorted_registrations:
    _accepts = converter.accepts(
        file_stream, stream_info, **_kwargs
    )
    if _accepts:
        res = converter.convert(
            file_stream, stream_info, **_kwargs
        )
    if res is not None:
        return res
            

# 归一化内容
res.text_content = "\n".join(
    [line.rstrip()
     for line in re.split(r"\r?\n",
         res.text_content)]
)
res.text_content = re.sub(
    r"\n{3,}", "\n\n",
    res.text_content
)
            

# 逐页判断是否是表单
with pdfplumber.open(pdf_bytes) as pdf:
    for page_idx, page in enumerate(pdf.pages):
        page_content = _extract_form_content(
            page
        )
        if page_content is not None:
            markdown_chunks.append(
                page_content
            )
        page.close()  # 立即释放内存
            

class DocxConverter(HtmlConverter):
    def convert(self, file_stream,
                 stream_info, **kwargs):
        pre_process_stream = pre_process_docx(
            file_stream
        )
        return self._html_converter.convert_string(
            mammoth.convert_to_html(
                pre_process_stream,
                style_map=style_map
            ).value,
            **kwargs,
        )
            

def _load_plugins():
    global _plugins
    if _plugins is not None:
        return _plugins  # 已加载过

    _plugins = []
    for entry_point in entry_points(
        group="markitdown.plugin"
    ):
        try:
            _plugins.append(
                entry_point.load()
            )
        except Exception:
            warn("Plugin failed to load")
    return _plugins
            

from markitdown import (
    MarkItDown, DocumentConverter,
    DocumentConverterResult, StreamInfo,
)

def register_converters(
    markitdown: MarkItDown, **kwargs
):
    markitdown.register_converter(
        RtfConverter()
    )

class RtfConverter(DocumentConverter):
    def accepts(self, file_stream,
                 stream_info, **kwargs) -> bool:
        ext = (stream_info.extension
               or "").lower()
        return ext in [".rtf"]

    def convert(self, file_stream,
                 stream_info, **kwargs):
        encoding = stream_info.charset
        data = file_stream.read()
                      .decode(encoding)
        return DocumentConverterResult(
            markdown=rtf_to_text(data)
        )
            

# Claude Desktop 配置
{
  "mcpServers": {
    "markitdown": {
      "command": "uvx",
      "args": ["markitdown-mcp"]
    }
  }
}
            

from markitdown from pathlib import Path

md = MarkItDown()
output_dir = Path("markdown_output")
output_dir.mkdir(exist_ok=True)

# 转换 ZIP 包内所有文件
result = md.convert("季度报告.zip")

# 或者遍历文件夹
for f in Path("reports/").glob("*.*"):
    if f.suffix in [".pdf", ".docx", ".xlsx"]:
        r = md.convert(str(f))
        out = output_dir / (f.stem + ".md")
        out.write_text(r.text_content,
                     encoding="utf-8")
        print(f"Converted: {f.name}")
            

# 提示文件类型（从管道读取时有用）
cat data.bin | markitdown -x .pdf

# 指定 MIME 类型
markitdown data.file -m application/pdf

# 列出所有已安装的插件
markitdown --list-plugins

# 启用插件转换
markitdown --use-plugins document.rtf

# 保留 base64 图片数据
markitdown slides.pptx --keep-data-uris