docx-mcp

Instructions

Implementation Reference

• src/docx_mcp/server.py:1114-1182 (handler)

  The 'extract_images' tool implementation, including the @app.tool() decorator and the core logic for extracting and optionally saving or base64-encoding images from a Word document.

  @app.tool()
  def extract_images(
      filepath: str,
      output_dir: Optional[str] = None,
      return_base64: bool = False,
  ) -> dict[str, Any]:
      """
      Extract all images from a Word document.
  
      Args:
          filepath: Path to the document
          output_dir: Directory to save extracted images (optional)
          return_base64: If True, return images as base64 encoded strings
  
      Returns:
          Dictionary with extracted images info and optionally base64 data
      """
      import base64
  
      logger.info("Extracting images", extra={"tool": "extract_images", "filepath": filepath})
  
      try:
          doc = safe_open_document(filepath)
          extracted = []
          image_index = 0
  
          for rel in doc.part.rels.values():
              if "image" in rel.target_ref:
                  image_part = rel.target_part
                  image_data = image_part.blob
                  filename = Path(image_part.partname).name
                  content_type = image_part.content_type
  
                  image_info = {
                      "index": image_index,
                      "filename": filename,
                      "content_type": content_type,
                      "size_bytes": len(image_data),
                  }
  
                  # Save to output directory if specified
                  if output_dir:
                      out_path = normalize_path(output_dir)
                      out_path.mkdir(parents=True, exist_ok=True)
                      image_file = out_path / filename
                      image_file.write_bytes(image_data)
                      image_info["saved_path"] = str(image_file)
  
                  # Return base64 if requested
                  if return_base64:
                      image_info["base64"] = base64.b64encode(image_data).decode("utf-8")
  
                  extracted.append(image_info)
                  image_index += 1
  
          logger.info(f"Extracted {len(extracted)} images", extra={"filepath": filepath})
  
          return {
              "status": "success",
              "filepath": filepath,
              "images": extracted,
              "count": len(extracted),
          }
      except DocxMcpError as e:
          logger.warning(e.message, extra={"tool": "extract_images", "error_code": e.error_code})
          return {"status": "error", "error": e.message, "error_code": e.error_code}
      except Exception as e:
          logger.error(f"Unexpected error extracting images: {str(e)}")
          return {"status": "error", "error": str(e)}