3

backend/app/processors/parsers/markitdown_parser.py

@@ -0,0 +1,30 @@
+"""MarkItDown 解析器（优先）。"""
+
+from app.core.logging import get_logger
+
+logger = get_logger(__name__)
+
+
+def parse_with_markitdown(file_content: bytes, filename: str) -> str | None:
+    """用 MarkItDown 解析文档，返回 Markdown 或 None（失败时）。"""
+    try:
+        from markitdown import MarkItDown
+
+        md = MarkItDown()
+        # MarkItDown 需要文件路径或文件对象，用临时文件
+        import tempfile
+        from pathlib import Path
+
+        suffix = Path(filename).suffix.lower()
+        with tempfile.NamedTemporaryFile(suffix=suffix, delete=False) as tmp:
+            tmp.write(file_content)
+            tmp_path = tmp.name
+
+        try:
+            result = md.convert(tmp_path)
+            return result.text_content if result and result.text_content else None
+        finally:
+            Path(tmp_path).unlink(missing_ok=True)
+    except Exception as e:
+        logger.warning("MarkItDown 解析失败 (%s): %s", filename, e)
+        return None