bugfix: 创建清洗任务时修改数据集状态；无法删除已在模板/运行任务的算子

* bugfix: 创建清洗任务时修改数据集状态；无法删除已在模板/运行任务的算子

runtime/ops/formatter/mineru_formatter/__init__.py

@@ -3,4 +3,4 @@
 from datamate.core.base_op import OPERATORS
 
 OPERATORS.register_module(module_name='MineruFormatter',
-                          module_path="ops.formatter.external_pdf_formatter.process")
+                          module_path="ops.formatter.mineru_formatter.process")

runtime/ops/formatter/mineru_formatter/process.py

@@ -25,6 +25,8 @@ class MineruFormatter(Mapper):
     def execute(self, sample: Dict[str, Any]) -> Dict[str, Any]:
         start = time.time()
         filename = sample[self.filename_key]
+        if not filename.lower().endswith(".pdf"):
+            return sample
         try:
             data = {"source_path": sample[self.filepath_key], "export_path": sample[self.export_path_key]}
             response = http_request(method="POST", url=self.pdf_extract_url, data=data)