新增DifyCompareTest.py文件以实现Dify新旧流程对比测试功能，并更新.gitignore以忽略新测试文件。

.gitignore

@@ -9,3 +9,4 @@ rag2_0/demo/Test.py
 data/excel/*.xlsx
 rag2_0/demo/ProfessionalTermAnalyzer.py
 data/logs/*
+rag2_0/dify/Test.py

rag2_0/dify/dify_client/client.py

@@ -457,3 +457,40 @@ class KnowledgeBaseClient(DifyClient):
         data = {"segment": segment_data}
         url = f"/datasets/{self._get_dataset_id()}/documents/{document_id}/segments/{segment_id}"
         return self._send_request("POST", url, json=data, **kwargs)
+
+    def retrieve(
+        self,
+        query: str,
+        retrieval_model: dict | None = None,
+        external_retrieval_model: dict | None = None,
+        **kwargs
+    ):
+        """
+        Retrieve from knowledge base.
+
+        :param query: Query keyword for retrieval
+        :param retrieval_model: Retrieval parameters (optional), example:
+            {
+                "search_method": "semantic_search",  # One of: "keyword_search", "semantic_search", "full_text_search", "hybrid_search"
+                "reranking_enable": True,  # Optional, whether to enable reranking
+                "reranking_mode": {  # Optional, required if reranking_enable is True
+                    "reranking_provider_name": "provider_name",
+                    "reranking_model_name": "model_name"
+                },
+                "weights": 0.5,  # Optional, weight for semantic search in hybrid search mode
+                "top_k": 10,  # Optional, number of results to return
+                "score_threshold_enabled": True,  # Optional, whether to enable score threshold
+                "score_threshold": 0.5  # Optional, score threshold value
+            }
+        :param external_retrieval_model: Unused field
+        :return: Response from the API
+        """
+        data = {"query": query}
+        if retrieval_model is not None:
+            data["retrieval_model"] = retrieval_model
+        if external_retrieval_model is not None:
+            data["external_retrieval_model"] = external_retrieval_model
+            
+        url = f"/datasets/{self._get_dataset_id()}/retrieve"
+        return self._send_request("POST", url, json=data, **kwargs)
+