0.2.6 enhance

0.2.6 enhance

server/knowledge_base/kb_doc_api.py

@@ -41,11 +41,13 @@ def search_docs(query: str = Body(..., description="用户输入", examples=["
     
     print(f"search_docs, query:{query}")
     docs = kb.search_docs(query, top_k, score_threshold)
+    if len(pre_doc) > 0:
+        if docs is not None:
+            docs.append(pre_doc[0])
+        else:
+            docs = pre_doc[0]
     data = [DocumentWithScore(**x[0].dict(), score=x[1]) for x in docs]
-    # i = 1
-    # for x in docs:
-    #     print(f"相似文档 {i}: {x}")
-    #     i = i+1
+
 
     return data
 

server/knowledge_base/kb_service/faiss_kb_service.py

@@ -63,7 +63,7 @@ class FaissKBService(KBService):
         print(f"do_search,top_k:{top_k},score_threshold:{score_threshold}")
         with self.load_vector_store().acquire() as vs:
             docs = vs.similarity_search_with_score(query, k=top_k, score_threshold=score_threshold)
-        print(f"do_search,docs:{docs}")
+        #print(f"do_search,docs:{docs}")
         return docs
 
     def do_add_doc(self,

server/knowledge_base/utils.py

@@ -68,7 +68,7 @@ def load_embeddings(model: str = EMBEDDING_MODEL, device: str = embedding_device
     from server.knowledge_base.kb_cache.base import embeddings_pool
     return embeddings_pool.load_embeddings(model=model, device=device)
 
-
+#PDFPlumberLoader
 LOADER_DICT = {"UnstructuredHTMLLoader": ['.html'],
                "UnstructuredMarkdownLoader": ['.md'],
                "CustomJSONLoader": [".json"],
@@ -302,6 +302,8 @@ class KnowledgeFile:
         text_splitter: TextSplitter = None,
     ):
         docs = docs or self.file2docs(refresh=refresh)
+        file_name_without_extension, file_extension = os.path.splitext(self.filepath)
+        print(f"filepath:{self.filepath},文件名拆分后：{file_name_without_extension},{file_extension}")
         if not docs:
             return []
         if self.ext not in [".csv"]:
@@ -314,13 +316,27 @@ class KnowledgeFile:
                     if doc.metadata:
                         doc.metadata["source"] = os.path.basename(self.filepath)
             else:
+                print(f"**********************docs2texts: text_splitter.split_documents(docs)")
+                outputfile = file_name_without_extension + "_source.txt"
+                with open(outputfile, 'w') as file:
+                    for doc in docs:
+                        file.write(doc.page_content)
                 docs = text_splitter.split_documents(docs)
                 
         #print(f"文档切分示例：{docs[0]}")
-        i = 0
-        for doc in docs:
-            print(f"**********切分段{i}：{doc}")
-            i = i+1
+        # print(f"KnowledgeFile: filepath:{self.filepath}")
+        # file_name_without_extension, file_extension = os.path.splitext(self.filepath)
+        # print("filepath:{self.filepath},文件名拆分后：{file_name_without_extension},{file_extension}")
+
+        i = 1
+        outputfile = file_name_without_extension + "_split.txt"
+        # 打开文件以写入模式
+        with open(outputfile, 'w') as file:
+            for doc in docs:
+                print(f"**********切分段{i}：{doc}")
+                file.write(f"分段{i}")
+                file.write(doc.page_content)
+                i = i+1
            
         if zh_title_enhance:
             docs = func_zh_title_enhance(docs)
@@ -407,7 +423,8 @@ if __name__ == "__main__":
     kb_file = KnowledgeFile(filename="test.txt", knowledge_base_name="samples")
     # kb_file.text_splitter_name = "RecursiveCharacterTextSplitter"
     docs = kb_file.file2docs()
-    pprint(docs[-1])
+    #pprint(docs[-1])
 
-    docs = kb_file.file2text()
-    pprint(docs[-1])
+    docs  = kb_file.docs2texts()
+    #docs = kb_file.file2text()
+    #pprint(docs[-1])

test.py

@@ -13,9 +13,9 @@ if __name__ == '__main__':
     # pprint(docs[-1])
 
     faissService = FaissKBService("test")
-    faissService.add_doc(KnowledgeFile("国网安徽信通公司安全准入实施要求_修订.docx", "test"))
+    faissService.add_doc(KnowledgeFile("电力电缆故障测寻车技术规范.docx", "test"))
     # faissService.delete_doc(KnowledgeFile("README.md", "test"))
     # faissService.do_drop_kb()
-    print(faissService.search_docs("准入手续的内容是什么？"))
+    #print(faissService.search_docs("准入手续的内容是什么？"))
 
 

webui.py

@@ -17,7 +17,7 @@ if __name__ == "__main__":
         menu_items={
             'Get Help': 'https://github.com/chatchat-space/Langchain-Chatchat',
             'Report a bug': "https://github.com/chatchat-space/Langchain-Chatchat/issues",
-            'About': f"""欢迎使用 Langchain-Chatchat WebUI {VERSION}！"""
+            'About': f"""欢迎使用 思极大模型 WebUI {VERSION}！"""
         }
     )
 
@@ -36,7 +36,8 @@ if __name__ == "__main__":
         st.image(
             os.path.join(
                 "img",
-                "logo-long-chatchat-trans-v2.png"
+                "siji.jpg"
+                #"logo-long-chatchat-trans-v2.png"
             ),
             use_column_width=True
         )