feat: 添加系统Embedding配置功能并优化文档解析 (#35)

在知识库模块中新增了获取和设置系统Embedding配置的API接口，支持动态配置Embedding模型的基础URL、模型名称和API Key。同时，优化了文档解析逻辑，使用系统配置的Embedding模型生成文本块的向量，并将图片与文本块关联存储。

api/db/services/dialog_service.py

@@ -162,10 +162,12 @@ def chat(dialog, messages, stream=True, **kwargs):
             prompt_config["system"] = prompt_config["system"].replace(
                 "{%s}" % p["key"], " ")
     
-    if len(questions) > 1 and prompt_config.get("refine_multiturn"):
-        questions = [full_question(dialog.tenant_id, dialog.llm_id, messages)]
-    else:
-        questions = questions[-1:]
+    # 不再使用多轮对话优化
+    # if len(questions) > 1 and prompt_config.get("refine_multiturn"):
+    #     questions = [full_question(dialog.tenant_id, dialog.llm_id, messages)]
+    # else:
+    #     questions = questions[-1:]
+    questions = questions[-1:]
 
     refine_question_ts = timer()
 
@@ -188,40 +190,50 @@ def chat(dialog, messages, stream=True, **kwargs):
         tenant_ids = list(set([kb.tenant_id for kb in kbs]))
 
         knowledges = []
-        if prompt_config.get("reasoning", False):
-            reasoner = DeepResearcher(chat_mdl,
-                                      prompt_config,
-                                      partial(retriever.retrieval, embd_mdl=embd_mdl, tenant_ids=tenant_ids, kb_ids=dialog.kb_ids, page=1, page_size=dialog.top_n, similarity_threshold=0.2, vector_similarity_weight=0.3))
         
-            for think in reasoner.thinking(kbinfos, " ".join(questions)):
-                if isinstance(think, str):
-                    thought = think
-                    knowledges = [t for t in think.split("\n") if t]
-                elif stream:
-                    yield think
-        else:
-            kbinfos = retriever.retrieval(" ".join(questions), embd_mdl, tenant_ids, dialog.kb_ids, 1, dialog.top_n,
-                                          dialog.similarity_threshold,
-                                          dialog.vector_similarity_weight,
-                                          doc_ids=attachments,
-                                          top=dialog.top_k, aggs=False, rerank_mdl=rerank_mdl,
-                                          rank_feature=label_question(" ".join(questions), kbs)
-                                          )
-            if prompt_config.get("tavily_api_key"):
-                tav = Tavily(prompt_config["tavily_api_key"])
-                tav_res = tav.retrieve_chunks(" ".join(questions))
-                kbinfos["chunks"].extend(tav_res["chunks"])
-                kbinfos["doc_aggs"].extend(tav_res["doc_aggs"])
-            if prompt_config.get("use_kg"):
-                ck = settings.kg_retrievaler.retrieval(" ".join(questions),
-                                                       tenant_ids,
-                                                       dialog.kb_ids,
-                                                       embd_mdl,
-                                                       LLMBundle(dialog.tenant_id, LLMType.CHAT))
-                if ck["content_with_weight"]:
-                    kbinfos["chunks"].insert(0, ck)
+        # 不再使用推理
+        # if prompt_config.get("reasoning", False):
+        #     reasoner = DeepResearcher(chat_mdl,
+        #                               prompt_config,
+        #                               partial(retriever.retrieval, embd_mdl=embd_mdl, tenant_ids=tenant_ids, kb_ids=dialog.kb_ids, page=1, page_size=dialog.top_n, similarity_threshold=0.2, vector_similarity_weight=0.3))
 
-            knowledges = kb_prompt(kbinfos, max_tokens)
+        #     for think in reasoner.thinking(kbinfos, " ".join(questions)):
+        #         if isinstance(think, str):
+        #             thought = think
+        #             knowledges = [t for t in think.split("\n") if t]
+        #         elif stream:
+        #             yield think
+        # else:
+        #     kbinfos = retriever.retrieval(" ".join(questions), embd_mdl, tenant_ids, dialog.kb_ids, 1, dialog.top_n,
+        #                                   dialog.similarity_threshold,
+        #                                   dialog.vector_similarity_weight,
+        #                                   doc_ids=attachments,
+        #                                   top=dialog.top_k, aggs=False, rerank_mdl=rerank_mdl,
+        #                                   rank_feature=label_question(" ".join(questions), kbs)
+        #                                   )
+        #     if prompt_config.get("tavily_api_key"):
+        #         tav = Tavily(prompt_config["tavily_api_key"])
+        #         tav_res = tav.retrieve_chunks(" ".join(questions))
+        #         kbinfos["chunks"].extend(tav_res["chunks"])
+        #         kbinfos["doc_aggs"].extend(tav_res["doc_aggs"])
+        #     if prompt_config.get("use_kg"):
+        #         ck = settings.kg_retrievaler.retrieval(" ".join(questions),
+        #                                                tenant_ids,
+        #                                                dialog.kb_ids,
+        #                                                embd_mdl,
+        #                                                LLMBundle(dialog.tenant_id, LLMType.CHAT))
+        #         if ck["content_with_weight"]:
+        #             kbinfos["chunks"].insert(0, ck)
+
+        #     knowledges = kb_prompt(kbinfos, max_tokens)
+        kbinfos = retriever.retrieval(" ".join(questions), embd_mdl, tenant_ids, dialog.kb_ids, 1, dialog.top_n,
+                                        dialog.similarity_threshold,
+                                        dialog.vector_similarity_weight,
+                                        doc_ids=attachments,
+                                        top=dialog.top_k, aggs=False, rerank_mdl=rerank_mdl,
+                                        rank_feature=label_question(" ".join(questions), kbs)
+                                        )
+        knowledges = kb_prompt(kbinfos, max_tokens)
 
     logging.debug(
         "{}->{}".format(" ".join(questions), "\n->".join(knowledges)))
@@ -255,6 +267,7 @@ def chat(dialog, messages, stream=True, **kwargs):
         nonlocal prompt_config, knowledges, kwargs, kbinfos, prompt, retrieval_ts, questions
 
         refs = []
+        image_markdowns = [] # 用于存储图片的 Markdown 字符串
         ans = answer.split("</think>")
         think = ""
         if len(ans) == 2:
@@ -262,6 +275,7 @@ def chat(dialog, messages, stream=True, **kwargs):
             answer = ans[1]
         if knowledges and (prompt_config.get("quote", True) and kwargs.get("quote", True)):
             answer = re.sub(r"##[ij]\$\$", "", answer, flags=re.DOTALL)
+            cited_chunk_indices = set() # 用于存储被引用的 chunk 索引
             if not re.search(r"##[0-9]+\$\$", answer):
                 answer, idx = retriever.insert_citations(answer,
                                                          [ck["content_ltks"]
@@ -271,12 +285,34 @@ def chat(dialog, messages, stream=True, **kwargs):
                                                          embd_mdl,
                                                          tkweight=1 - dialog.vector_similarity_weight,
                                                          vtweight=dialog.vector_similarity_weight)
+                cited_chunk_indices = idx # 获取 insert_citations 返回的索引
+           
             else:
                 idx = set([])
                 for r in re.finditer(r"##([0-9]+)\$\$", answer):
                     i = int(r.group(1))
                     if i < len(kbinfos["chunks"]):
                         idx.add(i)
+                cited_chunk_indices = idx # 获取从 ##...$$ 标记中提取的索引   
+
+            # 根据引用的 chunk 索引提取图像信息并生成 Markdown
+            cited_doc_ids = set()
+            processed_image_urls = set() # 避免重复添加同一张图片
+            print(f"DEBUG: cited_chunk_indices = {cited_chunk_indices}")
+            for i in cited_chunk_indices:
+                i_int = int(i)
+                if i_int < len(kbinfos["chunks"]):
+                    chunk = kbinfos["chunks"][i_int]
+                    cited_doc_ids.add(chunk["doc_id"])
+                    print(f"DEBUG: chunk = {chunk}")
+                    # 检查 chunk 是否有关联的 image_id (URL) 且未被处理过
+                    print(f"DEBUG: chunk_id={chunk.get('chunk_id', i_int)}, image_id={chunk.get('image_id')}")
+                    img_url = chunk.get("image_id")
+                    if img_url and img_url not in processed_image_urls:
+                        # 生成 Markdown 字符串，alt text 可以简单设为 "image" 或 chunk ID
+                        alt_text = f"image_chunk_{chunk.get('chunk_id', i_int)}"
+                        image_markdowns.append(f"\n![{alt_text}]({img_url})")
+                        processed_image_urls.add(img_url) # 标记为已处理
 
             idx = set([kbinfos["chunks"][int(i)]["doc_id"] for i in idx])
             recall_docs = [
@@ -290,6 +326,10 @@ def chat(dialog, messages, stream=True, **kwargs):
                 if c.get("vector"):
                     del c["vector"]
         
+        # 将图片的 Markdown 字符串追加到回答末尾
+        if image_markdowns:
+            answer += "".join(image_markdowns)
+
         if answer.lower().find("invalid key") >= 0 or answer.lower().find("invalid api") >= 0:
             answer += " Please set LLM API-Key in 'User Setting -> Model providers -> API-Key'"
         finish_chat_ts = timer()

management/server/routes/knowledgebases/routes.py

@@ -1,3 +1,4 @@
+import traceback
 from flask import Blueprint, request
 from services.knowledgebases.service import KnowledgebaseService
 from utils import success_response, error_response
@@ -132,13 +133,12 @@ def add_documents_to_knowledgebase(kb_id):
             )
         except Exception as service_error:
             print(f"[ERROR] 服务层错误详情: {str(service_error)}")
-            import traceback
+            
             traceback.print_exc()
             return error_response(str(service_error), code=500)
             
     except Exception as e:
         print(f"[ERROR] 路由层错误详情: {str(e)}")
-        import traceback
         traceback.print_exc()
         return error_response(str(e), code=500)
 
@@ -193,5 +193,50 @@ def get_parse_progress(doc_id):
             return error_response(result['error'], code=404)
         return success_response(data=result)
     except Exception as e:
-        current_app.logger.error(f"获取解析进度失败: {str(e)}")
+        print(f"获取解析进度失败: {str(e)}")
         return error_response("解析进行中，请稍后重试", code=202)
+
+# 获取系统 Embedding 配置路由
+@knowledgebase_bp.route('/system_embedding_config', methods=['GET'])
+def get_system_embedding_config_route():
+    """获取系统级 Embedding 配置的API端点"""
+    try:
+        config_data = KnowledgebaseService.get_system_embedding_config()
+        return success_response(data=config_data)
+    except Exception as e:
+        print(f"获取系统 Embedding 配置失败: {str(e)}")
+        return error_response(message=f"获取配置失败: {str(e)}", code=500) # 返回通用错误信息
+
+# 设置系统 Embedding 配置路由
+@knowledgebase_bp.route('/system_embedding_config', methods=['POST'])
+def set_system_embedding_config_route():
+    """设置系统级 Embedding 配置的API端点"""
+    try:
+        data = request.json
+        if not data:
+            return error_response('请求数据不能为空', code=400)
+
+        llm_name = data.get('llm_name', '').strip()
+        api_base = data.get('api_base', '').strip()
+        api_key = data.get('api_key', '').strip() # 允许空
+
+        if not llm_name or not api_base:
+            return error_response('模型名称和 API 地址不能为空', code=400)
+
+        # 调用服务层进行处理（包括连接测试和数据库操作）
+        success, message = KnowledgebaseService.set_system_embedding_config(
+            llm_name=llm_name,
+            api_base=api_base,
+            api_key=api_key
+        )
+
+        if success:
+            return success_response(message=message)
+        else:
+            # 如果服务层返回失败（例如连接测试失败或数据库错误），将消息返回给前端
+            return error_response(message=message, code=400) # 使用 400 表示操作失败
+
+    except Exception as e:
+        # 捕获路由层或未预料的服务层异常
+        print(f"设置系统 Embedding 配置失败: {str(e)}")
+        return error_response(message=f"设置配置时发生内部错误: {str(e)}", code=500)

management/server/services/knowledgebases/document_parser.py

@@ -5,6 +5,8 @@ import json
 import mysql.connector
 import time 
 import traceback
+import re
+import requests
 from io import BytesIO
 from datetime import datetime
 from elasticsearch import Elasticsearch
@@ -16,6 +18,7 @@ from magic_pdf.config.enums import SupportedPdfParseMethod
 from magic_pdf.data.read_api import read_local_office
 from utils import generate_uuid
 
+
 # 自定义tokenizer和文本处理函数，替代rag.nlp中的功能
 def tokenize_text(text):
     """将文本分词，替代rag_tokenizer功能"""
@@ -173,7 +176,7 @@ def get_text_from_block(block):
                         block_text += content
     return ' '.join(block_text.split())
 
-def perform_parse(doc_id, doc_info, file_info):
+def perform_parse(doc_id, doc_info, file_info, embedding_config):
     """
     执行文档解析的核心逻辑
 
@@ -189,6 +192,26 @@ def perform_parse(doc_id, doc_info, file_info):
     temp_image_dir = None
     start_time = time.time()
     middle_json_content = None # 初始化 middle_json_content
+    image_info_list = []  # 图片信息列表
+    
+    # 默认值处理
+    embedding_model_name = embedding_config.get("llm_name") if embedding_config and embedding_config.get("llm_name") else "bge-m3" # 默认模型
+    # 对模型名称进行处理
+    if embedding_model_name and '___' in embedding_model_name:
+        embedding_model_name = embedding_model_name.split('___')[0]
+    embedding_api_base = embedding_config.get("api_base") if embedding_config and embedding_config.get("api_base") else "http://localhost:8000" # 默认基础 URL
+    embedding_api_key = embedding_config.get("api_key") if embedding_config else None # 可能为 None 或空字符串
+    
+    # 构建完整的 Embedding API URL
+    if embedding_api_base:
+        if not embedding_api_base.startswith(('http://', 'https://')):
+            embedding_api_base = 'http://' + embedding_api_base
+        # 标准端点是 /embeddings
+        embedding_url = embedding_api_base.rstrip('/') + "/embeddings"
+    else:
+        embedding_url = None # 如果没有配置 Base URL，则无法请求
+
+    print(f"[Parser-INFO] 使用 Embedding 配置: URL='{embedding_url}', Model='{embedding_model_name}', Key={embedding_api_key}")
     
     try:
         kb_id = doc_info['kb_id']
@@ -330,8 +353,18 @@ def perform_parse(doc_id, doc_info, file_info):
             es_client.indices.create(
                 index=index_name,
                 body={
-                    "settings": {"number_of_replicas": 0}, # 单节点设为0
-                    "mappings": { "properties": { "doc_id": {"type": "keyword"}, "kb_id": {"type": "keyword"}, "content_with_weight": {"type": "text"} } } # 简化字段
+                    "settings": {"number_of_replicas": 0},
+                    "mappings": {
+                        "properties": {
+                            "doc_id": {"type": "keyword"},
+                            "kb_id": {"type": "keyword"},
+                            "content_with_weight": {"type": "text"},
+                            "q_1024_vec": {
+                                "type": "dense_vector",
+                                "dims": 1024
+                            }
+                        }
+                    }
                 }
             )
             print(f"[Parser-INFO] 创建Elasticsearch索引: {index_name}")
@@ -348,6 +381,42 @@ def perform_parse(doc_id, doc_info, file_info):
                 if not content or not content.strip():
                     continue
                 
+                # 过滤 markdown 特殊符号
+                content = re.sub(r"[!#\\$/]", "", content)
+                q_1024_vec = [] # 初始化为空列表
+                
+                # 获取embedding向量
+                try:
+                    # embedding_resp = requests.post(
+                    #     "http://localhost:8000/v1/embeddings",
+                    #     json={
+                    #         "model": "bge-m3",  # 你的embedding模型名
+                    #         "input": content
+                    #     },
+                    #     timeout=10
+                    # )
+                    headers = {"Content-Type": "application/json"}
+                    if embedding_api_key:
+                        headers["Authorization"] = f"Bearer {embedding_api_key}"
+
+                    embedding_resp = requests.post(
+                        embedding_url, # 使用动态构建的 URL
+                        headers=headers, # 添加 headers (包含可能的 API Key)
+                        json={
+                            "model": embedding_model_name,  # 使用动态获取或默认的模型名
+                            "input": content
+                        },
+                        timeout=15 # 稍微增加超时时间
+                    )
+                    
+                    embedding_resp.raise_for_status()
+                    embedding_data = embedding_resp.json()
+                    q_1024_vec = embedding_data["data"][0]["embedding"]
+                    print(f"[Parser-INFO] 获取embedding成功，长度: {len(q_1024_vec)}")
+                except Exception as e:
+                    print(f"[Parser-ERROR] 获取embedding失败: {e}")
+                    q_1024_vec = []
+                    
                 chunk_id = generate_uuid()
                 page_idx = 0  # 默认页面索引
                 bbox = [0, 0, 0, 0] # 默认 bbox
@@ -363,7 +432,6 @@ def perform_parse(doc_id, doc_info, file_info):
                     if processed_text_chunks == len(block_info_list) + 1: # 只在第一次耗尽时警告一次
                          print(f"[Parser-WARNING] middle_data 提供的块信息少于 content_list 中的文本块数量。后续文本块将使用默认 page/bbox。")
                         
-
                 try:
                     # 上传文本块到 MinIO
                     minio_client.put_object(
@@ -382,7 +450,6 @@ def perform_parse(doc_id, doc_info, file_info):
                     x1, y1, x2, y2 = bbox
                     bbox_reordered = [x1, x2, y1, y2]
 
-
                     es_doc = {
                         "doc_id": doc_id,
                         "kb_id": kb_id,
@@ -390,19 +457,19 @@ def perform_parse(doc_id, doc_info, file_info):
                         "title_tks": doc_info['name'],
                         "title_sm_tks": doc_info['name'],
                         "content_with_weight": content,
-                        "content_ltks": content_tokens,
-                        "content_sm_ltks": content_tokens,
+                        "content_ltks": " ".join(content_tokens), # 字符串类型
+                        "content_sm_ltks": " ".join(content_tokens),  # 字符串类型
                         "page_num_int": [page_idx + 1],
                         "position_int": [[page_idx + 1] + bbox_reordered], # 格式: [[page, x1, x2, y1, y2]]
                         "top_int": [1],
                         "create_time": current_time_es,
                         "create_timestamp_flt": current_timestamp_es,
                         "img_id": "",
-                        "q_1024_vec": [] # 向量字段留空
+                        "q_1024_vec": q_1024_vec
                     }
 
                     # 存储到Elasticsearch
-                    es_client.index(index=index_name, document=es_doc) # 使用 document 参数
+                    es_client.index(index=index_name, id=chunk_id, document=es_doc) # 使用 document 参数
 
                     chunk_count += 1
                     chunk_ids_list.append(chunk_id)
@@ -428,27 +495,95 @@ def perform_parse(doc_id, doc_info, file_info):
                 content_type = f"image/{img_ext[1:].lower()}"
                 if content_type == "image/jpg": content_type = "image/jpeg"
 
-                # try:
-                #     # 上传图片到MinIO (桶为kb_id)
-                #     minio_client.fput_object(
-                #         bucket_name=output_bucket,
-                #         object_name=img_key,
-                #         file_path=img_path_abs,
-                #         content_type=content_type
-                #     )
-                #     print(f"成功上传图片: {img_key}")
-                #     # 注意：设置公共访问权限可能需要额外配置MinIO服务器和存储桶策略
+                try:
+                    # 上传图片到MinIO (桶为kb_id)
+                    minio_client.fput_object(
+                        bucket_name=output_bucket,
+                        object_name=img_key,
+                        file_path=img_path_abs,
+                        content_type=content_type
+                    )
 
-                # except Exception as e:
-                #     print(f"[Parser-ERROR] 上传图片 {img_path_abs} 失败: {e}")
+                    # 设置图片的公共访问权限
+                    policy = {
+                        "Version": "2012-10-17",
+                        "Statement": [
+                            {
+                                "Effect": "Allow",
+                                "Principal": {"AWS": "*"},
+                                "Action": ["s3:GetObject"],
+                                "Resource": [f"arn:aws:s3:::{kb_id}/{img_key}"]
+                            }
+                        ]
+                    }
+                    minio_client.set_bucket_policy(kb_id, json.dumps(policy))
+
+                    print(f"成功上传图片: {img_key}")
+                    minio_endpoint = MINIO_CONFIG["endpoint"]
+                    use_ssl = MINIO_CONFIG.get("secure", False)
+                    protocol = "https" if use_ssl else "http"
+                    img_url = f"{protocol}://{minio_endpoint}/{output_bucket}/{img_key}"
+
+                    # 记录图片信息，包括URL和位置信息
+                    image_info = {
+                        "url": img_url,
+                        "position": processed_text_chunks  # 使用当前处理的文本块数作为位置参考
+                    }
+                    image_info_list.append(image_info)
+
+                    print(f"图片访问链接: {img_url}")
+
+                except Exception as e:
+                    print(f"[Parser-ERROR] 上传图片 {img_path_abs} 失败: {e}")
         
         # 打印匹配总结信息
         print(f"[Parser-INFO] 共处理 {processed_text_chunks} 个文本块。")
         if middle_block_idx < len(block_info_list):
              print(f"[Parser-WARNING] middle_data 中还有 {len(block_info_list) - middle_block_idx} 个提取的块信息未被使用。")
         
+        # 4. 更新文本块的图像信息 
+        if image_info_list and chunk_ids_list:
+            conn = None
+            cursor = None
+            try:
+                conn = _get_db_connection()
+                cursor = conn.cursor()
                 
-        # 4. 更新最终状态
+                # 为每个文本块找到最近的图片
+                for i, chunk_id in enumerate(chunk_ids_list):
+                    # 找到与当前文本块最近的图片
+                    nearest_image = None
+                    
+                    for img_info in image_info_list:
+                        # 计算文本块与图片的"距离"
+                        distance = abs(i - img_info["position"])  # 使用位置差作为距离度量
+                        # 如果文本块与图片的距离间隔小于10个块,则认为块与图片是相关的
+                        if distance < 10:
+                            nearest_image = img_info
+                    
+                    # 如果找到了最近的图片，则更新文本块的img_id
+                    if nearest_image:
+                        # 更新ES中的文档
+                        direct_update = {
+                            "doc": {
+                                "img_id": nearest_image["url"]
+                            }
+                        }
+                        es_client.update(index=index_name, id=chunk_id, body=direct_update, refresh=True)
+                        
+                        index_name = f"ragflow_{tenant_id}"
+                        
+                        print(f"[Parser-INFO] 更新文本块 {chunk_id} 的图片关联: {nearest_image['url']}")
+                
+            except Exception as e:
+                print(f"[Parser-ERROR] 更新文本块图片关联失败: {e}")
+            finally:
+                if cursor:
+                    cursor.close()
+                if conn:
+                    conn.close()
+
+        # 5. 更新最终状态
         process_duration = time.time() - start_time
         _update_document_progress(doc_id, progress=1.0, message="解析完成", status='1', run='3', chunk_count=chunk_count, process_duration=process_duration)
         _update_kb_chunk_count(kb_id, chunk_count) # 更新知识库总块数

management/server/services/knowledgebases/service.py

@@ -1,12 +1,15 @@
 import mysql.connector
 import json
 import threading 
+import requests
+import traceback
 from datetime import datetime
 from utils import generate_uuid
 from database import DB_CONFIG 
 # 解析相关模块
 from .document_parser import perform_parse, _update_document_progress
 
+
 class KnowledgebaseService:
     
     @classmethod
@@ -704,7 +707,8 @@ class KnowledgebaseService:
             _update_document_progress(doc_id, status='2', run='1', progress=0.0, message='开始解析')
 
             # 3. 调用后台解析函数
-            parse_result = perform_parse(doc_id, doc_info, file_info)
+            embedding_config = cls.get_system_embedding_config()
+            parse_result = perform_parse(doc_id, doc_info, file_info, embedding_config)
 
             # 4. 返回解析结果
             return parse_result
@@ -792,3 +796,199 @@ class KnowledgebaseService:
                 cursor.close()
             if conn:
                 conn.close()
+
+    # --- 获取最早用户 ID ---
+    @classmethod
+    def _get_earliest_user_tenant_id(cls):
+        """获取创建时间最早的用户的 ID (作为 tenant_id)"""
+        conn = None
+        cursor = None
+        try:
+            conn = cls._get_db_connection()
+            cursor = conn.cursor()
+            query = "SELECT id FROM user ORDER BY create_time ASC LIMIT 1"
+            cursor.execute(query)
+            result = cursor.fetchone()
+            if result:
+                return result[0] # 返回用户 ID
+            else:
+                print("警告: 数据库中没有用户！")
+                return None
+        except Exception as e:
+            print(f"查询最早用户时出错: {e}")
+            traceback.print_exc()
+            return None
+        finally:
+            if cursor:
+                cursor.close()
+            if conn and conn.is_connected():
+                conn.close()
+
+    # ---  测试 Embedding 连接 ---
+    @classmethod
+    def _test_embedding_connection(cls, base_url, model_name, api_key):
+        """
+        测试与自定义 Embedding 模型的连接 (使用 requests)。
+        """
+        print(f"开始测试连接: base_url={base_url}, model_name={model_name}")
+        try:
+            headers = {"Content-Type": "application/json"}
+            if api_key:
+                headers["Authorization"] = f"Bearer {api_key}"
+
+            payload = {"input": ["Test connection"], "model": model_name}
+
+            if not base_url.startswith(('http://', 'https://')):
+                 base_url = 'http://' + base_url
+            if not base_url.endswith('/'):
+                base_url += '/'
+            
+            endpoint = "embeddings"
+            current_test_url = base_url + endpoint
+            print(f"尝试请求 URL: {current_test_url}")
+            try:
+                response = requests.post(current_test_url, headers=headers, json=payload, timeout=15)
+                print(f"请求 {current_test_url} 返回状态码: {response.status_code}")
+
+                if response.status_code == 200:
+                    res_json = response.json()
+                    if ("data" in res_json and isinstance(res_json["data"], list) and len(res_json["data"]) > 0 and "embedding" in res_json["data"][0] and len(res_json["data"][0]["embedding"]) > 0) or \
+                        (isinstance(res_json, list) and len(res_json) > 0 and isinstance(res_json[0], list) and len(res_json[0]) > 0):
+                        print(f"连接测试成功: {current_test_url}")
+                        return True, "连接成功"
+                    else:
+                        print(f"连接成功但响应格式不正确于 {current_test_url}")
+                        
+            except Exception as json_e:
+                print(f"解析 JSON 响应失败于 {current_test_url}: {json_e}")
+                     
+            return False, "连接失败: 响应错误"
+
+        except Exception as e:
+            print(f"连接测试发生未知错误: {str(e)}")
+            traceback.print_exc()
+            return False, f"测试时发生未知错误: {str(e)}"
+
+    # --- 获取系统 Embedding 配置 ---
+    @classmethod
+    def get_system_embedding_config(cls):
+        """获取系统级（最早用户）的 Embedding 配置"""
+        tenant_id = cls._get_earliest_user_tenant_id()
+        if not tenant_id:
+            raise Exception("无法找到系统基础用户") # 在服务层抛出异常
+
+        conn = None
+        cursor = None
+        try:
+            conn = cls._get_db_connection()
+            cursor = conn.cursor(dictionary=True) # 使用字典游标方便访问列名
+            query = """
+                SELECT llm_name, api_key, api_base
+                FROM tenant_llm
+                WHERE tenant_id = %s
+                LIMIT 1
+            """
+            cursor.execute(query, (tenant_id,))
+            config = cursor.fetchone()
+       
+            if config:
+                llm_name = config.get("llm_name", "")
+                api_key = config.get("api_key", "")
+                api_base = config.get("api_base", "")
+                # 对模型名称进行处理
+                if llm_name and '___' in llm_name:
+                    llm_name = llm_name.split('___')[0]
+                # 如果有配置，返回
+                return {
+                    "llm_name": llm_name,
+                    "api_key": api_key,
+                    "api_base": api_base
+                }
+            else:
+                # 如果没有配置，返回空
+                return {
+                     "llm_name": "",
+                     "api_key": "",
+                     "api_base": ""
+                }
+        except Exception as e:
+            print(f"获取系统 Embedding 配置时出错: {e}")
+            traceback.print_exc()
+            raise Exception(f"获取配置时数据库出错: {e}") # 重新抛出异常
+        finally:
+            if cursor:
+                cursor.close()
+            if conn and conn.is_connected():
+                conn.close()
+
+    # --- 设置系统 Embedding 配置 ---
+    @classmethod
+    def set_system_embedding_config(cls, llm_name, api_base, api_key):
+        """设置系统级（最早用户）的 Embedding 配置"""
+        tenant_id = cls._get_earliest_user_tenant_id()
+        if not tenant_id:
+            raise Exception("无法找到系统基础用户")
+
+        # 执行连接测试
+        is_connected, message = cls._test_embedding_connection(
+            base_url=api_base,
+            model_name=llm_name,
+            api_key=api_key
+        )
+
+        if not is_connected:
+            # 返回具体的测试失败原因给调用者（路由层）处理
+            return False, f"连接测试失败: {message}"
+
+        return True, f"连接成功: {message}"
+        # 测试通过，保存或更新配置到数据库(先不保存，以防冲突)
+        # conn = None
+        # cursor = None
+        # try:
+        #     conn = cls._get_db_connection()
+        #     cursor = conn.cursor()
+
+        #     # 检查 TenantLLM 记录是否存在
+        #     check_query = """
+        #         SELECT id FROM tenant_llm
+        #         WHERE tenant_id = %s AND llm_name = %s
+        #     """
+        #     cursor.execute(check_query, (tenant_id, llm_name))
+        #     existing_config = cursor.fetchone()
+
+        #     now = datetime.now()
+        #     if existing_config:
+        #         # 更新记录
+        #         update_sql = """
+        #             UPDATE tenant_llm
+        #             SET api_key = %s, api_base = %s, max_tokens = %s, update_time = %s, update_date = %s
+        #             WHERE id = %s
+        #         """
+        #         update_params = (api_key, api_base, max_tokens, now, now.date(), existing_config[0])
+        #         cursor.execute(update_sql, update_params)
+        #         print(f"已更新 TenantLLM 记录 (ID: {existing_config[0]})")
+        #     else:
+        #         # 插入新记录
+        #         insert_sql = """
+        #             INSERT INTO tenant_llm (tenant_id, llm_factory, model_type, llm_name, api_key, api_base, max_tokens, create_time, create_date, update_time, update_date, used_tokens)
+        #             VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s)
+        #         """
+        #         insert_params = (tenant_id, "VLLM", "embedding", llm_name, api_key, api_base, max_tokens, now, now.date(), now, now.date(), 0) # used_tokens 默认为 0
+        #         cursor.execute(insert_sql, insert_params)
+        #         print(f"已创建新的 TenantLLM 记录")
+
+        #     conn.commit() # 提交事务
+        #     return True, "配置已成功保存"
+
+        # except Exception as e:
+        #     if conn:
+        #         conn.rollback() # 出错时回滚
+        #     print(f"保存系统 Embedding 配置时数据库出错: {e}")
+        #     traceback.print_exc()
+        #     # 返回 False 和错误信息给路由层
+        #     return False, f"保存配置时数据库出错: {e}"
+        # finally:
+            # if cursor:
+            #     cursor.close()
+            # if conn and conn.is_connected():
+            #     conn.close()

management/web/src/common/apis/kbs/knowledgebase.ts

@@ -77,3 +77,24 @@ export function addDocumentToKnowledgeBaseApi(data: {
     data: { file_ids: data.file_ids }
   })
 }
+
+// 获取系统 Embedding 配置
+export function getSystemEmbeddingConfigApi() {
+  return request({
+    url: "/api/v1/knowledgebases/system_embedding_config", // 确认 API 路径前缀是否正确
+    method: "get"
+  })
+}
+
+// 设置系统 Embedding 配置
+export function setSystemEmbeddingConfigApi(data: {
+  llm_name: string
+  api_base: string
+  api_key?: string
+}) {
+  return request({
+    url: "/api/v1/knowledgebases/system_embedding_config", // 确认 API 路径前缀是否正确
+    method: "post",
+    data
+  })
+}

management/web/src/pages/knowledgebase/index.vue

@@ -11,13 +11,15 @@ import {
   batchDeleteKnowledgeBaseApi,
   createKnowledgeBaseApi,
   deleteKnowledgeBaseApi,
-  getKnowledgeBaseListApi
+  getKnowledgeBaseListApi,
+  getSystemEmbeddingConfigApi,
+  setSystemEmbeddingConfigApi
 } from "@@/apis/kbs/knowledgebase"
 import { usePagination } from "@@/composables/usePagination"
-import { CaretRight, Delete, Plus, Refresh, Search, View } from "@element-plus/icons-vue"
+import { CaretRight, Delete, Plus, Refresh, Search, Setting, View } from "@element-plus/icons-vue"
 import axios from "axios"
 import { ElMessage, ElMessageBox } from "element-plus"
-import { onActivated, onBeforeUnmount, onMounted, reactive, ref, watch } from "vue"
+import { nextTick, onActivated, onBeforeUnmount, onDeactivated, onMounted, reactive, ref, watch } from "vue"
 import "element-plus/dist/index.css"
 import "element-plus/theme-chalk/el-message-box.css"
 import "element-plus/theme-chalk/el-message.css"
@@ -615,6 +617,110 @@ onMounted(() => {
 onActivated(() => {
   getTableData()
 })
+
+// 系统 Embedding 配置逻辑
+const configModalVisible = ref(false)
+const configFormRef = ref<FormInstance>() // 表单引用
+const configFormLoading = ref(false) // 表单加载状态
+const configSubmitLoading = ref(false) // 提交按钮加载状态
+
+const configForm = reactive({
+  llm_name: "",
+  api_base: "",
+  api_key: ""
+})
+
+// 简单的 URL 校验规则
+function validateUrl(rule: any, value: any, callback: any) {
+  if (!value) {
+    return callback(new Error("请输入模型 API 地址"))
+  }
+  // 允许 http, https 开头，允许 IP 地址和域名，允许端口
+  // 修正：允许路径，但不允许查询参数或片段
+  const urlPattern = /^(https?:\/\/)?([a-zA-Z0-9.-]+|\[[a-fA-F0-9:]+\])(:\d+)?(\/[^?#]*)?$/
+  if (!urlPattern.test(value)) {
+    callback(new Error("请输入有效的 Base URL (例如 http://host:port 或 https://domain/path)"))
+  } else {
+    callback()
+  }
+}
+
+const configFormRules = reactive({
+  llm_name: [{ required: true, message: "请输入模型名称", trigger: "blur" }],
+  api_base: [{ required: true, validator: validateUrl, trigger: "blur" }]
+  // api_key 不是必填项
+})
+
+// 显示配置模态框
+async function showConfigModal() {
+  configModalVisible.value = true
+  configFormLoading.value = true
+  // 重置表单可能需要在 nextTick 中执行，确保 DOM 更新完毕
+  await nextTick()
+  configFormRef.value?.resetFields() // 清空上次的输入和校验状态
+
+  try {
+    // 确认 API 函数名称是否正确，并添加类型断言
+    const res = await getSystemEmbeddingConfigApi() as ApiResponse<{ llm_name?: string, api_base?: string, api_key?: string }>
+    if (res.code === 0 && res.data) {
+      configForm.llm_name = res.data.llm_name || ""
+      configForm.api_base = res.data.api_base || ""
+      // 注意：API Key 通常不应在 GET 请求中返回，如果后端不返回，这里会是空字符串
+      configForm.api_key = res.data.api_key || ""
+    } else if (res.code !== 0) {
+      ElMessage.error(res.message || "获取配置失败")
+    } else {
+      // code === 0 但 data 为空，说明没有配置
+      console.log("当前未配置编码模型。")
+    }
+  } catch (error: any) {
+    ElMessage.error(error.message || "获取配置请求失败")
+    console.error("获取配置失败:", error)
+  } finally {
+    configFormLoading.value = false
+  }
+}
+
+// 处理模态框关闭
+function handleModalClose() {
+  // 可以在这里再次重置表单，以防用户未保存直接关闭
+  configFormRef.value?.resetFields()
+}
+
+// 处理配置提交
+async function handleConfigSubmit() {
+  if (!configFormRef.value) return
+  // 使用 .then() .catch() 处理 validate 的 Promise
+  configFormRef.value.validate().then(async () => {
+    // 验证通过
+    configSubmitLoading.value = true
+    try {
+      const payload = {
+        llm_name: configForm.llm_name.trim(),
+        api_base: configForm.api_base.trim(),
+        api_key: configForm.api_key
+      }
+      // 确认 API 函数名称是否正确，并添加类型断言
+      const res = await setSystemEmbeddingConfigApi(payload) as ApiResponse<any> // 使用类型断言并指定泛型参数为any
+      if (res.code === 0) {
+        ElMessage.success("连接验证成功！")
+        configModalVisible.value = false
+      } else {
+        // 后端应在 res.message 中返回错误信息，包括连接测试失败的原因
+        ElMessage.error(res.message || "连接验证失败")
+      }
+    } catch (error: any) {
+      ElMessage.error(error.message || "连接验证请求失败")
+      console.error("连接验证失败:", error)
+    } finally {
+      configSubmitLoading.value = false
+    }
+  }).catch((errorFields) => {
+    // 验证失败
+    console.log("表单验证失败!", errorFields)
+    // 这里不需要返回 false，validate 的 Promise reject 就表示失败了
+  })
+}
 </script>
 
 <template>
@@ -654,6 +760,12 @@ onActivated(() => {
               批量删除
             </el-button>
           </div>
+
+          <div>
+            <el-button type="primary" :icon="Setting" @click="showConfigModal">
+              编码模型配置
+            </el-button>
+          </div>
         </div>
 
         <div class="table-wrapper">
@@ -922,6 +1034,56 @@ onActivated(() => {
           </span>
         </template>
       </el-dialog>
+
+      <!-- 系统 Embedding 配置模态框 -->
+      <el-dialog
+        v-model="configModalVisible"
+        title="编码模型配置"
+        width="500px"
+        :close-on-click-modal="false"
+        @close="handleModalClose"
+        append-to-body
+      >
+        <el-form
+          ref="configFormRef"
+          :model="configForm"
+          :rules="configFormRules"
+          label-width="120px"
+          v-loading="configFormLoading"
+        >
+          <el-form-item label="模型名称" prop="llm_name">
+            <el-input v-model="configForm.llm_name" placeholder="请先在前台进行配置" disabled />
+            <div class="form-tip">
+              与模型服务中部署的名称一致
+            </div>
+          </el-form-item>
+          <el-form-item label="模型 API 地址" prop="api_base">
+            <el-input v-model="configForm.api_base" placeholder="请先在前台进行配置" disabled />
+            <div class="form-tip">
+              模型的 Base URL
+            </div>
+          </el-form-item>
+          <el-form-item label="API Key (可选)" prop="api_key">
+            <el-input v-model="configForm.api_key" type="password" show-password placeholder="请先在前台进行配置" disabled />
+            <div class="form-tip">
+              如果模型服务需要认证，请提供
+            </div>
+          </el-form-item>
+          <el-form-item>
+            <div style="color: #909399; font-size: 12px; line-height: 1.5;">
+              此配置将作为知识库解析时默认的 Embedding 模型。
+            </div>
+          </el-form-item>
+        </el-form>
+        <template #footer>
+          <span class="dialog-footer">
+            <el-button @click="configModalVisible = false">取消</el-button>
+            <el-button type="primary" @click="handleConfigSubmit" :loading="configSubmitLoading">
+              测试连接
+            </el-button>
+          </span>
+        </template>
+      </el-dialog>
     </div>
     <DocumentParseProgress
       :document-id="currentDocId"
@@ -952,7 +1114,8 @@ onActivated(() => {
 
 .toolbar-wrapper {
   display: flex;
-  justify-content: space-between;
+  justify-content: space-between; // 确保左右两边对齐
+  align-items: center; // 垂直居中对齐
   margin-bottom: 20px;
 }
 
@@ -1011,4 +1174,11 @@ onActivated(() => {
     text-align: center;
   }
 }
+
+.form-tip {
+  color: #909399;
+  font-size: 12px;
+  line-height: 1.5;
+  margin-top: 4px;
+}
 </style>

vllm/model_test.py

@@ -12,10 +12,12 @@ def test_embedding(model, text):
     )
     
     # 打印嵌入响应内容
-    print(f"Embedding response: {response}")
+    # print(f"Embedding response: {response}")
+    
+    result = response.data[0].embedding
 
     if response and response.data:
-        print(f"Embedding: {response.data[0].embedding}")
+        print(len(result))
     else:
         print("Failed to get embedding.")