删除 testfiles/output/163-06A0016-B01003_雷达-图解目录_content_list.json

filename_proceess_and_kgquery.py里面配置信息统一从app.py导入

chunk_text.py

@@ -898,7 +898,7 @@ def find_ship_info_by_hull(json_file_path, data):
     except json.JSONDecodeError:
         print("错误：JSON 文件格式不正确")
         return None
-def merge_short_slices(slices, min_length=30,filename="163-06A0014-B01001_发动机-维修手册.pdf"):
+def merge_short_slices(slices, min_length=30,filename="122-06A0014-B01003_雷达-使用说明书.pdf"):
     """ 
     合并过短的切片：
     - 如果某切片 content 长度 <= min_length，则将其合并到下一个切片的开头
@@ -908,6 +908,8 @@ def merge_short_slices(slices, min_length=30,filename="163-06A0014-B01001_发动
     Args:
         slices: [{"content": str, "positions": [...]}]
         min_length: 短切片的字符长度阈值（含）
+        filename: 用于实体提取的文件名
+    
     Returns:
         合并后的 slices 列表
     """
@@ -934,6 +936,7 @@ def merge_short_slices(slices, min_length=30,filename="163-06A0014-B01001_发动
                 positions = pending_positions + positions
                 pending_contents = []
                 pending_positions = []
+            
             result.append({
                 "content": content,
                 "positions": positions
@@ -953,35 +956,43 @@ def merge_short_slices(slices, min_length=30,filename="163-06A0014-B01001_发动
                 "content": merged_suffix,
                 "positions": pending_positions
             })
+
     try:
         final_result = get_entity(filename)
-        print(final_result)
         xinghao = find_ship_info_by_hull(SHIP_MODEL_NAME, final_result)
-        if xinghao.get('model_name'):
+        if xinghao is None:  # 确保 xinghao 为 None 时不会报错
-            model_name = xinghao['model_name']
+            xinghao = {"model_name": "", "ship_name": ""}
-        else:
+            print("未找到匹配的舰船信息")
+
+        # 安全处理 xinghao 为 None 的情况
         model_name = ""
-        # lower_entity = final_result.get("low_level")
+        if xinghao and 'model_name' in xinghao:
-        otehr_data = format_entity_text(final_result)
+            model_name = xinghao['model_name']
-        logger.info(f"文件名实体提取成功: {otehr_data}")
+            
+        other_data = format_entity_text(final_result)
+        logger.info(f"文件名实体提取成功: {other_data}")
     except Exception as e:
         logger.warning(f"文件名实体提取失败: {e}")
-        otehr_data = ""  # 建议加个兜底,避免后面 NameError
+        other_data = ""
+        model_name = ""  # 确保异常时 model_name 有定义
 
+    # 将提取的信息（如舰艇名、型号）注入到每个切片的开头
     for ins in result:
         content = ins['content']
-        if not otehr_data:
+        # 如果没有提取到有效数据，则跳过注入
-            continue
+            
         newline_idx = content.find('\n')
-        # 拼接成一个括号：otehr_data + 型号为xxx
+        info = other_data
-        info = otehr_data
         if model_name:
             info += f'，型号为{model_name}'
         suffix = f'({info})'
+        
+        # 将信息插入到第一行末尾
         if newline_idx == -1:
             ins['content'] = content + suffix
         else:
             ins['content'] = content[:newline_idx] + suffix + content[newline_idx:]
+
     return result
 def find_and_read_content_list(directory, original_filename, encoding='utf-8'):
     """
@@ -1133,8 +1144,3 @@ if __name__ == "__main__":
 
     for ins in slices_2:
         print(ins)
-
-
-
-
-

config.py

@@ -17,6 +17,7 @@ EMBEDDING_CONFIG = {
     "model": "bge-m3",
     "base_url": "http://192.168.0.46:59700/v1/embeddings",
     "api_key": "gpustack_dee9ca823290886c_5edfc86aeeeceb1e9ee5162941cb2cb5",
+    "base_url_v2" : "http://192.168.0.46:59700/v1",
 }
 
 # ==================== Neo4j 图数据库配置 ====================
@@ -75,3 +76,87 @@ API_OTHER_URLS = [
 ]
 
 SHIP_MODEL_NAME = "E:\ZKYNLP\Hjunproject\project0506\kgrag\ship_xinghao.json"
+TREE_JSON_PATH = r"E:\ZKYNLP\Hjunproject\project0506\kgrag\tree_data.json"
+root@70964b104c5b:/app# pwd
+/app
+root@70964b104c5b:/app# vim config.py 
+root@70964b104c5b:/app# cat config.py 
+# app/model_config.py
+
+import os
+
+# ==================== 大模型配置 ====================
+LLM_CONFIG = {
+    "model": "46-qwen3.5-35B",
+    "base_url": "http://192.168.0.46:59800/v1",
+    "api_key": "gpustack_dee9ca823290886c_5edfc86aeeeceb1e9ee5162941cb2cb5",
+    "temperature": 0.1,
+    "max_tokens": 55096,
+    "timeout": 30
+}
+
+# ==================== 嵌入模型配置 ====================
+EMBEDDING_CONFIG = {
+    "model": "bge-m3",
+    "base_url": "http://192.168.0.46:59700/v1/embeddings",
+    "api_key": "gpustack_dee9ca823290886c_5edfc86aeeeceb1e9ee5162941cb2cb5",
+    "base_url_v2" : "http://192.168.0.46:59700/v1",
+}
+
+# ==================== Neo4j 图数据库配置 ====================
+NEO4J_CONFIG = {
+    # 优先读取环境变量，如果没有则使用默认值
+    # 注意：本地开发通常用 localhost，部署时用服务名 neo4j
+    "uri": "neo4j://192.168.0.46:57687", 
+    "username":  "neo4j",
+    "password": "zdht123@",
+    "database":  "neo4j", # 新增 database 配置
+}
+
+# ==================== 检索与索引配置 ====================
+SEARCH_CONFIG = {
+    "name_property": "名称",
+    "fulltext_property": "fulltext",
+    "embedding_property": "embedding",
+    "search_label": "Searchable",
+    "vector_index_name": "global_searchable_embedding",
+    "fulltext_index_name": "global_searchable_content_search",
+    "fulltext_field_index_name": "global_searchable_fulltext_search",
+    "global_entity_embedding" : "global_entity_embedding",
+    "global_entity_content_search": "global_entity_content_search",
+    "excluded_business_labels": [
+        "Entity", "Chunk", "Document", "_Bloom_Perspective_", "Searchable"
+    ],
+    "vector_dimension": 1024,
+    "jieba_pos_whitelist": ["n", "nr", "ns", "nt", "nz", "vn", "eng"],
+    "embed_batch_size" : 64,
+    "embed_thread_num"  : 4,
+    "node_fetch_page" : 10000,
+}
+
+INDEXING_ULR = "http://192.168.0.46:59085/neo4j_indexing"
+
+URL = {
+    "indexing_url" : "http://192.168.0.46:59085/neo4j_indexing",
+    "prefix_url" : "http://192.168.0.46:59085",
+}
+
+API_URLS = [
+      "http://192.168.0.64:59988/analyze-pdf",
+    # "http://192.168.0.111:9977/analyze-pdf",
+    # "http://192.168.0.111:9978/analyze-pdf",
+    #"http://192.168.0.111:9979/analyze-pdf",
+    #"http://192.168.0.111:9980/analyze-pdf",
+    # "http://192.168.0.111:9975/analyze-pdf",
+]
+API_OTHER_URLS = [
+      "http://192.168.0.46:59988/analyze-otherfile",
+    # "http://192.168.0.111:9977/analyze-otherfile",
+    # "http://192.168.0.111:9978/analyze-otherfile",
+    #"http://192.168.0.111:9979/analyze-otherfile",
+    #"http://192.168.0.111:9980/analyze-otherfile",
+    # "http://192.168.0.111:9975/analyze-otherfile",
+]
+
+SHIP_MODEL_NAME = "/app/ship_xinghao.json"
+TREE_JSON_PATH = "/app/tree_data.json"

filename_proceess_and_kgquery.py

@@ -13,6 +13,7 @@ import json
 from modelsAPI.model_api import OpenaiAPI
 from openai import OpenAI
 from typing import Dict, List, Any
+from config import TREE_JSON_PATH
 
 app = FastAPI(title="PDF Upload Service")
 logging.basicConfig(level=logging.INFO, format="%(asctime)s | %(levelname)s | %(name)s | %(message)s")
@@ -540,8 +541,8 @@ def fetch_graph_sample1(driver):
 # 文件名处理方法
 # =============================
 
-BASE_DIR = os.path.dirname(os.path.abspath(__file__))
+# BASE_DIR = os.path.dirname(os.path.abspath(__file__))
-TREE_JSON_PATH = os.path.join(BASE_DIR, "tree_data.json")
+# TREE_JSON_PATH = os.path.join(BASE_DIR, "tree_data.json")
 
 
 def _load_tree_data() -> Dict:

graph_search/node_retrieval.py

@@ -117,9 +117,10 @@ class NodeRetrieval:
         tasks = []
         for label, entity, query_text, query_vector in zip(labels, entities, query_texts, query_vectors):
             try:
-                vector_index_name = label.lower() + "_vector"
+                # vector_index_name = label.lower() + "_vector"
-                fulltext_index_name = label.lower() + "_fulltext"
+                # fulltext_index_name = label.lower() + "_fulltext"
-                
+                vector_index_name = "global_searchable_embedding"
+                fulltext_index_name = "global_searchable_fulltext_search"
                 retriever = HybridRetriever(
                     self.driver,
                     vector_index_name=vector_index_name,

neo4j_indexing.py

@@ -35,8 +35,6 @@ EMBED_MAX_WORKERS = SEARCH_CONFIG["embed_thread_num"]  # embedding 批次并行
 NODE_FETCH_PAGE = SEARCH_CONFIG["node_fetch_page"]  # 单次拉取待生成 embedding 节点的分页大小
 
 # 全局索引名称（覆盖所有节点标签）
-GLOBAL_VECTOR_INDEX_NAME = SEARCH_CONFIG["global_entity_embedding"]
-GLOBAL_FULLTEXT_INDEX_NAME = SEARCH_CONFIG["global_entity_content_search"]
 DEFAULT_URI = NEO4J_CONFIG["uri"]
 DEFAULT_AUTH = (
     NEO4J_CONFIG["username"],
@@ -57,8 +55,6 @@ OLD_INDEX_NAMES: List[str] = [
     VECTOR_INDEX_NAME,
     FULLTEXT_INDEX_NAME,
     FULLTEXT_FIELD_INDEX_NAME,
-    GLOBAL_VECTOR_INDEX_NAME,
-    GLOBAL_FULLTEXT_INDEX_NAME,
 ]
 
 
@@ -192,390 +188,6 @@ def drop_index_without_constraint(driver):
             driver.execute_query(f"drop index {record['name']} if exists")
 
 
-# --------- 创建全局混合索引（覆盖所有节点标签） ---------
-def create_global_indexes(driver, force_refresh=False):
-    """
-    创建覆盖所有节点标签的全局向量索引和全文索引。
-    与按标签创建的索引互补，支持跨标签统一检索。
-
-    Args:
-        driver: Neo4j driver
-        force_refresh: True=先删除旧索引再重建，False=IF NOT EXISTS 跳过已有索引
-    """
-    if force_refresh:
-        for index_name in (GLOBAL_VECTOR_INDEX_NAME, GLOBAL_FULLTEXT_INDEX_NAME):
-            try:
-                driver.execute_query(f"DROP INDEX {index_name} IF EXISTS")
-                logger.info(f"强制刷新：已删除全局索引 {index_name}")
-            except Exception as e:
-                logger.warning(f"删除全局索引 {index_name} 时出错（可能不存在）: {e}")
-
-    # 全局向量索引（作用于所有节点的 embedding 属性）
-    try:
-        driver.execute_query(f"""
-            CREATE VECTOR INDEX {GLOBAL_VECTOR_INDEX_NAME} IF NOT EXISTS
-            FOR (n) ON n.embedding
-            OPTIONS {{indexConfig: {{
-                `vector.dimensions`: {vector_dim},
-                `vector.similarity_function`: 'cosine'
-            }}}}
-        """)
-        logger.info(f"全局向量索引 '{GLOBAL_VECTOR_INDEX_NAME}' 已就绪")
-    except Exception as e:
-        logger.error(f"创建全局向量索引失败: {e}")
-        raise
-
-    # 全局全文索引（作用于所有节点的 名称 和 内容 属性）
-    try:
-        driver.execute_query(f"""
-            CREATE FULLTEXT INDEX {GLOBAL_FULLTEXT_INDEX_NAME} IF NOT EXISTS
-            FOR (n) ON EACH [n.名称, n.内容]
-        """)
-        logger.info(f"全局全文索引 '{GLOBAL_FULLTEXT_INDEX_NAME}' 已就绪")
-    except Exception as e:
-        logger.error(f"创建全局全文索引失败: {e}")
-        raise
-
-
-# --------- 创建向量索引 ---------
-def vector_indexing(driver, label, force_refresh=False):
-    """
-    创建向量索引（如果节点没有 embedding，使用"名称"属性生成 embedding）
-
-    说明：
-      - 待生成 embedding 的节点通过分页循环全部处理，不再受单次 LIMIT 限制，
-        避免节点数超过一万时被静默丢弃。
-      - embedding 计算采用并行批次以缩短耗时。
-
-    Args:
-        driver: Neo4j driver
-        label: 节点标签
-        force_refresh: 是否强制刷新（如果索引已存在，force_refresh=False 时会跳过，True 时会先删除再创建）
-    """
-
-    # === 分页处理所有缺失 embedding 的节点 ===
-    # 每轮取一页未生成 embedding 的节点并写回，直到没有剩余节点为止。
-    # 这样即使待处理节点远超一万也能全部覆盖。
-    total_generated = 0
-    while True:
-        query_no_embedding = f"""
-            MATCH (n:{label})
-            WHERE n.embedding IS NULL
-              AND n.名称 IS NOT NULL
-            RETURN elementId(n) AS id, n.名称 AS 名称
-            LIMIT {NODE_FETCH_PAGE}
-        """
-        result_no_embedding = driver.execute_query(query_no_embedding)
-        nodes_to_process = [
-            (record["id"], record["名称"]) for record in result_no_embedding.records
-        ]
-
-        if not nodes_to_process:
-            break
-
-        logger.info(f"标签 {label} 本轮发现 {len(nodes_to_process)} 个节点需要生成 embedding")
-
-        ids, names = zip(*nodes_to_process)
-        ids = list(ids)
-        names = list(names)
-
-        # 并行计算 embedding
-        logger.info(f"开始为 {len(names)} 个节点计算 embedding（并行批次）...")
-        try:
-            all_embeddings = _compute_embeddings_parallel(names)
-        except Exception as e:
-            logger.error(f"批量嵌入向量计算失败: {e}")
-            raise
-
-        # 转换为 numpy 数组并做 L2 归一化
-        embeddings = np.array(all_embeddings)
-        norms = np.linalg.norm(embeddings, axis=1, keepdims=True)
-        norms = np.where(norms == 0, 1, norms)  # 避免除零
-        embeddings = embeddings / norms
-
-        # 保存 embedding 到节点
-        logger.info(f"写入 {len(ids)} 个节点的 embedding...")
-        upsert_vectors(
-            driver,
-            ids=ids,
-            embedding_property="embedding",
-            embeddings=embeddings,
-        )
-        total_generated += len(ids)
-        logger.info(f"本轮成功为 {len(ids)} 个节点生成并保存 embedding")
-
-        # 不足一页说明已处理完，提前结束
-        if len(nodes_to_process) < NODE_FETCH_PAGE:
-            break
-
-    if total_generated:
-        logger.info(f"标签 {label} 共生成 {total_generated} 个节点的 embedding")
-
-    # 检查是否有节点存在 embedding（包括刚生成的）
-    check_query = f"MATCH (n:{label}) WHERE n.embedding IS NOT NULL RETURN count(n) AS count"
-    result = driver.execute_query(check_query)
-    node_count = result.records[0]["count"] if result.records else 0
-
-    if node_count == 0:
-        logger.warning(f"标签 {label} 没有节点包含 embedding 属性，跳过向量索引创建")
-        return
-
-    # 创建向量索引
-    index_name = f"{label.lower()}_vector"
-
-    # 如果强制刷新且索引已存在，先删除它
-    if force_refresh:
-        try:
-            driver.execute_query(f"DROP INDEX {index_name} IF EXISTS")
-            logger.info(f"强制刷新模式：已删除旧索引 {index_name}")
-        except Exception as e:
-            logger.warning(f"删除索引 {index_name} 时出错（可能不存在）: {e}")
-
-    try:
-        create_vector_index(
-            driver,
-            name=index_name,  # 索引的唯一名称
-            label=label,  # 要索引的节点标签
-            embedding_property="embedding",  # 包含嵌入向量值的节点属性键
-            dimensions=vector_dim,  # 向量嵌入维度，1024与使用的 bge-m3 嵌入模型一致
-            similarity_fn="cosine",  # 向量相似度函数，可选 "euclidean" 或 "cosine"
-        )
-        logger.info(f"成功为标签 {label} 创建向量索引（共 {node_count} 个节点）")
-    except Exception as e:
-        # 如果是增量更新模式且索引已存在，这是正常的，只记录警告
-        if not force_refresh and "already exists" in str(e).lower():
-            logger.info(f"标签 {label} 的向量索引已存在，跳过创建")
-        else:
-            logger.error(f"为标签 {label} 创建向量索引失败: {e}")
-            raise
-
-
-# --------- 创建全文索引 ---------
-def fulltext_indexing(driver, label, force_refresh=False):
-    """
-    创建全文索引，并添加节点属性（自动获取所有属性）
-
-    全文索引：词 ---》 id
-    后续检索，分词后，根据词反向找到节点
-
-    Args:
-        driver: Neo4j driver
-        label: 节点标签
-        force_refresh: 是否强制刷新所有节点的全文索引（True=清空后重新生成，False=只处理缺失的节点）
-    """
-    # 排除内部属性
-    excluded_props = {"embedding", "fulltext", "elementId", "id"}
-
-    # 根据 force_refresh 决定查询条件
-    if force_refresh:
-        # 强制刷新：处理所有节点（包括已有 fulltext 的节点）
-        query = f"""MATCH (n:{label})
-                WITH n, elementId(n) AS id, properties(n) AS props
-                RETURN id, props"""
-    else:
-        # 增量更新：只处理没有 fulltext 的节点
-        query = f"""MATCH (n:{label})
-                WHERE n.fulltext IS NULL
-                WITH n, elementId(n) AS id, properties(n) AS props
-                RETURN id, props"""
-
-    # 执行查询
-    records = driver.execute_query(query).records
-
-    if not records:
-        if force_refresh:
-            logger.info(f"{label} 没有节点需要处理")
-        else:
-            logger.info(f"{label} 所有节点皆存在全文索引属性")
-        return
-
-    # 构建文本描述（使用所有属性）
-    record_list = []
-    for r in records:
-        node_id = r["id"]
-        props = r["props"]
-        # 过滤掉内部属性
-        filtered_props = {k: v for k, v in props.items() if k not in excluded_props}
-        # 使用所有属性构建文本
-        text = build_text_from_all_properties(filtered_props)
-        if text:
-            record_list.append({"id": node_id, "text": text})
-
-    if not record_list:
-        logger.info(f"{label} 没有可索引的文本内容")
-        return
-
-    # 创建全文索引（在写入数据之前创建索引）
-    try:
-        create_fulltext_index(
-            driver,
-            name=f"{label.lower()}_fulltext",  # 索引的唯一名称
-            label=label,  # 要创建索引的节点标签
-            node_properties=["fulltext"],  # 要创建全文索引的节点属性列表
-        )
-    except Exception as e:
-        logger.warning(f"创建全文索引可能已存在（继续处理）: {e}")
-
-    # 文本分词，作为全文索引属性
-    logger.info(f"计算 {label} ({len(record_list)}) 的全文索引")
-    pattern = re.compile(r"[a-zA-Z0-9\u4e00-\u9fa5]+")  # 匹配英文字母、数字和中文字符
-    fulltext_tuple_list = []
-    for id_, text in [(r["id"], r["text"]) for r in record_list]:
-        # 使用 jieba 分词
-        words = jieba.lcut(text)
-        # 过滤并连接
-        filtered_words = [
-            word.strip()
-            for word in words
-            if word.strip() and pattern.fullmatch(word.strip())
-        ]
-        fulltext_value = " ".join(filtered_words)
-        if fulltext_value:
-            fulltext_tuple_list.append((id_, fulltext_value))
-
-    if not fulltext_tuple_list:
-        logger.warning(f"{label} 分词后没有有效内容")
-        return
-
-    ids, fulltexts = zip(*fulltext_tuple_list)
-    ids = list(ids)
-    fulltexts = list(fulltexts)
-
-    # 按 elementId 添加全文索引属性
-    logger.info(f"写入 {label} ({len(fulltexts)}) 的全文索引")
-    insert_batch_size = 1000
-    for i in range(0, len(fulltext_tuple_list), insert_batch_size):
-        batch_rows = [
-            {"id": id_, "fulltext": ft}
-            for id_, ft in zip(
-                ids[i: i + insert_batch_size],
-                fulltexts[i: i + insert_batch_size],
-            )
-        ]
-
-        # UNWIND：将列表数据展开为多行记录
-        driver.execute_query(
-            "UNWIND $rows AS row "  # 将传入的 rows 列表展开，每项作为一行数据，命名为 row
-            "MATCH (n) "
-            "WHERE elementId(n) = row.id "
-            "SET n.fulltext = row.fulltext ",
-            {"rows": batch_rows},
-        )
-
-
-def create_all_indexes(neo4j_uri=None, neo4j_username=None, neo4j_password=None, force_refresh=False, driver=None):
-    """
-    创建所有索引的主函数（可被外部调用）
-
-    注意：Neo4j 连接配置优先从 .env 文件中读取（NEO4J_URI, NEO4J_USERNAME, NEO4J_PASSWORD）
-    如果 .env 文件中没有配置，则使用参数中的值（如果提供）或默认值
-
-    Args:
-        neo4j_uri: Neo4j URI（可选，优先级低于 .env，提供 driver 时忽略）
-        neo4j_username: Neo4j 用户名（可选，优先级低于 .env，提供 driver 时忽略）
-        neo4j_password: Neo4j 密码（可选，优先级低于 .env，提供 driver 时忽略）
-        force_refresh: 是否强制刷新所有索引（True=清空重建，False=增量更新）
-        driver: 可选的 Neo4j driver 实例，提供则复用外部 driver
-
-    Returns:
-        dict: 包含执行结果的字典
-    """
-    use_external_driver = driver is not None
-
-    if not use_external_driver:
-        # 优先从 .env 文件读取配置，如果 .env 中没有则使用参数或默认值
-        neo4j_uri = os.getenv("NEO4J_URI") or neo4j_uri or "bolt://192.168.0.46:57687"
-        neo4j_username = os.getenv("NEO4J_USERNAME") or neo4j_username or "neo4j"
-        neo4j_password = os.getenv("NEO4J_PASSWORD") or neo4j_password or "zdht123@"
-        logger.info(f"连接到 Neo4j: {neo4j_uri}")
-    else:
-        logger.info("使用外部提供的 Neo4j driver")
-
-    # 定义内部函数来执行索引创建逻辑（避免代码重复）
-    def _execute_indexing(driver):
-        # 1、根据 force_refresh 决定是否清空索引和约束
-        if force_refresh:
-            logger.info("强制刷新模式：清空所有约束...")
-            drop_constraint(driver)
-            logger.info("强制刷新模式：清空所有索引...")
-            drop_index_without_constraint(driver)
-            logger.info("强制刷新模式：清空所有节点的 fulltext 属性...")
-            driver.execute_query("MATCH (n) WHERE n.fulltext IS NOT NULL REMOVE n.fulltext")
-        else:
-            logger.info("增量更新模式：保留现有索引和约束")
-
-        # 2、自动获取所有节点标签并创建向量索引
-        logger.info("=" * 50)
-        logger.info("开始创建向量索引...")
-        logger.info("=" * 50)
-
-        all_labels = get_all_labels(driver)
-        logger.info(f"发现 {len(all_labels)} 个节点标签: {all_labels}")
-
-        vector_index_errors = []
-        for label in all_labels:
-            logger.info(f"处理标签: {label}")
-            try:
-                vector_indexing(driver, label, force_refresh=force_refresh)
-            except Exception as e:
-                error_msg = f"为标签 {label} 创建向量索引时出错: {e}"
-                logger.error(error_msg)
-                vector_index_errors.append(error_msg)
-                continue
-
-        # 3、创建全文索引（自动获取所有属性）
-        logger.info("=" * 50)
-        logger.info("开始创建全文索引...")
-        logger.info("=" * 50)
-
-        fulltext_index_errors = []
-        for label in all_labels:
-            logger.info(f"处理标签: {label}")
-            try:
-                fulltext_indexing(driver, label, force_refresh=force_refresh)
-            except Exception as e:
-                error_msg = f"为标签 {label} 创建全文索引时出错: {e}"
-                logger.error(error_msg)
-                fulltext_index_errors.append(error_msg)
-                continue
-
-        logger.info("=" * 50)
-        logger.info("所有索引创建完成！")
-        logger.info("=" * 50)
-
-        has_errors = len(vector_index_errors) > 0 or len(fulltext_index_errors) > 0
-        mode_text = "强制刷新" if force_refresh else "增量更新"
-        return {
-            "success": not has_errors,
-            "message": f"{mode_text}模式：索引创建完成" if not has_errors
-            else f"{mode_text}模式：索引创建完成，但部分标签出现错误",
-            "mode": "force_refresh" if force_refresh else "incremental",
-            "labels": all_labels,
-            "vector_index_errors": vector_index_errors,
-            "fulltext_index_errors": fulltext_index_errors,
-        }
-
-    try:
-        if use_external_driver:
-            # 使用外部 driver，直接执行，不关闭 driver
-            return _execute_indexing(driver)
-        else:
-            # 使用内部 driver，使用上下文管理器自动关闭
-            with GraphDatabase.driver(neo4j_uri, auth=(neo4j_username, neo4j_password)) as driver:
-                return _execute_indexing(driver)
-    except Exception as e:
-        error_msg = f"创建索引过程中发生错误: {e}"
-        logger.error(error_msg, exc_info=True)
-        return {
-            "success": False,
-            "message": error_msg,
-            "mode": "error",
-            "labels": [],
-            "vector_index_errors": [],
-            "fulltext_index_errors": [],
-        }
-
-
 def print_neo4j_version(driver) -> None:
     """打印当前 Neo4j 实例的版本和版本类型。"""
     with driver.session() as session:
@@ -795,13 +407,12 @@ build_hyrid_indexes = build_hybrid_indexes
 
 
 if __name__ == "__main__":
-    # 可以通过命令行参数或环境变量控制是否强制刷新
-    import sys
 
-    force_refresh = "--force" in sys.argv or os.getenv("FORCE_REFRESH", "false").lower() == "true"
+    result = build_hybrid_indexes(
+        drop_old=True
+    )
 
-    result = create_all_indexes(force_refresh=force_refresh)
     if result["success"]:
-        logger.info(f"索引创建成功完成（模式: {result['mode']}）")
+        logger.info("混合索引创建成功")
     else:
-        logger.error(f"索引创建完成但有错误: {result['message']}")
+        logger.error(result["message"])