From f41ab11dfd6a06f823a1881c3f5783cf7d3739e0 Mon Sep 17 00:00:00 2001
From: HuangHai <10402852@qq.com>
Date: Mon, 23 Jun 2025 18:57:40 +0800
Subject: [PATCH] 'commit'

---
 dsRag/Config/Config.py   |  8 ++++++
 dsRag/Config/__init__.py |  0
 dsRag/T3_LinkEs.py       | 22 ++++----------
 dsRag/T4_VectorSave.py   | 62 ++++++++++++++++++++++++++++++++++++++++
 4 files changed, 76 insertions(+), 16 deletions(-)
 create mode 100644 dsRag/Config/Config.py
 create mode 100644 dsRag/Config/__init__.py
 create mode 100644 dsRag/T4_VectorSave.py

diff --git a/dsRag/Config/Config.py b/dsRag/Config/Config.py
new file mode 100644
index 00000000..a93be2c4
--- /dev/null
+++ b/dsRag/Config/Config.py
@@ -0,0 +1,8 @@
+# Elasticsearch配置
+ES_CONFIG = {
+    "hosts": "https://10.10.14.206:9200",
+    "basic_auth": ("elastic", "your_password"),
+    "verify_certs": False,
+    "ssl_show_warn": False,
+    "default_index": "knowledge_base"
+}
\ No newline at end of file
diff --git a/dsRag/Config/__init__.py b/dsRag/Config/__init__.py
new file mode 100644
index 00000000..e69de29b
diff --git a/dsRag/T3_LinkEs.py b/dsRag/T3_LinkEs.py
index 0d8b509c..9bcef1cd 100644
--- a/dsRag/T3_LinkEs.py
+++ b/dsRag/T3_LinkEs.py
@@ -1,23 +1,13 @@
+from Config.Config import ES_CONFIG
 from elasticsearch import Elasticsearch
-import ssl
 import warnings
-from urllib3.exceptions import InsecureRequestWarning
-
-# Suppress only the single InsecureRequestWarning from urllib3
-warnings.filterwarnings('ignore', category=InsecureRequestWarning)
-warnings.filterwarnings('once', category=DeprecationWarning)
-warnings.filterwarnings('once', category=UserWarning)
-
-context = ssl.create_default_context()
-context.check_hostname = False
-context.verify_mode = ssl.CERT_NONE
 
+# 初始化ES连接
 es = Elasticsearch(
-    hosts="https://10.10.14.206:9200",
-    basic_auth=("elastic", "jv9h8uwRrRxmDi1dq6u8"),
-    ssl_context=context,
-    verify_certs=False,
-    ssl_show_warn=False  # This will suppress the Elasticsearch warning
+    hosts=ES_CONFIG["hosts"],
+    basic_auth=ES_CONFIG["basic_auth"],
+    verify_certs=ES_CONFIG["verify_certs"],
+    ssl_show_warn=ES_CONFIG["ssl_show_warn"]
 )
 
 try:
diff --git a/dsRag/T4_VectorSave.py b/dsRag/T4_VectorSave.py
new file mode 100644
index 00000000..e66fb5e9
--- /dev/null
+++ b/dsRag/T4_VectorSave.py
@@ -0,0 +1,62 @@
+from Config.Config import ES_CONFIG
+from elasticsearch import Elasticsearch
+from T2_Txt2Vec import text_to_embedding
+import datetime
+import warnings
+
+# 初始化ES连接
+es = Elasticsearch(
+    hosts=ES_CONFIG["hosts"],
+    basic_auth=ES_CONFIG["basic_auth"],
+    verify_certs=ES_CONFIG["verify_certs"],
+    ssl_show_warn=ES_CONFIG["ssl_show_warn"]
+)
+
+# 修改create_vector_index和save_to_es函数中使用ES_CONFIG["default_index"]
+def create_vector_index(index_name="knowledge_base"):
+    """创建带有向量字段的索引"""
+    mapping = {
+        "mappings": {
+            "properties": {
+                "text": {"type": "text", "analyzer": "ik_max_word"},
+                "vector": {
+                    "type": "dense_vector",
+                    "dims": 768,  # 需与text2vec模型维度一致
+                    "index": True,
+                    "similarity": "cosine"
+                },
+                "timestamp": {"type": "date"}
+            }
+        }
+    }
+    
+    try:
+        if es.indices.exists(index=index_name):
+            es.indices.delete(index=index_name)
+        es.indices.create(index=index_name, body=mapping)
+        print(f"索引 {index_name} 创建成功")
+    except Exception as e:
+        print(f"创建索引失败: {str(e)}")
+        raise
+
+def save_to_es(text, index_name="knowledge_base"):
+    """将文本向量化后保存到ES"""
+    vector = text_to_embedding(text)
+    doc = {
+        "text": text,
+        "vector": vector,
+        "timestamp": datetime.datetime.now().isoformat()
+    }
+    
+    try:
+        res = es.index(index=index_name, document=doc)
+        print(f"文档已保存，ID: {res['_id']}")
+        return res
+    except Exception as e:
+        print(f"保存到ES失败: {str(e)}")
+        raise
+
+# 使用示例
+if __name__ == "__main__":
+    create_vector_index()  # 首次运行前执行
+    save_to_es("如何更换支付宝绑定银行卡")
\ No newline at end of file