feat(kb): Enhance knowledge base with new collections and search functionality

- Added new configuration for knowledge base root directory in `.env.example` and `settings.py`.
- Implemented `find_products` method in `CoreReadOnlyRepository` for fuzzy product search based on user queries.
- Introduced `search_cs_knowledge` function in `rag_service.py` to facilitate semantic search across new `fin_*` collections.
- Updated document parsing to support Markdown and YAML front-matter for knowledge base entries.
- Created multiple new FAQ and policy documents in the `data/kb_collections` directory to enrich the knowledge base.

This update significantly improves the knowledge retrieval capabilities for customer service interactions, ensuring more relevant and accurate responses.

scripts/kb/test_search.py

@@ -1,6 +1,11 @@
-"""快速测试 Milvus 检索是否正常工作。"""
+"""快速测试 fin_* Milvus 检索（需先 build_collections 灌库）。"""
+from __future__ import annotations
+
 import sys
-sys.path.insert(0, "d:/金融系统")
+from pathlib import Path
+
+ROOT = Path(__file__).resolve().parents[2]
+sys.path.insert(0, str(ROOT))
 
 from app.tool.embedding_tool import get_embedder
 from app.tool.milvus_tool import get_milvus_client
@@ -8,29 +13,23 @@ from app.tool.milvus_tool import get_milvus_client
 e = get_embedder()
 m = get_milvus_client()
 
-# 测试产品检索
-v = e.embed("R3中风险产品有哪些")
-hits = m.search("fin_product", v, top_k=3)
-print("=== 产品检索 ===")
-for h in hits:
-    score = h.get("score", 0)
-    text = h.get("chunk_text", "")[:100]
-    print(f"score={score:.4f} | {text}")
+try:
+    v = e.embed("R3中风险产品有哪些")
+    hits = m.search("fin_product", v, top_k=3)
+    print("=== 产品检索 ===")
+    for h in hits:
+        print(f"score={h.get('score', 0):.4f} | {(h.get('chunk_text') or '')[:100]}")
 
-# 测试 FAQ 检索
-v2 = e.embed("开户需要什么材料")
-hits2 = m.search("fin_faq", v2, top_k=3)
-print("\n=== FAQ 检索 ===")
-for h in hits2:
-    score = h.get("score", 0)
-    text = h.get("chunk_text", "")[:100]
-    print(f"score={score:.4f} | {text}")
+    v2 = e.embed("开户需要什么材料")
+    hits2 = m.search("fin_faq", v2, top_k=3)
+    print("\n=== FAQ 检索 ===")
+    for h in hits2:
+        print(f"score={h.get('score', 0):.4f} | {(h.get('chunk_text') or '')[:100]}")
 
-# 测试政策检索
-v3 = e.embed("投资者适当性匹配规则")
-hits3 = m.search("fin_policy", v3, top_k=3)
-print("\n=== 政策检索 ===")
-for h in hits3:
-    score = h.get("score", 0)
-    text = h.get("chunk_text", "")[:100]
-    print(f"score={score:.4f} | {text}")
+    v3 = e.embed("投资者适当性匹配规则")
+    hits3 = m.search("fin_policy", v3, top_k=3)
+    print("\n=== 政策检索 ===")
+    for h in hits3:
+        print(f"score={h.get('score', 0):.4f} | {(h.get('chunk_text') or '')[:100]}")
+finally:
+    m.close()