From c15cbcf9ce0d5d25a5b52e3f1881535c06706586 Mon Sep 17 00:00:00 2001
From: QCQCQC <1220204124@zust.edu.cn>
Date: Tue, 6 May 2025 11:22:20 +0800
Subject: [PATCH] =?UTF-8?q?=E8=BF=94=E5=9B=9E=E5=80=BC=E6=9B=B4=E5=AE=8C?=
 =?UTF-8?q?=E5=96=84?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 app/__pycache__/router.cpython-310.pyc | Bin 3405 -> 3825 bytes
 app/router.py                          |  23 +++++++++++++++++++----
 2 files changed, 19 insertions(+), 4 deletions(-)

diff --git a/app/__pycache__/router.cpython-310.pyc b/app/__pycache__/router.cpython-310.pyc
index 69b816929dc32f86288548ac801ab00e42533d68..db4f25afeef17fdf9c3efddbf573f7c232133118 100644
GIT binary patch
delta 688
zcmYjNJ8u&~5Z>AIIp_2Fe17ENJV22U!URP~NC<(IhoFUmq%t7u;@vnVXCLft0HZY)
zn3jeu0Exo+2?+iJ6{0_h9!e?-8pe(zG17eV&3xaEM*D5?_lh6qaxTGm_0BsuDtz(-
z8!ni$_D@I~C;c8Sn&b2zSpWQCW(07<T+bZl%jnKv-o%ub!F<WQgxlt-dv4GoeP0Ed
zYLB@UGNiZ2gQfelpYI2%r1PWlF&RR?fIOJ;qAuv-7U`F?Kblo@x-^7i0P^^rc1E*0
ztLD{$c89>SL#pO=Zd^}xYoB=^?TL3&I}a-?$NblM!tyM5LxE^#hg1vZg}Xfvt9a(G
z6hhS&1uRK9Y_$Y$g^D+1yVH(B(N?`Ah{7Hh|4r(CCsfS?&KeU%Cc$Zc$a_y?!DJI~
z-D!-ug}H+%0|ZSNRLB4et&n|+2U&N-77Ai>4D$$8NZ{FooxlRl5~mi0ojn$A^sti@
zAt2AsNFIvjL4yksi}hL*H^WG7AUUz_R%v44S;>sxPq>i0x+olcDJ=8VId7J-wM&U5
zxk_yLC~7NlWvs0uB({i;x-7Bxqc~JaO2$G}%Z6ur+uJDD-6M&gGWmeI!}g{uJDIBZ
zkauN0gX1a9HRDbv?oM0tBfFBmi(BNix8Cxqm9pl*qaHXAK#Ecrvoug}{}wYTDt679
c_dT_XeltkVno@2pUv$B<mcjD^@aD?!55$PDv;Y7A

delta 272
zcmWMdy-EW?7`$)qH+#FAlbnU&L<2!MEP@CY5)et1QtZ;m;RSP;7~MOIgKNY`_!<Sl
z<PjuV_yTs4-Y1Bqm6hN(!_2@i^RxNcR8uLrL68^z=v}_3z=jyN>=n4^IS;UlbLSU2
zi$`|`(8a^rWnIzQCG|-8hk>uPJ80nnZ?`LN)~Ng<tYpgXStU|`!$`%GIR`Lqjw<ie
zh$XK0ap&8juWpgWre32M@I0Gj<3wjgIZ%_NI3EquvvE=mbsn$dw#by;r`?yWr?*#W
ztnf{gcrBy76KeAvGK&e|%mWS)3m^cCHNhh)uz*oI!2c{q_i-tg(8p)ht2a0ZGXfz3
H5Mc=aIE^~e

diff --git a/app/router.py b/app/router.py
index 999f6c0..47c3367 100644
--- a/app/router.py
+++ b/app/router.py
@@ -94,13 +94,28 @@ def search_error(query: ErrorQuery):
     for idx, score in zip(indices[0], sim[0]):
         db_id = index_to_id[idx]
         db_entry = error_memory[db_id]
+
+        # 分层匹配得分
         layer_score = compute_layered_similarity_sco(user_vectors, db_entry["layer_vectors"])
+
+        # 逐层匹配，提取每一层的关键字
+        matched_keywords = []
+        for user_layer, db_layer in zip(user_layers, db_entry["layers"]):
+            # 这里简单取两个文本的公共单词作为关键字（你可以换成更复杂的匹配方法）
+            user_tokens = set(user_layer.split())
+            db_tokens = set(db_layer.split())
+            common_tokens = user_tokens & db_tokens
+            matched_keywords.append(list(common_tokens))
+
         results.append({
             "db_id": db_id,
-            "similarity": round(layer_score, 4),
-            "matched_layers": db_entry["layers"]
+            "aggregate_similarity": round(float(score), 4),    # 聚合索引匹配得分
+            "layer_similarity": round(layer_score, 4),         # 分层匹配得分
+            "matched_layers": db_entry["layers"],              # 匹配到的 traceback 每一层
+            "matched_keywords": matched_keywords               # 每一层匹配到的关键字列表
         })
 
-    # 还需要根据相似度排序
-    results.sort(key=lambda x: x["similarity"], reverse=True)
+    # 先按分层得分排，再按聚合得分排
+    results.sort(key=lambda x: (x["layer_similarity"], x["aggregate_similarity"]), reverse=True)
+
     return results