Fix Solution link generation: preserve files data in lookup

- Prefer TOML data (with files) over API data (without files) in lookup
- Avoid adding duplicate problems from API when local TOML exists
- Ensure Solution links are added when problems have solution files
- Fix issue where API data overwrote TOML data in problems lookup

Author: lufftw

tools/ai-markmap-agent/src/leetcode_api.py

@@ -7,6 +7,7 @@
 from __future__ import annotations
 
 import json
+import re
 from pathlib import Path
 from typing import Any, Dict
 
@@ -87,8 +88,29 @@ def merge_leetcode_api_data(
                 problem_data["difficulty"] = api_data.get("difficulty", "")
 
     # 添加 API 中有但本地沒有的問題（作為參考）
+    # 注意：只添加真正本地沒有的問題，避免覆蓋本地 TOML 資料
     for api_id, api_data in api_problems.items():
-        if api_id not in merged:
+        # 檢查是否已經存在（可能以不同的 key 格式存在，如 "0001_two_sum"）
+        # 標準化 api_id 以便比較
+        normalized_api_id = api_id
+        if isinstance(api_id, str) and api_id.isdigit():
+            normalized_api_id = api_id.zfill(4)
+        
+        # 檢查是否已經存在（檢查所有可能的 key 格式）
+        exists = False
+        for key in merged.keys():
+            # 檢查 key 是否匹配（可能是 "0001" 或 "0001_two_sum"）
+            if key == normalized_api_id or key == api_id:
+                exists = True
+                break
+            # 檢查 key 是否以 ID 開頭（如 "0001_two_sum"）
+            if isinstance(key, str):
+                match = re.match(r'^(\d+)_', key)
+                if match and match.group(1).zfill(4) == normalized_api_id:
+                    exists = True
+                    break
+        
+        if not exists:
             # 只添加基本資訊，標記為 API 來源
             merged[api_id] = {
                 "id": api_id,

tools/ai-markmap-agent/src/post_processing.py

@@ -100,16 +100,41 @@ def _build_problems_lookup(self, problems: dict[str, Any]) -> dict[str, dict]:
                 if problem_id_str.isdigit():
                     # Store as 4-digit format: "0011"
                     normalized_id = problem_id_str.zfill(4)
-                    lookup[normalized_id] = value
+                    # Only store if not exists, or if current value has files (prefer TOML data over API data)
+                    if normalized_id not in lookup:
+                        lookup[normalized_id] = value
+                    else:
+                        # Prefer value with files (TOML data) over value without files (API data)
+                        existing = lookup[normalized_id]
+                        existing_has_files = bool(existing.get("files", {}).get("solution"))
+                        current_has_files = bool(value.get("files", {}).get("solution"))
+                        if current_has_files and not existing_has_files:
+                            lookup[normalized_id] = value
 
                     # Store as integer string (no leading zeros): "11"
                     int_id = str(int(problem_id_str))
                     if int_id != normalized_id:
-                        lookup[int_id] = value
+                        if int_id not in lookup:
+                            lookup[int_id] = value
+                        else:
+                            # Prefer value with files
+                            existing = lookup[int_id]
+                            existing_has_files = bool(existing.get("files", {}).get("solution"))
+                            current_has_files = bool(value.get("files", {}).get("solution"))
+                            if current_has_files and not existing_has_files:
+                                lookup[int_id] = value
 
                     # Also store original format if different
                     if problem_id_str != normalized_id and problem_id_str != int_id:
-                        lookup[problem_id_str] = value
+                        if problem_id_str not in lookup:
+                            lookup[problem_id_str] = value
+                        else:
+                            # Prefer value with files
+                            existing = lookup[problem_id_str]
+                            existing_has_files = bool(existing.get("files", {}).get("solution"))
+                            current_has_files = bool(value.get("files", {}).get("solution"))
+                            if current_has_files and not existing_has_files:
+                                lookup[problem_id_str] = value
 
                     # Debug: Store sample for first few problems
                     if len(debug_samples) < 3: