v1.5: SSL auto-retry, 3-tier rune recommendations, ApexLol data scraper

- Add SSL EOF auto-retry (2 retries with 2s delay) for proxy connectivity issues

- Enhance prompt to recommend 3 augment builds (best/alternative/different style)

- Add ApexLol data scraper module for champion augment data

- Add data management module (apexlol_data.py)

- Update model to gemini-3.1-flash-lite-preview

- Update .gitignore to exclude test files

Author: MJ33520

.gitignore

@@ -5,3 +5,8 @@ screenshots/
 gh.exe
 lcu_client.py
 .env
+apexlol_cache/
+test_scraper.py
+test_filter.py
+test_filter2.py
+test_models.py

apexlol_data.py

@@ -0,0 +1,232 @@
+# -*- coding: utf-8 -*-
+"""ARAM 助手 - ApexLol 数据查询模块
+
+管理本地缓存的 apexlol.info 数据，提供英雄联动信息查询。
+"""
+
+import os
+import json
+import time
+import logging
+
+log = logging.getLogger("ARAM")
+
+# ==================== 英雄名称映射 ====================
+# 中文标题 -> 英文 ID（爬取时自动构建，这里是常用别名的手动补充）
+CHAMPION_ALIASES = {
+    # 常用简称 -> ID
+    "卡特": "Katarina", "火男": "Brand", "剑圣": "MasterYi",
+    "大嘴": "KogMaw", "小法": "Veigar", "女枪": "MissFortune",
+    "老鼠": "Twitch", "锤石": "Thresh", "狗头": "Nasus",
+    "牛头": "Alistar", "猴子": "MonkeyKing", "蛮王": "Tryndamere",
+    "石头人": "Malphite", "机器人": "Blitzcrank", "稻草人": "Fiddlesticks",
+    "酒桶": "Gragas", "皇子": "JarvanIV", "螳螂": "Khazix",
+    "人马": "Hecarim", "薇恩": "Vayne", "亚索": "Yasuo",
+    "永恩": "Yone", "艾克": "Ekko", "EZ": "Ezreal",
+    "ez": "Ezreal", "VN": "Vayne", "vn": "Vayne",
+    "ADC": None,  # 通用标签不映射
+}
+
+# 全局缓存
+_cache = None
+_name_to_id = None
+
+
+def _build_name_map(data: dict) -> dict:
+    """从缓存数据构建 名称 -> ID 的映射表。"""
+    name_map = {}
+
+    # 从 champion_list 构建
+    for champ in data.get("champion_list", []):
+        champ_id = champ["id"]
+        cn_title = champ.get("cn_title", "")
+
+        # 中文标题 -> ID（如 "不祥之刃" -> "Katarina"）
+        if cn_title:
+            name_map[cn_title] = champ_id
+
+        # 英文 ID 本身（大小写不敏感）
+        name_map[champ_id.lower()] = champ_id
+
+    # 从 champions 数据构建
+    for champ_id, info in data.get("champions", {}).items():
+        cn_title = info.get("cn_title", "")
+        if cn_title:
+            name_map[cn_title] = champ_id
+            
+        cn_name = info.get("cn_name", "")
+        if cn_name:
+            name_map[cn_name] = champ_id
+
+    # 加入手动别名
+    for alias, cid in CHAMPION_ALIASES.items():
+        if cid:
+            name_map[alias] = cid
+
+    return name_map
+
+
+def load_cache(cache_dir: str) -> dict:
+    """加载本地缓存数据。"""
+    global _cache, _name_to_id
+
+    cache_file = os.path.join(cache_dir, "apexlol_data.json")
+    if not os.path.exists(cache_file):
+        log.warning("[ApexLol] 缓存文件不存在")
+        return {}
+
+    try:
+        with open(cache_file, "r", encoding="utf-8") as f:
+            _cache = json.load(f)
+        _name_to_id = _build_name_map(_cache)
+        log.info(f"[ApexLol] ✅ 已加载缓存 ({len(_cache.get('champions', {}))} 英雄)")
+        return _cache
+    except Exception as e:
+        log.error(f"[ApexLol] 加载缓存失败: {e}")
+        return {}
+
+
+def is_cache_valid(cache_dir: str, ttl_days: int = 7) -> bool:
+    """检查缓存是否有效（文件存在且未过期）。"""
+    cache_file = os.path.join(cache_dir, "apexlol_data.json")
+    if not os.path.exists(cache_file):
+        return False
+
+    try:
+        mtime = os.path.getmtime(cache_file)
+        age_days = (time.time() - mtime) / 86400
+        return age_days < ttl_days
+    except Exception:
+        return False
+
+
+def resolve_champion_id(name: str) -> str | None:
+    """将英雄名（中文/英文/别名）解析为英文 ID。"""
+    global _name_to_id
+    if _name_to_id is None:
+        return None
+
+    # 精确匹配
+    if name in _name_to_id:
+        return _name_to_id[name]
+
+    # 大小写不敏感匹配
+    lower = name.lower()
+    if lower in _name_to_id:
+        return _name_to_id[lower]
+
+    # 模糊匹配：包含关系
+    for key, cid in _name_to_id.items():
+        if name in key or key in name:
+            return cid
+
+    return None
+
+
+def lookup_champion(name: str) -> str:
+    """查询单个英雄的海克斯联动分析文本。
+
+    Args:
+        name: 英雄名（中文标题、英文 ID 或别名）
+
+    Returns:
+        格式化的联动分析文本，未找到返回空字符串
+    """
+    global _cache
+    if not _cache:
+        return ""
+
+    champ_id = resolve_champion_id(name)
+    if not champ_id:
+        return ""
+
+    champ_data = _cache.get("champions", {}).get(champ_id)
+    if not champ_data or not champ_data.get("synergies"):
+        return ""
+
+    cn_title = champ_data.get("cn_title", champ_id)
+    lines = [f"【{cn_title}({champ_id})的海克斯联动数据 - 来源: apexlol.info】"]
+
+    for s in champ_data["synergies"]:
+        hex_names = " + ".join(s.get("hex_names", []))
+        rating = s.get("rating", "")
+        tiers = " / ".join(s.get("hex_tiers", []))
+        tag = s.get("tag", "")
+
+        header_parts = []
+        if rating:
+            header_parts.append(f"[{rating}级]")
+        header_parts.append(hex_names)
+        if tiers:
+            header_parts.append(f"({tiers})")
+        if tag:
+            header_parts.append(f"- {tag}")
+
+        lines.append(f"\n▸ {' '.join(header_parts)}")
+        lines.append(f"  {s.get('analysis', '')}")
+
+    return "\n".join(lines)
+
+
+def lookup_champions(names: list[str], highlight_mine: str = None) -> str:
+    """批量查询多个英雄的联动数据，拼接为 Gemini 可用的参考文本。
+
+    Args:
+        names: 英雄名列表
+        highlight_mine: 我的英雄名（会放在最前面并额外标注）
+
+    Returns:
+        拼接后的参考文本
+    """
+    sections = []
+
+    # 先处理"我的英雄"
+    if highlight_mine:
+        my_data = lookup_champion(highlight_mine)
+        if my_data:
+            sections.append(f"=== ⭐ 我的英雄 ===\n{my_data}")
+
+    # 处理其余英雄
+    for name in names:
+        if highlight_mine and name == highlight_mine:
+            continue
+        data = lookup_champion(name)
+        if data:
+            sections.append(data)
+
+    if not sections:
+        return ""
+
+    header = (
+        "📚 以下是来自 apexlol.info 的海克斯联动分析数据（专业玩家社区贡献的机制级分析）。\n"
+        "请重点参考这些数据来推荐海克斯符文，特别注意隐藏的联动机制（如某技能能触发特定符文效果等）。\n"
+        "注意：这些数据是参考建议，请结合当前阵容做出最佳判断。\n"
+        "=" * 60
+    )
+
+    return header + "\n\n" + "\n\n".join(sections)
+
+
+def get_cache_info(cache_dir: str) -> dict:
+    """获取缓存状态信息。"""
+    cache_file = os.path.join(cache_dir, "apexlol_data.json")
+    if not os.path.exists(cache_file):
+        return {"exists": False}
+
+    try:
+        mtime = os.path.getmtime(cache_file)
+        age_hours = (time.time() - mtime) / 3600
+        size_mb = os.path.getsize(cache_file) / (1024 * 1024)
+
+        with open(cache_file, "r", encoding="utf-8") as f:
+            data = json.load(f)
+
+        return {
+            "exists": True,
+            "age_hours": round(age_hours, 1),
+            "size_mb": round(size_mb, 2),
+            "champion_count": len(data.get("champions", {})),
+            "scraped_at": data.get("meta", {}).get("scraped_at", "unknown"),
+        }
+    except Exception:
+        return {"exists": False}