log/2026-06-update.log

@@ -49,4 +49,3 @@
 [2026-06-24 20:53:59] [WARN] 未能获取新的更新内容
 [2026-06-25 09:01:50] [WARN] 未能获取新的更新内容
 [2026-06-25 20:55:17] [WARN] 未能获取新的更新内容
-[2026-06-26 09:04:28] [WARN] 未能获取新的更新内容

main.py

@@ -1,4 +1,3 @@
-import base64
 import os
 import re
 import smtplib
@@ -42,36 +41,28 @@ def _extract_urls(summary):
     return urls, decoded
 
 
-_NODE_SCHEME_RE = re.compile(r"(?:vmess|vless|trojan|ss|ssr|hysteria2?|tuic)://")
+def _pick_url(urls, mode):
+    if mode == "v2ray":
+        for suffix in (".txt", ".json"):
+            for url in urls:
+                if url.lower().endswith(suffix):
+                    return url
+    if mode == "clash":
+        for suffix in (".yaml", ".yml"):
+            for url in urls:
+                if url.lower().endswith(suffix):
+                    return url
+    return ""
 
 
-def _b64decode(text):
-    compact = re.sub(r"\s+", "", text)
-    if not compact:
-        return ""
-    try:
-        # binascii.Error 是 ValueError 的子类，统一捕获即可
-        raw = base64.b64decode(compact + "=" * (-len(compact) % 4))
-    except ValueError:
-        return ""
-    return raw.decode("utf-8", "ignore")
-
-
-def _detect_kind(text):
-    """根据下载内容判断订阅类型：'clash' / 'v2ray'，无法识别返回 None。"""
-    sample = text.strip()
-    if not sample:
-        return None
-    # clash 配置为 YAML，包含 proxies/proxy-groups 字段
-    if re.search(r"^(?:proxies|proxy-groups)\s*:", sample, re.MULTILINE):
-        return "clash"
-    # v2ray 订阅为节点 URI 列表，可能是明文或 base64 编码
-    if _NODE_SCHEME_RE.search(sample):
-        return "v2ray"
-    decoded = _b64decode(sample)
-    if decoded and _NODE_SCHEME_RE.search(decoded):
-        return "v2ray"
-    return None
+def _pick_urls(urls, mode):
+    matched = []
+    suffixes = (".txt", ".json") if mode == "v2ray" else (".yaml", ".yml")
+    for suffix in suffixes:
+        for url in urls:
+            if url.lower().endswith(suffix) and url not in matched:
+                matched.append(url)
+    return matched
 
 
 def _download_with_retry(urls):
@@ -117,25 +108,19 @@ def _build_session():
     return session
 
 
-def _classify_subscriptions(session, urls):
-    """逐个下载候选链接并按内容判断类型，返回 {'v2ray': (req, url), 'clash': (req, url)}。"""
-    found = {}
+def _download_candidates(session, urls):
+    if not urls:
+        return None, None
     for url in urls:
-        if "v2ray" in found and "clash" in found:
-            break
         try:
             req = session.get(url, verify=False, timeout=20)
         except requests.RequestException as e:
             write_log(f"请求失败：{url} - {e}", "WARN")
             continue
-        if req.status_code not in ok_code:
-            write_log(f"请求失败：{url} - {req.status_code}", "WARN")
-            continue
-        kind = _detect_kind(req.text)
-        if kind and kind not in found:
-            found[kind] = (req, url)
-            write_log(f"识别到 {kind} 订阅：{url}", "INFO")
-    return found
+        if req.status_code in ok_code:
+            return req, url
+        write_log(f"请求失败：{url} - {req.status_code}", "WARN")
+    return None, urls[0]
 
 def get_subscribe_url():
     dirs = './subscribe'
@@ -176,40 +161,62 @@ def get_subscribe_url():
         write_log("暂时没有可用的订阅更新", "WARN")
         return
 
-    urls, _ = _extract_urls(summary)
+    urls, decoded_summary = _extract_urls(summary)
 
-    # 链接已无固定后缀，需下载内容后再判断是 v2ray 还是 clash
-    classified = _classify_subscriptions(session, urls)
+    v2ray_url = _pick_url(urls, "v2ray")
+    clash_url = _pick_url(urls, "clash")
+    v2ray_candidates = _pick_urls(urls, "v2ray")
+    clash_candidates = _pick_urls(urls, "clash")
+
+    # 兼容旧页面结构，通用提取失败时再尝试历史规则
+    if not v2ray_url:
+        v2ray_list = re.findall(r">V2Ray/XRay -&gt; (.*?)</span>", summary)
+        if not v2ray_list:
+            v2ray_list = re.findall(r">V2Ray/XRay -> (.*?)</span>", decoded_summary)
+        if any(v2ray_list):
+            v2ray_url = v2ray_list[-1].replace('amp;', '')
+            if v2ray_url not in v2ray_candidates:
+                v2ray_candidates.append(v2ray_url)
+
+    if not clash_url:
+        clash_list = re.findall(r">clash -&gt; (.*?)</span>", summary)
+        if not clash_list:
+            clash_list = re.findall(r">clash -> (.*?)</span>", decoded_summary)
+        if any(clash_list) and not clash_list[-1].startswith("订阅地址生成失败"):
+            clash_url = clash_list[-1].replace('amp;', '')
+            if clash_url not in clash_candidates:
+                clash_candidates.append(clash_url)
 
     # 获取普通订阅链接
-    v2ray_entry = classified.get("v2ray")
-    if v2ray_entry:
-        v2ray_req, _ = v2ray_entry
-        update_list.append(f"v2ray: {v2ray_req.status_code}")
-        with open(dirs + '/v2ray.txt', 'w', encoding="utf-8") as f:
-            f.write(v2ray_req.text)
-    else:
-        cache_file = dirs + '/v2ray.txt'
-        if os.path.exists(cache_file) and os.path.getsize(cache_file) > 0:
-            update_list.append("v2ray: cache")
-            write_log("未获取到 v2ray 订阅，已保留本地缓存", "WARN")
+    if v2ray_url:
+        v2ray_req, used_v2ray_url = _download_candidates(session, v2ray_candidates)
+        if not v2ray_req:
+            cache_file = dirs + '/v2ray.txt'
+            if os.path.exists(cache_file) and os.path.getsize(cache_file) > 0:
+                update_list.append("v2ray: cache")
+                write_log(f"获取 v2ray 订阅失败，已保留本地缓存：{used_v2ray_url}", "WARN")
+            else:
+                write_log(f"获取 v2ray 订阅失败：{used_v2ray_url}", "WARN")
         else:
-            write_log("未获取到 v2ray 订阅", "WARN")
+            update_list.append(f"v2ray: {v2ray_req.status_code}")
+            with open(dirs + '/v2ray.txt', 'w', encoding="utf-8") as f:
+                f.write(v2ray_req.text)
 
     # 获取clash订阅链接
-    clash_entry = classified.get("clash")
-    if clash_entry:
-        clash_req, _ = clash_entry
-        update_list.append(f"clash: {clash_req.status_code}")
-        with open(dirs + '/clash.yml', 'w', encoding="utf-8") as f:
-            f.write(clash_req.content.decode("utf-8"))
-    else:
-        cache_file = dirs + '/clash.yml'
-        if os.path.exists(cache_file) and os.path.getsize(cache_file) > 0:
-            update_list.append("clash: cache")
-            write_log("未获取到 clash 订阅，已保留本地缓存", "WARN")
+    if clash_url and not clash_url.startswith("订阅地址生成失败"):
+        clash_req, used_clash_url = _download_candidates(session, clash_candidates)
+        if not clash_req:
+            cache_file = dirs + '/clash.yml'
+            if os.path.exists(cache_file) and os.path.getsize(cache_file) > 0:
+                update_list.append("clash: cache")
+                write_log(f"获取 clash 订阅失败，已保留本地缓存：{used_clash_url}", "WARN")
+            else:
+                write_log(f"获取 clash 订阅失败：{used_clash_url}", "WARN")
         else:
-            write_log("未获取到 clash 订阅", "WARN")
+            update_list.append(f"clash: {clash_req.status_code}")
+            with open(dirs + '/clash.yml', 'w', encoding="utf-8") as f:
+                clash_content = clash_req.content.decode("utf-8")
+                f.write(clash_content)
     if update_list:
         file_pat = re.compile(r"v2ray\.txt|clash\.yml")
         if file_pat.search(os.popen("git status").read()):