fix(v2.0.1): 修复反爬误判 + Wayback 兜底未触发两个 bug

Bug 1: Wayback 兜底未触发 (search.py fetch_page)

- 根因: Cloudflare JS 质询页常返回 HTTP 200 (非 403), 主抓取 result 非 None

- _should_try_fallback 在 result 非 None 时直接返回 False, 跳过兜底

- 反爬检测在兜底判断之后执行, 错过兜底入口

- 修复: 反爬检测提前到兜底判断之前, anti_bot_detected=True 也触发 Wayback

- Wayback 结果重新做反爬检测 (防御性)

Bug 2: WAF 指纹库误判正常内容 (search.py WAF_FINGERPRINTS)

- 根因: 裸公司名 (cloudflare/akamai) 和宽泛词 (captcha/challenge/dd-) 做全文匹配

- DataCamp 文章引用 cloudflare.com 文档链接 -> 误判为 cloudflare WAF

- 'coding challenges' 正常内容 -> 误判为 generic 反爬

- Wayback 归档正文被误判, 兜底返回的有效内容被丢弃

- 修复: 移除裸公司名和宽泛词, 改用技术标识符 (cf-ray/incap_ses/bm_sz 等)

- 通用文案用完整短语 (please complete the captcha) 替代单词

- 增加 <title> 标签精准检测 (反爬页 title 是特征文案, 误判率极低)

- 新增 Anubis 反爬系统检测 (anubis_challenge/miserere)

真实测试验证 (search.metona.cn 实例):

- v2.0.0: --fetch 3 全部失败 (3 ERR: cloudflare/generic, Wayback 未触发)

- v2.0.1: --fetch 3 全部成功 (3 OK: 38100/93442/1945 chars, UA 轮换绕过 Cloudflare)

测试: 458 个全部通过 (新增 7 个测试覆盖修复行为)

README.md

@@ -297,7 +297,7 @@ python scripts/fetch.py -u https://example.com \
 - JSON Lines 流式输出（`--stream`，含 `error` 事件类型）
 - 进度事件（`--progress`，含 `instance_try`/`instance_ok`/`instance_fail`，JSON Lines 到 stderr）
 - batch 模式统一 schema（`status` 字段区分成功/失败）
-- 451 个单元+集成测试
+- 458 个单元+集成测试
 
 ## 跨 Agent 兼容性
 
@@ -325,7 +325,7 @@ pip install pytest
 pytest -q
 ```
 
-451 个测试覆盖：缓存操作、认证解析、域名过滤、Markdown 转换、搜索逻辑、集成流程、日志配置、HTML 回退、自动抓取、健康检查、输出格式化、实例解析、并行搜索、CLI 端到端、错误码分类、流式输出、进度事件、配置文件认证、schema_version、recovery_hint、batch 统一 schema、--dump-schema、UTF-8 stdout 强制、Windows APPDATA 路径、v2.0.0 浏览器指纹头、WAF 反爬检测、Wayback Machine 兜底、自适应限流。
+458 个测试覆盖：缓存操作、认证解析、域名过滤、Markdown 转换、搜索逻辑、集成流程、日志配置、HTML 回退、自动抓取、健康检查、输出格式化、实例解析、并行搜索、CLI 端到端、错误码分类、流式输出、进度事件、配置文件认证、schema_version、recovery_hint、batch 统一 schema、--dump-schema、UTF-8 stdout 强制、Windows APPDATA 路径、v2.0.0 浏览器指纹头、WAF 反爬检测（v2.0.1 收窄误判 + title 精准检测）、Wayback Machine 兜底（v2.0.1 修复 HTTP 200 反爬页触发）、自适应限流。
 
 ## 项目结构
 

SKILL.md

@@ -1,7 +1,7 @@
 ---
 name: searxng-use-cli
 description: Use when you need to search the web via your OWN SearXNG instance (no public-instance discovery). 3 CLI scripts + a shared common.py module — execute privacy-respecting searches against a user-supplied instance (with multi-instance failover, 5xx/429 retry, auto-fetch) or via SEARXNG_INSTANCE env / config file, fetch/extract readable text or markdown from web pages. Zero-config replacement for proprietary search APIs.
-version: 2.0.0
+version: 2.0.1
 author: Metona Team
 license: MIT
 platforms: [linux, macos, windows]
@@ -69,8 +69,8 @@ SearXNG is a privacy-respecting metasearch engine that aggregates results from 7
 - **Split timeouts** — `timeout=(connect, read)` tuple (5s connect, 15s read by default). Previously a single 15s total timeout wasted already-established connections on large pages
 - **Retry-After compliance** — 429/503 responses read the `Retry-After` header (numeric seconds or HTTP date) and wait at least that long before retrying. Non-compliance triggers harsher rate limits
 - **Capped backoff** — `compute_backoff_delay()` caps at 60s (was uncapped: 1.5*2^10 = 1536s would hang the process)
-- **WAF fingerprint library** — `_detect_anti_bot()` identifies Cloudflare / Imperva / PerimeterX / DataDome / Akamai / generic challenges via full-document scan (was first 2000 chars only). Returns `waf_type` for AI-agent decisioning
+- **WAF fingerprint library** — `_detect_anti_bot()` identifies Cloudflare / Imperva / PerimeterX / DataDome / Akamai / Anubis / generic challenges via `<title>` tag matching (most precise) + full-document scan of technical identifiers (cookie/header/JS variable names, not bare vendor names). v2.0.1 narrowed broad keywords (e.g. bare `cloudflare`/`captcha`/`challenge`) that caused false positives on normal articles, and added title-tag detection. Returns `waf_type` for AI-agent decisioning
-- **Wayback Machine fallback** — 404/403/timeout automatically retries via `https://web.archive.org/web/2/<url>` (latest snapshot). Default ON; `--no-fallback` disables. Independent 10s timeout so Wayback slowness never blocks the main flow
+- **Wayback Machine fallback** — 404/403/timeout AND anti-bot-blocked pages automatically retry via `https://web.archive.org/web/2/<url>` (latest snapshot). v2.0.1 fixed a bug where HTTP 200 anti-bot challenge pages (Cloudflare returns 200 for JS challenges) bypassed the fallback trigger. Default ON; `--no-fallback` disables. Independent 10s timeout so Wayback slowness never blocks the main flow
 - **Adaptive throttling** — `AdaptiveThrottle` state machine: 3 consecutive failures → double delay + halve concurrency; 5 consecutive successes → gradual recovery; 429 → global pause 30s. Thread-safe
 - **readability-lite extraction** — when `<article>`/`<main>`/content-class `<div>` are all missing, `_readability_lite()` picks the highest text-density node (text chars / tag count + `<p>` weighting), avoiding nav/sidebar/footer noise
 - **`--fetch-report`** — structured per-URL report to stderr after `--fetch N`: status, WAF type, fallback used, char count, plus adaptive throttle stats and a JSON summary line

scripts/_config.py

@@ -7,6 +7,6 @@ Retry settings and shared HTTP utilities now live in ``common.py`` so that
 both ``search.py`` and ``fetch.py`` share one consistent implementation.
 """
 
-VERSION = "2.0.0"
+VERSION = "2.0.1"
 SCHEMA_VERSION = "1.0"
 USER_AGENT = f"searxng-cli/{VERSION}"

scripts/search.py

@@ -13,6 +13,7 @@ import json
 import logging
 import os
 import random
+import re
 import sys
 import threading
 import time
@@ -855,9 +856,35 @@ def fetch_page(url: str, timeout: int = 10, auth_headers: dict = None,
     except Exception as e:
         error_msg = str(e) if str(e) else e.__class__.__name__
 
-    # Wayback 兜底：主抓取失败或被反爬拦截时尝试
+    # 反爬检测（v2.0.0 增强：全文档扫描 + WAF 指纹库）
+    # 必须在 Wayback 兜底判断之前执行：Cloudflare 质询页常返回 HTTP 200，
+    # 此时 result 非 None 但内容是反爬页，必须识别出来才能触发兜底。
+    anti_bot_detected = False
+    waf_type = None
+    if result is not None:
+        content = result.content
+        content_type = result.content_type or ""
+        is_html = ("html" in content_type.lower() or
+                   content.strip().startswith("<!") or
+                   content.strip().startswith("<htm"))
+        if is_html:
+            waf_type = _detect_anti_bot(content)
+            if waf_type:
+                anti_bot_detected = True
+
+    # Wayback 兜底触发条件（v2.0.1 修复）：
+    #   1. 主抓取抛异常且错误暗示 404/403/超时（_should_try_fallback）
+    #   2. 主抓取"成功"但被反爬拦截（anti_bot_detected=True）
+    # 原 v2.0.0 bug：仅条件 1 触发兜底，条件 2 漏网（Cloudflare 200 质询页）。
     fallback_used = None
-    if fallback_enabled and _should_try_fallback(result, error_msg):
+    need_fallback = False
+    if fallback_enabled:
+        if anti_bot_detected:
+            need_fallback = True
+        elif result is None and _should_try_fallback(result, error_msg):
+            need_fallback = True
+
+    if need_fallback:
         wb_result = _try_wayback_fallback(url, timeout=timeout,
                                           auth_headers=auth_headers,
                                           max_retries=max_retries,
@@ -866,6 +893,18 @@ def fetch_page(url: str, timeout: int = 10, auth_headers: dict = None,
             result = wb_result
             error_msg = None
             fallback_used = "wayback"
+            # Wayback 结果重新做反爬检测（防御性：Wayback 快照极少是反爬页）
+            anti_bot_detected = False
+            waf_type = None
+            content = result.content
+            content_type = result.content_type or ""
+            is_html = ("html" in content_type.lower() or
+                       content.strip().startswith("<!") or
+                       content.strip().startswith("<htm"))
+            if is_html:
+                waf_type = _detect_anti_bot(content)
+                if waf_type:
+                    anti_bot_detected = True
 
     # 仍然失败
     if result is None:
@@ -877,25 +916,10 @@ def fetch_page(url: str, timeout: int = 10, auth_headers: dict = None,
             "fallback_used": None,
         }
 
-    content = result.content
+    # 反爬仍被检测到（Wayback 也无能为力或兜底被禁用）
-    content_type = result.content_type
-    final_url = result.final_url
-
-    is_html = ("html" in content_type.lower() or
-               content.strip().startswith("<!") or
-               content.strip().startswith("<htm"))
-
-    # 反爬检测（v2.0.0 增强：全文档扫描 + WAF 指纹库）
-    anti_bot_detected = False
-    waf_type = None
-    if is_html:
-        waf_type = _detect_anti_bot(content)
-        if waf_type:
-            anti_bot_detected = True
-
     if anti_bot_detected:
         return {
-            "url": url, "final_url": final_url, "status": "error",
+            "url": url, "final_url": result.final_url, "status": "error",
             "error": f"Bot protection detected ({waf_type})",
             "text": "", "text_length": 0, "truncated": False,
             "anti_bot_detected": True, "waf_type": waf_type,
@@ -905,7 +929,7 @@ def fetch_page(url: str, timeout: int = 10, auth_headers: dict = None,
     text = extract_text(content) if is_html else content
     return {
         "url": url,
-        "final_url": final_url,
+        "final_url": result.final_url,
         "status": "ok",
         "content_type": content_type,
         "text": text,
@@ -971,59 +995,105 @@ def _try_wayback_fallback(url: str, timeout: int = 10,
         return None
 
 
-# ----- 反爬检测（v2.0.0 增强版）-----
+# ----- 反爬检测（v2.0.1 收窄误判 + title 精准检测）-----
 # WAF 指纹库：每项 = (waf_type, [指示词])
-# 指示词在页面 HTML/body/headers 中出现即判定为该 WAF。
+#
-# 顺序按检测优先级：专用指纹在前，通用指纹在后。
+# v2.0.1 修复：v2.0.0 用裸公司名（cloudflare/akamai）和宽泛词（captcha/
+# challenge/dd-）做全文匹配，导致正常文章（如引用 Cloudflare 文档、讨论
+# "coding challenges" 的文章、Wayback 归档正文）被误判为反爬页，Wayback
+# 兜底返回的有效内容也被丢弃。收窄原则：
+#   1. 专用指纹只用 WAF 厂商的技术标识符（cookie 名/HTTP header 名/JS 变量名）
+#      ——这些不会出现在文章正文里
+#   2. 通用文案用完整短语而非单词（如 "please complete the captcha" 而非
+#      "captcha"），避免正常内容误判
+#   3. 增加 <title> 标签检测——反爬页 title 是特征文案，最精准
 WAF_FINGERPRINTS = [
     ("cloudflare", [
+        # Cloudflare 技术标识符（cookie/header/JS 变量名，不会出现在正文）
         "cf-ray", "cf-chl-bypass", "cf-mitigated",
-        "cloudflare", "cf-browser-verification",
+        "cf-browser-verification", "cf-error-details", "cf-error-code",
-        "attention required! | cloudflare", "just a moment",
+        # 质询页特征文案（足够具体，正常内容不会完整出现）
-        "checking your browser before accessing",
+        "just a moment", "checking your browser before accessing",
+        "attention required! | cloudflare",
+        "enable javascript and cookies to continue",
     ]),
     ("imperva", [
+        # Incapsula cookie/技术标识符
         "incap_ses", "visid_incap", "incap_ses_",
-        "imperva", "incapsula",
+        "incapsula incident id", "request unsuccessful. incapsula",
-        "request unsuccessful. incapsula incident id",
+        "visit denied by incapsula",
     ]),
     ("perimeterx", [
-        "_px", "px-captcha", "pxhd", "pxcts", "pxcookie",
+        # PerimeterX 专有标识符（_px 太短会匹配 CSS 类名，已移除）
-        "perimeterx", "press & hold to confirm you are a human",
+        "px-captcha", "pxhd", "pxcts", "pxcookie",
+        "_pxff", "_pxhd",
+        "press & hold to confirm you are a human",
     ]),
     ("datadome", [
-        "datadome", "dd-", "data-dome",
+        # DataDome 专有标识符（dd- 太宽泛会匹配 dd-class 等，已移除）
-        "protected by datadome",
+        "datadome", "data-dome",
+        "protected by datadome", "datadome-bot-protect",
     ]),
     ("akamai", [
-        "akamai", "bm_sz", "_abck",
+        # Akamai Bot Manager cookie/标识符（akamai 裸名会匹配正文引用，已移除）
-        "reference #", "akamaighost",
+        "bm_sz", "_abck", "akamaighost", "akamai-bot-manager",
+        "ak_bmsc",
     ]),
-    # 通用反爬指示词（无明确 WAF 归属）
+    # 通用反爬指示词（v2.0.1 收窄：完整短语而非单词，避免正文误判）
     ("generic", [
-        "captcha", "challenge", "verify you are human",
+        "verify you are human", "verify that you are human",
-        "making sure you're not a bot",
+        "making sure you're not a bot", "are you a robot",
-        "please enable javascript", "enable javascript to continue",
+        "robot or human", "human verification",
-        "ddos protection", "access denied",
+        "please complete the captcha", "complete the security check",
+        "please enable javascript to continue",
+        "enable javascript to continue",
+        "ddos protection by", "access denied - sucuri",
         "you have been blocked", "unusual traffic from your computer",
-        "robot or human", "are you a robot",
         "pardon our interruption", "we'll be right back",
+        "bot protection", "anti-bot protection",
+        "anubis_challenge", "miserere",  # Anubis 反爬系统（拦截 AI 爬虫）
     ]),
 ]
 
+# <title> 标签检测：反爬页 title 通常是特征文案，比全文扫描更精准。
+# key = title 中的特征子串（小写），value = 对应 WAF 类型
+_TITLE_ANTI_BOT_SIGNATURES = {
+    "just a moment": "cloudflare",
+    "attention required": "cloudflare",
+    "access denied": "generic",
+    "are you a robot": "generic",
+    "robot check": "generic",
+    "human verification": "generic",
+    "please verify you are human": "generic",
+    "security check": "generic",
+    "verify you are human": "generic",
+}
+
+_TITLE_RE = re.compile(r"<title[^>]*>(.*?)</title>", re.IGNORECASE | re.DOTALL)
+
 
 def _detect_anti_bot(content: str) -> str:
     """检测反爬页面，返回 WAF 类型或 None。
 
-    v2.0.0 改进：
+    v2.0.1 改进：
-      * 全文档扫描（去除 2000 字符限制——大页面反爬页可能在前 2000 字之外）
+      * 增加 <title> 标签精准检测（反爬页 title 是特征文案，误判率极低）
-      * WAF 指纹库覆盖 Cloudflare/Imperva/PerimeterX/DataDome/Akamai/通用
+      * 收窄 WAF 指纹库关键词（移除裸公司名和宽泛词，改用技术标识符 + 完整短语）
-      * 返回具体 WAF 类型而非布尔值，让 AI Agent 可决策
+      * 保留全文档扫描（v2.0.0 改进，大页面反爬页可能在前 2000 字之外）
 
-    性能：全文档 lower() 一次，对 5MB 页面约 5ms，可接受。
+    检测顺序：title 标签（最精准）→ 全文档指纹扫描（技术标识符 + 文案短语）。
+
+    性能：全文档 lower() 一次 + title regex 一次，对 5MB 页面约 5ms，可接受。
     """
     if not content:
         return None
+    # 1. <title> 标签检测（最精准，误判率极低）
+    title_match = _TITLE_RE.search(content)
+    if title_match:
+        title_lower = title_match.group(1).strip().lower()
+        for sig, waf_type in _TITLE_ANTI_BOT_SIGNATURES.items():
+            if sig in title_lower:
+                return waf_type
+    # 2. 全文档指纹扫描（收窄后的技术标识符 + 完整文案）
     lower = content.lower()
     for waf_type, indicators in WAF_FINGERPRINTS:
         for ind in indicators:

tests/test_anti_bot.py

@@ -85,6 +85,7 @@ def test_detect_akamai_reference():
 # ----- Generic detection -----
 
 def test_detect_generic_captcha():
+    """v2.0.1：收窄为完整短语 'please complete the captcha'，避免正文误判。"""
     html = "<html><body>Please complete the CAPTCHA</body></html>"
     assert _detect_anti_bot(html) == "generic"
 
@@ -95,7 +96,15 @@ def test_detect_generic_verify_human():
 
 
 def test_detect_generic_access_denied():
-    html = "<html><body>Access Denied</body></html>"
+    """v2.0.1：裸 'access denied' 从全文档扫描移除（避免权限文章误判），
+    改由 <title> 标签检测覆盖。反爬页 title 常为 'Access Denied'。"""
+    html = "<html><head><title>Access Denied</title></head><body>Access Denied</body></html>"
+    assert _detect_anti_bot(html) == "generic"
+
+
+def test_detect_generic_access_denied_sucuri():
+    """v2.0.1：Sucuri WAF 的完整文案在全文档扫描中保留。"""
+    html = "<html><body>Access Denied - Sucuri Website Firewall</body></html>"
     assert _detect_anti_bot(html) == "generic"
 
 
@@ -130,7 +139,8 @@ def test_detect_anti_bot_beyond_2000_chars():
 def test_detect_anti_bot_large_page_end():
     """反爬关键词在文档末尾也能检测到。"""
     padding = "y" * 5000
-    html = f"<html><body>{padding}captcha</body></html>"
+    # v2.0.1：用完整短语而非裸 captcha，避免误判
+    html = f"<html><body>{padding}please complete the captcha</body></html>"
     assert _detect_anti_bot(html) == "generic"
 
 
@@ -149,14 +159,34 @@ def test_detect_anti_bot_empty_content():
 
 
 def test_detect_anti_bot_article_mentions_captcha_in_context():
-    """文章讨论 captcha 但不是反爬页（上下文判断的局限——接受误报）。
+    """v2.0.1 修复：文章讨论 captcha 但不是反爬页，不再误判。
 
-    注意：当前实现是关键词匹配，无法区分"讨论 captcha 的文章"和
+    v2.0.0 用裸 "captcha" 做全文匹配，导致"讨论 captcha 工作原理的文章"
-    "captcha 拦截页"。这是已知局限，测试记录此行为。
+    被误判为反爬页。v2.0.1 收窄为 "please complete the captcha" 等完整短语，
+    正常讨论 captcha 的文章不再触发。
     """
     html = "<html><body><p>This article explains how CAPTCHA works.</p></body></html>"
-    # 关键词匹配会误报为 generic
+    # v2.0.1：收窄后不再误判
-    assert _detect_anti_bot(html) == "generic"
+    assert _detect_anti_bot(html) is None
+
+
+def test_detect_anti_bot_article_mentions_cloudflare_in_context():
+    """v2.0.1 修复：文章引用 Cloudflare 文档链接，不再误判为 cloudflare WAF。
+
+    v2.0.0 用裸 "cloudflare" 做全文匹配，Wayback 归档的 DataCamp 文章
+    正文里有 cloudflare.com 链接，被误判为反爬页导致兜底失败。
+    """
+    html = ("<html><body><article>"
+            "<p>Learn more at <a href='https://www.cloudflare.com/learn/'>Cloudflare</a></p>"
+            "<p>Join our daily coding challenges!</p>"
+            "</article></body></html>")
+    assert _detect_anti_bot(html) is None
+
+
+def test_detect_anti_bot_article_mentions_challenge_in_context():
+    """v2.0.1 修复：文章含 'coding challenge' 等正常内容，不再误判。"""
+    html = "<html><body><p>Daily 5-minute coding challenges.</p></body></html>"
+    assert _detect_anti_bot(html) is None
 
 
 # ----- Priority: specialized WAF before generic -----
@@ -197,7 +227,8 @@ def test_waf_fingerprints_all_have_indicators():
 
 def test_is_blocked_page_delegates_to_detect():
     """_is_blocked_page 应委托给 _detect_anti_bot。"""
-    assert _is_blocked_page("<html>captcha</html>") is True
+    # v2.0.1：用完整反爬短语而非裸 captcha
+    assert _is_blocked_page("<html><body>please complete the captcha</body></html>") is True
     assert _is_blocked_page("<html>normal content</html>") is False
 
 

tests/test_fetch_auto.py

@@ -46,9 +46,12 @@ def test_blocked_page_checks_first_2000_chars():
     v2.0.0 changed this to full-document scanning because large anti-bot
     pages (e.g. Cloudflare challenges with big JS blobs) may place the
     telltale keyword beyond the 2000-char boundary.
+
+    v2.0.1: 用完整短语 'please complete the captcha' 替代裸 'captcha'，
+    避免正常内容误判。
     """
     padding = "x" * 2500
-    html = f"<html>{padding}captcha</html>"
+    html = f"<html>{padding}please complete the captcha</html>"
     # v2.0.0: now detected (was: not detected)
     assert _is_blocked_page(html)
 

tests/test_wayback_fallback.py

@@ -208,22 +208,67 @@ def test_fetch_page_normal_page_no_anti_bot():
 
 
 def test_fetch_page_anti_bot_triggers_wayback():
-    """被反爬拦截后也应尝试 Wayback（_should_try_fallback 防御性检查）。
+    """被反爬拦截后应尝试 Wayback 兜底（v2.0.1 修复）。
 
-    当前实现：主抓取成功返回反爬页内容 → result 非 None → 不触发兜底。
+    v2.0.0 bug：主抓取"成功"返回反爬页（HTTP 200 + Cloudflare 质询）时，
-    此测试记录此行为：反爬页被当作"成功抓取"返回，在 fetch_page 内部检测。
+    result 非 None 导致 _should_try_fallback 返回 False，Wayback 永不触发。
+    v2.0.1 修复：反爬检测提前到兜底判断之前，反爬阳性也触发 Wayback。
+
+    本测试 mock fetch_url 两次调用：
+      1. 主抓取 → 返回 Cloudflare 质询页
+      2. Wayback 兜底 → 返回正常归档页
     """
     cf_html = "<html><body>Just a moment... cf-ray: 123</body></html>"
+    wb_html = "<html><body><p>Archived article content.</p></body></html>"
+    cf_result = _make_result(content=cf_html, final_url="https://example.com")
+    wb_result = _make_result(content=wb_html,
+                             final_url="https://web.archive.org/web/2024/https://example.com")
     with patch("search.fetch_url",
-               return_value=_make_result(content=cf_html)):
+               side_effect=[cf_result, wb_result]) as mock_fetch:
         result = fetch_page("https://example.com", fallback_enabled=True)
-    # 反爬页被检测到，标记为 error + anti_bot_detected
+    # Wayback 兜底成功，反爬标记清除，status=ok
+    assert result["status"] == "ok"
+    assert result["anti_bot_detected"] is False
+    assert result["waf_type"] is None
+    assert result["fallback_used"] == "wayback"
+    assert "Archived article content." in result["text"]
+    # 确认 fetch_url 被调用两次：主抓取 + Wayback
+    assert mock_fetch.call_count == 2
+    # 第二次调用应该是 Wayback URL
+    assert "web.archive.org/web/2/" in mock_fetch.call_args_list[1][0][0]
+
+
+def test_fetch_page_anti_bot_wayback_also_blocked():
+    """主抓取反爬 + Wayback 也反爬/失败 → 最终返回 error。
+
+    Wayback 兜底返回 None（失败）时，保留主抓取的反爬检测结果。
+    """
+    cf_html = "<html><body>Just a moment... cf-ray: 123</body></html>"
+    cf_result = _make_result(content=cf_html, final_url="https://example.com")
+    with patch("search.fetch_url",
+               side_effect=[cf_result, RuntimeError("wayback timeout")]):
+        result = fetch_page("https://example.com", fallback_enabled=True)
+    # Wayback 失败，保留反爬 error
     assert result["status"] == "error"
     assert result["anti_bot_detected"] is True
-    # 但不会触发 Wayback（因为主抓取"成功"了，只是内容是反爬页）
+    assert result["waf_type"] == "cloudflare"
     assert result["fallback_used"] is None
 
 
+def test_fetch_page_anti_bot_fallback_disabled():
+    """--no-fallback 时反爬页直接返回 error，不尝试 Wayback。"""
+    cf_html = "<html><body>Just a moment... cf-ray: 123</body></html>"
+    with patch("search.fetch_url",
+               return_value=_make_result(content=cf_html)) as mock_fetch:
+        result = fetch_page("https://example.com", fallback_enabled=False)
+    assert result["status"] == "error"
+    assert result["anti_bot_detected"] is True
+    assert result["waf_type"] == "cloudflare"
+    assert result["fallback_used"] is None
+    # 只调用一次（主抓取），没有 Wayback
+    assert mock_fetch.call_count == 1
+
+
 def test_fetch_page_passes_referer():
     """referer 透传给 fetch_url。"""
     with patch("search.fetch_url", return_value=_make_result()) as mock_fetch: