crawlo 1.2.7__py3-none-any.whl → 1.2.9__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.

Potentially problematic release.

This version of crawlo might be problematic. Click here for more details.

Files changed (221) hide show
  1. crawlo/__init__.py +63 -61
  2. crawlo/__version__.py +1 -1
  3. crawlo/cli.py +75 -75
  4. crawlo/commands/__init__.py +14 -14
  5. crawlo/commands/check.py +594 -594
  6. crawlo/commands/genspider.py +151 -151
  7. crawlo/commands/help.py +138 -138
  8. crawlo/commands/list.py +155 -155
  9. crawlo/commands/run.py +323 -323
  10. crawlo/commands/startproject.py +436 -436
  11. crawlo/commands/stats.py +187 -187
  12. crawlo/commands/utils.py +186 -186
  13. crawlo/config.py +312 -312
  14. crawlo/config_validator.py +277 -251
  15. crawlo/core/__init__.py +2 -2
  16. crawlo/core/engine.py +366 -365
  17. crawlo/core/processor.py +40 -40
  18. crawlo/core/scheduler.py +256 -251
  19. crawlo/crawler.py +1103 -1100
  20. crawlo/data/__init__.py +5 -5
  21. crawlo/data/user_agents.py +194 -107
  22. crawlo/downloader/__init__.py +273 -266
  23. crawlo/downloader/aiohttp_downloader.py +226 -228
  24. crawlo/downloader/cffi_downloader.py +245 -256
  25. crawlo/downloader/httpx_downloader.py +259 -259
  26. crawlo/downloader/hybrid_downloader.py +212 -212
  27. crawlo/downloader/playwright_downloader.py +402 -402
  28. crawlo/downloader/selenium_downloader.py +472 -472
  29. crawlo/event.py +11 -11
  30. crawlo/exceptions.py +81 -81
  31. crawlo/extension/__init__.py +39 -39
  32. crawlo/extension/health_check.py +141 -141
  33. crawlo/extension/log_interval.py +57 -57
  34. crawlo/extension/log_stats.py +81 -81
  35. crawlo/extension/logging_extension.py +43 -43
  36. crawlo/extension/memory_monitor.py +104 -104
  37. crawlo/extension/performance_profiler.py +133 -133
  38. crawlo/extension/request_recorder.py +107 -107
  39. crawlo/filters/__init__.py +154 -154
  40. crawlo/filters/aioredis_filter.py +234 -234
  41. crawlo/filters/memory_filter.py +269 -269
  42. crawlo/items/__init__.py +23 -23
  43. crawlo/items/base.py +21 -21
  44. crawlo/items/fields.py +52 -52
  45. crawlo/items/items.py +104 -104
  46. crawlo/middleware/__init__.py +21 -21
  47. crawlo/middleware/default_header.py +132 -132
  48. crawlo/middleware/download_delay.py +104 -104
  49. crawlo/middleware/middleware_manager.py +136 -136
  50. crawlo/middleware/offsite.py +114 -114
  51. crawlo/middleware/proxy.py +386 -368
  52. crawlo/middleware/request_ignore.py +86 -86
  53. crawlo/middleware/response_code.py +163 -163
  54. crawlo/middleware/response_filter.py +136 -136
  55. crawlo/middleware/retry.py +124 -124
  56. crawlo/middleware/simple_proxy.py +65 -0
  57. crawlo/mode_manager.py +211 -211
  58. crawlo/network/__init__.py +21 -21
  59. crawlo/network/request.py +379 -338
  60. crawlo/network/response.py +359 -359
  61. crawlo/pipelines/__init__.py +21 -21
  62. crawlo/pipelines/bloom_dedup_pipeline.py +157 -157
  63. crawlo/pipelines/console_pipeline.py +39 -39
  64. crawlo/pipelines/csv_pipeline.py +316 -316
  65. crawlo/pipelines/database_dedup_pipeline.py +223 -223
  66. crawlo/pipelines/json_pipeline.py +218 -218
  67. crawlo/pipelines/memory_dedup_pipeline.py +115 -115
  68. crawlo/pipelines/mongo_pipeline.py +131 -131
  69. crawlo/pipelines/mysql_pipeline.py +317 -317
  70. crawlo/pipelines/pipeline_manager.py +62 -62
  71. crawlo/pipelines/redis_dedup_pipeline.py +167 -167
  72. crawlo/project.py +290 -315
  73. crawlo/queue/pqueue.py +37 -37
  74. crawlo/queue/queue_manager.py +379 -378
  75. crawlo/queue/redis_priority_queue.py +306 -306
  76. crawlo/settings/__init__.py +7 -7
  77. crawlo/settings/default_settings.py +216 -220
  78. crawlo/settings/setting_manager.py +163 -122
  79. crawlo/spider/__init__.py +639 -639
  80. crawlo/stats_collector.py +59 -59
  81. crawlo/subscriber.py +129 -129
  82. crawlo/task_manager.py +30 -30
  83. crawlo/templates/crawlo.cfg.tmpl +10 -10
  84. crawlo/templates/project/__init__.py.tmpl +3 -3
  85. crawlo/templates/project/items.py.tmpl +17 -17
  86. crawlo/templates/project/middlewares.py.tmpl +118 -118
  87. crawlo/templates/project/pipelines.py.tmpl +96 -96
  88. crawlo/templates/project/settings.py.tmpl +261 -288
  89. crawlo/templates/project/settings_distributed.py.tmpl +174 -157
  90. crawlo/templates/project/settings_gentle.py.tmpl +95 -100
  91. crawlo/templates/project/settings_high_performance.py.tmpl +125 -134
  92. crawlo/templates/project/settings_minimal.py.tmpl +30 -0
  93. crawlo/templates/project/settings_simple.py.tmpl +96 -98
  94. crawlo/templates/project/spiders/__init__.py.tmpl +5 -5
  95. crawlo/templates/run.py.tmpl +47 -45
  96. crawlo/templates/spider/spider.py.tmpl +143 -143
  97. crawlo/tools/__init__.py +200 -182
  98. crawlo/tools/anti_crawler.py +268 -268
  99. crawlo/tools/authenticated_proxy.py +240 -240
  100. crawlo/{cleaners → tools}/data_formatter.py +225 -225
  101. crawlo/tools/data_validator.py +180 -180
  102. crawlo/tools/date_tools.py +290 -36
  103. crawlo/tools/distributed_coordinator.py +388 -387
  104. crawlo/{cleaners → tools}/encoding_converter.py +127 -126
  105. crawlo/tools/request_tools.py +83 -0
  106. crawlo/tools/retry_mechanism.py +224 -221
  107. crawlo/tools/scenario_adapter.py +262 -262
  108. crawlo/{cleaners → tools}/text_cleaner.py +232 -232
  109. crawlo/utils/__init__.py +35 -35
  110. crawlo/utils/batch_processor.py +259 -259
  111. crawlo/utils/controlled_spider_mixin.py +439 -439
  112. crawlo/utils/db_helper.py +343 -343
  113. crawlo/utils/enhanced_error_handler.py +356 -356
  114. crawlo/utils/env_config.py +142 -142
  115. crawlo/utils/error_handler.py +123 -123
  116. crawlo/utils/func_tools.py +82 -82
  117. crawlo/utils/large_scale_config.py +286 -286
  118. crawlo/utils/large_scale_helper.py +344 -344
  119. crawlo/utils/log.py +187 -128
  120. crawlo/utils/performance_monitor.py +285 -285
  121. crawlo/utils/queue_helper.py +175 -175
  122. crawlo/utils/redis_connection_pool.py +351 -351
  123. crawlo/utils/redis_key_validator.py +198 -198
  124. crawlo/utils/request.py +267 -267
  125. crawlo/utils/request_serializer.py +218 -218
  126. crawlo/utils/spider_loader.py +61 -61
  127. crawlo/utils/system.py +11 -11
  128. crawlo/utils/tools.py +4 -4
  129. crawlo/utils/url.py +39 -39
  130. {crawlo-1.2.7.dist-info → crawlo-1.2.9.dist-info}/METADATA +1011 -764
  131. crawlo-1.2.9.dist-info/RECORD +219 -0
  132. examples/__init__.py +7 -7
  133. tests/DOUBLE_CRAWLO_PREFIX_FIX_REPORT.md +81 -81
  134. tests/__init__.py +7 -7
  135. tests/advanced_tools_example.py +275 -275
  136. tests/authenticated_proxy_example.py +107 -237
  137. tests/cleaners_example.py +160 -160
  138. tests/config_validation_demo.py +143 -103
  139. tests/controlled_spider_example.py +205 -205
  140. tests/date_tools_example.py +180 -180
  141. tests/debug_pipelines.py +67 -0
  142. tests/dynamic_loading_example.py +523 -523
  143. tests/dynamic_loading_test.py +104 -104
  144. tests/env_config_example.py +133 -133
  145. tests/error_handling_example.py +171 -171
  146. tests/redis_key_validation_demo.py +130 -130
  147. tests/request_params_example.py +151 -0
  148. tests/response_improvements_example.py +144 -144
  149. tests/test_advanced_tools.py +148 -148
  150. tests/test_all_redis_key_configs.py +145 -145
  151. tests/test_authenticated_proxy.py +141 -141
  152. tests/test_cleaners.py +54 -54
  153. tests/test_comprehensive.py +146 -146
  154. tests/test_config_consistency.py +80 -80
  155. tests/test_config_merge.py +153 -0
  156. tests/test_config_validator.py +182 -193
  157. tests/test_crawlo_proxy_integration.py +109 -173
  158. tests/test_date_tools.py +123 -123
  159. tests/test_default_header_middleware.py +158 -158
  160. tests/test_distributed.py +65 -0
  161. tests/test_double_crawlo_fix.py +207 -207
  162. tests/test_double_crawlo_fix_simple.py +124 -124
  163. tests/test_download_delay_middleware.py +221 -221
  164. tests/test_downloader_proxy_compatibility.py +268 -268
  165. tests/test_dynamic_downloaders_proxy.py +124 -124
  166. tests/test_dynamic_proxy.py +92 -92
  167. tests/test_dynamic_proxy_config.py +146 -146
  168. tests/test_dynamic_proxy_real.py +109 -109
  169. tests/test_edge_cases.py +303 -303
  170. tests/test_enhanced_error_handler.py +270 -270
  171. tests/test_env_config.py +121 -121
  172. tests/test_error_handler_compatibility.py +112 -112
  173. tests/test_final_validation.py +153 -153
  174. tests/test_framework_env_usage.py +103 -103
  175. tests/test_integration.py +169 -357
  176. tests/test_item_dedup_redis_key.py +122 -122
  177. tests/test_mode_consistency.py +51 -51
  178. tests/test_offsite_middleware.py +221 -221
  179. tests/test_parsel.py +29 -29
  180. tests/test_performance.py +327 -327
  181. tests/test_proxy_api.py +264 -264
  182. tests/test_proxy_health_check.py +32 -32
  183. tests/test_proxy_middleware.py +121 -121
  184. tests/test_proxy_middleware_enhanced.py +216 -216
  185. tests/test_proxy_middleware_integration.py +136 -136
  186. tests/test_proxy_middleware_refactored.py +185 -0
  187. tests/test_proxy_providers.py +56 -56
  188. tests/test_proxy_stats.py +19 -19
  189. tests/test_proxy_strategies.py +59 -59
  190. tests/test_queue_manager_double_crawlo.py +173 -173
  191. tests/test_queue_manager_redis_key.py +176 -176
  192. tests/test_random_user_agent.py +73 -0
  193. tests/test_real_scenario_proxy.py +195 -195
  194. tests/test_redis_config.py +28 -28
  195. tests/test_redis_connection_pool.py +294 -294
  196. tests/test_redis_key_naming.py +181 -181
  197. tests/test_redis_key_validator.py +123 -123
  198. tests/test_redis_queue.py +224 -224
  199. tests/test_request_ignore_middleware.py +182 -182
  200. tests/test_request_params.py +112 -0
  201. tests/test_request_serialization.py +70 -70
  202. tests/test_response_code_middleware.py +349 -349
  203. tests/test_response_filter_middleware.py +427 -427
  204. tests/test_response_improvements.py +152 -152
  205. tests/test_retry_middleware.py +241 -241
  206. tests/test_scheduler.py +252 -252
  207. tests/test_scheduler_config_update.py +133 -133
  208. tests/test_simple_response.py +61 -61
  209. tests/test_telecom_spider_redis_key.py +205 -205
  210. tests/test_template_content.py +87 -87
  211. tests/test_template_redis_key.py +134 -134
  212. tests/test_tools.py +159 -153
  213. tests/test_user_agents.py +97 -0
  214. tests/tools_example.py +260 -257
  215. tests/verify_distributed.py +117 -0
  216. crawlo/cleaners/__init__.py +0 -61
  217. crawlo/utils/date_tools.py +0 -290
  218. crawlo-1.2.7.dist-info/RECORD +0 -209
  219. {crawlo-1.2.7.dist-info → crawlo-1.2.9.dist-info}/WHEEL +0 -0
  220. {crawlo-1.2.7.dist-info → crawlo-1.2.9.dist-info}/entry_points.txt +0 -0
  221. {crawlo-1.2.7.dist-info → crawlo-1.2.9.dist-info}/top_level.txt +0 -0
crawlo/data/__init__.py CHANGED
@@ -1,6 +1,6 @@
1
- #!/usr/bin/env python
2
- # -*- coding: utf-8 -*-
3
- """
4
- Crawlo 数据模块
5
- 包含用户代理、字典等数据文件
1
+ #!/usr/bin/env python
2
+ # -*- coding: utf-8 -*-
3
+ """
4
+ Crawlo 数据模块
5
+ 包含用户代理、字典等数据文件
6
6
  """
crawlo/data/user_agents.py CHANGED
@@ -1,108 +1,195 @@
1
- #!/usr/bin/env python
2
- # -*- coding: utf-8 -*-
3
- """
4
- User-Agent列表
5
- 包含各种设备和浏览器的User-Agent字符串，用于爬虫伪装
6
- """
7
-
8
- # 桌面浏览器User-Agent
9
- DESKTOP_USER_AGENTS = [
10
- # Chrome
11
- "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
12
- "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
13
- "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
14
- "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
15
-
16
- # Firefox
17
- "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:125.0) Gecko/20100101 Firefox/125.0",
18
- "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:125.0) Gecko/20100101 Firefox/125.0",
19
- "Mozilla/5.0 (X11; Linux i686; rv:125.0) Gecko/20100101 Firefox/125.0",
20
- "Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:125.0) Gecko/20100101 Firefox/125.0",
21
-
22
- # Safari
23
- "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.4 Safari/605.1.15",
24
- "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/16.6 Safari/605.1.15",
25
-
26
- # Edge
27
- "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36 Edg/124.0.2478.80",
28
- "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36 Edg/123.0.2420.40",
29
-
30
- # Internet Explorer
31
- "Mozilla/5.0 (Windows NT 10.0; WOW64; Trident/7.0; rv:11.0) like Gecko",
32
- ]
33
-
34
- # 移动设备User-Agent
35
- MOBILE_USER_AGENTS = [
36
- # iPhone
37
- "Mozilla/5.0 (iPhone; CPU iPhone OS 17_4 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.4 Mobile/15E148 Safari/604.1",
38
- "Mozilla/5.0 (iPhone; CPU iPhone OS 16_6 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/16.6 Mobile/15E148 Safari/604.1",
39
-
40
- # iPad
41
- "Mozilla/5.0 (iPad; CPU OS 17_4 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.4 Mobile/15E148 Safari/604.1",
42
- "Mozilla/5.0 (iPad; CPU OS 16_6 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/16.6 Mobile/15E148 Safari/604.1",
43
-
44
- # Android
45
- "Mozilla/5.0 (Linux; Android 14; SM-S911B) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.6367.118 Mobile Safari/537.36",
46
- "Mozilla/5.0 (Linux; Android 13; SM-S908B) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.6367.118 Mobile Safari/537.36",
47
- "Mozilla/5.0 (Linux; Android 12; Pixel 6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.6367.118 Mobile Safari/537.36",
48
- "Mozilla/5.0 (Linux; Android 11; Pixel 5) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.6367.118 Mobile Safari/537.36",
49
-
50
- # Android平板
51
- "Mozilla/5.0 (Linux; Android 13; SM-X906C Build/TP1A.220624.014; wv) AppleWebKit/537.36 (KHTML, like Gecko) Version/4.0 Chrome/124.0.6367.118 Safari/537.36",
52
- ]
53
-
54
- # 爬虫/机器人User-Agent (用于测试)
55
- BOT_USER_AGENTS = [
56
- "Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)",
57
- "Mozilla/5.0 (compatible; Bingbot/2.0; +http://www.bing.com/bingbot.htm)",
58
- "Mozilla/5.0 (compatible; Baiduspider/2.0; +http://www.baidu.com/search/spider.html)",
59
- "Mozilla/5.0 (compatible; Yahoo! Slurp; http://help.yahoo.com/help/us/ysearch/slurp)",
60
- ]
61
-
62
- # 所有User-Agent的组合列表
63
- ALL_USER_AGENTS = DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
64
-
65
- # 按设备类型分类的User-Agent字典
66
- USER_AGENTS_BY_TYPE = {
67
- "desktop": DESKTOP_USER_AGENTS,
68
- "mobile": MOBILE_USER_AGENTS,
69
- "bot": BOT_USER_AGENTS,
70
- "all": ALL_USER_AGENTS,
71
- }
72
-
73
- def get_user_agents(device_type="all"):
74
- """
75
- 获取指定类型的User-Agent列表
76
-
77
- Args:
78
- device_type (str): 设备类型，可选值: "desktop", "mobile", "bot", "all"
79
-
80
- Returns:
81
- list: User-Agent字符串列表
82
- """
83
- return USER_AGENTS_BY_TYPE.get(device_type, ALL_USER_AGENTS)
84
-
85
- def get_random_user_agent(device_type="all"):
86
- """
87
- 获取随机User-Agent
88
-
89
- Args:
90
- device_type (str): 设备类型，可选值: "desktop", "mobile", "bot", "all"
91
-
92
- Returns:
93
- str: 随机User-Agent字符串
94
- """
95
- import random
96
- user_agents = get_user_agents(device_type)
97
- return random.choice(user_agents) if user_agents else ""
98
-
99
- # 导出常用的User-Agent列表
100
- __all__ = [
101
- "DESKTOP_USER_AGENTS",
102
- "MOBILE_USER_AGENTS",
103
- "BOT_USER_AGENTS",
104
- "ALL_USER_AGENTS",
105
- "USER_AGENTS_BY_TYPE",
106
- "get_user_agents",
107
- "get_random_user_agent"
1
+ #!/usr/bin/env python
2
+ # -*- coding: utf-8 -*-
3
+ """
4
+ User-Agent列表
5
+ 包含各种设备和浏览器的User-Agent字符串，用于爬虫伪装
6
+ """
7
+
8
+ # 桌面浏览器User-Agent
9
+ DESKTOP_USER_AGENTS = [
10
+ # Chrome (最新版本)
11
+ "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36",
12
+ "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36",
13
+ "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36",
14
+ "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36",
15
+
16
+ # Chrome (较新版本)
17
+ "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36",
18
+ "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36",
19
+ "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36",
20
+ "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36",
21
+
22
+ # Chrome (常用版本)
23
+ "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36",
24
+ "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36",
25
+ "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36",
26
+ "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36",
27
+
28
+ # Firefox (最新版本)
29
+ "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:136.0) Gecko/20100101 Firefox/136.0",
30
+ "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:136.0) Gecko/20100101 Firefox/136.0",
31
+ "Mozilla/5.0 (X11; Linux i686; rv:136.0) Gecko/20100101 Firefox/136.0",
32
+ "Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:136.0) Gecko/20100101 Firefox/136.0",
33
+
34
+ # Firefox (较新版本)
35
+ "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:135.0) Gecko/20100101 Firefox/135.0",
36
+ "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:135.0) Gecko/20100101 Firefox/135.0",
37
+ "Mozilla/5.0 (X11; Linux i686; rv:135.0) Gecko/20100101 Firefox/135.0",
38
+ "Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:135.0) Gecko/20100101 Firefox/135.0",
39
+
40
+ # Safari (最新版本)
41
+ "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/18.4 Safari/605.1.15",
42
+ "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.6 Safari/605.1.15",
43
+
44
+ # Safari (较新版本)
45
+ "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.5 Safari/605.1.15",
46
+ "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.4 Safari/605.1.15",
47
+
48
+ # Edge (最新版本)
49
+ "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36 Edg/136.0.2917.92",
50
+ "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36 Edg/135.0.2916.12",
51
+
52
+ # Edge (较新版本)
53
+ "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36 Edg/135.0.2915.72",
54
+ "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36 Edg/134.0.2914.89",
55
+
56
+ # Opera
57
+ "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36 OPR/122.0.0.0",
58
+ "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36 OPR/122.0.0.0",
59
+ ]
60
+
61
+ # 移动设备User-Agent
62
+ MOBILE_USER_AGENTS = [
63
+ # iPhone (最新版本)
64
+ "Mozilla/5.0 (iPhone; CPU iPhone OS 18_4 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/18.4 Mobile/15E148 Safari/604.1",
65
+ "Mozilla/5.0 (iPhone; CPU iPhone OS 17_6 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.6 Mobile/15E148 Safari/604.1",
66
+
67
+ # iPhone (较新版本)
68
+ "Mozilla/5.0 (iPhone; CPU iPhone OS 17_5 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.5 Mobile/15E148 Safari/604.1",
69
+ "Mozilla/5.0 (iPhone; CPU iPhone OS 17_4 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.4 Mobile/15E148 Safari/604.1",
70
+
71
+ # iPad (最新版本)
72
+ "Mozilla/5.0 (iPad; CPU OS 18_4 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/18.4 Mobile/15E148 Safari/604.1",
73
+ "Mozilla/5.0 (iPad; CPU OS 17_6 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.6 Mobile/15E148 Safari/604.1",
74
+
75
+ # iPad (较新版本)
76
+ "Mozilla/5.0 (iPad; CPU OS 17_5 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.5 Mobile/15E148 Safari/604.1",
77
+ "Mozilla/5.0 (iPad; CPU OS 17_4 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.4 Mobile/15E148 Safari/604.1",
78
+
79
+ # Android (最新版本)
80
+ "Mozilla/5.0 (Linux; Android 15; SM-S921B) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
81
+ "Mozilla/5.0 (Linux; Android 14; SM-S911B) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
82
+ "Mozilla/5.0 (Linux; Android 14; Pixel 8) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
83
+
84
+ # Android (较新版本)
85
+ "Mozilla/5.0 (Linux; Android 13; SM-S908B) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
86
+ "Mozilla/5.0 (Linux; Android 13; Pixel 7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
87
+ "Mozilla/5.0 (Linux; Android 12; Pixel 6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
88
+
89
+ # Android平板
90
+ "Mozilla/5.0 (Linux; Android 14; SM-X916C Build/UP1A.231005.007; wv) AppleWebKit/537.36 (KHTML, like Gecko) Version/4.0 Chrome/136.0.6367.118 Safari/537.36",
91
+ "Mozilla/5.0 (Linux; Android 13; SM-X906C Build/TP1A.220624.014; wv) AppleWebKit/537.36 (KHTML, like Gecko) Version/4.0 Chrome/136.0.6367.118 Safari/537.36",
92
+
93
+ # 其他移动浏览器
94
+ "Mozilla/5.0 (Linux; Android 14; Pixel 8) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Mobile Safari/537.36 EdgA/136.0.2917.92",
95
+ "Mozilla/5.0 (iPhone; CPU iPhone OS 17_6 like Mac OS X) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/17.6 Mobile/15E148 Safari/604.1 EdgiOS/136.2917.92",
96
+ ]
97
+
98
+ # 爬虫/机器人User-Agent (用于测试)
99
+ BOT_USER_AGENTS = [
100
+ "Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)",
101
+ "Mozilla/5.0 (compatible; Bingbot/2.0; +http://www.bing.com/bingbot.htm)",
102
+ "Mozilla/5.0 (compatible; Baiduspider/2.0; +http://www.baidu.com/search/spider.html)",
103
+ "Mozilla/5.0 (compatible; Yahoo! Slurp; http://help.yahoo.com/help/us/ysearch/slurp)",
104
+ "Mozilla/5.0 (compatible; DuckDuckBot/1.1; +http://duckduckgo.com/duckduckbot.html)",
105
+ "Mozilla/5.0 (compatible; YandexBot/3.0; +http://yandex.com/bots)",
106
+ "Mozilla/5.0 (compatible; facebookexternalhit/1.1; +http://www.facebook.com/externalhit_uatext.php)",
107
+ "Mozilla/5.0 (compatible; Twitterbot/1.0)",
108
+ ]
109
+
110
+ # 按浏览器分类的User-Agent
111
+ CHROME_USER_AGENTS = [
112
+ ua for ua in DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
113
+ if "Chrome" in ua and "Edg" not in ua and "OPR" not in ua
114
+ ]
115
+
116
+ FIREFOX_USER_AGENTS = [
117
+ ua for ua in DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
118
+ if "Firefox" in ua
119
+ ]
120
+
121
+ SAFARI_USER_AGENTS = [
122
+ ua for ua in DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
123
+ if "Safari" in ua and "Chrome" not in ua
124
+ ]
125
+
126
+ EDGE_USER_AGENTS = [
127
+ ua for ua in DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
128
+ if "Edg" in ua
129
+ ]
130
+
131
+ OPERA_USER_AGENTS = [
132
+ ua for ua in DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
133
+ if "OPR" in ua
134
+ ]
135
+
136
+ # 所有User-Agent的组合列表
137
+ ALL_USER_AGENTS = DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
138
+
139
+ # 按设备类型分类的User-Agent字典
140
+ USER_AGENTS_BY_TYPE = {
141
+ "desktop": DESKTOP_USER_AGENTS,
142
+ "mobile": MOBILE_USER_AGENTS,
143
+ "bot": BOT_USER_AGENTS,
144
+ "all": ALL_USER_AGENTS,
145
+ "chrome": CHROME_USER_AGENTS,
146
+ "firefox": FIREFOX_USER_AGENTS,
147
+ "safari": SAFARI_USER_AGENTS,
148
+ "edge": EDGE_USER_AGENTS,
149
+ "opera": OPERA_USER_AGENTS,
150
+ }
151
+
152
+
153
+ def get_user_agents(device_type="all"):
154
+ """
155
+ 获取指定类型的User-Agent列表
156
+
157
+ Args:
158
+ device_type (str): 设备类型，可选值: "desktop", "mobile", "bot", "all", "chrome", "firefox", "safari", "edge", "opera"
159
+
160
+ Returns:
161
+ list: User-Agent字符串列表
162
+ """
163
+ return USER_AGENTS_BY_TYPE.get(device_type, ALL_USER_AGENTS)
164
+
165
+
166
+ def get_random_user_agent(device_type="all"):
167
+ """
168
+ 获取随机User-Agent
169
+
170
+ Args:
171
+ device_type (str): 设备类型，可选值: "desktop", "mobile", "bot", "all", "chrome", "firefox", "safari", "edge", "opera"
172
+
173
+ Returns:
174
+ str: 随机User-Agent字符串
175
+ """
176
+ import random
177
+ user_agents = get_user_agents(device_type)
178
+ return random.choice(user_agents) if user_agents else ""
179
+
180
+
181
+ # 导出常用的User-Agent列表
182
+ __all__ = [
183
+ "DESKTOP_USER_AGENTS",
184
+ "MOBILE_USER_AGENTS",
185
+ "BOT_USER_AGENTS",
186
+ "CHROME_USER_AGENTS",
187
+ "FIREFOX_USER_AGENTS",
188
+ "SAFARI_USER_AGENTS",
189
+ "EDGE_USER_AGENTS",
190
+ "OPERA_USER_AGENTS",
191
+ "ALL_USER_AGENTS",
192
+ "USER_AGENTS_BY_TYPE",
193
+ "get_user_agents",
194
+ "get_random_user_agent"
108
195
  ]