PyPI - crawlo - Versions diffs - 1.3.3__py3-none-any.whl → 1.3.4__py3-none-any.whl - Mend

crawlo 1.3.3py3-none-any.whl → 1.3.4py3-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 (279) hide show

crawlo/__init__.py +87 -63
crawlo/__version__.py +1 -1
crawlo/cli.py +75 -75
crawlo/commands/__init__.py +14 -14
crawlo/commands/check.py +594 -594
crawlo/commands/genspider.py +151 -151
crawlo/commands/help.py +138 -138
crawlo/commands/list.py +155 -155
crawlo/commands/run.py +341 -323
crawlo/commands/startproject.py +436 -436
crawlo/commands/stats.py +187 -187
crawlo/commands/utils.py +196 -196
crawlo/config.py +312 -312
crawlo/config_validator.py +277 -277
crawlo/core/__init__.py +46 -2
crawlo/core/engine.py +439 -365
crawlo/core/processor.py +40 -40
crawlo/core/scheduler.py +257 -256
crawlo/crawler.py +639 -1167
crawlo/data/__init__.py +5 -5
crawlo/data/user_agents.py +194 -194
crawlo/downloader/__init__.py +273 -273
crawlo/downloader/aiohttp_downloader.py +228 -226
crawlo/downloader/cffi_downloader.py +245 -245
crawlo/downloader/httpx_downloader.py +259 -259
crawlo/downloader/hybrid_downloader.py +212 -212
crawlo/downloader/playwright_downloader.py +402 -402
crawlo/downloader/selenium_downloader.py +472 -472
crawlo/event.py +11 -11
crawlo/exceptions.py +81 -81
crawlo/extension/__init__.py +39 -39
crawlo/extension/health_check.py +141 -141
crawlo/extension/log_interval.py +57 -57
crawlo/extension/log_stats.py +81 -81
crawlo/extension/logging_extension.py +61 -52
crawlo/extension/memory_monitor.py +104 -104
crawlo/extension/performance_profiler.py +133 -133
crawlo/extension/request_recorder.py +107 -107
crawlo/factories/__init__.py +28 -0
crawlo/factories/base.py +69 -0
crawlo/factories/crawler.py +104 -0
crawlo/factories/registry.py +85 -0
crawlo/filters/__init__.py +154 -154
crawlo/filters/aioredis_filter.py +257 -234
crawlo/filters/memory_filter.py +269 -269
crawlo/framework.py +292 -0
crawlo/initialization/__init__.py +40 -0
crawlo/initialization/built_in.py +426 -0
crawlo/initialization/context.py +142 -0
crawlo/initialization/core.py +194 -0
crawlo/initialization/phases.py +149 -0
crawlo/initialization/registry.py +146 -0
crawlo/items/__init__.py +23 -23
crawlo/items/base.py +23 -22
crawlo/items/fields.py +52 -52
crawlo/items/items.py +104 -104
crawlo/logging/__init__.py +38 -0
crawlo/logging/config.py +97 -0
crawlo/logging/factory.py +129 -0
crawlo/logging/manager.py +112 -0
crawlo/middleware/__init__.py +21 -21
crawlo/middleware/default_header.py +132 -132
crawlo/middleware/download_delay.py +104 -104
crawlo/middleware/middleware_manager.py +135 -135
crawlo/middleware/offsite.py +123 -123
crawlo/middleware/proxy.py +386 -386
crawlo/middleware/request_ignore.py +86 -86
crawlo/middleware/response_code.py +163 -163
crawlo/middleware/response_filter.py +136 -136
crawlo/middleware/retry.py +124 -124
crawlo/middleware/simple_proxy.py +65 -65
crawlo/mode_manager.py +212 -187
crawlo/network/__init__.py +21 -21
crawlo/network/request.py +379 -379
crawlo/network/response.py +359 -359
crawlo/pipelines/__init__.py +21 -21
crawlo/pipelines/bloom_dedup_pipeline.py +156 -156
crawlo/pipelines/console_pipeline.py +39 -39
crawlo/pipelines/csv_pipeline.py +316 -316
crawlo/pipelines/database_dedup_pipeline.py +222 -222
crawlo/pipelines/json_pipeline.py +218 -218
crawlo/pipelines/memory_dedup_pipeline.py +115 -115
crawlo/pipelines/mongo_pipeline.py +131 -131
crawlo/pipelines/mysql_pipeline.py +318 -318
crawlo/pipelines/pipeline_manager.py +76 -75
crawlo/pipelines/redis_dedup_pipeline.py +166 -166
crawlo/project.py +327 -325
crawlo/queue/pqueue.py +43 -37
crawlo/queue/queue_manager.py +503 -379
crawlo/queue/redis_priority_queue.py +326 -306
crawlo/settings/__init__.py +7 -7
crawlo/settings/default_settings.py +321 -225
crawlo/settings/setting_manager.py +214 -198
crawlo/spider/__init__.py +657 -639
crawlo/stats_collector.py +73 -59
crawlo/subscriber.py +129 -129
crawlo/task_manager.py +139 -30
crawlo/templates/crawlo.cfg.tmpl +10 -10
crawlo/templates/project/__init__.py.tmpl +3 -3
crawlo/templates/project/items.py.tmpl +17 -17
crawlo/templates/project/middlewares.py.tmpl +118 -118
crawlo/templates/project/pipelines.py.tmpl +96 -96
crawlo/templates/project/settings.py.tmpl +168 -267
crawlo/templates/project/settings_distributed.py.tmpl +167 -180
crawlo/templates/project/settings_gentle.py.tmpl +167 -61
crawlo/templates/project/settings_high_performance.py.tmpl +168 -131
crawlo/templates/project/settings_minimal.py.tmpl +66 -35
crawlo/templates/project/settings_simple.py.tmpl +165 -102
crawlo/templates/project/spiders/__init__.py.tmpl +10 -6
crawlo/templates/run.py.tmpl +34 -38
crawlo/templates/spider/spider.py.tmpl +143 -143
crawlo/templates/spiders_init.py.tmpl +10 -0
crawlo/tools/__init__.py +200 -200
crawlo/tools/anti_crawler.py +268 -268
crawlo/tools/authenticated_proxy.py +240 -240
crawlo/tools/data_formatter.py +225 -225
crawlo/tools/data_validator.py +180 -180
crawlo/tools/date_tools.py +289 -289
crawlo/tools/distributed_coordinator.py +388 -388
crawlo/tools/encoding_converter.py +127 -127
crawlo/tools/network_diagnostic.py +365 -0
crawlo/tools/request_tools.py +82 -82
crawlo/tools/retry_mechanism.py +224 -224
crawlo/tools/scenario_adapter.py +262 -262
crawlo/tools/text_cleaner.py +232 -232
crawlo/utils/__init__.py +34 -34
crawlo/utils/batch_processor.py +259 -259
crawlo/utils/class_loader.py +26 -0
crawlo/utils/controlled_spider_mixin.py +439 -439
crawlo/utils/db_helper.py +343 -343
crawlo/utils/enhanced_error_handler.py +356 -356
crawlo/utils/env_config.py +142 -142
crawlo/utils/error_handler.py +165 -124
crawlo/utils/func_tools.py +82 -82
crawlo/utils/large_scale_config.py +286 -286
crawlo/utils/large_scale_helper.py +344 -344
crawlo/utils/log.py +44 -200
crawlo/utils/performance_monitor.py +285 -285
crawlo/utils/queue_helper.py +175 -175
crawlo/utils/redis_connection_pool.py +388 -351
crawlo/utils/redis_key_validator.py +198 -198
crawlo/utils/request.py +267 -267
crawlo/utils/request_serializer.py +225 -218
crawlo/utils/spider_loader.py +61 -61
crawlo/utils/system.py +11 -11
crawlo/utils/tools.py +4 -4
crawlo/utils/url.py +39 -39
{crawlo-1.3.3.dist-info → crawlo-1.3.4.dist-info}/METADATA +1126 -1020
crawlo-1.3.4.dist-info/RECORD +278 -0
examples/__init__.py +7 -7
tests/__init__.py +7 -7
tests/advanced_tools_example.py +275 -275
tests/authenticated_proxy_example.py +107 -107
tests/baidu_performance_test.py +109 -0
tests/baidu_test.py +60 -0
tests/cleaners_example.py +160 -160
tests/comprehensive_framework_test.py +213 -0
tests/comprehensive_test.py +82 -0
tests/comprehensive_testing_summary.md +187 -0
tests/config_validation_demo.py +142 -142
tests/controlled_spider_example.py +205 -205
tests/date_tools_example.py +180 -180
tests/debug_configure.py +70 -0
tests/debug_framework_logger.py +85 -0
tests/debug_log_levels.py +64 -0
tests/debug_pipelines.py +66 -66
tests/distributed_test.py +67 -0
tests/distributed_test_debug.py +77 -0
tests/dynamic_loading_example.py +523 -523
tests/dynamic_loading_test.py +104 -104
tests/env_config_example.py +133 -133
tests/error_handling_example.py +171 -171
tests/final_command_test_report.md +0 -0
tests/final_comprehensive_test.py +152 -0
tests/final_validation_test.py +183 -0
tests/framework_performance_test.py +203 -0
tests/optimized_performance_test.py +212 -0
tests/performance_comparison.py +246 -0
tests/queue_blocking_test.py +114 -0
tests/queue_test.py +90 -0
tests/redis_key_validation_demo.py +130 -130
tests/request_params_example.py +150 -150
tests/response_improvements_example.py +144 -144
tests/scrapy_comparison/ofweek_scrapy.py +139 -0
tests/scrapy_comparison/scrapy_test.py +134 -0
tests/simple_command_test.py +120 -0
tests/simple_crawlo_test.py +128 -0
tests/simple_log_test.py +58 -0
tests/simple_optimization_test.py +129 -0
tests/simple_spider_test.py +50 -0
tests/simple_test.py +48 -0
tests/test_advanced_tools.py +148 -148
tests/test_all_commands.py +231 -0
tests/test_all_redis_key_configs.py +145 -145
tests/test_authenticated_proxy.py +141 -141
tests/test_batch_processor.py +179 -0
tests/test_cleaners.py +54 -54
tests/test_component_factory.py +175 -0
tests/test_comprehensive.py +146 -146
tests/test_config_consistency.py +80 -80
tests/test_config_merge.py +152 -152
tests/test_config_validator.py +182 -182
tests/test_controlled_spider_mixin.py +80 -0
tests/test_crawlo_proxy_integration.py +108 -108
tests/test_date_tools.py +123 -123
tests/test_default_header_middleware.py +158 -158
tests/test_distributed.py +65 -65
tests/test_double_crawlo_fix.py +207 -207
tests/test_double_crawlo_fix_simple.py +124 -124
tests/test_download_delay_middleware.py +221 -221
tests/test_downloader_proxy_compatibility.py +268 -268
tests/test_dynamic_downloaders_proxy.py +124 -124
tests/test_dynamic_proxy.py +92 -92
tests/test_dynamic_proxy_config.py +146 -146
tests/test_dynamic_proxy_real.py +109 -109
tests/test_edge_cases.py +303 -303
tests/test_enhanced_error_handler.py +270 -270
tests/test_enhanced_error_handler_comprehensive.py +246 -0
tests/test_env_config.py +121 -121
tests/test_error_handler_compatibility.py +112 -112
tests/test_factories.py +253 -0
tests/test_final_validation.py +153 -153
tests/test_framework_env_usage.py +103 -103
tests/test_framework_logger.py +67 -0
tests/test_framework_startup.py +65 -0
tests/test_integration.py +169 -169
tests/test_item_dedup_redis_key.py +122 -122
tests/test_large_scale_config.py +113 -0
tests/test_large_scale_helper.py +236 -0
tests/test_mode_change.py +73 -0
tests/test_mode_consistency.py +51 -51
tests/test_offsite_middleware.py +221 -221
tests/test_parsel.py +29 -29
tests/test_performance.py +327 -327
tests/test_performance_monitor.py +116 -0
tests/test_proxy_api.py +264 -264
tests/test_proxy_health_check.py +32 -32
tests/test_proxy_middleware.py +121 -121
tests/test_proxy_middleware_enhanced.py +216 -216
tests/test_proxy_middleware_integration.py +136 -136
tests/test_proxy_middleware_refactored.py +184 -184
tests/test_proxy_providers.py +56 -56
tests/test_proxy_stats.py +19 -19
tests/test_proxy_strategies.py +59 -59
tests/test_queue_empty_check.py +42 -0
tests/test_queue_manager_double_crawlo.py +173 -173
tests/test_queue_manager_redis_key.py +176 -176
tests/test_random_user_agent.py +72 -72
tests/test_real_scenario_proxy.py +195 -195
tests/test_redis_config.py +28 -28
tests/test_redis_connection_pool.py +294 -294
tests/test_redis_key_naming.py +181 -181
tests/test_redis_key_validator.py +123 -123
tests/test_redis_queue.py +224 -224
tests/test_request_ignore_middleware.py +182 -182
tests/test_request_params.py +111 -111
tests/test_request_serialization.py +70 -70
tests/test_response_code_middleware.py +349 -349
tests/test_response_filter_middleware.py +427 -427
tests/test_response_improvements.py +152 -152
tests/test_retry_middleware.py +241 -241
tests/test_scheduler.py +252 -252
tests/test_scheduler_config_update.py +133 -133
tests/test_simple_response.py +61 -61
tests/test_telecom_spider_redis_key.py +205 -205
tests/test_template_content.py +87 -87
tests/test_template_redis_key.py +134 -134
tests/test_tools.py +159 -159
tests/test_user_agents.py +96 -96
tests/tools_example.py +260 -260
tests/untested_features_report.md +139 -0
tests/verify_debug.py +52 -0
tests/verify_distributed.py +117 -117
tests/verify_log_fix.py +112 -0
crawlo-1.3.3.dist-info/RECORD +0 -219
tests/DOUBLE_CRAWLO_PREFIX_FIX_REPORT.md +0 -82
{crawlo-1.3.3.dist-info → crawlo-1.3.4.dist-info}/WHEEL +0 -0
{crawlo-1.3.3.dist-info → crawlo-1.3.4.dist-info}/entry_points.txt +0 -0
{crawlo-1.3.3.dist-info → crawlo-1.3.4.dist-info}/top_level.txt +0 -0

crawlo/data/__init__.py CHANGED Viewed

@@ -1,6 +1,6 @@
-#!/usr/bin/env python
-# -*- coding: utf-8 -*-
-"""
-Crawlo 数据模块
-包含用户代理、字典等数据文件
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+"""
+Crawlo 数据模块
+包含用户代理、字典等数据文件
 """

crawlo/data/user_agents.py CHANGED Viewed

@@ -1,195 +1,195 @@
-#!/usr/bin/env python
-# -*- coding: utf-8 -*-
-"""
-User-Agent列表
-包含各种设备和浏览器的User-Agent字符串，用于爬虫伪装
-"""
-# 桌面浏览器User-Agent
-DESKTOP_USER_AGENTS = [
-    # Chrome (最新版本)
-    "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36",
-    "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36",
-    "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",
-    "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36",
-    # Chrome (较新版本)
-    "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36",
-    "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36",
-    "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",
-    "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36",
-    # Chrome (常用版本)
-    "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36",
-    "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36",
-    "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",
-    "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36",
-    # Firefox (最新版本)
-    "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:136.0) Gecko/20100101 Firefox/136.0",
-    "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:136.0) Gecko/20100101 Firefox/136.0",
-    "Mozilla/5.0 (X11; Linux i686; rv:136.0) Gecko/20100101 Firefox/136.0",
-    "Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:136.0) Gecko/20100101 Firefox/136.0",
-    # Firefox (较新版本)
-    "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:135.0) Gecko/20100101 Firefox/135.0",
-    "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:135.0) Gecko/20100101 Firefox/135.0",
-    "Mozilla/5.0 (X11; Linux i686; rv:135.0) Gecko/20100101 Firefox/135.0",
-    "Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:135.0) Gecko/20100101 Firefox/135.0",
-    # Safari (最新版本)
-    "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",
-    "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",
-    # Safari (较新版本)
-    "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",
-    "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",
-    # Edge (最新版本)
-    "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",
-    "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",
-    # Edge (较新版本)
-    "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",
-    "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",
-    # Opera
-    "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",
-    "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",
-]
-# 移动设备User-Agent
-MOBILE_USER_AGENTS = [
-    # iPhone (最新版本)
-    "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",
-    "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",
-    # iPhone (较新版本)
-    "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",
-    "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",
-    # iPad (最新版本)
-    "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",
-    "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",
-    # iPad (较新版本)
-    "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",
-    "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",
-    # Android (最新版本)
-    "Mozilla/5.0 (Linux; Android 15; SM-S921B) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
-    "Mozilla/5.0 (Linux; Android 14; SM-S911B) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
-    "Mozilla/5.0 (Linux; Android 14; Pixel 8) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
-    # Android (较新版本)
-    "Mozilla/5.0 (Linux; Android 13; SM-S908B) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
-    "Mozilla/5.0 (Linux; Android 13; Pixel 7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
-    "Mozilla/5.0 (Linux; Android 12; Pixel 6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
-    # Android平板
-    "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",
-    "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",
-    # 其他移动浏览器
-    "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",
-    "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",
-]
-# 爬虫/机器人User-Agent (用于测试)
-BOT_USER_AGENTS = [
-    "Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)",
-    "Mozilla/5.0 (compatible; Bingbot/2.0; +http://www.bing.com/bingbot.htm)",
-    "Mozilla/5.0 (compatible; Baiduspider/2.0; +http://www.baidu.com/search/spider.html)",
-    "Mozilla/5.0 (compatible; Yahoo! Slurp; http://help.yahoo.com/help/us/ysearch/slurp)",
-    "Mozilla/5.0 (compatible; DuckDuckBot/1.1; +http://duckduckgo.com/duckduckbot.html)",
-    "Mozilla/5.0 (compatible; YandexBot/3.0; +http://yandex.com/bots)",
-    "Mozilla/5.0 (compatible; facebookexternalhit/1.1; +http://www.facebook.com/externalhit_uatext.php)",
-    "Mozilla/5.0 (compatible; Twitterbot/1.0)",
-]
-# 按浏览器分类的User-Agent
-CHROME_USER_AGENTS = [
-    ua for ua in DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
-    if "Chrome" in ua and "Edg" not in ua and "OPR" not in ua
-]
-FIREFOX_USER_AGENTS = [
-    ua for ua in DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
-    if "Firefox" in ua
-]
-SAFARI_USER_AGENTS = [
-    ua for ua in DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
-    if "Safari" in ua and "Chrome" not in ua
-]
-EDGE_USER_AGENTS = [
-    ua for ua in DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
-    if "Edg" in ua
-]
-OPERA_USER_AGENTS = [
-    ua for ua in DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
-    if "OPR" in ua
-]
-# 所有User-Agent的组合列表
-ALL_USER_AGENTS = DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
-# 按设备类型分类的User-Agent字典
-USER_AGENTS_BY_TYPE = {
-    "desktop": DESKTOP_USER_AGENTS,
-    "mobile": MOBILE_USER_AGENTS,
-    "bot": BOT_USER_AGENTS,
-    "all": ALL_USER_AGENTS,
-    "chrome": CHROME_USER_AGENTS,
-    "firefox": FIREFOX_USER_AGENTS,
-    "safari": SAFARI_USER_AGENTS,
-    "edge": EDGE_USER_AGENTS,
-    "opera": OPERA_USER_AGENTS,
-}
-def get_user_agents(device_type="all"):
-    """
-    获取指定类型的User-Agent列表
-    Args:
-        device_type (str): 设备类型，可选值: "desktop", "mobile", "bot", "all", "chrome", "firefox", "safari", "edge", "opera"
-    Returns:
-        list: User-Agent字符串列表
-    """
-    return USER_AGENTS_BY_TYPE.get(device_type, ALL_USER_AGENTS)
-def get_random_user_agent(device_type="all"):
-    """
-    获取随机User-Agent
-    Args:
-        device_type (str): 设备类型，可选值: "desktop", "mobile", "bot", "all", "chrome", "firefox", "safari", "edge", "opera"
-    Returns:
-        str: 随机User-Agent字符串
-    """
-    import random
-    user_agents = get_user_agents(device_type)
-    return random.choice(user_agents) if user_agents else ""
-# 导出常用的User-Agent列表
-__all__ = [
-    "DESKTOP_USER_AGENTS",
-    "MOBILE_USER_AGENTS",
-    "BOT_USER_AGENTS",
-    "CHROME_USER_AGENTS",
-    "FIREFOX_USER_AGENTS",
-    "SAFARI_USER_AGENTS",
-    "EDGE_USER_AGENTS",
-    "OPERA_USER_AGENTS",
-    "ALL_USER_AGENTS",
-    "USER_AGENTS_BY_TYPE",
-    "get_user_agents",
-    "get_random_user_agent"
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+"""
+User-Agent列表
+包含各种设备和浏览器的User-Agent字符串，用于爬虫伪装
+"""
+# 桌面浏览器User-Agent
+DESKTOP_USER_AGENTS = [
+    # Chrome (最新版本)
+    "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36",
+    "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36",
+    "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",
+    "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Safari/537.36",
+    # Chrome (较新版本)
+    "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36",
+    "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36",
+    "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",
+    "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36",
+    # Chrome (常用版本)
+    "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36",
+    "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36",
+    "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",
+    "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36",
+    # Firefox (最新版本)
+    "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:136.0) Gecko/20100101 Firefox/136.0",
+    "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:136.0) Gecko/20100101 Firefox/136.0",
+    "Mozilla/5.0 (X11; Linux i686; rv:136.0) Gecko/20100101 Firefox/136.0",
+    "Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:136.0) Gecko/20100101 Firefox/136.0",
+    # Firefox (较新版本)
+    "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:135.0) Gecko/20100101 Firefox/135.0",
+    "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:135.0) Gecko/20100101 Firefox/135.0",
+    "Mozilla/5.0 (X11; Linux i686; rv:135.0) Gecko/20100101 Firefox/135.0",
+    "Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:135.0) Gecko/20100101 Firefox/135.0",
+    # Safari (最新版本)
+    "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",
+    "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",
+    # Safari (较新版本)
+    "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",
+    "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",
+    # Edge (最新版本)
+    "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",
+    "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",
+    # Edge (较新版本)
+    "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",
+    "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",
+    # Opera
+    "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",
+    "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",
+]
+# 移动设备User-Agent
+MOBILE_USER_AGENTS = [
+    # iPhone (最新版本)
+    "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",
+    "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",
+    # iPhone (较新版本)
+    "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",
+    "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",
+    # iPad (最新版本)
+    "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",
+    "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",
+    # iPad (较新版本)
+    "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",
+    "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",
+    # Android (最新版本)
+    "Mozilla/5.0 (Linux; Android 15; SM-S921B) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
+    "Mozilla/5.0 (Linux; Android 14; SM-S911B) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
+    "Mozilla/5.0 (Linux; Android 14; Pixel 8) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
+    # Android (较新版本)
+    "Mozilla/5.0 (Linux; Android 13; SM-S908B) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
+    "Mozilla/5.0 (Linux; Android 13; Pixel 7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
+    "Mozilla/5.0 (Linux; Android 12; Pixel 6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.6367.118 Mobile Safari/537.36",
+    # Android平板
+    "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",
+    "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",
+    # 其他移动浏览器
+    "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",
+    "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",
+]
+# 爬虫/机器人User-Agent (用于测试)
+BOT_USER_AGENTS = [
+    "Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)",
+    "Mozilla/5.0 (compatible; Bingbot/2.0; +http://www.bing.com/bingbot.htm)",
+    "Mozilla/5.0 (compatible; Baiduspider/2.0; +http://www.baidu.com/search/spider.html)",
+    "Mozilla/5.0 (compatible; Yahoo! Slurp; http://help.yahoo.com/help/us/ysearch/slurp)",
+    "Mozilla/5.0 (compatible; DuckDuckBot/1.1; +http://duckduckgo.com/duckduckbot.html)",
+    "Mozilla/5.0 (compatible; YandexBot/3.0; +http://yandex.com/bots)",
+    "Mozilla/5.0 (compatible; facebookexternalhit/1.1; +http://www.facebook.com/externalhit_uatext.php)",
+    "Mozilla/5.0 (compatible; Twitterbot/1.0)",
+]
+# 按浏览器分类的User-Agent
+CHROME_USER_AGENTS = [
+    ua for ua in DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
+    if "Chrome" in ua and "Edg" not in ua and "OPR" not in ua
+]
+FIREFOX_USER_AGENTS = [
+    ua for ua in DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
+    if "Firefox" in ua
+]
+SAFARI_USER_AGENTS = [
+    ua for ua in DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
+    if "Safari" in ua and "Chrome" not in ua
+]
+EDGE_USER_AGENTS = [
+    ua for ua in DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
+    if "Edg" in ua
+]
+OPERA_USER_AGENTS = [
+    ua for ua in DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
+    if "OPR" in ua
+]
+# 所有User-Agent的组合列表
+ALL_USER_AGENTS = DESKTOP_USER_AGENTS + MOBILE_USER_AGENTS
+# 按设备类型分类的User-Agent字典
+USER_AGENTS_BY_TYPE = {
+    "desktop": DESKTOP_USER_AGENTS,
+    "mobile": MOBILE_USER_AGENTS,
+    "bot": BOT_USER_AGENTS,
+    "all": ALL_USER_AGENTS,
+    "chrome": CHROME_USER_AGENTS,
+    "firefox": FIREFOX_USER_AGENTS,
+    "safari": SAFARI_USER_AGENTS,
+    "edge": EDGE_USER_AGENTS,
+    "opera": OPERA_USER_AGENTS,
+}
+def get_user_agents(device_type="all"):
+    """
+    获取指定类型的User-Agent列表
+    Args:
+        device_type (str): 设备类型，可选值: "desktop", "mobile", "bot", "all", "chrome", "firefox", "safari", "edge", "opera"
+    Returns:
+        list: User-Agent字符串列表
+    """
+    return USER_AGENTS_BY_TYPE.get(device_type, ALL_USER_AGENTS)
+def get_random_user_agent(device_type="all"):
+    """
+    获取随机User-Agent
+    Args:
+        device_type (str): 设备类型，可选值: "desktop", "mobile", "bot", "all", "chrome", "firefox", "safari", "edge", "opera"
+    Returns:
+        str: 随机User-Agent字符串
+    """
+    import random
+    user_agents = get_user_agents(device_type)
+    return random.choice(user_agents) if user_agents else ""
+# 导出常用的User-Agent列表
+__all__ = [
+    "DESKTOP_USER_AGENTS",
+    "MOBILE_USER_AGENTS",
+    "BOT_USER_AGENTS",
+    "CHROME_USER_AGENTS",
+    "FIREFOX_USER_AGENTS",
+    "SAFARI_USER_AGENTS",
+    "EDGE_USER_AGENTS",
+    "OPERA_USER_AGENTS",
+    "ALL_USER_AGENTS",
+    "USER_AGENTS_BY_TYPE",
+    "get_user_agents",
+    "get_random_user_agent"
 ]

crawlo 1.3.3__py3-none-any.whl → 1.3.4__py3-none-any.whl

Potentially problematic release.

crawlo 1.3.3py3-none-any.whl → 1.3.4py3-none-any.whl