PyPI - crawlo - Versions diffs - 1.2.3__py3-none-any.whl → 1.2.5__py3-none-any.whl - Mend

crawlo 1.2.3py3-none-any.whl → 1.2.5py3-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 (222) hide show

crawlo/__init__.py +61 -61
crawlo/__version__.py +1 -1
crawlo/cleaners/__init__.py +60 -60
crawlo/cleaners/data_formatter.py +225 -225
crawlo/cleaners/encoding_converter.py +125 -125
crawlo/cleaners/text_cleaner.py +232 -232
crawlo/cli.py +88 -81
crawlo/commands/__init__.py +14 -14
crawlo/commands/check.py +594 -594
crawlo/commands/genspider.py +151 -151
crawlo/commands/help.py +144 -142
crawlo/commands/list.py +155 -155
crawlo/commands/run.py +323 -292
crawlo/commands/startproject.py +436 -417
crawlo/commands/stats.py +187 -187
crawlo/commands/utils.py +186 -186
crawlo/config.py +312 -312
crawlo/config_validator.py +251 -251
crawlo/core/__init__.py +2 -2
crawlo/core/engine.py +354 -354
crawlo/core/processor.py +40 -40
crawlo/core/scheduler.py +143 -143
crawlo/crawler.py +1110 -1027
crawlo/data/__init__.py +5 -5
crawlo/data/user_agents.py +107 -107
crawlo/downloader/__init__.py +266 -266
crawlo/downloader/aiohttp_downloader.py +220 -220
crawlo/downloader/cffi_downloader.py +256 -256
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 +37 -37
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 +43 -43
crawlo/extension/memory_monitor.py +104 -104
crawlo/extension/performance_profiler.py +133 -133
crawlo/extension/request_recorder.py +107 -107
crawlo/filters/__init__.py +154 -154
crawlo/filters/aioredis_filter.py +280 -280
crawlo/filters/memory_filter.py +269 -269
crawlo/items/__init__.py +23 -23
crawlo/items/base.py +21 -21
crawlo/items/fields.py +52 -52
crawlo/items/items.py +104 -104
crawlo/middleware/__init__.py +21 -21
crawlo/middleware/default_header.py +131 -131
crawlo/middleware/download_delay.py +104 -104
crawlo/middleware/middleware_manager.py +135 -135
crawlo/middleware/offsite.py +114 -114
crawlo/middleware/proxy.py +367 -367
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/mode_manager.py +211 -211
crawlo/network/__init__.py +21 -21
crawlo/network/request.py +338 -338
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 +317 -317
crawlo/pipelines/pipeline_manager.py +61 -61
crawlo/pipelines/redis_dedup_pipeline.py +165 -165
crawlo/project.py +279 -187
crawlo/queue/pqueue.py +37 -37
crawlo/queue/queue_manager.py +337 -337
crawlo/queue/redis_priority_queue.py +298 -298
crawlo/settings/__init__.py +7 -7
crawlo/settings/default_settings.py +217 -226
crawlo/settings/setting_manager.py +122 -122
crawlo/spider/__init__.py +639 -639
crawlo/stats_collector.py +59 -59
crawlo/subscriber.py +129 -129
crawlo/task_manager.py +30 -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 +324 -325
crawlo/templates/project/settings_distributed.py.tmpl +154 -121
crawlo/templates/project/settings_gentle.py.tmpl +127 -94
crawlo/templates/project/settings_high_performance.py.tmpl +149 -151
crawlo/templates/project/settings_simple.py.tmpl +102 -68
crawlo/templates/project/spiders/__init__.py.tmpl +5 -5
crawlo/templates/{project/run.py.tmpl → run.py.tmpl} +47 -45
crawlo/templates/spider/spider.py.tmpl +143 -143
crawlo/tools/__init__.py +182 -182
crawlo/tools/anti_crawler.py +268 -268
crawlo/tools/authenticated_proxy.py +240 -240
crawlo/tools/data_validator.py +180 -180
crawlo/tools/date_tools.py +35 -35
crawlo/tools/distributed_coordinator.py +386 -386
crawlo/tools/retry_mechanism.py +220 -220
crawlo/tools/scenario_adapter.py +262 -262
crawlo/utils/__init__.py +35 -35
crawlo/utils/batch_processor.py +259 -259
crawlo/utils/controlled_spider_mixin.py +439 -439
crawlo/utils/date_tools.py +290 -290
crawlo/utils/db_helper.py +343 -343
crawlo/utils/enhanced_error_handler.py +356 -356
crawlo/utils/env_config.py +105 -105
crawlo/utils/error_handler.py +123 -123
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 +128 -128
crawlo/utils/performance_monitor.py +285 -285
crawlo/utils/queue_helper.py +175 -175
crawlo/utils/redis_connection_pool.py +334 -334
crawlo/utils/redis_key_validator.py +198 -198
crawlo/utils/request.py +267 -267
crawlo/utils/request_serializer.py +218 -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-1.2.3.dist-info → crawlo-1.2.5.dist-info}/METADATA +764 -692
crawlo-1.2.5.dist-info/RECORD +206 -0
examples/__init__.py +7 -7
tests/DOUBLE_CRAWLO_PREFIX_FIX_REPORT.md +81 -81
tests/__init__.py +7 -7
tests/advanced_tools_example.py +275 -275
tests/authenticated_proxy_example.py +236 -236
tests/cleaners_example.py +160 -160
tests/config_validation_demo.py +102 -102
tests/controlled_spider_example.py +205 -205
tests/date_tools_example.py +180 -180
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/redis_key_validation_demo.py +130 -130
tests/response_improvements_example.py +144 -144
tests/test_advanced_tools.py +148 -148
tests/test_all_redis_key_configs.py +145 -145
tests/test_authenticated_proxy.py +141 -141
tests/test_cleaners.py +54 -54
tests/test_comprehensive.py +146 -146
tests/test_config_validator.py +193 -193
tests/test_crawlo_proxy_integration.py +172 -172
tests/test_date_tools.py +123 -123
tests/test_default_header_middleware.py +158 -158
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_env_config.py +121 -121
tests/test_error_handler_compatibility.py +112 -112
tests/test_final_validation.py +153 -153
tests/test_framework_env_usage.py +103 -103
tests/test_integration.py +356 -356
tests/test_item_dedup_redis_key.py +122 -122
tests/test_offsite_middleware.py +221 -221
tests/test_parsel.py +29 -29
tests/test_performance.py +327 -327
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_providers.py +56 -56
tests/test_proxy_stats.py +19 -19
tests/test_proxy_strategies.py +59 -59
tests/test_queue_manager_double_crawlo.py +173 -173
tests/test_queue_manager_redis_key.py +176 -176
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_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 +241 -241
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 +153 -153
tests/tools_example.py +257 -257
crawlo-1.2.3.dist-info/RECORD +0 -222
examples/aiohttp_settings.py +0 -42
examples/curl_cffi_settings.py +0 -41
examples/default_header_middleware_example.py +0 -107
examples/default_header_spider_example.py +0 -129
examples/download_delay_middleware_example.py +0 -160
examples/httpx_settings.py +0 -42
examples/multi_downloader_proxy_example.py +0 -81
examples/offsite_middleware_example.py +0 -55
examples/offsite_spider_example.py +0 -107
examples/proxy_spider_example.py +0 -166
examples/request_ignore_middleware_example.py +0 -51
examples/request_ignore_spider_example.py +0 -99
examples/response_code_middleware_example.py +0 -52
examples/response_filter_middleware_example.py +0 -67
examples/tong_hua_shun_settings.py +0 -62
examples/tong_hua_shun_spider.py +0 -170
{crawlo-1.2.3.dist-info → crawlo-1.2.5.dist-info}/WHEEL +0 -0
{crawlo-1.2.3.dist-info → crawlo-1.2.5.dist-info}/entry_points.txt +0 -0
{crawlo-1.2.3.dist-info → crawlo-1.2.5.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

@@ -1,108 +1,108 @@
-#!/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
-    "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
-    "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
-    "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",
-    "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
-    # Firefox
-    "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:125.0) Gecko/20100101 Firefox/125.0",
-    "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:125.0) Gecko/20100101 Firefox/125.0",
-    "Mozilla/5.0 (X11; Linux i686; rv:125.0) Gecko/20100101 Firefox/125.0",
-    "Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:125.0) Gecko/20100101 Firefox/125.0",
-    # Safari
-    "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",
-    "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",
-    # Edge
-    "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",
-    "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",
-    # Internet Explorer
-    "Mozilla/5.0 (Windows NT 10.0; WOW64; Trident/7.0; rv:11.0) like Gecko",
-]
-# 移动设备User-Agent
-MOBILE_USER_AGENTS = [
-    # iPhone
-    "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",
-    "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",
-    # iPad
-    "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",
-    "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",
-    # Android
-    "Mozilla/5.0 (Linux; Android 14; SM-S911B) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.6367.118 Mobile Safari/537.36",
-    "Mozilla/5.0 (Linux; Android 13; SM-S908B) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.6367.118 Mobile Safari/537.36",
-    "Mozilla/5.0 (Linux; Android 12; Pixel 6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.6367.118 Mobile Safari/537.36",
-    "Mozilla/5.0 (Linux; Android 11; Pixel 5) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.6367.118 Mobile Safari/537.36",
-    # Android平板
-    "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",
-]
-# 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)",
-]
-# 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,
-}
-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"
-    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"
-    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",
-    "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
+    "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
+    "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
+    "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",
+    "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36",
+    # Firefox
+    "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:125.0) Gecko/20100101 Firefox/125.0",
+    "Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:125.0) Gecko/20100101 Firefox/125.0",
+    "Mozilla/5.0 (X11; Linux i686; rv:125.0) Gecko/20100101 Firefox/125.0",
+    "Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:125.0) Gecko/20100101 Firefox/125.0",
+    # Safari
+    "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",
+    "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",
+    # Edge
+    "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",
+    "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",
+    # Internet Explorer
+    "Mozilla/5.0 (Windows NT 10.0; WOW64; Trident/7.0; rv:11.0) like Gecko",
+]
+# 移动设备User-Agent
+MOBILE_USER_AGENTS = [
+    # iPhone
+    "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",
+    "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",
+    # iPad
+    "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",
+    "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",
+    # Android
+    "Mozilla/5.0 (Linux; Android 14; SM-S911B) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.6367.118 Mobile Safari/537.36",
+    "Mozilla/5.0 (Linux; Android 13; SM-S908B) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.6367.118 Mobile Safari/537.36",
+    "Mozilla/5.0 (Linux; Android 12; Pixel 6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.6367.118 Mobile Safari/537.36",
+    "Mozilla/5.0 (Linux; Android 11; Pixel 5) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.6367.118 Mobile Safari/537.36",
+    # Android平板
+    "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",
+]
+# 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)",
+]
+# 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,
+}
+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"
+    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"
+    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",
+    "ALL_USER_AGENTS",
+    "USER_AGENTS_BY_TYPE",
+    "get_user_agents",
+    "get_random_user_agent"
 ]

crawlo 1.2.3__py3-none-any.whl → 1.2.5__py3-none-any.whl

Potentially problematic release.

crawlo 1.2.3py3-none-any.whl → 1.2.5py3-none-any.whl