github_mirrors/ua-parser-js
3
0
Fork 1
mirror of https://github.com/faisalman/ua-parser-js.git synced 2025-09-27 16:08:47 +03:00
Code Issues Packages Projects Releases Wiki Activity

[submodule:extensions] Add new bots: AI2Bot, aiHitBot, anthropic-ai, cohere-ai, Diffbot, ImagesiftBot, magpie-crawler, Omgilibot, Screaming Frog SEO Spider, Seznambot, Teoma, Timpibot, VelenPublicWebCrawler, Vercelbot, Webzio-Extended, YouBot

Browse Source
This commit is contained in:
Faisal Salman 2024-11-16 18:40:02 +07:00
parent 2181559b01
commit 2b125c8de4
3 changed files with 141 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
src/extensions/ua-parser-extensions.js
View File

@ -53,7 +53,8 @@ const Crawlers = Object.freeze({
// OpenAI's SearchGPT - https://platform.openai.com/docs/bots
// PerplexityBot - https://perplexity.ai/perplexitybot
// SemrushBot - http://www.semrush.com/bot.html
/((?:ahrefs|amazon|bing|cc|dot|duckduck|exa|facebook|gpt|mj12|mojeek|oai-search|perplexity|semrush)bot)\/([\w\.]+)/i,
// SeznamBot - http://napoveda.seznam.cz/seznambot-intro
/((?:ahrefs|amazon|bing|cc|dot|duckduck|exa|facebook|gpt|mj12|mojeek|oai-search|perplexity|semrush|seznam)bot)\/([\w\.-]+)/i,
// Applebot - http://apple.com/go/applebot
/(applebot(?:-extended)?)\/([\w\.]+)/i,
@ -62,7 +63,7 @@ const Crawlers = Object.freeze({
/(baiduspider)[-imagevdonsfcpr]{0,6}\/([\w\.]+)/i,
// ClaudeBot (Anthropic)
/(claude(?:bot|-web))\/([\w\.]+)/i,
/(claude(?:bot|-web)|anthropic-ai)\/?([\w\.]*)/i,
// Coc Coc Bot - https://help.coccoc.com/en/search-engine
/(coccocbot-(?:image|web))\/([\w\.]+)/i,
@ -89,8 +90,8 @@ const Crawlers = Object.freeze({
// Yeti (Naver)
/(yeti)\/([\w\.]+)/i,
// YisouSpider
/(yisouspider)\/?([\w\.]*)/i
// aiHitBot / Cohere-AI / Diffbot / Magpie-Crawler / Omgilibot / Webzio-Extended / Screaming Frog SEO Spider / Timpibot / VelenPublicWebCrawler / YisouSpider / YouBot
/((?:aihit|diff|timpi|you)bot|cohere-ai|omgili(?:bot)?|(?:magpie-|velenpublicweb)crawler|webzio-extended|(?:screaming frog seo |yisou)spider)\/?([\w\.]*)/i
],
[NAME, VERSION, [TYPE, CRAWLER]],
@ -99,13 +100,15 @@ const Crawlers = Object.freeze({
// Google Bots
/((?:adsbot|apis|mediapartners)-google(?:-mobile)?|google-?(?:other|cloudvertexbot|extended|safety))/i,
// AI2Bot - https://allenai.org/crawler
// Bytespider
// DataForSeoBot - https://dataforseo.com/dataforseo-bot
// Huawei AspiegelBot / PetalBot https://aspiegel.com/petalbot
// ImagesiftBot - https://imagesift.com/about
// Qihoo 360Spider
// TurnitinBot - https://www.turnitin.com/robot/crawlerinfo.html
// Yahoo! Slurp - http://help.yahoo.com/help/us/ysearch/slurp
/(360spider-?(?:image|video)?|bytespider|(?:aspiegel|dataforseo|petal|turnitin)bot|(?=yahoo! )slurp)/i
/\b(360spider-?(?:image|video)?|bytespider|(?:ai2|aspiegel|dataforseo|imagesift|petal|turnitin)bot|teoma|(?=yahoo! )slurp)/i
],
[NAME, [TYPE, CRAWLER]]
]
@ -238,8 +241,8 @@ const Fetchers = Object.freeze({
],
[NAME, VERSION, [TYPE, FETCHER]],
// Google Bots / Snapchat
[/(feedfetcher-google|google(?:-read-aloud|producer)|(?=bot; )snapchat)/i],
// Google Bots / Snapchat / Vercelbot
[/(vercelbot|feedfetcher-google|google(?:-read-aloud|producer)|(?=bot; )snapchat)/i],
[NAME, [TYPE, FETCHER]],
]
});

122
test/specs/browser-crawlers.json
View File

@ -49,6 +49,26 @@
"type" : "crawler"
}
},
{
"desc" : "AI2Bot",
"ua" : "Mozilla/5.0 (compatible) AI2Bot (+https://www.allenai.org/crawler)",
"expect" :
{
"name" : "AI2Bot",
"version" : "undefined",
"type" : "crawler"
}
},
{
"desc" : "aiHitBot",
"ua" : "Mozilla/5.0 (compatible; aiHitBot/2.9; +https://www.aihitdata.com/about)",
"expect" :
{
"name" : "aiHitBot",
"version" : "2.9",
"type" : "crawler"
}
},
{
"desc" : "Applebot",
"ua" : "Mozilla/5.0 (iPhone; CPU iPhone OS 8_1 like Mac OS X) AppleWebKit/600.1.4 (KHTML, like Gecko) Version/8.0 Mobile/12B410 Safari/600.1.4 (Applebot/0.1;+http://www.apple.com/go/applebot)",
@ -131,7 +151,7 @@
},
{
"desc" : "DataForSEO",
"ua" : "Mozilla/5.0 (compatible; DataForSeoBot; +https://dataforseo.com/dataforseo-bot) ",
"ua" : "Mozilla/5.0 (compatible; DataForSeoBot; +https://dataforseo.com/dataforseo-bot)",
"expect" :
{
"name" : "DataForSeoBot",
@ -139,6 +159,16 @@
"type" : "crawler"
}
},
{
"desc" : "Diffbot",
"ua" : "Diffbot/0.1",
"expect" :
{
"name" : "Diffbot",
"version" : "0.1",
"type" : "crawler"
}
},
{
"desc" : "Dotbot",
"ua" : "Mozilla/5.0 (compatible; DotBot/1.2; +https://opensiteexplorer.org/dotbot; help@moz.com)",
@ -329,6 +359,26 @@
"type" : "crawler"
}
},
{
"desc" : "ImagesiftBot",
"ua" : "Mozilla/5.0 (compatible; ImagesiftBot; +imagesift.com)",
"expect" :
{
"name" : "ImagesiftBot",
"version" : "undefined",
"type" : "crawler"
}
},
{
"desc" : "magpie-crawler",
"ua" : "magpie-crawler/1.1 (robots-txt-checker; +http://www.brandwatch.net)",
"expect" :
{
"name" : "magpie-crawler",
"version" : "1.1",
"type" : "crawler"
}
},
{
"desc" : "Meta-ExternalAgent",
"ua" : "meta-externalagent/1.1 (+https://developers.facebook.com/docs/sharing/webmasters/crawler)",
@ -360,6 +410,26 @@
"type" : "crawler"
}
},
{
"desc" : "Omgili",
"ua" : "omgili/0.5 +https://omgili.com",
"expect" :
{
"name" : "omgili",
"version" : "0.5",
"type" : "crawler"
}
},
{
"desc" : "Omgilibot",
"ua" : "omgilibot/0.3 +http://www.omgili.com/Crawler.html",
"expect" :
{
"name" : "omgilibot",
"version" : "0.3",
"type" : "crawler"
}
},
{
"desc" : "OpenAI Search",
"ua" : "Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko); compatible; OAI-SearchBot/1.0; +https://openai.com/searchbot",
@ -410,6 +480,36 @@
"type" : "crawler"
}
},
{
"desc" : "SeznamBot",
"ua" : "Mozilla/5.0 (compatible; SeznamBot/4.0-RC1; +http://napoveda.seznam.cz/seznambot-intro/)",
"expect" :
{
"name" : "SeznamBot",
"version" : "4.0-RC1",
"type" : "crawler"
}
},
{
"desc" : "Teoma",
"ua" : "Mozilla/2.0 (compatible; Ask Jeeves/Teoma; +http://sp.ask.com/docs/about/tech_crawling.html)",
"expect" :
{
"name" : "Teoma",
"version" : "undefined",
"type" : "crawler"
}
},
{
"desc" : "Timpibot",
"ua" : "Timpibot/0.8 (+http://www.timpi.io)",
"expect" :
{
"name" : "Timpibot",
"version" : "0.8",
"type" : "crawler"
}
},
{
"desc" : "TurnitinBot",
"ua" : "TurnitinBot (https://turnitin.com/robot/crawlerinfo.html)",
@ -420,6 +520,16 @@
"type" : "crawler"
}
},
{
"desc" : "VelenPublicWebCrawler",
"ua" : "Mozilla/5.0 (compatible; VelenPublicWebCrawler/1.0; +https://velen.io)",
"expect" :
{
"name" : "VelenPublicWebCrawler",
"version" : "1.0",
"type" : "crawler"
}
},
{
"desc" : "Yahoo! Japan",
"ua" : "Y!J-BRW/1.0 (https://www.yahoo-help.jp/app/answers/detail/p/595/a_id/42716)",
@ -469,5 +579,15 @@
"version" : "undefined",
"type" : "crawler"
}
},
{
"desc" : "YouBot",
"ua" : "YouBot (+http://www.you.com)",
"expect" :
{
"name" : "YouBot",
"version" : "undefined",
"type" : "crawler"
}
}
]

10
test/specs/browser-fetchers.json
View File

@ -118,5 +118,15 @@
"version" : "2.0",
"type" : "fetcher"
}
},
{
"desc" : "Vercelbot",
"ua" : "Vercelbot (+https://vercel.com)",
"expect" :
{
"name" : "Vercelbot",
"version" : "undefined",
"type" : "fetcher"
}
}
]