1.2.912 - NodeJS update, performance improvement, adaptation to changes in recaptchas

[Support Artur]

We have completed the transition to NodeJS as the main engine for scrapers and present a new stable version 1.2.912 with support for NodeJS 14.2.0. This update combines many improvements, including increased performance, reduced memory consumption, a completely new network stack, as well as support for native NodeJS modules, allowing you to use the full power of the npmjs directory in A-Parser

Also, this update includes changes in working with ReCaptcha2 in the Google scraper, our team was one of the first to find a solution to circumvent the new version of the recaptcha and tested it together with the RuCaptcha service, for which they have a separate respect. At the moment, the correct captcha bypass has been tested with RuCaptcha, Anti-Captcha, XEvil and CapMonster.

In addition, many optimizations were made in the core of A-Parser, and performance was significantly increased when using a large number of tasks or large proxy lists. The scraper Rank::CMS has been completely rewritten and stabilized, support for the new apps.json format and support for user rules have been added.

Improvements

• NodeJS updated to v14.0.0, v8 to 8.1
• Added support for the data-s parameter in recaptures for SE::Google, also added the ReCaptcha2 pass proxy option
  
• Increased thread limit to 10,000 for Windows OS
• Significantly improved performance with a large number of active proxies and / or jobs, completely rewritten the stack for working with proxies, optimized work with large lists
• Added new scraper Rank::KeysSo
• Completely rewritten in JS SE::Yahoo::Suggest[parser], [parser]Rank::Alexa::API and Rank::Archive
• Improved performance when using regular expressions, as well as improved compatibility
• In SE::Google::KeywordPlanner added automatic token retrieval
  
• In SE::Bing added the ability to scrap links to cached pages, as well as the ability to scrap mobile results
  
• In the scraper Util::ReCaptcha2, when choosing the provider Capmonster or Xevil it is now optional to specify the Provider url
• In SE::Google::Trends added the ability to specify an arbitrary date range
• In Rank::CMS added the choice of a regular engine and support for its own file with features
• In SE::Yandex::ByImage added option Don't scrape if no other sizes, which allows you to disable the collection of results if the desired image is not in other sizes
• [NodeJS] Fixed this.cookies.getAll()
  
• [NodeJS] Added protection against endless loops and long regulars
• [JS scrapers] Added follow_meta_refresh option for this.request
• [JS scrapers] Added bypass_cloudflare option for this.request
• [JS scrapers] Underscore replaced by Lodash
• [JS scrapers] Added a mark in the log when calling other scrapers
• [JS scrapers] Using the previous proxy after a request to another scraper
• [JS scrapers] Added destroy() method

Corrections due to changes in the issuance

• Many fixes in SE::Google
• Fixed SE::Youtube, incl. scraping by tags
• Fixed collection of links in Shop::eBay
• Fixed phone scraping in Maps::Google
• Fixed work with captchas in SE::Yandex::ByImage
  
• In Rank::Social::Signal the variable $facebook_comment was deleted due to irrelevance
• SE::Startpage, Rank::Linkpad, Social::Instagram::post, SE::Yandex::Translate

Corrections

• Fixed a bug due to which the selected proxy checker was ignored
• Fixed work of Decode HTML entities and Extract domain functions in Result Constructor
• Fixed problem with encoding detection
  
• Fixed error using $tools.query
• Fixed bug in Rank::MajesticSEO in which all attempts were used in the absence of results
• Fixed work of http2
• Fixed a bug when the scraper crashes due to the inability to write in alive.txt
• Fixed captcha capturing in SE::Yandex::Register and Check::RosKomNadzor
• Fixed the difference in requests sent via Net::HTTP and JS
• Fixed bug in SE::Yahoo
• Bugs fixed in Rank::CMS when choosing an application without a category
• [NodeJS] Fixed calculation of scraper code execution time
• [JS scrapers] When the body is empty, the content-length header was not transmitted when posting a request
• [JS scrapers] Fixed work of CloudFlare bypass
• [JS scrapers] Fixed work with sessions
• [JS scrapers] Fixed work with overrides for this.parser.request
  
• [JS scrapers] Fixed error in encoding detection in JS scrapers