UBlock LLC UBlock Versions Save

• Removed donation request
• Fixed issue of Rate Us link which was not working for Firefox

• Improvements in assets management
• Add Rate Us in pop-up menu
• Change the short description to "A fast, effective, and free ad blocker."

• Added the phrase "- free ad blocker" after uBlock

• Handle more complex cases of Extended CSS selectors
• Added support for Adguard :not operator
• Added support for using Extended CSS selectors in Style filter rule like mycinema.pro#$#[class]:matches-css(width: 336px):matches-css(height: 280px):matches-css(min-height: 280px):has(>ins.adsbygoogle) { visibility: hidden!important; }
• Added support for two types of syntax from uBO
  • +js
  • +style

• Added support for following AdGuard pseudo-classes
  • :matches-css
  • :matches-css-before
  • :matches-css-after
• Removed obsolete filter list / Updated filter list links / Added new filter list / Changed location of filter list
• Fixed issues to make uBlock compatible with Chrome 49
• Doing parsing and validation of Extended CSS at the time of filter list data processing. It saves time when actually applying extended CSS

• Translating Adguard CSS rules into uBlock supported compatible rule syntax.
  • -ext-has to :-abp-has()
  • :has to :-abp-has()
  • -ext-contains to :-abp-contains()
  • :contains to :-abp-contains()
• Adding support for Adguard Cosmetic Css Rule
• Start using chrome.tabs.insertCSS API for CSS rules injection which is applicable for Firefox >= 53 and Chrome >= 66.
• Code Improvements

• Added support for following filter options
  • abort-on-property-read
  • abort-on-property-write
  • abort-current-inline-script
• Fixed issue
• Fixed regression testing issues

Updated Translation.

Performance and memory related changes

1. Special treatment for hostname-anchored (||) filters having wildcard characters. These filters are categorized into three types. The purpose is to convert regex based filter internally into plain string based filters.
   1. Filters having a single occurrence of pattern ^* after the hostname and no other wildcard character on another side of this pattern. ||.cloudfront.net^F2P_SideBar_Banner_EN.jpg ||2mdn.net^/1x1image.jpg$image ||2mdn.^/1x1image.jpg$image
   2. Filters having a single occurrence of wildcard character * after the hostname and no other wildcard character on another side of it. ||*.cloudfront.net/wp-content/F2P_SideBar_Banner_EN.jpg ||cloudfront.net/wp-content/F2P_SideBar_Banner_EN.jpg ||www.google./aclk?^/am=&adurl= ||i.com.com/wp-content/*F2P_SideBar_Banner_EN.jpg
   3. Other filters ||cloud*.net^$third-party
      @@||doubleclick.net/adi/mlb.mlb/*^free_agent_tracker_12^$subdocument,domain=mlb.com
2. Now hostname-anchored (||) filters matches hostname in URL accurately. ||*-aaa.net^$third-party
3. Filters having a single occurrence of wildcard character * converted internally into plain string based filters. @@/cdn-cgi/pe/bag2?googleadservices.com%2Fpagead%2Fconversion.js$xmlhttprequest,domain=ethica.net.au|factom.org|gogoonhold.com.au /cdn-cgi/pe/bag2?r.qualitypublishers.com
4. Using singleton object in case of any match URL filter. $script,third-party,domain=0dt.net|123videos.tv
5. Ignore token if it is preceded by * characters in Token generate logic

Made blocking 3x times faster

1. Special handling is given to filters which had to match against document origin only. The filter will only be considered if it satisfies the given criteria.

   • Should be in the form of /czf*. (without token) or |https:// or |http:// or *
   • Should have domain= option
   • Should not have a negated domain in its domain= option
   • Should not have csp= option

   examples: |http://$image,script,subdocument,third-party,xmlhttprequest,domain=dwindly.io |https://$image,other,script,stylesheet,third-party,domain=movies123.xyz $websocket,domain=povw1deo.com|povwideo.net|powvideo.net /czf*.$image,domain=100percentfedup.com

   These filters had to be matched against each network request which was causing extra time. Now, these filters will only be considered when document origin matches the domain mentioned in the filter domain= option.

2. Added google keyword to badTokens list

3. Replaced substr with startWith method for string pattern matching

4. Validated regular expression based filter

5. Code optimizations