How to create your own ad filters
In diesem Artikel wird erklärt, wie Sie benutzerdefinierte Filterregeln für die Verwendung in AdGuard-Produkten erstellen können. To test your rules, you can download the AdGuard app
A filter is a set of filtering rules applied to specific content, such as banners or pop-ups. AdGuard has a list of standard filters created by our team. We constantly improve and update them, striving to meet the needs of most of our users.
At the same time, AdGuard allows you to create your own custom filters using the same types of rules that we have in our filters.
To describe the syntax of our filtering rules, we use Augmented BNF for Syntax Specifications, but we do not always strictly follow this specification.
Originally, the AdGuard's syntax was based on the syntax of Adblock Plus rules. Later, we extended it with new types of rules for better ad filtering. Some parts of this article about the rules common both to AdGuard and ABP were taken from the Adblock Plus guide on how to write filters.
Comments
Any line that starts with an exclamation mark is a comment. In the list of rules it is displayed in gray color. AdGuard will ignore this line, so you can write anything you want. Comments are usually placed above the rules and used to describe what a rule does.
Zum Beispiel:
! This is the comment. Below this line, there is an actual filtering rule.
||example.org^
Examples
Blocking by domain name
This rule blocks:
http://example.org/ad1.gif
http://subdomain.example.org/ad1.gif
https://ads.example.org:8000/
This rule does not block:
http://ads.example.org.us/ad1.gif
http://example.com/redirect/http://ads.example.org/
By default, such rules do not work for document requests. This means that the ||example.org^
rule will block a request made to example.org
when you try to navigate to this domain from another website, but if you type example.org
into the address bar and try to navigate to it, the website will open. To block the document request, you will need to use a rule with the $document
modifier: ||example.org^$document
.
Blocking exact address
This rule blocks:
http://example.org/
This rule does not block:
https://example.org/banner/img
Basic rule modifiers
Filtering rules support numerous modifiers that allow you to fine-tune the rule behavior. Here is an example of a rule with some simple modifiers.
This rule blocks:
http://example.org/script.js
if this script is loaded fromexample.com
.
This rule does not block:
https://example.org/script.js
if this script is loaded fromexample.org
.https://example.org/banner.png
because it is not a script.
Unblocking an address
This rule unblocks:
http://example.org/banner.png
even if there is a blocking rule for this address.
Blocking rules with $important
modifier can override exceptions.
Unblocking an entire website
This rule unblocks
- It disables all cosmetic rules on
example.com
. - It unblocks all requests sent from this website even if there is are blocking rules matching these requests.
Cosmetic rule
Cosmetic rules are based on using a special language named CSS, which every browser understands. Basically, it adds a new CSS style to the website which purpose is to hide particular elements. You can learn more about CSS in general here.
AdGuard extends CSS and lets filters developers handle much more complicated cases. However, to use these extended rules, you need to be fluent in regular CSS.
Popular CSS selectors
Name | CSS selector | Description |
---|---|---|
ID selector | #banners | Matches all elements with id attribute equal to banners . |
Class selector | .banners | Matches all elements with class attribute containing banners . |
Attribute selector | div[class="banners"] | Matches all div elements with class attribute exactly equal to banners . |
Attribute substring selector | div[class^="advert1"] | Matches all div elements which class attribute starts with the advert1 string. |
Attribute substring selector | div[class$="banners_ads"] | Matches all div elements which class attribute ends with the banners_ads string. |
Attribute substring selector | a[href^="http://example.com/"] | Matches all links that are loaded from http://example.com/ domain. |
Attribute selector | a[href="http://example.com/"] | Matches all links to exactly the http://example.com/ address. |
Restrictions and limitations
Trusted filters
Some rules can be used only in trusted filters. This category includes:
- filter lists created by the AdGuard team,
- custom filter lists installed as
trusted
, - user rules.
AdGuard-Inhaltsblocker
AdGuard Content Blocker is an extension for Samsung and Yandex browsers that can be installed from Google Play. It is not to be confused with the fully functional AdGuard for Android that can only be downloaded from our website. Unfortunately, AdGuard Content Blocker capabilities are limited by what the browsers allow and they only support an old Adblock Plus filters syntax:
- Basic blocking rules with the following modifiers:
$domain
,$third-party
, content-type modifiers. - Basic exception rules with the following modifiers:
$document
,$elemhide
. - Basic element hiding rules with no extended CSS support.
Because of the limitations above AdGuard Content Blocker will not be mentioned in the compatibility notes.
Basic rules
The most simple rules are so-called Basic rules. They are used to block requests to specific URLs. Or to unblock it, if there is a special marker "@@" at the beginning of the rule. The basic principle for this type of rules is quite simple: you have to specify the address and additional parameters that limit or expand the rule scope.
Basic rules for blocking requests are applied only to sub-requests. That means they will not block the loading of the page unless it is explicitly specified with a $document
modifier.
Browser detects a blocked request as completed with an error.
Rules shorter than 4 characters are considered incorrect and will be ignored.
Basic rule syntax
rule = ["@@"] pattern [ "$" modifiers ]
modifiers = [modifier0, modifier1[, ...[, modifierN]]]
pattern
— an address mask. Every request URL is collated to this mask. In the template, you can also use the special characters described below. Note that AdGuard truncates URLs to a length of 4096 characters in order to speed up matching and avoid issues with ridiculously long URLs.@@
— a marker that is used in rules of exception. To turn off filtering for a request, start your rule with this marker.modifiers
— parameters that "clarify" the basic rule. Some of them limit the rule scope and some can completely change they way it works.
Special characters
*
— a wildcard character. It is used to represent any set of characters. This can also be an empty string or a string of any length.||
— an indication to apply the rule to the specified domain and its subdomains. With this character, you do not have to specify a particular protocol and subdomain in address mask. It means that||
stands forhttp://*.
,https://*.
,ws://*.
,wss://*.
at once.^
— a separator character mark. Separator character is any character, but a letter, a digit, or one of the following:_
-
.
%
. In this example separator characters are shown in bold:http:
//
example.com
/?
t=1
&
t2=t3
. The end of the address is also accepted as separator.|
— a pointer to the beginning or the end of address. The value depends on the character placement in the mask. For example, a ruleswf|
corresponds tohttp://example.com/annoyingflash.swf
, but not tohttp://example.com/swf/index.html
.|http://example.org
corresponds tohttp://example.org
, but not tohttp://domain.com?url=http://example.org
.
|
, ||
, ^
can only be used with rules that have a URL pattern. For example, ||example.com##.advert
is incorrect and will be ignored by the blocker.
We also recommend to get acquainted with the Adblock Plus filter cheatsheet, for better understanding of how such rules should be made.
Regular expressions support
If you want even more flexibility in making rules, you can use Regular expressions instead of a default simplified mask with special characters.
Rules with regular expressions work more slowly, therefore it is recommended to avoid them or to limit their scope to specific domains.
If you want a blocker to determine a regular expression, the pattern
has to look like this:
pattern = "/" regexp "/"
For example, /banner\d+/$third-party
this rule will apply the regular expression banner\d+
to all third-party requests. Exclusion rule with regular expression looks like this: @@/banner\d+/
.
AdGuard Safari and AdGuard for iOS do not fully support regular expressions because of Content Blocking API restrictions (look for "The Regular expression format" section).
Wildcard support for TLD (top-level domains)
Wildcard characters are supported for TLDs of the domains in patterns of cosmetic, HTML filtering and JavaScript rules.
For cosmetic rules, e.g. example.*##.banner
, multiple domains are matched due to the part .*
, i.e. example.com
, sub.example.net
, example.co.uk
, etc.
For basic rules the described logic is applicable only for the domains specified in $domain
modifier, e.g. ||*/banners/*$image,domain=example.*
.
In AdGuard for Windows, Mac, Android, and AdGuard Browser Extension rules with wildcard .*
match any public suffix (or eTLD). But for AdGuard for Safari and iOS the supported list of top-level domains is limited due to Safari limitations.
Rules with wildcard for TLD are not supported by AdGuard Content Blocker.
Basic rule examples
||example.com/ads/*
— a simple rule, which corresponds to addresses likehttp://example.com/ads/banner.jpg
and evenhttp://subdomain.example.com/ads/otherbanner.jpg
.||example.org^$third-party
— this rule blocks third-party requests toexample.org
and its subdomains.@@||example.com$document
— general exception rule. It completely disables filtering forexample.com
and all subdomains. There is a number of modifiers which can be used in exception rules. For more details, please follow the link below.
Basic rule modifiers
The features described in this section are intended for experienced users. They extend capabilities of "Basic rules", but in order to use them you need to have a basic understanding of the way your browser works.
You can change the behavior of a "basic rule" by using additional modifiers. Modifiers should be located in the end of the rule after a $
sign and be separated by commas.
Example:
||domain.com^$popup,third-party
Basic modifiers
The following modifiers are the most simple and frequently used. Basically, they just limit the scope of rule application.
Modifikator \ Produkte | CoreLibs apps | AdGuard for Chromium | AdGuard for Firefox | AdGuard für iOS | AdGuard für Safari | AdGuard-Inhaltsblocker |
---|---|---|---|---|---|---|
$app | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ |
$denyallow | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
$domain | ✅ | ✅ | ✅ | ✅ * | ✅ * | ✅ |
$header | ✅ | 🧩 ** | 🧩 ** | ❌ | ❌ | ❌ |
$important | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
$match-case | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
$method | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
$popup | ✅ *** | ✅ | ✅ | ✅ *** | ✅ *** | ❌ |
$third-party | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
$to | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
- ✅ — fully supported
- ✅ * — supported, but reliability may vary or limitations may occur; check the modifier description for more details
- 🧩 — may already be implemented in nightly or beta versions but is not yet supported in release versions
- ❌ — not supported
$app
This modifier lets you narrow the rule coverage down to a specific application (or a list of applications). This might be not too important on Windows and Mac, but this is very important on mobile devices where some of the filtering rules must be app-specific.
- Android — use the app package name, e.g.
org.example.app
. - Windows — use the process name, e.g.
chrome.exe
. - Mac — use the bundle ID or the process name, e.g.
com.google.Chrome
.
For Mac, you can find out the bundle ID or the process name of the app by viewing the respective request details in the Filtering log.
Examples
||baddomain.com^$app=org.example.app
— a rule to block requests that match the specified mask and are sent from theorg.example.app
Android app.||baddomain.com^$app=org.example.app1|org.example.app2
— the same rule but it works for bothorg.example.app1
andorg.example.app2
apps.
If you want the rule not to be applied to certain apps, start the app name with the ~
sign.
||baddomain.com^$app=~org.example.app
— a rule to block requests that match the specified mask and are sent from any app except for theorg.example.app
.||baddomain.com^$app=~org.example.app1|~org.example.app2
— same as above, but now two apps are excluded:org.example.app1
andorg.example.app2
.
Apps in the modifier value cannot have a wildcard, e.g. $app=com.*.music
. Rules with such modifier are considered invalid.
- Only AdGuard for Windows, Mac, Android are technically capable of using rules with
$app
modifier. - On Windows the process name is case-insensitive starting with AdGuard for Windows with CoreLibs v1.12 or later.
$denyallow
$denyallow
modifier allows to avoid creating additional rules when it is needed to disable a certain rule for specific domains. $denyallow
matches only target domains and not referrer domains.
Adding this modifier to a rule is equivalent to excluding the domains by the rule's matching pattern or to adding the corresponding exclusion rules. To add multiple domains to one rule, use the |
character as a separator.
Examples
This rule:
*$script,domain=a.com|b.com,denyallow=x.com|y.com
is equivalent to this one:
/^(?!.*(x.com|y.com)).*$/$script,domain=a.com|b.com
or to the combination of these three:
*$script,domain=a.com|b.com
@@||x.com$script,domain=a.com|b.com
@@||y.com$script,domain=a.com|b.com
- The rule's matching pattern cannot target any specific domains, e.g. it cannot start with
||
. - Domains in the modifier value cannot be negated, e.g.
$denyallow=~x.com
, or have a wildcard TLD, e.g.$denyallow=x.*
, or be a regular expression, e.g.$denyallow=/\.(com\|org)/
. $denyallow
cannot be used together with$to
. It can be expressed with inverted$to
:$denyallow=a.com|b.com
is equivalent to$to=~a.com|~b.com
.
The rules which violate these restrictions are considered invalid.
Rules with $denyallow
modifier are not supported by AdGuard for iOS, Safari, and AdGuard Content Blocker.
$domain
$domain
limits the rule scope to requests made from the specified domains and their subdomains (as indicated by the Referer HTTP header).
Syntax
The modifier is a list of one or more expressions separated by the |
symbol, each of which is matched against a domain in a particular way depending on its type (see below).
domains = ["~"] entry_0 ["|" ["~"] entry_1 ["|" ["~"]entry_2 ["|" ... ["|" ["~"]entry_N]]]]
entry_i = ( regular_domain / any_tld_domain / regexp )
regular_domain
— a regular domain name (domain.com
). Corresponds the specified domain and its subdomains. It is matched lexicographically.any_tld_domain
— a domain name ending with a wildcard character as a public suffix, e.g. forexample.*
it isco.uk
inexample.co.uk
. Corresponds to the specified domain and its subdomains with any public suffix. It is matched lexicographically.regexp
— a regular expression, starts and ends with/
. The pattern works the same way as in the basic URL rules, but the characters/
,$
, and|
must be escaped with\
.
Rules with $domain
modifier as regular_domain
or any_tld_domain
are supported by all AdGuard products.
Examples
Just $domain
:
||baddomain.com^$domain=example.org
blocks requests that match the specified mask, and are sent from domainexample.org
or its subdomains.||baddomain.com^$domain=example.org|example.com
— the same rule, but it works for bothexample.org
andexample.com
.
If you want the rule not to be applied to certain domains, start a domain name with ~
sign.
$domain
and negation ~
:
||baddomain.com^$domain=example.org
blocks requests that match the specified mask and are sent from theexample.org
domain or its subdomains.||baddomain.com^$domain=example.org|example.com
— the same rule, but it works for bothexample.org
andexample.com
.||baddomain.com^$domain=~example.org
blocks requests matching the pattern sent from any domain exceptexample.org
and its subdomains.||baddomain.com^$domain=example.org|~foo.example.org
blocks requests sent fromexample.org
and its subdomains, except the subdomainfoo.example.org
.||baddomain.com^$domain=/(^\|.+\.)example\.(com\|org)\$/
blocks requests sent fromexample.org
andexample.com
domains and all their subdomains.||baddomain.com^$domain=~a.com|~b.*|~/(^\|.+\.)c\.(com\|org)\$/
blocks requests sent from any domains, excepta.com
,b
with any public suffix (b.com
,b.co.uk
, etc.),c.com
,c.org
, and their subdomains.
$domain
modifier matching target domain:
In some cases the $domain
modifier can match not only the referrer domain, but also the target domain. This happens when all the following conditions are met:
- The request has the
document
content type - The rule pattern does not match any particular domains
- The rule pattern does not contain regular expressions
- The
$domain
modifier contains only excluded domains, e.g.$domain=~example.org|~example.com
The following predicate should be satisfied to perform a target domain matching:
1 AND ((2 AND 3) OR 4)
That is, if the modifier $domain
contains only excluded domains, then the rule does not need to meet the second and third conditions to match the target domain against the modifier $domain
.
If some of the conditions above are not met but the rule contains $cookie
or $csp
modifier, the target domain will still be matched.
If the referrer matches a rule with $domain
that explicitly excludes the referrer domain, then the rule will not be applied even if the target domain also matches the rule. This affects rules with $cookie
and $csp
modifiers, too.
Examples
*$cookie,domain=example.org|example.com
will block cookies for all requests to and fromexample.org
andexample.com
.*$document,domain=example.org|example.com
will block all requests to and fromexample.org
andexample.com
.
In the following examples it is implied that requests are sent from http://example.org/page
(the referrer) and the target URL is http://targetdomain.com/page
.
page$domain=example.org
will be matched, as it matches the referrer domain.page$domain=targetdomain.com
will be matched, as it matches the target domain and satisfies all requirements mentioned above.||*page$domain=targetdomain.com
will not be matched, as the pattern||*page
may match specific domains, e.g.example.page
.||*page$domain=targetdomain.com,cookie
will be matched because the rule contains$cookie
modifier despite the pattern||*page
may match specific domains./banner\d+/$domain=targetdomain.com
will not be matched as it contains a regular expression.page$domain=targetdomain.com|~example.org
will not be matched because the referrer domain is explicitly excluded.
$domain
modifier limitations
Safari does not support the simultaneous use of allowed and disallowed domains, so rules like ||baddomain.com^$domain=example.org|~foo.example.org
will not work in AdGuard for iOS and AdGuard for Safari.
Rules with regular expressions in the $domain
modifier are supported by AdGuard for Windows, AdGuard for Mac, and AdGuard for Android with CoreLibs v1.11 or later, and AdGuard Browser Extension with TSUrlFilter v3.0.0 or later.
In AdGuard for Windows, Mac and Android with CoreLibs v1.12 or later the $domain
modifier can be alternatively spelled as $from
.
$header
The $header
modifier allows matching the HTTP response having a specific header with (optionally) a specific value.
Syntax
$header "=" h_name [":" h_value]
h_value = string / regexp
where:
h_name
— required, an HTTP header name. It is matched case-insensitively.h_value
— optional, an HTTP header value matching expression, it may be one of the following:string
— a sequence of characters. It is matched against the header value lexicographically;regexp
— a regular expression, starts and ends with/
. The pattern works the same way as in the basic URL rules, but the characters/
,$
and,
must be escaped with\
.
The modifier part, ":" h_value
, may be omitted. In that case, the modifier matches the header name only.
Examples
||example.com^$header=set-cookie:foo
blocks requests whose responses have theSet-Cookie
header with the value matchingfoo
literally.||example.com^$header=set-cookie
blocks requests whose responses have theSet-Cookie
header with any value.@@||example.com^$header=set-cookie:/foo\, bar\$/
unblocks requests whose responses have theSet-Cookie
header with value matching thefoo, bar$
regular expression.@@||example.com^$header=set-cookie
unblocks requests whose responses have aSet-Cookie
header with any value.
$header
modifier limitations
- The
$header
modifier can be matched only when headers are received. So if the request is blocked or redirected at an earlier stage, the modifier cannot be applied. - In Adguard Browser Extension, the
$header
modifier is only compatible with$csp
,$removeheader
,$important
, and$badfilter
.
Rules with the $header
modifier are supported by AdGuard for Windows, AdGuard for Mac, and AdGuard for Android with CoreLibs v1.11 or later, and AdGuard Browser Extension with TSUrlFilter v3.0.0 or later.
$important
The $important
modifier applied to a rule increases its priority over rules without the identical modifier. Even over basic exception rules.
Go to rules priorities for more details.
Examples
! blocking rule will block all requests despite of the exception rule
||example.org^$important
@@||example.org^
! if the exception rule also has `$important` modifier, it will prevail and requests won't be blocked
||example.org^$important
@@||example.org^$important
$match-case
This modifier defines a rule which applies only to addresses that match the case. Default rules are case-insensitive.
Examples
*/BannerAd.gif$match-case
— this rule will blockhttp://example.com/BannerAd.gif
, but nothttp://example.com/bannerad.gif
.
Rules with the $match-case
are supported by AdGuard for iOS and AdGuard for Safari with SafariConverterLib v2.0.43 or later.
All other products already support this modifier.
$method
This modifier limits the rule scope to requests that use the specified set of HTTP methods. Negated methods are allowed. The methods must be specified in all lowercase characters, but are matched case-insensitively. To add multiple methods to one rule, use the vertical bar |
as a separator.
Examples
||evil.com^$method=get|head
blocks only GET and HEAD requests toevil.com
.||evil.com^$method=~post|~put
blocks any requests toevil.com
except POST or PUT.@@||evil.com$method=get
unblocks only GET requests toevil.com
.@@||evil.com$method=~post
unblocks any requests toevil.com
except POST.
Rules with mixed negated and not negated values are considered invalid. So, for example, the rule ||evil.com^$method=get|~head
will be rejected.
Rules with $method
modifier are supported by AdGuard for Windows, Mac, and Android with CoreLibs v1.12 or later and AdGuard Browser Extension for Chrome, Firefox, and Edge with TSUrlFilter v2.1.1 or later.
$popup
AdGuard will try to close the browser tab with any address that matches a blocking rule with this modifier. Please note that not all the tabs can be closed.
Examples
||domain.com^$popup
— if you try to go tohttp://domain.com/
from any page in the browser, a new tab in which specified site has to be opened will be closed by this rule.
$popup
modifier limitations
- The
$popup
modifier works best in AdGuard Browser Extension. - In AdGuard for iOS and AdGuard for Safari,
$popup
rules simply block the page right away. - In AdGuard for Windows, AdGuard for Mac, and AdGuard for Android, the
$popup
modifier may not detect a popup in some cases and it will not be blocked. The$popup
modifier applies thedocument
content type with a special flag which is passed to a blocking page. Blocking page itself can do some checks and close the window if it is really a popup. Otherwise, page should be loaded. It can be combined with other request type modifiers, such as$third-party
and$important
.
Rules with the $popup
modifier are not supported by AdGuard Content Blocker.
$third-party
A restriction of third-party and own requests. A third-party request is a request from a different domain. For example, a request to example.org
from domain.com
is a third-party request.
To be considered as such, a third-party request should meet one of the following conditions:
- Its referrer is not a subdomain of the target domain or vice versa. For example, a request to
subdomain.example.org
fromexample.org
is not a third-party request - Its
Sec-Fetch-Site
header is set tocross-site
Examples
$third-party
:
||domain.com^$third-party
— this rule applies to all domains, exceptdomain.com
and its subdomains. An example of a third-party request:http://example.org/banner.jpg
.
If there is a $~third-party
modifier, the rule is only applied to the requests that are not from third parties. Which means, they have to be sent from the same domain.
$~third-party
:
||domain.com$~third-party
— this rule is applied exclusively todomain.com
. Example of a non third-party request:http://domain.com/icon.ico
.
You may use a shorter name (alias) instead of using the full modifier name: $3p
.
$to
$to
limits the rule scope to requests made to the specified domains and their subdomains. To add multiple domains to one rule, use the |
character as a separator.
Examples
/ads$to=evil.com|evil.org
blocks any request toevil.com
orevil.org
and their subdomains with a path matching/ads
./ads$to=~not.evil.com|evil.com
blocks any request toevil.com
and its subdomains, with a path matching/ads
, except requests tonot.evil.com
and its subdomains./ads$to=~good.com|~good.org
blocks any request with a path matching/ads
, except requests togood.com
orgood.org
and their subdomains.
$denyallow
cannot be used together with $to
. It can be expressed with inverted $to
: $denyallow=a.com|b.com
is equivalent to $to=~a.com|~b.com
.
Rules with the $to
modifier are supported by AdGuard for Windows, Mac, and Android with CoreLibs v1.12 or later and AdGuard Browser Extension with TSUrlFilter v2.1.3 or later.
Content-type modifiers
There is a set of modifiers, which can be used to limit the rule's application area to certain type of content. These modifiers can also be combined to cover, for example, both images and scripts.
There is a big difference in how AdGuard determines the content type on different platforms. For AdGuard Browser Extension, content type for every request is provided by the browser. AdGuard for Windows, Mac, and Android use the following method: first, the apps try to determine the type of the request by the Sec-Fetch-Dest
request header or by the filename extension. If the request is not blocked at this stage, the type will be determined using the Content-Type
header at the beginning of the server response.
Examples of content-type modifiers
||example.org^$image
— corresponds to all images fromexample.org
.||example.org^$script,stylesheet
— corresponds to all the scripts and styles fromexample.org
.||example.org^$~image,~script,~stylesheet
— corresponds to all requests toexample.org
except for the images, scripts and styles.
Modifikator \ Produkte | CoreLibs apps | AdGuard for Chromium | AdGuard for Firefox | AdGuard für iOS | AdGuard für Safari | AdGuard-Inhaltsblocker |
---|---|---|---|---|---|---|
$document | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
$font | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
$image | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
$media | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
$object | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
$other | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
$ping | ✅ * | ✅ | ✅ | ❌ | ❌ | ✅ |
$script | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
$stylesheet | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
$subdocument | ✅ * | ✅ | ✅ | ✅ | ✅ | ❌ |
$websocket | ✅ | ✅ | ✅ | ✅ * | ✅ * | ✅ |
$xmlhttprequest | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
$webrtc 🚫 | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ |
$object-subrequest 🚫 | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ |
- ✅ — fully supported
- ✅ * — supported, but reliability may vary or limitations may occur; check the modifier description for more details
- ❌ — not supported
- 🚫 — removed and no longer supported
$document
The rule corresponds to the main frame document requests, i.e. HTML documents that are loaded in the browser tab. It does not match iframes, there is a $subdocument
modifier for these.
By default, AdGuard does not block the requests that are loaded in the browser tab (e.g. "main frame bypass"). The idea is not to prevent pages from loading as the user clearly indicated that they want this page to be loaded. However, if the $document
modifier is specified explicitly, AdGuard does not use that logic and prevents the page load. Instead, it responds with a "blocking page".
If this modifier is used with an exclusion rule (@@
), it completely disables blocking on corresponding pages. It is equivalent to using $elemhide
, $content
, $urlblock
, $jsinject
, $extension
modifiers simultaneously.
Examples
@@||example.com^$document
completely disables filtering on all pages atexample.com
and all subdomains.||example.com^$document
blocks HTML document request toexample.com
with a blocking page.||example.com^$document,redirect=noopframe
redirects HTML document request toexample.com
to an empty html document.||example.com^$document,removeparam=test
removestest
query parameter from HTML document request toexample.com
.||example.com^$document,replace=/test1/test2/
replacestest1
withtest2
in HTML document request toexample.com
.
You may use a shorter name (alias) instead of using the full modifier name: $doc
.
$font
The rule corresponds to requests for fonts, e.g. .woff
filename extension.
$image
The rule corresponds to images requests.
$media
The rule corresponds to requests for media files — music and video, e.g. .mp4
files.
$object
The rule corresponds to browser plugins resources, e.g. Java or Flash.
$other
The rule applies to requests for which the type has not been determined or does not match the types listed above.
$ping
The rule corresponds to requests caused by either navigator.sendBeacon()
or the ping
attribute on links.
AdGuard for Windows, Mac, and Android often cannot accurately detect navigator.sendBeacon()
. Using $ping
is not recommended in the filter lists that are supposed to be used by CoreLibs-based AdGuard products.
Rules with $ping
modifier are not supported by AdGuard for Safari and iOS.
$script
The rule corresponds to script requests, e.g. javascript, vbscript.
$stylesheet
The rule corresponds to CSS files requests.
You may use a shorter name (alias) instead of using the full modifier name: $css
.
$subdocument
The rule corresponds to requests for built-in pages — HTML tags frame
and iframe
.
Examples
||example.com^$subdocument
blocks built-in page requests (frame
andiframe
) toexample.com
and all its subdomains anywhere.||example.com^$subdocument,domain=domain.com
blocks built-in page requests (frame
иiframe
) toexample.com
(and its subdomains) fromdomain.com
and all its subdomains.
You may use a shorter name (alias) instead of using the full modifier name: $frame
.
In AdGuard for Windows, Mac, and Android subdocuments are being detected by the Sec-Fetch-Dest header if it is present. Otherwise, some main pages may be treated as subdocuments.
Rules with $subdocument
modifier are not supported by AdGuard Content Blocker.
$websocket
The rule applies only to WebSocket connections.
$websocket
modifier is supported in all AdGuard products except AdGuard Content Blocker. As for AdGuard for Safari and AdGuard for iOS, it's supported on devices with macOS Monterey (version 12) and iOS 16 or higher.
$xmlhttprequest
The rule applies only to ajax requests (requests sent via javascript object XMLHttpRequest
).
You may use a shorter name (alias) instead of using the full modifier name: $xhr
.
AdGuard for Windows, Mac, Android when filtering older browsers cannot accurately detect this type and sometimes detects it as $other
or $script
. They can only reliably detect this content type when filtering modern browsers that support Fetch metadata request headers.
$object-subrequest
(removed)
$object-subrequest
modifier is removed and is no longer supported. Rules with it are considered as invalid. The rule corresponds to requests by browser plugins (it is usually Flash).
$webrtc
(removed)
This modifier is removed and is no longer supported. Rules with it are considered as invalid. If you need to suppress WebRTC, consider using the nowebrtc
scriptlet.
The rule applies only to WebRTC connections.
Examples
||example.com^$webrtc,domain=example.org
blocks webRTC connections toexample.com
fromexample.org
.@@*$webrtc,domain=example.org
disables the RTC wrapper forexample.org
.
Exception rules modifiers
Exception rules disable the other basic rules for the addresses to which they correspond. They begin with a @@
mark. All the basic modifiers listed above can be applied to them and they also have a few special modifiers.
We recommend to get acquainted with the Adblock Plus filter cheatsheet, for better understanding of how exception rules should be made.
Modifikator \ Produkte | CoreLibs apps | AdGuard for Chromium | AdGuard for Firefox | AdGuard für iOS | AdGuard für Safari | AdGuard-Inhaltsblocker |
---|---|---|---|---|---|---|
$content | ✅ | ❌ | ✅ | ❌ | ❌ | ❌ |
$elemhide | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
$extension | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ |
$jsinject | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
$stealth | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
$urlblock | ✅ | ✅ | ✅ | ✅ * | ✅ * | ❌ |
$genericblock | ✅ | ✅ | ✅ | ✅ * | ✅ * | ❌ |
$generichide | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
$specifichide | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
- ✅ — fully supported
- ✅ * — supported, but reliability may vary or limitations may occur; check the modifier description for more details
- ❌ — not supported
$content
Disables HTML filtering, $hls
, $replace
, and $jsonprune
rules on the pages that match the rule.
Examples
@@||example.com^$content
disables all content modifying rules on pages atexample.com
and all its subdomains.
$elemhide
Disables any cosmetic rules on the pages matching the rule.
Examples
@@||example.com^$elemhide
disables all cosmetic rules on pages atexample.com
and all subdomains.
You may use a shorter name (alias) instead of using the full modifier name: $ehide
.
$extension
Disables specific userscripts or all userscripts for a given domain.
Syntax
$extension[="userscript_name1"[|"userscript_name2"[|"userscript_name3"[...]]]]
userscript_name(i)
stands for a specific userscript name to be disabled by the modifier. The modifier can contain any number of userscript names or none. In the latter case the modifier disables all the userscripts.
Userscript names usually contain spaces or other special characters, which is why you should enclose the name in quotes. Both single ('
) and double ("
) ASCII quotes are supported. Multiple userscript names should be separated with a pipe (|
). However, if a userscript name is a single word without any special characters, it can be used without quotes.
You can also exclude a userscript by adding a ~
character before it. In this case, the userscript will not be disabled by the modifier.
$extension=~"userscript name"
When excluding a userscript, you must place ~
outside the quotes.
If a userscript's name includes quotes ("
), commas (,
), or pipes (|
), they must be escaped with a backslash (\
).
$extension="userscript name\, with \"quote\""
Examples
@@||example.com^$extension="AdGuard Assistant"
disables theAdGuard Assistant
userscript onexample.com
website.@@||example.com^$extension=MyUserscript
disables theMyUserscript
userscript onexample.com
website.@@||example.com^$extension='AdGuard Assistant'|'Popup Blocker'
disables bothAdGuard Assistant
andPopup Blocker
userscripts onexample.com
website.@@||example.com^$extension=~"AdGuard Assistant"
disables all user scripts onexample.com
website, exceptAdGuard Assistant
.@@||example.com^$extension=~"AdGuard Assistant"|~"Popup Blocker"
disables all user scripts onexample.com
website, exceptAdGuard Assistant
andPopup Blocker
.@@||example.com^$extension
no userscript will work on webpages onexample.com
.@@||example.com^$extension="AdGuard \"Assistant\""
disables theAdGuard "Assistant"
userscript onexample.com
website.
- Only AdGuard for Windows, Mac, Android are technically capable of using rules with
$extension
modifier. - Rules with
$extension
modifier with specific userscript name are supported by AdGuard for Windows, Mac, and Android with CoreLibs v1.13 or later.
$jsinject
Forbids adding of javascript code to the page. You can read about scriptlets and javascript rules further.
Examples
@@||example.com^$jsinject
disables javascript on pages atexample.com
and all subdomains.
$stealth
Disables the Stealth Mode module for all corresponding pages and requests.
Syntax
$stealth [= opt1 [| opt2 [| opt3 [...]]]]
opt(i)
stand for certain Stealth Mode options disabled by the modifier. The modifier can contain any number of specific options (see below) or none. In the latter case the modifier disables all the Stealth Mode features.
The list of the available modifier options:
searchqueries
disables Hide your search queries optiondonottrack
disables Ask websites not to track you option3p-cookie
disables Self-destructing third-party cookies option1p-cookie
disables Self-destructing first-party cookies option3p-cache
disables Disable cache for third-party requests option3p-auth
disables Block third-party Authorization header optionwebrtc
disables Block WebRTC optionpush
disables Block Push API optionlocation
disables Block Location API optionflash
disables Block Flash optionjava
disables Block Java optionreferrer
disables Hide Referer from third parties optionuseragent
disables Hide your User-Agent optionip
disables Hide your IP address optionxclientdata
disables Remove X-Client-Data header from HTTP requests optiondpi
disables Protect from DPI option
Examples
@@||example.com^$stealth
disables Stealth Mode forexample.com
(and subdomains) requests, except for blocking cookies and hiding tracking parameters (see below).@@||domain.com^$script,stealth,domain=example.com
disables Stealth Mode only for script requests todomain.com
(and its subdomains) onexample.com
and all its subdomains.@@||example.com^$stealth=3p-cookie|dpi
disables blocking third-party cookies and DPI fooling measures forexample.com
.
Blocking cookies and removing tracking parameters is achieved by using rules with the $cookie
, $urltransform
and $removeparam
modifiers. Exception rules that contain only the $stealth
modifier will not do these things. If you want to completely disable all Stealth mode features for a given domain, you must include all three modifiers: @@||example.org^$stealth,removeparam,cookie
.
- Modifier options must be lowercase, i.e.
$stealth=DPI
will be rejected. - Modifier options cannot be negated, i.e.
$stealth=~3p-cookie
will be rejected. - AdGuard Browser Extension supports only
searchqueries
,donottrack
,referrer
,xclientdata
,1p-cookie
and3p-cookie
options.
- Stealth Mode is available in AdGuard for Windows, AdGuard for Mac, AdGuard for Android, and AdGuard Browser Extension. All other products will ignore the rules with
$stealth
modifier. - Rules with
$stealth
modifier with specific options are supported by AdGuard for Windows, AdGuard for Mac, and AdGuard for Android with CoreLibs v1.10 or later, and AdGuard Browser Extension with TSUrlFilter v3.0.0 or later.
$urlblock
Disables blocking of all requests sent from the pages matching the rule and disables all $cookie
rules.
Examples
@@||example.com^$urlblock
— any requests sent from the pages atexample.com
and all subdomains are not going to be blocked.
In AdGuard for iOS and AdGuard for Safari, rules with $urlblock
work as $document exclusion — unblock everything.
Rules with $urlblock
modifier are not supported by AdGuard Content Blocker.
Generic rules
Before we can proceed to the next modifiers, we have to make a definition of generic rules. The rule is generic if it is not limited to specific domains. Wildcard character *
is supported as well.
For example, these rules are generic:
###banner
*###banner
#@#.adsblock
*#@#.adsblock
~domain.com###banner
||domain.com^
||domain.com^$domain=~example.com
And these are not:
domain.com###banner
||domain.com^$domain=example.com
$genericblock
Disables generic basic rules on pages that correspond to exception rule.
Examples
@@||example.com^$genericblock
disables generic basic rules on any pages atexample.com
and all subdomains.
In AdGuard for iOS and AdGuard for Safari, rules with $genericblock
work as $document exclusion — unblock everything.
Rules with $genericblock
modifier are not supported by AdGuard Content Blocker.
$generichide
Disables all generic cosmetic rules on pages that correspond to the exception rule.
Examples
@@||example.com^$generichide
disables generic cosmetic rules on any pages atexample.com
and its subdomains.
You may use a shorter name (alias) instead of using the full modifier name: $ghide
.
specifichide
Disables all specific element hiding and CSS rules, but not general ones. Has an opposite effect to $generichide
.
Examples
@@||example.org^$specifichide
disablesexample.org##.banner
but not##.banner
.
You may use a shorter name (alias) instead of using the full modifier name: $shide
.
All cosmetic rules — not just specific ones — can be disabled by $elemhide
modifier.
Rules with $specifichide
modifier are not supported by AdGuard for iOS, AdGuard for Safari, and AdGuard Content Blocker.
Advanced capabilities
These modifiers are able to completely change the behavior of basic rules.
Modifikator \ Produkte | CoreLibs apps | AdGuard for Chromium | AdGuard for Firefox | AdGuard für iOS | AdGuard für Safari | AdGuard-Inhaltsblocker |
---|---|---|---|---|---|---|
$all | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
$badfilter | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
$cookie | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
$csp | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
$hls | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ |
$inline-font | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
$inline-script | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
$jsonprune | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ |
$xmlprune | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ |
$network | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ |
$permissions | ✅ * | 🧩 | 🧩 * | ❌ | ❌ | ❌ |
$redirect | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
$redirect-rule | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
$referrerpolicy | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ |
$removeheader | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
$removeparam | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
$replace | ✅ | ❌ | ✅ | ❌ | ❌ | ❌ |
$urltransform | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ |
noop | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
$empty 👎 | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
$mp4 👎 | ✅ | ✅ | ✅ | ❌ | ❌ | ❌ |
- ✅ — fully supported
- ✅ * — supported, but reliability may vary or limitations may occur; check the modifier description for more details
- 🧩 — may already be implemented in nightly or beta versions but is not yet supported in release versions
- ❌ — not supported
- 👎 — deprecated; still supported but will be removed in the future
$all
$all
modifier is made of all content-types modifiers and $popup
. E.g. rule ||example.org^$all
is converting into rule:
||example.org^$document,subdocument,font,image,media,object,other,ping,script,stylesheet,websocket,xmlhttprequest,popup
This modifier cannot be used as an exception with the @@
mark.
Rules with $all
modifier are not supported by AdGuard Content Blocker.
$badfilter
The rules with the $badfilter
modifier disable other basic rules to which they refer. It means that the text of the disabled rule should match the text of the $badfilter
rule (without the $badfilter
modifier).
Examples
||example.com$badfilter
disables||example.com
||example.com$image,badfilter
disables||example.com$image
@@||example.com$badfilter
disables@@||example.com
||example.com$domain=domain.com,badfilter
disables||example.com$domain=domain.com
Rules with $badfilter
modifier can disable other basic rules for specific domains if they fulfill the following conditions:
- The rule has a
$domain
modifier - The rule does not have a negated domain
~
in$domain
modifier value
In that case, the $badfilter
rule will disable the corresponding rule for domains specified in both the $badfilter
and basic rules. Please note that wildcard-TLD logic works here as well.
Examples
/some$domain=example.com|example.org|example.io
is disabled forexample.com
by/some$domain=example.com,badfilter
/some$domain=example.com|example.org|example.io
is disabled forexample.com
andexample.org
by/some$domain=example.com|example.org,badfilter
/some$domain=example.com|example.org
and/some$domain=example.io
are disabled completely by/some$domain=example.com|example.org|example.io,badfilter
/some$domain=example.com|example.org|example.io
is disabled completely by/some$domain=example.*,badfilter
/some$domain=example.*
is disabled forexample.com
andexample.org
by/some$domain=example.com|example.org,badfilter
/some$domain=example.com|example.org|example.io
is NOT disabled forexample.com
by/some$domain=example.com|~example.org,badfilter
because the value of$domain
modifier contains a negated domain
Rules with $badfilter
modifier are not supported by AdGuard Content Blocker.
$cookie
The $cookie
modifier completely changes rule behavior. Instead of blocking a request, this modifier makes AdGuard suppress or modify the Cookie
and Set-Cookie
headers.
Multiple rules matching a single request
In case if multiple $cookie
rules match a single request, we will apply each of them one by one.
Syntax
$cookie [= name[; maxAge = seconds [; sameSite = strategy ]]]
where:
name
— optional, string or regular expression to match cookie name.seconds
— number of seconds for current time to offset the expiration date of cookie.strategy
— string for Same-Site strategy to make cookie use.
For example,
||example.org^$cookie=NAME;maxAge=3600;sameSite=lax
every time AdGuard encounters a cookie called NAME
in a request to example.org
, it will do the following:
- Set its expiration date to current time plus
3600
seconds - Makes the cookie use Same-Site "lax" strategy.
Escaping special characters
If regular expression name
is used for matching, two characters must be escaped: comma ,
and dollar sign $
. Use backslash \
escape each of them. For example, escaped comma looks like this: \,
.
Examples
||example.org^$cookie
blocks all cookies set byexample.org
; this is an equivalent to settingmaxAge=0
$cookie=__cfduid
blocks CloudFlare cookie everywhere$cookie=/__utm[a-z]/
blocks Google Analytics cookies everywhere||facebook.com^$third-party,cookie=c_user
prevents Facebook from tracking you even if you are logged in
There are two methods to deactivate $cookie
rules: the primary method involves using an exception marked with @@
— @@||example.org^$cookie
. The alternative method employs a $urlblock
exception (also included under the $document
exception alias — $elemhide,jsinject,content,urlblock,extension
). Here's how it works:
@@||example.org^$cookie
unblocks all cookies set byexample.org
@@||example.org^$urlblock
unblocks all cookies set byexample.org
and disables blocking of all requests sent fromexample.org
@@||example.org^$cookie=concept
unblocks a single cookie namedconcept
@@||example.org^$cookie=/^_ga_/
unblocks every cookie that matches the regular expression
$cookie
rules support three types of modifiers: $domain
, $~domain
, $important
, $third-party
, and $~third-party
.
Rules with $cookie
modifier are not supported by AdGuard Content Blocker, AdGuard for iOS, and AdGuard for Safari.
$csp
This modifier completely changes the rule behavior. If it is applied to a rule, the rule will not block the matching request. Response headers will be modified instead.
In order to use this type of rules, it is required to have the basic understanding of the Content Security Policy security layer.
For the requests matching a $csp
rule, we will strengthen response security policy by enhancing the content security policy, similar to the content security policy of the $csp
modifier contents. $csp
rules are applied independently from any other rule type. Only document-level exceptions can influence it (see the examples section), but no other basic rules.
Multiple rules matching a single request
In case if multiple $csp
rules match a single request, we will apply each of them.
Syntax
$csp
value syntax is similar to the Content Security Policy header syntax.
$csp
value can be empty in the case of exception rules. See examples section below.
Examples
||example.org^$csp=frame-src 'none'
blocks all frames on example.org and its subdomains.@@||example.org/page/*$csp=frame-src 'none'
disables all rules with the$csp
modifier exactly matchingframe-src 'none'
on all the pages matching the rule pattern. For instance, the rule above.@@||example.org/page/*$csp
disables all the$csp
rules on all the pages matching the rule pattern.||example.org^$csp=script-src 'self' 'unsafe-eval' http: https:
disables inline scripts on all the pages matching the rule pattern.@@||example.org^$document
or@@||example.org^$urlblock
disables all the$csp
rules on all the pages matching the rule pattern.
- There are a few characters forbidden in the
$csp
value:,
,$
. $csp
rules support limited list of modifiers:$domain
,$important
,$subdocument
.- Rules with
report-*
directives are considered invalid.
Rules with $csp
modifier are not supported by AdGuard Content Blocker, AdGuard for iOS and AdGuard for Safari.
$hls
$hls
rules modify the response of a matching request. They are intended as a convenient way to remove segments from HLS playlists (RFC 8216).
The word "segment" in this document means either a "Media Segment" or a "playlist" as part of a "Master Playlist": $hls
rules do not distinguish between a "Master Playlist" and a "Media Playlist".
Syntax
||example.org^$hls=urlpattern
removes segments whose URL matches the URL patternurlpattern
. The pattern works just like the one in basic URL rules, however, the characters/
,$
and,
must be escaped with\
insideurlpattern
.||example.org^$hls=/regexp/options
removes segments where the URL or one of the tags (for certain options, if present) is matched by the regular expressionregexp
. Availableoptions
are:t
— instead of testing the segment's URL, test each of the segment's tags against the regular expression. A segment with a matching tag is removed;i
— make the regular expression case-insensitive.
The characters /
, $
and ,
must be escaped with \
inside regexp
.
Exceptions
Basic URL exceptions shall not disable rules with $hls
modifier. They can be disabled as described below:
@@||example.org^$hls
disables all$hls
rules for responses from URLs matching||example.org^
.@@||example.org^$hls=text
disables all$hls
rules with the value of the$hls
modifier equal totext
for responses from URLs matching||example.org^
.
$hls
rules can also be disabled by $document
, $content
and $urlblock
exception rules.
When multiple $hls
rules match the same request, their effect is cumulative.
Examples
||example.org^$hls=\/videoplayback^?*&source=dclk_video_ads
removes all segments with the matching URL.||example.org^$hls=/\/videoplayback\/?\?.*\&source=dclk_video_ads/i
achieves more or less the same with a regular expression instead of a URL pattern.||example.org^$hls=/#UPLYNK-SEGMENT:.*\,ad/t
removes all segments which have the matching tag.
Anatomy of an HLS playlist
A quick summary of the specification:
- An HLS playlist is a collection of text lines
- A line may be empty, a comment (starts with
#
), a tag (also starts with#
, can only be recognized by name) or a URL - A URL line is called a "segment"
- Tags may apply to a single segment, i.e. the first URL line after the tag, to all segments following the tag and until the tag with the same name, or to the whole playlist
Some points specific to the operation of $hls
rules:
- When a segment is removed, all of the tags that apply only to that segment are also removed
- When there is a tag that applies to multiple segments, and all of those segments are removed, the tag is also removed
- Since there is no way to recognize different kinds of tags by syntax, we recognize all of the tags specified by the RFC, plus some non-standard tags that we have seen in the field. Any lines starting with
#
and not recognized as a tag are passed through without modification, and are not matched against the rules - Tags will not be matched if they apply to the entire playlist, and
$hls
rrules cannot be used to remove them, as these rule types are intended for segment removals. If you know what you are doing, you can use$replace
rules to remove or rewrite just a single tag from the playlist
An example of a transformation done by the rules:
Original response
#EXTM3U
#EXT-X-TARGETDURATION:10
#EXTINF,5
preroll.ts
#UPLYNK-SEGMENT:abc123,ad
#UPLYNK-KEY:aabb1122
#EXT-X-DISCONTINUITY
#EXTINF,10
01.ts
#EXTINF,10
02.ts
#UPLYNK-SEGMENT:abc123,segment
#UPLYNK-KEY:ccdd2233
#EXT-X-DISCONTINUITY
#EXTINF,10
01.ts
#EXTINF,10
02.ts
#EXT-X-ENDLIST
Applied rules
||example.org^$hls=preroll
||example.org^$hls=/#UPLYNK-SEGMENT:.*\,ad/t
Modified response
#EXTM3U
#EXT-X-TARGETDURATION:10
#UPLYNK-SEGMENT:abc123,segment
#UPLYNK-KEY:ccdd2233
#EXT-X-DISCONTINUITY
#EXTINF,10
01.ts
#EXTINF,10
02.ts
#EXT-X-ENDLIST
$hls
rules are only allowed in trusted filters.$hls
rules are compatible with the modifiers$domain
,$third-party
,$app
,$important
,$match-case
, and$xmlhttprequest
only.$hls
rules only apply to HLS playlists, which are UTF-8 encoded text starting with the line#EXTM3U
. Any other response will not be modified by these rules.$hls
rules do not apply if the size of the original response is more than 10 MB.
Rules with the $hls
modifier are supported by AdGuard for Windows, Mac, and Android with CoreLibs v1.10 or later.
$inline-script
The $inline-script
modifier is designed to block inline JavaScript embedded into the web page, using Content Security Policy (CSP). It improves security and privacy by preventing application of inline ads or potentially malicious scripts. The rule ||example.org^$inline-script
is converting into the CSP-syntax rule:
||example.org^$csp=script-src 'self' 'unsafe-eval' http: https: data: blob: mediastream: filesystem:
$inline-font
The $inline-font
modifier is designed to block inline fonts embedded into the web page, using Content Security Policy (CSP). It improves security and privacy by preventing application of inline fonts that could be used for data collection and fingerprinting. The rule ||example.org^$inline-font
is converting into the CSP-syntax rule:
||example.org^$csp=font-src 'self' 'unsafe-eval' http: https: data: blob: mediastream: filesystem:
$jsonprune
$jsonprune
rules modify the response to a matching request by removing JSON items that match a modified JSONPath expression. They do not modify responses which are not valid JSON documents.
In AdGuard for Windows, Mac, and Android with CoreLibs v1.11 or later, $jsonprune
also supports modifying JSONP (padded JSON) documents.
Syntax
||example.org^$jsonprune=expression
removes items that match the modified JSONPath expressionexpression
from the response.
Due to the way rule parsing works, the characters $
and ,
must be escaped with \
inside expression
.
The modified JSONPath syntax has the following differences from the original:
- Script expressions are not supported
- The supported filter expressions are:
?(has <key>)
— true if the current object has the specified key?(key-eq <key> <value>)
— true if the current object has the specified key, and its value is equal to the specified value?(key-substr <key> <value>)
— true if the specified value is a substring of the value of the specified key of the current object
- Whitespace outside of double- or single-quoted strings has no meaning
- Both double- and single-quoted strings can be used
- Expressions ending with
..
are not supported - Multiple array slices can be specified in square brackets
There are various online tools that make working with JSONPath expressions more convenient:
https://www.site24x7.com/tools/jsonpath-finder-validator.html https://jsonpathfinder.com/ https://jsonpath.com/
Keep in mind, though, that all JSONPath implementations have unique features/quirks and are subtly incompatible with each other.
Exceptions
Basic URL exceptions shall not disable rules with the $jsonprune
modifier. They can be disabled as described below:
@@||example.org^$jsonprune
disables all$jsonprune
rules for responses from URLs matching||example.org^
.@@||example.org^$jsonprune=text
disables all$jsonprune
rules with the value of the$jsonprune
modifier equal totext
for responses from URLs matching||example.org^
.
$jsonprune
rules can also be disabled by $document
, $content
and $urlblock
exception rules.
When multiple $jsonprune
rules match the same request, they are sorted in lexicographical order, the first rule is applied to the original response, and each of the remaining rules is applied to the result of applying the previous one.
Examples
||example.org^$jsonprune=\$..[one\, "two three"]
removes all occurrences of the keys "one" and "two three" anywhere in the JSON document.
Input
{
"one": 1,
"two": {
"two three": 23,
"four five": 45
}
}
Output
{
"two": {
"four five": 45
}
}
||example.org^$jsonprune=\$.a[?(has ad_origin)]
removes all children ofa
that have anad_origin
key.
Input
{
"a": [
{
"ad_origin": "https://example.org",
"b": 42
},
{
"b": 1234
}
]
}
Output
{
"a": [
{
"b": 1234
}
]
}
||example.org^$jsonprune=\$.*.*[?(key-eq 'Some key' 'Some value')]
removes all items that are at nesting level 3 and have a property "Some key" equal to "Some value".
Input
{
"a": {"b": {"c": {"Some key": "Some value"}, "d": {"Some key": "Other value"}}},
"e": {"f": [{"Some key": "Some value"}, {"Some key": "Other value"}]}
}
Output
{
"a": {"b": {"d": {"Some key": "Other value"}}},
"e": {"f": [{"Some key": "Other value"}]}
}
Nested JSONPath expressions
In AdGuard for Windows, Mac and Android with CoreLibs v1.11 or later, JSONPath expressions may be used as keys in filter expressions.
||example.org^$jsonprune=\$.elems[?(has "\$.a.b.c")]
removes all children ofelems
which have a property selectable by the JSONPath expression$.a.b.c
.
Input
{
"elems": [
{
"a": {"b": {"c": 123}},
"k": "v"
},
{
"d": {"e": {"f": 123}},
"k1": "v1"
}
]
}
Output
{
"elems": [
{
"d": {"e": {"f": 123}},
"k1": "v1"
}
]
}
||example.org^$jsonprune=\$.elems[?(key-eq "\$.a.b.c" "abc")]
removes all children ofelems
which have a property selectable by the JSONPath expression$.a.b.c
with a value equal to"abc"
.
Input
{
"elems": [
{
"a": {"b": {"c": 123}},
},
{
"a": {"b": {"c": "abc"}}
}
]
}
Output
{
"elems": [
{
"a": {"b": {"c": 123}}
}
]
}
$jsonprune
rules are only compatible with these modifiers:$domain
,$third-party
,$app
,$important
,$match-case
, and$xmlhttprequest
.$jsonprune
rules do not apply if the size of the original response is greater than 10 MB.
Rules with the $jsonprune
modifier are supported by AdGuard for Windows, Mac, and Android with CoreLibs v1.10 or later.
$xmlprune
$xmlprune
rules modify the response to a matching request by removing XML items that match an XPath 1.0 expression. The expression must return a node-set. $xmlprune
rules do not modify responses which are not well-formed XML documents.
Syntax
||example.org^$xmlprune=expression
removes items that match the XPath expressionexpression
from the response.
Due to the way rule parsing works, the characters $
and ,
must be escaped with \
inside expression
.
Exceptions
Basic URL exceptions shall not disable rules with the $xmlprune
modifier. They can be disabled as described below:
@@||example.org^$xmlprune
disables all$xmlprune
rules for responses from URLs matching||example.org^
.@@||example.org^$xmlprune=text
disables all$xmlprune
rules with the value of the$xmlprune
modifier equal totext
for responses from URLs matching||example.org^
.
$xmlprune
rules can also be disabled by $document
, $content
and $urlblock
exception rules.
When multiple $xmlprune
rules match the same request, they are applied in lexicographical order.
Examples
||example.org^$xmlprune=/bookstore/book[position() mod 2 = 1]
removes odd-numbered books from the bookstore.
Input
<?xml version="1.0" encoding="UTF-8"?>
<bookstore>
<book category="cooking">
<title lang="en">Everyday Italian</title>
<author>Giada De Laurentiis</author>
<year>2005</year>
<price>30.00</price>
</book>
<book category="children">
<title lang="en">Harry Potter</title>
<author>J K. Rowling</author>
<year>2005</year>
<price>29.99</price>
</book>
<book category="web">
<title lang="en">XQuery Kick Start</title>
<author>James McGovern</author>
<author>Per Bothner</author>
<author>Kurt Cagle</author>
<author>James Linn</author>
<author>Vaidyanathan Nagarajan</author>
<year>2003</year>
<price>49.99</price>
</book>
<book category="web">
<title lang="en">Learning XML</title>
<author>Erik T. Ray</author>
<year>2003</year>
<price>39.95</price>
</book>
</bookstore>
Output
<?xml version="1.0" encoding="UTF-8"?>
<bookstore>
<book category="children">
<title lang="en">Harry Potter</title>
<author>J K. Rowling</author>
<year>2005</year>
<price>29.99</price>
</book>
<book category="web">
<title lang="en">Learning XML</title>
<author>Erik T. Ray</author>
<year>2003</year>
<price>39.95</price>
</book>
</bookstore>
||example.org^$xmlprune=/bookstore/book[year = 2003]
removes books from the year 2003 from the bookstore.
Input
<?xml version="1.0" encoding="UTF-8"?>
<bookstore>
<book category="cooking">
<title lang="en">Everyday Italian</title>
<author>Giada De Laurentiis</author>
<year>2005</year>
<price>30.00</price>
</book>
<book category="children">
<title lang="en">Harry Potter</title>
<author>J K. Rowling</author>
<year>2005</year>
<price>29.99</price>
</book>
<book category="web">
<title lang="en">XQuery Kick Start</title>
<author>James McGovern</author>
<author>Per Bothner</author>
<author>Kurt Cagle</author>
<author>James Linn</author>
<author>Vaidyanathan Nagarajan</author>
<year>2003</year>
<price>49.99</price>
</book>
<book category="web">
<title lang="en">Learning XML</title>
<author>Erik T. Ray</author>
<year>2003</year>
<price>39.95</price>
</book>
</bookstore>
Output
<?xml version="1.0" encoding="UTF-8"?>
<bookstore>
<book category="cooking">
<title lang="en">Everyday Italian</title>
<author>Giada De Laurentiis</author>
<year>2005</year>
<price>30.00</price>
</book>
<book category="children">
<title lang="en">Harry Potter</title>
<author>J K. Rowling</author>
<year>2005</year>
<price>29.99</price>
</book>
</bookstore>
||example.org^$xmlprune=//*/@*
removes all attributes from all elements.
Input
<?xml version="1.0" encoding="UTF-8"?>
<bookstore location="cy">
<book category="cooking">
<title lang="en">Everyday Italian</title>
<author>Giada De Laurentiis</author>
<year>2005</year>
<price>30.00</price>
</book>
</bookstore>
Output
<?xml version="1.0" encoding="UTF-8"?>
<bookstore>
<book>
<title>Everyday Italian</title>
<author>Giada De Laurentiis</author>
<year>2005</year>
<price>30.00</price>
</book>
</bookstore>
$xmlprune
rules are only compatible with these modifiers:$domain
,$third-party
,$app
,$important
,$match-case
, and$xmlhttprequest
.$xmlprune
rules do not apply if the size of the original response is greater than 10 MB.
Rules with the $xmlprune
modifier are supported by AdGuard for Windows, Mac, and Android with CoreLibs v1.15 or later.
$network
This is basically a Firewall-like rule allowing to fully block or unblock access to a specified remote address.
$network
rules match IP addresses only! You cannot use them to block or unblock access to a domain.- To match an IPv6 address, you have to use the collapsed syntax, e.g. use
[2001:4860:4860::8888]$network
instead of[2001:4860:4860:0:0:0:0:8888]$network
. - An allowlist
$network
rule makes AdGuard bypass data to the matching endpoint, hence there will be no further filtering at all. - If the IP part starts and ends with
/
character, it is treated as a regular expression.
We recommend to get acquainted with this article for better understanding of regular expressions.
The $network
modifier can only be used in rules together with the $app
and $important
modifiers, and not with any other modifiers.
Examples
174.129.166.49:3478^$network
blocks access to174.129.166.49:3478
(but not to174.129.166.49:34788
).[2001:4860:4860::8888]:443^$network
blocks access to[2001:4860:4860::8888]:443
.174.129.166.49$network
blocks access to174.129.166.49:*
.@@174.129.166.49$network
makes AdGuard bypass data to the endpoint. No other rules will be applied./.+:3[0-9]{4}/$network
blocks access to any port from 30000 to 39999./8.8.8.(:?8|4)/$network
blocks access to both8.8.8.8
and8.8.8.4
.
Only AdGuard for Windows, Mac, and Android are technically capable of using rules with $network
modifier.
$permissions
This modifier completely changes the rule behavior. If it is applied to a rule, the rule will not block the matching request. Response headers will be modified instead.
In order to use this type of rules, it is required to have a basic understanding of the Permissions Policy security layer.
For the requests matching a $permissions
rule, AdGuard strengthens response's permissions policy by adding an additional permission policy equal to the $permissions
modifier contents. $permissions
rules are applied independently from any other rule type. Only document-level exceptions can influence it (see the examples section), but no other basic rules.
Syntax
$permissions
value syntax is identical to that of the Permissions-Policy
header syntax with the following exceptions:
- A comma that separates multiple features MUST be escaped — see examples below.
- A pipe character (
|
) can be used to separate features instead of a comma.
The list of available directives is available here.
$permissions
value can be empty in the case of exception rules — see examples below.
Examples
||example.org^$permissions=autoplay=()
disallows autoplay media requested through theHTMLMediaElement
interface acrossexample.org
.@@||example.org/page/*$permissions=autoplay=()
disables all rules with the$permissions
modifier exactly matchingautoplay=()
on all the pages matching the rule pattern. For instance, the rule above. It is important to note that the exception rule only takes effect in the case of an exact value match. For example, if you want to disable the rule$permissions=a=()\,b=()
, you need exception rule@@$permissions=a=()\,b=()
, and not@@$permissions=b=()\,a=()
, nor@@$permissions=b=()
becauseb=()\,a=()
orb=()
does not match witha=()\,b=()
.@@||example.org/page/*$permissions
disables all the$permissions
rules on all the pages matching the rule pattern.$domain=example.org|example.com,permissions=storage-access=()\, camera=()
disallows using the Storage Access API to request access to unpartitioned cookies and using video input devices acrossexample.org
andexample.com
.$domain=example.org|example.com,permissions=storage-access=()|camera=()
does the same — a|
can be used to separate the features instead of an escaped comma.@@||example.org^$document
or@@||example.org^$urlblock
disables all the$permission
rules on all the pages matching the rule pattern.
$permissions
rules only take effect for main frame and sub frame requests. This means they are applied when a page is loaded or when an iframe is loaded.
If there are multiple $permissions
rules that match the same request, multiple Permissions-Policy
headers will be added to the response for each rule with their $permissions
value. So if you have two rules: ||example.org^$permissions=autoplay=()
and ||example.org^$permissions=geolocation=()\,camera=()
that match the same request, the response will contain two Permissions-Policy
headers: autoplay=()
and geolocation=()\,camera=()
.
$permissions
modifier limitations
Firefox ignores the Permissions-Policy
header. For more information, see this issue.
- Characters forbidden in the
$permissions
value:$
. $permissions
is compatible with a limited set of modifiers:$domain
,$important
,$subdocument
, and content-type modifiers.$permissions
rules that do not have any content-type modifiers will match only requests where content type isdocument
.
- Rules with the
$permissions
modifier are supported by AdGuard for Windows, Mac, and Android with CoreLibs v1.11 or later, and AdGuard Browser Extension with TSUrlFilter v3.0.0 or later. - Pipe separator
|
instead of escaped comma is supported by AdGuard for Windows, Mac, and Android with CoreLibs v1.14 or later, and AdGuard Browser Extension with TSUrlFilter v3.0.0 or later.
$redirect
AdGuard is able to redirect web requests to a local "resource".
Syntax
AdGuard uses the same filtering rule syntax as uBlock Origin. Also, it is compatible with ABP $rewrite=abp-resource
modifier.
$redirect
is a modifier for the basic filtering rules so rules with this modifier support all other basic modifiers like $domain
, $third-party
, $script
, etc.
The value of the $redirect
modifier must be the name of the resource that will be used for redirection.
Disabling $redirect
rules
||example.org/script.js$script,redirect=noopjs
— this rule redirects all requests toexample.org/script.js
to the resource namednoopjs
.||example.org/test.mp4$media,redirect=noopmp4-1s
— this rule redirects all requests toexample.org/test.mp4
to the resource namednoopmp4-1s
.@@||example.org^$redirect
will disable all$redirect
rules for URLs that match||example.org^
.@@||example.org^$redirect=nooptext
will disable all rules with$redirect=nooptext
for any request that matches||example.org^
.
More information on redirects and their usage is available on GitHub.
Priorities of $redirect
rules
$redirect
rules have higher priority than regular basic blocking rules. This means that if there is a basic blocking rule, the $redirect
rule will override it. Allowlist rules with @@
mark have higher priority than $redirect
rules. If a basic rule with the $important
modifier and the $redirect
rule matches the same URL, the latter is overridden unless it's also marked as $important
.
In short: $important
> @@
> $redirect
> basic rules
.
Go to rules priorities for more details.
- Rules with
$redirect
modifier are not supported by AdGuard Content Blocker, AdGuard for iOS and AdGuard for Safari. $redirect
in uBlock Origin supports specifying priority, e.g.$redirect=noopjs:42
. AdGuard does not support it and instead just discards the priority postfix.
$redirect-rule
This is basically an alias to $redirect
since it has the same "redirection" values and the logic is almost similar. The difference is that $redirect-rule
is applied only in the case when the target request is blocked by a different basic rule.
Go to rules priorities for more details.
Negating $redirect-rule
works exactly the same way as for regular $redirect
rules. Even more than that, @@||example.org^$redirect
will negate both $redirect
and $redirect-rule
rules.
Examples
||example.org/script.js
||example.org^$redirect-rule=noopjs
In this case, only requests to example.org/script.js
will be "redirected" to noopjs
. All other requests to example.org
will be kept intact.
Rules with $redirect-rule
modifier are not supported by AdGuard Content Blocker, AdGuard for iOS, and AdGuard for Safari.
$referrerpolicy
These rules allow overriding of a page's referrer policy. Responses to matching requests will have all of their Referrer-Policy
headers replaced with a single header with the value equal to the matching rule's modifier value. If the response carries an HTML document with a <meta name="referrer"...
tag, the content
attribute of the tag will also be replaced with the modifier value.
An exception rule with a modifier value disables the blocking rule with the same modifier value. An exception rule without a modifier value disables all matched referrer-policy rules.
If a request matches multiple $referrerpolicy
rules not disabled by exceptions, only one of them (it is not specified which one) is applied. $referrerpolicy
rules without specified content-type modifiers apply to $document
and $subdocument
content types.
Examples
||example.com^$referrerpolicy=unsafe-url
overrides the referrer policy forexample.com
withunsafe-url
.@@||example.com^$referrerpolicy=unsafe-url
disables the previous rule.@@||example.com/abcd.html^$referrerpolicy
disables all$referrerpolicy
rules onexample.com/abcd.html
.
Rules with the $referrerpolicy
modifier are supported by AdGuard for Windows, Mac, and Android with CoreLibs v1.12 or later.
$removeheader
Rules with $removeheader
modifier are intended to remove headers from HTTP requests and responses. The initial motivation for this rule type is to be able to get rid of the Refresh
header which is often used to redirect users to an undesirable location. However, this is not the only case where this modifier can be useful.
Just like $csp
, $redirect
, $removeparam
, and $cookie
, this modifier exists independently, rules with it do not depend on the regular basic rules, i.e. regular exception or blocking rules will not affect it. By default, it only affects response headers. However, you can also change it to remove headers from HTTP requests as well.
Syntax
Basic syntax
||example.org^$removeheader=header-name
removes a response header calledheader-name
||example.org^$removeheader=request:header-name
removes a request header calledheader-name
$removeheader
is case-insensitive, but we suggest always using lower case.
Negating $removeheader
This type of rules works pretty much the same way it works with $csp
and $redirect
modifiers.
Use @@
to negate $removeheader
:
@@||example.org^$removeheader
negates all$removeheader
rules for URLs that match||example.org^
.@@||example.org^$removeheader=header
negates the rule with$removeheader=header
for any request matching||example.org^
.
$removeheader
rules can also be disabled by $document
and $urlblock
exception rules. But basic exception rules without modifiers will not do that. For example, @@||example.com^
will not disable $removeheader=p
for requests to example.com
, but @@||example.com^$urlblock
will.
In case of multiple $removeheader
rules matching a single request, we will apply each of them one by one.
Examples
||example.org^$removeheader=refresh
removesRefresh
header from all HTTP responses returned byexample.org
and its subdomains.||example.org^$removeheader=request:x-client-data
removesX-Client-Data
header from all HTTP requests.Next block of rules removes
Refresh
andLocation
headers from all HTTP responses returned byexample.org
save for requests toexample.org/path/*
, for which no headers will be removed:||example.org^$removeheader=refresh
||example.org^$removeheader=location
@@||example.org/path/$removeheader
This type of rules can be used only in trusted filters.
In order to avoid compromising the security
$removeheader
cannot remove headers from the list below:access-control-allow-origin
access-control-allow-credentials
access-control-allow-headers
access-control-allow-methods
access-control-expose-headers
access-control-max-age
access-control-request-headers
access-control-request-method
origin
timing-allow-origin
allow
cross-origin-embedder-policy
cross-origin-opener-policy
cross-origin-resource-policy
content-security-policy
content-security-policy-report-only
expect-ct
feature-policy
origin-isolation
strict-transport-security
upgrade-insecure-requests
x-content-type-options
x-download-options
x-frame-options
x-permitted-cross-domain-policies
x-powered-by
x-xss-protection
public-key-pins
public-key-pins-report-only
sec-websocket-key
sec-websocket-extensions
sec-websocket-accept
sec-websocket-protocol
sec-websocket-version
p3p
sec-fetch-mode
sec-fetch-dest
sec-fetch-site
sec-fetch-user
referrer-policy
content-type
content-length
accept
accept-encoding
host
connection
transfer-encoding
upgrade
$removeheader
rules are only compatible with$domain
,$third-party
,$app
,$important
,$match-case
, and content type modifiers such as$script
and$stylesheet
. The rules which have any other modifiers are considered invalid and will be discarded.
Rules with $removeheader
modifier are supported by AdGuard for Windows, Mac, and Android, and AdGuard Browser Extension for Chrome, Firefox, and Edge.
$removeparam
$queryprune
is an alias of $removeparam
. Since $queryprune
is deprecated, avoid using it and use $removeparam
instead.
Rules with $removeparam
modifier are intended to strip query parameters from requests' URLs. Please note that such rules are only applied to GET
, HEAD
, OPTIONS
, and sometimes POST
requests.
$removeparam
rules that do not have any content type modifiers will match only requests where content type is document
.
Syntax
Basic syntax
$removeparam=param
removes query parameter with the nameparam
from URLs of any request, e.g. a request tohttp://example.com/page?param=1&another=2
will be transformed intohttp://example.com/page?another=2
.
Rules with $removeparam
modifier are supported by AdGuard for Windows, Mac and, Android with CoreLibs v1.7 or later and AdGuard Browser Extension v3.6 or later.
Regular expressions
You can also use regular expressions to match query parameters and/or their values:
$removeparam=/regexp/[options]
— removes query parameters that matches theregexp
regular expression from URLs of any request. Unlike basic syntax, it means "remove query parameters normalized to aname=value
string which match theregexp
regular expression".[options]
here is the list of regular expression options. At the moment, the only supported option isi
which makes matching case-insensitive.
Escaping special characters
Do not forget to escape special characters like ,
, /
and $
in the regular expressions. Use \
character for that purpose. For example, an escaped comma should look like this: \,
.
Regexp-type rules target both name and value of the parameter. To minimize mistakes, it is safer to start every regexp with /^
unless you specifically target parameter values.
We will try to detect and ignore unescaped $
automatically using a simple rule of thumb — it is not an options delimiter if all three are true:
- It looks like
$/
- There is another slash character
/
to the left of it - There is another unescaped dollar sign
$
to the left of that slash character
Remove all query parameters
Specify naked $removeparam
to remove all query parameters:
||example.org^$removeparam
— removes all query parameters from URLs matching||example.org^
.
Inversion
Use ~
to apply inversion:
$removeparam=~param
— removes all query parameters with the name different fromparam
.$removeparam=~/regexp/
— removes all query parameters that do not match theregexp
regular expression.
Negating $removeparam
This sort of rules work pretty much the same way it works with $csp
and $redirect
modifiers.
Use @@
to negate $removeparam
:
@@||example.org^$removeparam
negates all$removeparam
rules for URLs that match||example.org^
.@@||example.org^$removeparam=param
negates the rule with$removeparam=param
for any request matching||example.org^
.@@||example.org^$removeparam=/regexp/
negates the rule with$removeparam=/regexp/
for any request matching||example.org^
.
Multiple rules matching a single request
In the case when multiple $removeparam
rules match a single request, each of them will be applied one by one.
Examples
$removeparam=/^(utm_source|utm_medium|utm_term)=/
$removeparam=/^(utm_content|utm_campaign|utm_referrer)=/
@@||example.com^$removeparam
With these rules some UTM parameters will be stripped out from any request, except that requests to example.com
will not be stripped at all, e.g. http://google.com/page?utm_source=s&utm_referrer=fb.com&utm_content=img
will be transformed to http://google.com/page
, but http://example.com/page?utm_source=s&utm_referrer=fb.com&utm_content=img
will not be affected by the blocking rule.
$removeparam=utm_source
removesutm_source
query parameter from all requests.$removeparam=/utm_.*/
removes allutm_* query
parameters from URL queries of any request, e.g. a request tohttp://example.com/page?utm_source=test
will be transformed tohttp://example.com/page
.$removeparam=/^utm_source=campaign$/
removesutm_source
query parameter with the value equal tocampaign
. It does not touch otherutm_source
parameters.
Negating one $removeparam
rule and replacing it with a different rule
$removeparam=/^(gclid|yclid|fbclid)=/
@@||example.com^$removeparam=/^(gclid|yclid|fbclid)=/
||example.com^$removeparam=/^(yclid|fbclid)=/
With these rules, Google, Yandex, and Facebook Click IDs will be removed from all requests. There is one exception: Google Click ID (gclid) will not be removed from requests to example.com.
Negating $removeparam
for all parameters
$removeparam=/^(utm_source|utm_medium|utm_term)=/
$removeparam=/^(utm_content|utm_campaign|utm_referrer)=/
@@||example.com^$removeparam
With these rules, specified UTM parameters will be removed from any request save for requests to example.org
.
$removeparam
rules can also be disabled by $document
and $urlblock
exception rules. But basic exception rules without modifiers do not do that. For example, @@||example.com^
will not disable $removeparam=p
for requests to example.com, but @@||example.com^$urlblock
will.
- Rules with
$removeparam
modifier can be used only in trusted filters. $removeparam
rules are compatible with basic modifiers, content-type modifiers, and with$important
and$app
modifiers. Rules with any other modifiers are considered invalid and will be discarded.
- Rules with
$removeparam
modifier are supported by AdGuard for Windows, Mac, and Android and AdGuard Browser Extension for Chrome, Firefox, and Edge. $removeparam
syntax for regular expressions is supported by AdGuard Browser Extension v4.0 and AdGuard for Windows, Mac, and Android with CoreLibs v1.8 or later.POST
request types are supported only by AdGuard for Windows, Mac, and Android with CoreLibs v1.10 or later, and AdGuard Browser Extension with TSWebExtension v0.4.6 or later.
$replace
This modifier completely changes the rule behavior. If it is applied, the rule will not block the request. The response is going to be modified instead.
You will need some knowledge of regular expressions to use $replace
modifier.
Features
$replace
rules apply to any text response, but will not apply to binary (media
,image
,object
, etc.).$replace
rules do not apply if the size of the original response is more than 10 MB.$replace
rules have a higher priority than other basic rules (including exception rules). So if a request corresponds to two different rules one of which has the$replace
modifier, this rule will be applied.- Document-level exception rules with
$content
or$document
modifiers do disable$replace
rules for requests matching them. - Other document-level exception rules (
$generichide
,$elemhide
or$jsinject
modifiers) are applied alongside$replace
rules. It means that you can modify the page content with a$replace
rule and disable cosmetic rules there at the same time.
$replace
value can be empty in the case of exception rules. See examples section for further information.
Multiple rules matching a single request
In case if multiple $replace
rules match a single request, we will apply each of them. The order is defined alphabetically.
Syntax
In general, $replace
syntax is similar to replacement with regular expressions in Perl.
replace = "/" regexp "/" replacement "/" modifiers
regexp
— a regular expression.replacement
— a string that will be used to replace the string corresponding toregexp
.modifiers
— a regular expression flags. For example,i
— insensitive search, ors
— single-line mode.
In the $replace
value, two characters must be escaped: comma ,
and dollar sign $
. Use backslash \
for it. For example, an escaped comma looks like this: \,
.
Examples
||example.org^$replace=/(<VAST[\s\S]*?>)[\s\S]*<\/VAST>/\$1<\/VAST>/i
There are three parts in this rule:
regexp
—(<VAST(.|\s)*?>)(.|\s)*<\/VAST>
;replacement
—\$1<\/VAST>
where$
is escaped;modifiers
—i
for insensitive search.
You can see how this rule works here: http://regexr.com/3cesk
Multiple $replace
rules
||example.org^$replace=/X/Y/
||example.org^$replace=/Z/Y/
@@||example.org/page/*$replace=/Z/Y/
- Both rule 1 and 2 will be applied to all requests sent to
example.org
. - Rule 2 is disabled for requests matching
||example.org/page/
, but rule 1 still works!
Disabling $replace
rules
@@||example.org^$replace
will disable all$replace
rules matching||example.org^
.@@||example.org^$document
or@@||example.org^$content
will disable all$replace
rules originated from pages ofexample.org
including the page itself.
Rules with $replace
modifier can be used only in trusted filters.
Rules with $replace
modifier are supported by AdGuard for Windows, Mac, and Android and AdGuard Browser Extension for Firefox. Such rules do not work in extensions for other browsers because they are unable to modify content on the network level.
urltransform
The $urltransform
rules allow you to modify the request URL by replacing the text matched by the regular expression.
Features
$urltransform
rules apply to any request URL text.$urltransform
rules can also modify the query part of the URL.$urltransform
will not be applied if the original URL is blocked by other rules.$urltransform
will be applied before$removeparam
rules.
The $urltransform
value can be empty for exception rules.
Multiple rules matching a single request
If multiple $urltransform
rules match a single request, we will apply each of them. The order is defined alphabetically.
Syntax
$urltransform
syntax is similar to replacement with regular expressions in Perl.
urltransform = "/" regexp "/" replacement "/" modifiers
regexp
— a regular expression.replacement
— a string that will be used to replace the string corresponding toregexp
.modifiers
— a regular expression flags. For example,i
— insensitive search, ors
— single-line mode.
In the $urltransform
value, two characters must be escaped: the comma ,
and the dollar sign $
. Use the backslash character \
for this. For example, an escaped comma looks like this: \,
.
Examples
||example.org^$urltransform=/(pref\/).*\/(suf)/\$1\$2/i
There are three parts in this rule:
regexp
—(pref\/).*\/(suf)
;replacement
—\$1\$2
where$
is escaped;modifiers
—i
for insensitive search.
Multiple $urltransform
rules
||example.org^$urltransform=/X/Y/
||example.org^$urltransform=/Z/Y/
@@||example.org/page/*$urltransform=/Z/Y/
- Both rule 1 and 2 will be applied to all requests sent to
example.org
. - Rule 2 is disabled for requests matching
||example.org/page/
, but rule 1 still works!
Re-matching rules after transforming the URL
If the $urltransform
rule is applied to a request, all the rules will be re-evaluated against the new URL.
E.g., with the following rules:
||example.com^$urltransform=/firstpath/secondpath/
||example.com/secondpath^
the request to https://example.com/firstpath
will be blocked before it is sent.
However, $urltransform
rules will not be re-applied in this case to avoid infinite recursion, e.g., with the following rules:
||example.com/firstpath^$urltransform=/firstpath/secondpath/
||example.com/secondpath^$urltransform=/secondpath/firstpath/
the request to https://example.com/fisrtpath
will be transformed to https://example.com/secondpath
and the second rule will not be applied.
Disabling $urltransform
rules
@@||example.org^$urltransform
will disable all$urltransform
rules matching||example.org^
.@@||example.org^$urltransform=/Z/Y/
will disable the rule with$urltransform=/Z/Y/
for any request matching||example.org^
.
$urltransform
rules can also be disabled by $document
and $urlblock
exception rules. But basic exception rules without modifiers do not do that. For example, @@||example.com^
will not disable $urltransform=/X/Y/
for requests to example.com, but @@||example.com^$urlblock
will.
Rules with the $urltransform
modifier can be used only in trusted filters.
Rules with the $urltransform
modifier are supported by AdGuard for Windows, AdGuard for Mac, and AdGuard for Android with CoreLibs v1.15 or later.
noop
noop
modifier does nothing and can be used solely to increase rules' readability. It consists of a sequence of underscore characters (_
) of any length and can appear in a rule as many times as needed.
Examples
||example.com$_,removeparam=/^ss\\$/,_,image
||example.com$replace=/bad/good/,___,~third-party
Rules with noop
modifier are not supported by AdGuard Content Blocker.
$empty
(deprecated)
This modifier is deprecated in favor of the $redirect
modifier. Rules with $empty
are still supported and being converted into $redirect=nooptext
now but the support shall be removed in the future.
Usually, blocked requests look like a server error to browser. If you use $empty
modifier, AdGuard will emulate a blank response from the server with200 OK
status.
Examples
||example.org^$empty
returns an empty response to all requests toexample.org
and all subdomains.
Rules with $empty
modifier are not supported by AdGuard Content Blocker, AdGuard for iOS, and AdGuard for Safari.
$mp4
(deprecated)
This modifier is deprecated in favor of the $redirect
modifier. Rules with $mp4
are still supported and being converted into $redirect=noopmp4-1s,media
now but the support shall be removed in the future.
As a response to blocked request AdGuard returns a short video placeholder.
Examples
||example.com/videos/$mp4
blocks all video downloads from||example.com/videos/*
and changes the response to a video placeholder.
Rules with $mp4
modifier are not supported by AdGuard Content Blocker, AdGuard for iOS, and AdGuard for Safari.
Rule priorities
Each rule has its own priority, which is necessary when several rules match the request and the filtering engine needs to select one of them. Priority is measured by a positive integer.
When two rules with the same priority match the same request, it depends on the filtering engine implementation which one will be selected.
The concept of rule priorities becomes increasingly important in light of Manifest V3 as the existing rules need to be converted to declarativeNetRequest rules.
Priority calculation
To calculate priority, we've categorized modifiers into different groups. These groups are ranked based on their priority, from lowest to highest. A modifier that significantly narrows the scope of a rule adds more weight to its total priority. Conversely, if a rule applies to a broader range of requests, its priority decreases.
It's worth noting that there are cases where a single-parameter modifier has a higher priority than multi-parameter ones. For instance, in the case of $domain=example.com|example.org
, a rule that includes two domains has a slightly broader effective area than a rule with one specified domain, therefore its priority is lower.
The base priority of any rule is 1. If the calculated priority is a floating-point number, it will be rounded up to the smallest integer greater than or equal to the calculated priority.
- The concept of priority has been introduced in TSUrlFilter v2.1.0 and CoreLibs v1.13. Before that AdGuard didn't have any special priority computation algorithm and collisions handling could be different depending on AdGuard product and version.
- AdGuard for iOS, Safari, and AdGuard Content Blocker rely on the browsers implementation and they cannot follow the rules specified here.
Modifier aliases (1p
, 3p
, etc.) are not included in these categories, however, they are utilized within the engine to compute the rule priority.
Basic modifiers, the presence of each adds 1 to the priority
$app
with negated applications using~
,$denyallow
,$domain
with negated domains using~
,$match-case
,$method
with negated methods using~
,$third-party
,$to
,- restricted content-types with
~
.
When dealing with a negated domain, app, method, or content-type, we add 1 point for the existence of the modifier itself, regardless of the quantity of negated domains or content-types. This is because the rule's scope is already infinitely broad. Put simply, by prohibiting multiple domains, content-types, methods or apps, the scope of the rule becomes only minimally smaller.
Defined content-type modifiers, defined methods, defined headers, $all, $popup, specific exceptions
All allowed content types:
$document
,$font
,$image
,$media
,$object
,$other
,$ping
,$script
,$stylesheet
,$subdocument
,$websocket
,$xmlhttprequest
;
This also includes rules that implicitly add all content types:
$all
;
Or rules that implicitly add the modifier $document
:
Or some specific exceptions that implicitly add $document,subdocument
:
Or allowed methods via $method
.
Or rules with $header
.
The presence of any content-type modifiers adds (50 + 50 / N)
, where N
is the number of modifiers present, for example: ||example.com^$image,script
will add 50 + 50 / 2 = 50 + 25 = 75
to the total weight of the rule.
The $all
also belongs to this category, because it implicitly adds all content type modifiers, e.g., $document,subdocument,image,script,media,<etc>
+ $popup
.
The $popup
also belongs to this category, because it implicitly adds the modifier $document
. Similarly, specific exceptions add $document,subdocument
.
If there is a $method
modifier in the rule with allowed methods it adds (50 + 50 / N)
, where N
is the number of methods allowed, for example: ||example.com^$method=GET|POST|PUT
will add 50 + 50 / 3 = 50 + 16.6 = 67
to the total weight of the rule.
If there is a $header
modifier in the rule it adds 50
.
$domain
or $app
with allowed domains or applications
Specified domains through $domain
or specified applications through $app
add 100 + 100 / N
, where N
is the number of modifier values for example: ||example.com^$domain=example.com|example.org|example.net
will add 100 + 100 / 3 = 134.3 = 135
or ||example.com^$app=org.example.app1|org.example.app2
will add 100 + 100 / 2 = 151
or ||example.com^$domain=example.com,app=org.example.app1|org.example.app2
will add 100 + 100/1
($domain part) and 100 + 100/2
($app part), totaling 350
.
Modifier values that are regexps or tld will be interpreted as normal entries of the form example.com
and counted one by one, for example: ||example.com^$domain=example.*
will add 100 + 100 / 1 = 200
or ||example.com^$domain=example.*|adguard.*
will add 100 + 100 / 2 = 150
.
$redirect
rules
Each of which adds 10^3
to rule priority.
Specific exceptions
Each of which adds 10^4
to the priority.
As well as exception with $document modifier
: because it's an alias for $elemhide,content,jsinject,urlblock,extension
. It will add 10^4
for each modifier from the top list, 10^4 * 5
in total.
In addition, each of these exceptions implicitly adds the two allowed content-type modifiers $document,subdocument
.
Allowlist rules
Modifier @@
adds 10^5
to rule priority.
$important
rules
Modifier $important
adds 10^6
to rule priority.
Rules for which there is no priority weight
Other modifiers, which are supposed to perform additional post- or pre-processing of requests, do not add anything to the rules priority.
Examples
||example.com^
Weight of the rule without modifiers:
1
.||example.com^$match-case
Rule weight: base weight + weight of the modifier from category 1:
1 + 1 = 2
.||example.org^$removeparam=p
Rule weight: base weight + 0, since $removeparam is not involved in the priority calculation:
1 + 0 = 1
.||example.org^$document,redirect=nooptext
Rule weight: base weight + allowed content type, category 3 + $redirect from category 6:
1 + (100 + 100 / 1) + 1000 = 1201
.@@||example.org^$removeparam=p,document
Rule weight: base weight + allowlist rule, category 5 + 0 because $removeparam is not involved in the priority calculation + allowed content type, category 2:
1 + 10000 + 0 + (50 + 50 / 1) = 10101
.@@||example.com/ad/*$domain=example.org|example.net,important
Rule weight: base weight + allowlist rule, category 5 + important rule, category 7 + allowed domains, category 3:
1 + 10000 + 1000000 + (100 + 100 / 2) = 1010152
.@@||example.org^$document
without additional modifiers is an alias for@@||example.com^$elemhide,content,jsinject,urlblock,extension
Rule weight: base weight + specific exceptions, category 4 + two allowed content types (document and subdocument), category 2:
1 + 10000 * 4 + (50 + 50 / 2) = 40076
.*$script,domain=a.com,denyallow=x.com|y.com
Rule weight: base weight + allowed content type, category 2 + allowed domain, category 3 + denyallow, category 1:
1 + (50 + 50/1) + (100 + 100 / 1) + 1 = 303
.||example.com^$all
— alias to||example.com^$document,subdocument,image,script,media,etc. + $popup
Rule weight: base weight + popup (category 1) + allowed content types (category 2):
1 + 1 + (50 + 50/12) = 55
.
Non-basic rules
However, basic rules may not be enough to block ads. Sometimes you need to hide an element or change part of the HTML code of a web page without breaking anything. The rules described in this section are created specifically for this purpose.
Categories \ Products | CoreLibs apps | AdGuard for Chromium | AdGuard for Firefox | AdGuard für iOS | AdGuard für Safari | AdGuard-Inhaltsblocker |
---|---|---|---|---|---|---|
Element hiding | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
CSS rules | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
Extended CSS | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
HTML filtering | ✅ | ❌ | ✅ | ❌ | ❌ | ❌ |
JavaScript | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
Scriptlets | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
- ✅ — fully supported
- ❌ — not supported
Cosmetic rules
Work with non-basic rules requires the basic knowledge of HTML and CSS. So, if you want to learn how to make such rules, we recommend to get acquainted with this documentation.
Element hiding rules
Element hiding rules are used to hide the elements of web pages. It is similar to applying { display: none; }
style to selected element.
Element hiding rules may operate differently depending on the platform.
Syntax
rule = [domains] "##" selector
domains = [domain0, domain1[, ...[, domainN]]]
selector
— CSS selector, defines the elements to be hidden.domains
— domain restriction for the rule.
If you want to limit the rule application area to certain domains, just enter them separated with commas. For example: example.org,example.com##selector
.
This rule will be also applied to all subdomains of example.org
and example.com
.
If you want the rule not to be applied to certain domains, start a domain name with ~
sign. For example: ~example.org##selector
.
You can use both approaches in a single rule. For example, example.org,~subdomain.example.org##domain
will work for example.org
and all subdomains, except subdomain.example.org
.
Element hiding rules are not dependent on each other. If there is a rule example.org##selector
in the filter and you add ~example.org##selector
both rules will be applied independently.
Examples
example.com##div.textad
— hides adiv
with the classtextad
atexample.com
and all subdomains.example.com,example.org###adblock
— hides an element with attributeid
equalsadblock
atexample.com
,example.org
and all subdomains.~example.com##.textad
— hides an element with the classtextad
at all domains, exceptexample.com
and its subdomains.
Limitations
Safari does not support both permitted and restricted domains. So the rules like example.org,~foo.example.org##.textad
are invalid in AdGuard for Safari.
Exceptions
Exceptions can disable some rules on particular domains. They are very similar to usual exception rules, but instead of ##
you have to use #@#
.
For example, there is a rule in filter:
##.textad
If you want to disable it for example.com
, you can create an exception rule:
example.com#@#.textad
Sometimes, it may be necessary to disable all restriction rules. For example, to conduct tests. To do this, use the exclusion rule without specifying a domain. It will completely disable matching CSS elemhide rule on ALL domains:
#@#.textad
The same can be achieved by adding this rule:
*#@#.textad
We recommend to use this kind of exceptions only if it is not possible to change the hiding rule itself. In other cases it is better to change the original rule, using domain restrictions.
CSS rules
Sometimes, simple hiding of an element is not enough to deal with advertising. For example, blocking an advertising element can just break the page layout. In this case AdGuard can use rules that are much more flexible than hiding rules. With these rules you can basically add any CSS styles to the page.
Syntax
rule = [domains] "#$#" selector "{" style "}"
domains = [domain0, domain1[, ...[, domainN]]]
selector
— CSS selector, that defines the elements we want to apply the style to.domains
— domain restriction for the rule. Same principles as in element hiding rules.style
— CSS style, that we want to apply to selected elements.
Examples
example.com#$#body { background-color: #333!important; }
This rule will apply a style background-color: #333!important;
to the body
element at example.com
and all subdomains.
Exceptions
Just like with element hiding, there is a type of rules that disable the selected CSS style rule for particular domains. Exception rule syntax is almost the same, you just have to change #$#
to #@$#
.
For example, there is a rule in filter:
#$#.textad { visibility: hidden; }
If you want to disable it for example.com
, you can create an exception rule:
example.com#@$#.textad { visibility: hidden; }
We recommend to use this kind of exceptions only if it is not possible to change the CSS rule itself. In other cases it is better to change the original rule, using domain restrictions.
Styles that lead to loading any resource are forbidden. Basically, it means that you cannot use any <url>
type of value in the style.
CSS rules are not supported by AdGuard Content Blocker.
CSS rules may operate differently depending on the platform.
Extended CSS selectors
- Limitations
- Pseudo-class
:has()
- Pseudo-class
:contains()
- Pseudo-class
:matches-css()
- Pseudo-class
:matches-attr()
- Pseudo-class
:matches-property()
- Pseudo-class
:xpath()
- Pseudo-class
:nth-ancestor()
- Pseudo-class
:upward()
- Pseudo-class
:remove()
and pseudo-propertyremove
- Pseudo-class
:is()
- Pseudo-class
:not()
- Pseudo-class
:if-not()
(removed)
CSS 3.0 is not always enough to block ads. To solve this problem AdGuard extends CSS capabilities by adding support for the new pseudo-elements. We have developed a separate open-source library for non-standard element selecting and applying CSS styles with extended properties.
The idea of extended capabilities is an opportunity to match DOM elements with selectors based on their own representation (style, text content, etc.) or relations with other elements. There is also an opportunity to apply styles with non-standard CSS properties.
Application area
Extended selectors can be used in any cosmetic rule, whether they are element hiding rules or CSS rules.
Rules with extended CSS selectors are not supported by AdGuard Content Blocker.
Syntax
Regardless of the CSS pseudo-classes you are using in the rule, you can use special markers to force applying these rules by ExtendedCss. It is recommended to use these markers for all extended CSS cosmetic rules so that it was easier to find them.
The syntax for extended CSS rules:
#?#
— for element hiding,#@?#
— for exceptions#$?#
— for CSS rules,#@$?#
— for exceptions
We strongly recommend using these markers any time when you use an extended CSS selector.
Examples
example.org#?#div:has(> a[target="_blank"][rel="nofollow"])
— this rule blocks alldiv
elements containing a child node that has a link with the attributes[target="_blank"][rel="nofollow"]
. The rule applies only toexample.org
and its subdomains.example.com#$?#h3:contains(cookies) { display: none!important; }
— this rule sets the styledisplay: none!important
to allh3
elements that contain the wordcookies
. The rule applies only toexample.com
and all its subdomains.example.net#?#.banner:matches-css(width: 360px)
— this rule blocks all.banner
elements with the style propertywidth: 360px
. The rule applies only toexample.net
and its subdomains.example.net#@?#.banner:matches-css(width: 360px)
— this rule will disable the previous rule.
You can apply standard CSS selectors using the ExtendedCss library by using the rule marker #?#
, e.g. #?#div.banner
.
Learn more about how to debug extended selectors.
Some pseudo-classes do not require selector before it. Still adding the universal selector *
makes an extended selector easier to read, even though it has no effect on the matching behavior. So selector #block :has(> .inner)
works exactly like #block *:has(> .inner)
but second one is more obvious.
Pseudo-class names are case-insensitive, e.g. :HAS()
works as :has()
. Still the lower-case names are used commonly.
ExtendedCss Limitations
Specific pseudo-class may have its own limitations:
:has()
,:xpath()
,:nth-ancestor()
,:upward()
,:is()
,:not()
, and:remove()
.
Pseudo-class :has()
Draft CSS 4.0 specification describes the :has()
pseudo-class. Unfortunately, it is not yet supported by all popular browsers.
Rules with the :has()
pseudo-class must use the native implementation of :has()
if they use ##
marker and if it is possible, i.e. with no other extended selectors inside. To force applying of ExtendedCss rules with :has()
, use #?#
/#$?#
marker explicitly.
Compatibility with other pseudo-classes
Synonyms :-abp-has()
is supported by ExtendedCss for better compatibility.
:if()
is no longer supported as a synonym for :has()
.
Syntax
[target]:has(selector)
target
— optional, standard or extended CSS selector, can be skipped for checking any elementselector
— required, standard or extended CSS selector
The pseudo-class :has()
selects the target
elements that fit to the selector
. Also the selector
can start with a combinator.
A selector list can be set in selector
as well. In this case all selectors in the list are being matched for now. In the future it will be fixed for <forgiving-relative-selector-list>
as argument.
:has()
limitations
Usage of the :has()
pseudo-class is restricted for some cases (2, 3):
- disallow
:has()
inside the pseudos accepting only compound selectors; - disallow
:has()
after regular pseudo-elements.
Native :has()
pseudo-class does not allow :has()
, :is()
, :where()
inside :has()
argument to avoid increasing the :has()
invalidation complexity (case 1). But ExtendedCss did not have such limitation earlier and filter lists already contain such rules, so we have not added this limitation to ExtendedCss and allow to use :has()
inside :has()
as it was possible before. To use it, just force ExtendedCss usage by setting #?#
/#$?#
rule marker.
Native implementation does not allow any usage of :scope
inside the :has()
argument ([1], [2]). Still, there are some such rules in filter lists: div:has(:scope a)
which we continue to support by simply converting them to div:has(> a)
, as it used to be done previously.
Examples
div:has(.banner)
selects all div
elements which include an element with the banner
class:
<!-- HTML code -->
<div>Not selected</div>
<div>Selected
<span class="banner">inner element</span>
</div>
div:has(> .banner)
selects all div
elements which include an banner
class element as a direct child of div
:
<!-- HTML code -->
<div>Not selected</div>
<div>Selected
<p class="banner">child element</p>
</div>
div:has(+ .banner)
selects all div
elements preceding banner
class element which immediately follows the div
and both are children of the same parent:
<!-- HTML code -->
<div>Not selected</div>
<div>Selected</div>
<p class="banner">adjacent sibling</p>
<span>Not selected</span>
div:has(~ .banner)
selects all div
elements preceding banner
class element which follows the div
but not necessarily immediately and both are children of the same parent:
<!-- HTML code -->
<div>Not selected</div>
<div>Selected</div>
<span>Not selected</span>
<p class="banner">general sibling</p>
div:has(span, .banner)
selects all div
elements which include both span
element and banner
class element:
<!-- HTML code -->
<div>Not selected</div>
<div>Selected
<span>child span</span>
<p class="banner">child .banner</p>
</div>
Backward compatible syntax for :has()
is supported but not recommended.
Pseudo-class :contains()
The :contains()
pseudo-class principle is very simple: it allows to select the elements that contain specified text or which content matches a specified regular expression. Regexp flags are supported.
The :contains()
pseudo-class uses the textContent
element property for matching, not the innerHTML
.
Compatibility with other pseudo-classes
Synonyms :-abp-contains()
and :has-text()
are supported for better compatibility.
Syntax
[target]:contains(match)
target
— optional, standard or extended CSS selector, can be skipped for checking any elementmatch
— required, string or regular expression for matching elementtextContent
. Regular expression flags are supported.
Examples
For such DOM:
<!-- HTML code -->
<div>Not selected</div>
<div id="match">Selected as IT contains "banner"</div>
<div>Not selected <div class="banner"></div></div>
the element div#match
can be selected by any of these extended selectors:
! plain text
div:contains(banner)
! regular expression
div:contains(/as .* banner/)
! regular expression with flags
div:contains(/it .* banner/gi)
Only the div
with id=match
is selected because the next element does not contain any text, and banner
is a part of code, not a text.
Backward compatible syntax for :contains()
is supported but not recommended.
Pseudo-class :matches-css()
The :matches-css()
pseudo-class allows to match the element by its current style properties. The work of the pseudo-class is based on using the Window.getComputedStyle()
method.
Syntax
[target]:matches-css([pseudo-element, ] property: pattern)
target
— optional, standard or extended CSS selector, can be skipped for checking any elementpseudo-element
— optional, valid standard pseudo-element, e.g.before
,after
,first-line
, etc.property
— required, a name of CSS property to check the element forpattern
— required, a value pattern that is using the same simple wildcard matching as in the basic url filtering rules OR a regular expression. For this type of matching, AdGuard always does matching in a case-insensitive manner. In the case of a regular expression, the pattern looks like/regexp/
.
Special characters escaping and unescaping
For non-regexp patterns (
,)
,[
,]
must be unescaped, e.g. :matches-css(background-image:url(data:*))
.
For regexp patterns \
should be escaped, e.g. :matches-css(background-image: /^url\\("data:image\\/gif;base64.+/)
.
Examples
For such DOM:
<!-- HTML code -->
<style type="text/css">
#matched::before {
content: "Block me"
}
</style>
<div id="matched"></div>
<div id="not-matched"></div>
the div
elements with pseudo-element ::before
and with specified content
property can be selected by any of these extended selectors:
! string pattern
div:matches-css(before, content: block me)
! string pattern with wildcard
div:matches-css(before, content: block*)
! regular expression pattern
div:matches-css(before, content: /block me/)
Regexp patterns do not support flags.
Obsolete pseudo-classes :matches-css-before()
and :matches-css-after()
are no longer recommended but still are supported for better compatibility.
Backward compatible syntax for :matches-css()
is supported but not recommended.
Pseudo-class :matches-attr()
The :matches-attr()
pseudo-class allows selecting an element by its attributes, especially if they are randomized.
Syntax
[target]:matches-attr("name"[="value"])
target
— optional, standard or extended CSS selector, can be skipped for checking any elementname
— required, simple string or string with wildcard or regular expression for attribute name matchingvalue
— optional, simple string or string with wildcard or regular expression for attribute value matching
Escaping special characters
For regexp patterns "
and \
should be escaped, e.g. div:matches-attr(class=/[\\w]{5}/)
.
Examples
div:matches-attr("ad-link")
selects the element div#target1
:
<!-- HTML code -->
<div id="target1" ad-link="1random23-banner_240x400"></div>
div:matches-attr("data-*"="adBanner")
selects the element div#target2
:
<!-- HTML code -->
<div id="target2" data-1random23="adBanner"></div>
div:matches-attr(*unit*=/^click$/)
selects the element div#target3
:
<!-- HTML code -->
<div id="target3" random123-unit094="click"></div>
*:matches-attr("/.{5,}delay$/"="/^[0-9]*$/")
selects the element #target4
:
<!-- HTML code -->
<div>
<inner-random23 id="target4" nt4f5be90delay="1000"></inner-random23>
</div>
Regexp patterns do not support flags.
Pseudo-class :matches-property()
The :matches-property()
pseudo-class allows selecting an element by matching its properties.
Syntax
[target]:matches-property("name"[="value"])
target
— optional, standard or extended CSS selector, can be skipped for checking any elementname
— required, simple string or string with wildcard or regular expression for element property name matchingvalue
— optional, simple string or string with wildcard or regular expression for element property value matching
Escaping special characters
For regexp patterns "
and \
must be escaped, e.g. div:matches-property(prop=/[\\w]{4}/)
.
Regexp patterns are supported in name
for any property in chain, e.g. prop./^unit[\\d]{4}$/.type
.
Examples
An element with such properties:
divProperties = {
id: 1,
check: {
track: true,
unit_2random1: true,
},
memoizedProps: {
key: null,
tag: 12,
_owner: {
effectTag: 1,
src: 'ad.com',
},
},
};
can be selected by any of these extended selectors:
div:matches-property(check.track)
div:matches-property("check./^unit_.{4,8}$/")
div:matches-property("check.unit_*"=true)
div:matches-property(memoizedProps.key="null")
div:matches-property(memoizedProps._owner.src=/ad/)
To check properties of a specific element, do the following:
- Inspect the page element or select it in
Elements
tab of browser DevTools - Run
console.dir($0)
inConsole
tab
Regexp patterns do not support flags.
Pseudo-class :xpath()
The :xpath()
pseudo-class allows selecting an element by evaluating an XPath expression.
Syntax
[target]:xpath(expression)
target
- optional, standard or extended CSS selectorexpression
— required, valid XPath expression
:xpath()
limitations
target
can be omitted so it is optional. For any other pseudo-class that would mean "apply to all DOM nodes", but in case of :xpath()
it just means "apply to the whole document", and such applying slows elements selecting significantly. That's why rules like #?#:xpath(expression)
are limited to looking inside the body
tag. For example, rule #?#:xpath(//div[@data-st-area=\'Advert\'])
is parsed as #?#body:xpath(//div[@data-st-area=\'Advert\'])
.
Extended selectors with defined target
as any selector — *:xpath(expression)
— can still be used but it is not recommended, so target
should be specified instead.
Works properly only at the end of selector, except for pseudo-class :remove().
Examples
:xpath(//*[@class="banner"])
selects the element div#target1
:
<!-- HTML code -->
<div id="target1" class="banner"></div>
:xpath(//*[@class="inner"]/..)
selects the element div#target2
:
<!-- HTML code -->
<div id="target2">
<div class="inner"></div>
</div>
Pseudo-class :nth-ancestor()
The :nth-ancestor()
pseudo-class allows to lookup the nth ancestor relative to the previously selected element.
subject:nth-ancestor(n)
subject
— required, standard or extended CSS selectorn
— required, number >= 1 and < 256, distance to the needed ancestor from the element selected bysubject
Syntax
subject:nth-ancestor(n)
subject
— required, standard or extended CSS selectorn
— required, number >= 1 and < 256, distance to the needed ancestor from the element selected bysubject
:nth-ancestor()
limitations
The :nth-ancestor()
pseudo-class is not supported inside the argument of the :not()
pseudo-class.
Examples
For such DOM:
<!-- HTML code -->
<div id="target1">
<div class="child"></div>
<div id="target2">
<div>
<div>
<div class="inner"></div>
</div>
</div>
</div>
</div>
.child:nth-ancestor(1)
selects the element div#target1
, div[class="inner"]:nth-ancestor(3)
selects the element div#target2
.
Pseudo-class :upward()
The :upward()
pseudo-class allows to lookup the ancestor relative to the previously selected element.
Syntax
subject:upward(ancestor)
subject
— required, standard or extended CSS selectorancestor
— required, specification for the ancestor of the element selected bysubject
, can be set as:- number >= 1 and < 256 for distance to the needed ancestor, same as
:nth-ancestor()
- standard CSS selector for matching closest ancestor
- number >= 1 and < 256 for distance to the needed ancestor, same as
:upward()
limitations
The :upward()
pseudo-class is not supported inside the argument of the :not()
pseudo-class.
Examples
For such DOM:
<!-- HTML code -->
<div id="target1" data="true">
<div class="child"></div>
<div id="target2">
<div>
<div>
<div class="inner"></div>
</div>
</div>
</div>
</div>
.inner:upward(div[data])
selects the element div#target1
, .inner:upward(div[id])
selects the element div#target2
, .child:upward(1)
selects the element div#target1
, .inner:upward(3)
selects the element div#target2
.
Pseudo-class :remove()
and pseudo-property remove
Sometimes, it is necessary to remove a matching element instead of hiding it or applying custom styles. In order to do it, you can use the :remove()
pseudo-class as well as the remove
pseudo-property.
Pseudo-class :remove()
can be placed only at the end of a selector.
Syntax
! pseudo-class
selector:remove()
! pseudo-property
selector { remove: true; }
selector
— required, standard or extended CSS selector
:remove()
and remove
limitations
The :remove()
pseudo-class is limited to work properly only at the end of selector.
For applying the :remove()
pseudo-class to any element, the universal selector *
should be used. Otherwise such extended selector may be considered as invalid, e.g. .banner > :remove()
is not valid for removing any child element of banner
class element, so it should look like .banner > *:remove()
.
If the :remove()
pseudo-class or the remove
pseudo-property is used, all style properties are ignored except for the debug
pseudo-property.
Examples
div.banner:remove()
div:has(> div[ad-attr]):remove()
div:contains(advertisement) { remove: true; }
div[class]:has(> a > img) { remove: true; }
Rules with the remove
pseudo-property must use #$?#
marker: $
for CSS-style rule syntax, ?
for ExtendedCss syntax.
Pseudo-class :is()
The :is()
pseudo-class allows to match any element that can be selected by any of selectors passed to it. Invalid selectors are skipped and the pseudo-class deals with valid ones with no error thrown. Our implementation of the native :is()
pseudo-class.
Syntax
[target]:is(selectors)
target
— optional, standard or extended CSS selector, can be skipped for checking any elementselectors
— forgiving selector list of standard or extended selectors. For extended selectors, only compound selectors are supported, not complex.
:is()
limitations
Rules with the :is()
pseudo-class must use the native implementation of :is()
if rules use ##
marker and it is possible, i.e. with no other extended selectors inside. To force applying ExtendedCss rules with :is()
, use #?#
/#$?#
marker explicitly.
If the :is()
pseudo-class argument selectors
is an extended selector, due to the way how the :is()
pseudo-class is implemented in ExtendedCss v2.0, it is impossible to apply it to the top DOM node which is html
, i.e. #?#html:is(<extended-selectors>)
does not work. So if target
is not defined or defined as the universal selector *
, the extended pseudo-class applying is limited to html
's children, e.g. rules #?#:is(...)
and #?#*:is(...)
are parsed as #?#html *:is(...)
. Please note that there is no such limitation for a standard selector argument, i.e. #?#html:is(.locked)
works fine.
Complex selectors with extended pseudo-classes are not supported as selectors
argument for :is()
pseudo-class, only compound ones are allowed. Check examples below for more details.
Examples
#container *:is(.inner, .footer)
selects only the element div#target1
:
<!-- HTML code -->
<div id="container">
<div data="true">
<div>
<div id="target1" class="inner"></div>
</div>
</div>
</div>
Due to limitations :is(*:not([class]) > .banner)'
does not work but :is(*:not([class]):has(> .banner))
can be used instead of it to select the element div#target2
:
<!-- HTML code -->
<span class="span">text</span>
<div id="target2">
<p class="banner">inner paragraph</p>
</div>
Pseudo-class :not()
The :not()
pseudo-class allows to select elements which are not matched by selectors passed as argument. Invalid argument selectors are not allowed and error is to be thrown. Our implementation of the :not()
pseudo-class.
Syntax
[target]:not(selectors)
target
— optional, standard or extended CSS selector, can be skipped for checking any elementselectors
— list of standard or extended selectors
:not()
limitations
Rules with the :not()
pseudo-class must use the native implementation of :not()
if rules use ##
marker and it is possible, i.e. with no other extended selectors inside. To force applying ExtendedCss rules with :not()
, use #?#
/#$?#
marker explicitly.
If the :not()
pseudo-class argument selectors
is an extended selector, due to the way how the :not()
pseudo-class is implemented in ExtendedCss v2.0, it is impossible to apply it to the top DOM node which is html
, i.e. #?#html:not(<extended-selectors>)
does not work. So if target
is not defined or defined as the universal selector *
, the extended pseudo-class applying is limited to html
's children, e.g. rules #?#:not(...)
and #?#*:not(...)
are parsed as #?#html *:not(...)
. Please note that there is no such limitation for a standard selector argument, i.e. #?#html:not(.locked)
works fine.
The :not()
is considered as a standard CSS pseudo-class inside the argument of the :upward()
pseudo-class because :upward()
supports only standard selectors.
"Up-looking" pseudo-classes which are :nth-ancestor()
and :upward()
are not supported inside selectors
argument for :not()
pseudo-class.
Examples
#container > *:not(h2, .text)
selects only the element div#target1
:
<!-- HTML code -->
<div id="container">
<h2>Header</h2>
<div id="target1"></div>
<span class="text">text</span>
</div>
Pseudo-class :if-not()
(removed)
The :if-not()
pseudo-class is removed and is no longer supported. Rules with it are considered as invalid.
This pseudo-class was basically a shortcut for :not(:has())
. It was supported by ExtendedCss for better compatibility with some filters subscriptions.
Cosmetic rules priority
The way element hiding and CSS rules are applied is platform-specific.
In AdGuard for Windows, Mac, and Android, we use a stylesheet injected into the page. The priority of cosmetic rules is the same as any other websites' CSS stylesheet. But there is a limitation: element hiding and CSS rules cannot override inline styles. In such cases, it is recommended to use extended selectors or HTML filtering.
In AdGuard Browser Extension, the so called "user stylesheets" are used. They have higher priority than even the inline styles.
Extended CSS selectors use JavaScript to work and basically add an inline style themselves, therefore they can override any style.
HTML filtering rules
In most cases, the basis and cosmetic rules are enough to filter ads. But sometimes it is necessary to change the HTML-code of the page itself before it is loaded. This is when you need filtering rules for HTML content. They allow to indicate the HTML elements to be cut out before the browser loads the page.
HTML filtering rules are supported by AdGuard for Windows, Mac, Android, and AdGuard Browser Extension for Firefox. Such rules do not work in extensions for other browsers because they are unable to modify content on network level.
Syntax
selector = [tagName] [attributes] [pseudoClasses]
combinator = ">"
rule = [domains] "$$" selector *(combinator selector)
domains = [domain0, domain1[, ...[, domainN]]]
attributes = "[" name0 = value0 "]" "[" name1 = value2 "]" ... "[" nameN = valueN "]"
pseudoClasses = pseudoClass *pseudoClass
pseudoClass = ":" pseudoName [ "(" pseudoArgs ")" ]
tagName
— name of the element in lower case, for example,div
orscript
.domains
— domain restriction for the rule. Same principles as in element hiding rule syntax.attributes
— a list of attributes that limit the selection of elements.name
— attribute name,value
— substring, that is contained in attribute value.pseudoName
— the name of a pseudo-class.pseudoArgs
— the arguments of a function-style pseudo-class.combinator
— an operator that works similarly to the CSS child combinator: that is, theselector
on the right of thecombinator
will only match an element whose direct parent matches theselector
on the left of thecombinator
.
Examples
HTML code:
<script data-src="/banner.js"></script>
Rule:
example.org$$script[data-src="banner"]
This rule removes all script
elements with the attribute data-src
containing the substring banner
. The rule applies only to example.org
and all its subdomains.
Special attributes
In addition to usual attributes, which value is every element checked for, there is a set of special attributes that change the way a rule works. Below there is a list of these attributes:
tag-content
This special attribute may become unsupported in the future. Prefer using the :contains()
pseudo-class where it is available.
This is the most frequently used special attribute. It limits selection with those elements whose innerHTML code contains the specified substring.
You must use ""
to escape "
, for instance: $$script[tag-content="alert(""this is ad"")"]
For example, take a look at this HTML code:
<script type="text/javascript">
document.write('<div>banner text</div>" />');
</script>
Following rule will delete all script
elements with a banner
substring in their code:
$$script[tag-content="banner"]
The tag-content
special attribute must not appear in a selector to the left of a >
combinator.
wildcard
This special attribute may become unsupported in the future. Prefer using the :contains()
pseudo-class where it is available.
This special attribute works almost like tag-content
and allows you to check the innerHTML code of the document. Rule will check if HTML code of the element fits to the search pattern.
You must use ""
to escape "
, for instance: $$script[wildcard=""banner""]
For example: $$script[wildcard="*banner*text*"]
It checks if the element code contains the two consecutive substrings banner
and text
.
The wildcard
special attribute must not appear in a selector to the left of a >
combinator.
max-length
This special attribute may become unsupported in the future. Prefer using the :contains()
pseudo-class with a regular expression where it is available.
Specifies the maximum length for content of HTML element. If this parameter is set and the content length exceeds the value, a rule does not apply to the element.
Default value
If this parameter is not specified, the max-length
is considered to be 8192.
Zum Beispiel:
$$div[tag-content="banner"][max-length="400"]
This rule will remove all the div
elements, whose code contains the substring banner
and the length of which does not exceed 400
characters.
The max-length
special attribute must not appear in a selector to the left of a >
combinator.
min-length
This special attribute may become unsupported in the future. Prefer using the :contains()
pseudo-class with a regular expression where it is available.
Specifies the minimum length for content of HTML element. If this parameter is set and the content length is less than preset value, a rule does not apply to the element.
Zum Beispiel:
$$div[tag-content="banner"][min-length="400"]
This rule will remove all the div
elements, whose code contains the substring banner
and the length of which exceeds 400
characters.
The min-length
special attribute must not appear in a selector to the left of a >
combinator.
Pseudo-classes
:contains()
Syntax
:contains(unquoted text)
or
:contains(/reg(ular )?ex(pression)?/)
:-abp-contains()
and :has-text()
are synonyms for :contains()
.
The :contains()
pseudo-class is supported by AdGuard for Windows, Mac, and Android with CoreLibs v1.13 or later.
Requires that the inner HTML of the element contains the specified text or matches the specified regular expression.
A :contains()
pseudo-class must not appear in a selector to the left of a >
combinator.
Exceptions
Similar to hiding rules, there is a special type of rules that disable the selected HTML filtering rule for particular domains. The syntax is the same, you just have to change $$
to $@$
.
For example, there is a rule in filter:
$$script[tag-content="banner"]
If you want to disable it for example.com
, you can create an exception rule:
example.com$@$script[tag-content="banner"]
Sometimes, it may be necessary to disable all restriction rules. For example, to conduct tests. To do this, use the exclusion rule without specifying a domain.
$@$script[tag-content="banner"]
We recommend to use this kind of exceptions only if it is not possible to change the hiding rule itself. In other cases it is better to change the original rule, using domain restrictions.
JavaScript rules
AdGuard supports a special type of rules that allows you to inject any javascript code to websites pages.
We strongly recommend using scriptlets instead of JavaScript rules whenever possible. JS rules are supposed to help with debugging, but as a long-time solution a scriptlet rule should be used.
Syntax
rule = [domains] "#%#" script
domains
— domain restriction for the rule. Same principles as in element hiding rules.script
— arbitrary javascript code in one string.
Examples
example.org#%#window.__gaq = undefined;
executes the codewindow.__gaq = undefined;
on all pages atexample.org
and all subdomains.
Exceptions
Similar to hiding rules, there is a special type of rules that disable the selected javascript rule for particular domains. The syntax is the same, you just have to change #%#
to #@%#
.
For example, there is a rule in filter:
#%#window.__gaq = undefined;
If you want to disable it for example.com
, you can create an exception rule:
example.com#@%#window.__gaq = undefined;
Sometimes, it may be necessary to disable all restriction rules. For example, to conduct tests. To do this, use the exclusion rule without specifying a domain.
#@%#window.__gaq = undefined;
We recommend to use this kind of exceptions only if it is not possible to change the hiding rule itself. In other cases it is better to change the original rule, using domain restrictions.
JavaScript rules can be used only in trusted filters.
JavaScript rules are not supported by AdGuard Content Blocker.
Scriptlet rules
Scriptlet is a JavaScript function that provides extended capabilities for content blocking. These functions can be used in a declarative manner in AdGuard filtering rules.
AdGuard supports a lot of different scriptlets. In order to achieve cross-blocker compatibility, we also support syntax of uBO and ABP.
Syntax
rule = [domains] "#%#//scriptlet(" scriptletName arguments ")"
scriptletName
— required, a name of the scriptlet from AdGuard's Scriptlets libraryarguments
— optional, a list ofstring
arguments (no other types of arguments are supported)
Examples
example.org#%#//scriptlet("abort-on-property-read", "alert")
This rule will be applied to example.org
and subdomains pages and will execute the abort-on-property-read
scriptlet with the alert
parameter.
Learn more about how to debug scriptlets.
More information about scriptlets can be found on GitHub.
Scriptlet rules are not supported by AdGuard Content Blocker.
Trusted scriptlets
Trusted scriptlets are scriptlets with extended functionality. It means the same syntax and restrictions. Trusted scriptlet names are prefixed with trusted-
, e.g. trusted-set-cookie
, to be easily distinguished from common scriptlets.
Trusted scriptlets are not compatible with other ad blockers except AdGuard.
Trusted scriptlets rules can be used only in trusted filters.
Trusted scriptlets rules are not supported by AdGuard Content Blocker.
Learn more about how to debug scriptlets.
More information about trusted scriptlets can be found on GitHub.
Modifiers for non-basic type of rules
Each rule can be modified using the modifiers described in the following paragraphs.
Syntax {#non-basic-rules-modifiers-syntax}
rule = "[$" modifiers "]" [rule text]
modifiers = modifier0[, modifier1[, ...[, modifierN]]]
modifier
— set of the modifiers described below.rule text
— a rule to be modified.
For example, [$domain=example.com,app=test_app]##selector
.
In the modifiers values, the following characters must be escaped: [
, ]
, ,
, and \
(unless it is used for the escaping). Use \
to escape them. For example, an escaped bracket looks like this: \]
.
Modifikator \ Produkte | CoreLibs apps | AdGuard for Chromium | AdGuard for Firefox | AdGuard für iOS | AdGuard für Safari | AdGuard-Inhaltsblocker |
---|---|---|---|---|---|---|
$app | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ |
$domain | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
$path | ✅ | ✅ | ✅ | ✅ | ✅ | ❌ |
$url | ✅ | 🧩 * | 🧩 * | ❌ | ❌ | ❌ |
- ✅ — fully supported
- 🧩 — may already be implemented in nightly or beta versions but is not yet supported in release versions
- ❌ — not supported
$app
$app
modifier lets you narrow the rule coverage down to a specific application or a list of applications. The modifier's behavior and syntax perfectly match the corresponding basic rules $app
modifier.
Examples
[$app=org.example.app]example.com##.textad
hides adiv
with the classtextad
atexample.com
and all subdomains in requests sent from theorg.example.app
Android app.[$app=~org.example.app1|~org.example.app2]example.com##.textad
hides adiv
with the classtextad
atexample.com
and all subdomains in requests sent from any app exceptorg.example.app1
andorg.example.app2
.[$app=com.apple.Safari]example.org#%#//scriptlet('prevent-setInterval', 'check', '!300')
applies scriptletprevent-setInterval
only in Safari browser on Mac.[$app=org.example.app]#@#.textad
disables all##.textad
rules for all domains while usingorg.example.app
.
Such rules with $app
modifier are supported by AdGuard for Windows, Mac, and Android.
$domain
$domain
modifier limits the rule application area to a list of domains and their subdomains. The modifier's behavior and syntax perfectly match the corresponding basic rules $domain
modifier.
Examples
[$domain=example.com]##.textad
— hides adiv
with the classtextad
atexample.com
and all subdomains.[$domain=example.com|example.org]###adblock
— hides an element with attributeid
equalsadblock
atexample.com
,example.org
and all subdomains.[$domain=~example.com]##.textad
— this rule hidesdiv
elements of the classtextad
for all domains, exceptexample.com
and its subdomains.
There are 2 ways to specify domain restrictions for non-basic rules:
- the "classic" way is to specify domains before rule mask and attributes:
example.com##.textad
; - the modifier approach is to specify domains via
$domain
modifier:[$domain=example.com]##.textad
.
But rules with mixed style domains restriction are considered invalid. So, for example, the rule [$domain=example.org]example.com##.textad
will be ignored.
Such rules with $domain
modifier are supported by AdGuard for Windows, Mac, Android, and AdGuard Browser Extension for Chrome, Firefox, Edge.
$path
$path
modifier limits the rule application area to specific locations or pages on websites.
Syntax
$path ["=" pattern]
pattern
— optional, a path mask to which the rule is restricted. Its syntax and behavior are pretty much the same as with the pattern for basic rules. You can also use special characters, except for ||
, which does not make any sense in this case (see examples below).
If pattern
is not set for $path
, rule will apply only on the main page of website.
$path
modifier matches the query string as well.
$path
modifier supports regular expressions in the same way basic rules do.
Examples
[$path=page.html]##.textad
hides adiv
with the classtextad
at/page.html
or/page.html?<query>
or/sub/page.html
or/another_page.html
[$path=/page.html]##.textad
hides adiv
with the classtextad
at/page.html
or/page.html?<query>
or/sub/page.html
of any domain but not at/another_page.html
[$path=|/page.html]##.textad
hides adiv
with the classtextad
at/page.html
or/page.html?<query>
of any domain but not at/sub/page.html
[$path=/page.html|]##.textad
hides adiv
with the classtextad
at/page.html
or/sub/page.html
of any domain but not at/page.html?<query>
[$path=/page*.html]example.com##.textad
hides adiv
with the classtextad
at/page1.html
or/page2.html
or any other path matching/page<...>.html
ofexample.com
[$path]example.com##.textad
hides adiv
with the classtextad
at the main page ofexample.com
[$domain=example.com,path=/page.html]##.textad
hides adiv
with the classtextad
atpage.html
ofexample.com
and all subdomains but not atanother_page.html
[$path=/\\/(sub1|sub2)\\/page\\.html/]##.textad
hides adiv
with the classtextad
at both/sub1/page.html
and/sub2/page.html
of any domain (please note the escaped special characters)
Rules with $path
modifier are not supported by AdGuard Content Blocker.
$url
$url
modifier limits the rule application area to URLs matching the specified mask.
Syntax
url = pattern
where pattern
is pretty much the same as pattern
of the basic rules assuming that some characters must be escaped. The special characters and regular expressions are supported as well.
Examples
[$url=||example.com/content/*]##div.textad
hides adiv
with the classtextad
at addresses likehttps://example.com/content/article.html
and evenhttps://subdomain.example.com/content/article.html
.[$url=||example.org^]###adblock
hides an element with attributeid
equal toadblock
atexample.org
and its subdomains.[$url=/\[a-z\]+\\.example\\.com^/]##.textad
hidesdiv
elements of the classtextad
for all domains matching the regular expression[a-z]+\.example\.com^
.
$url
modifier limitations
Rules with the $url
modifier are supported by AdGuard for Windows, Mac, and Android with CoreLibs v1.11 or later, and AdGuard Browser Extension with TSUrlFilter v3.0.0 or later.
Information for filters maintainers
If you maintain a third-party filter that is known to AdGuard, you might be interested in the information presented in this section. Please note that hints will be applied to registered filters only. The filter is considered to be registered and known by AdGuard, if it is present in the known filters index. If you want your filter to be registered, please file an issue to AdguardFilters repo.
Preprocessor directives
We provide preprocessor directives that can be used by filters maintainers to improve compatibility with different ad blockers and provide:
- including a file
- applying rules conditionally by ad blocker type
- content blocker specifying for rules applying in Safari
Any mistake in a preprocessor directive will lead to AdGuard failing the filter update in the same way as if the filter URL was unavailable.
Preprocessor directives can be used in the user rules or in the custom filters.
Including a file
The !#include
directive allows to include contents of a specified file into the filter. It supports only files from the same origin to make sure that the filter maintainer is in control of the specified file. The included file can also contain pre- directives (even other !#include
directives). Ad blockers should consider the case of recursive !#include
and implement a protection mechanism.
Syntax
!#include file_path
where file_path
is a same origin absolute or relative file path to be included.
The files must originate from the same domain but may be located in a different folder.
If included file is not found or unavailable, the whole filter update should fail.
Same-origin limitation should be disabled for local custom filters.
Examples
Filter URL: https://example.org/path/filter.txt
! Valid (same origin):
!#include https://example.org/path/includedfile.txt
!
! Valid (relative path):
!#include /includedfile.txt
!#include ../path2/includedfile.txt
!
! Invalid (another origin):
!#include https://domain.com/path/includedfile.txt
Conditions
Filters maintainers can use conditions to supply different rules depending on the ad blocker type. A conditional directive beginning with an !#if
directive must explicitly be terminated with an !#endif
directive. Conditions support all basic logical operators.
There are two possible scenarios:
When an ad blocker encounters an
!#if
directive and no!#else
directive, it will compile the code between!#if
and!#endif
directives only if the specified condition is true.If there is an
!#else
directive, the code between!#if
and!#else
will be compiled if the condition is true; otherwise, the code between!#else
and!#endif
will be compiled.
Whitespaces matter. !#if
is a valid directive, while !# if
is not.
Syntax
!#if (conditions)
rules_list
!#endif
or
!#if (conditions)
true_conditions_rules_list
!#else
false_conditions_rules_list
!#endif
where:
!#if (conditions)
— start of the block when conditions are trueconditions
— just like in some popular programming languages, preprocessor conditions are based on constants declared by ad blockers. Authors of ad blockers define on their own what exact constants they declare. Possible values:adguard
always declared; shows maintainers that this is one of AdGuard products; should be enough in 95% of cases- product-specific constants for cases when you need a rule to work (or not work — then
!
should be used before constant) in a specific product only:adguard_app_windows
— AdGuard for Windowsadguard_app_mac
— AdGuard for Macadguard_app_android
— AdGuard for Androidadguard_app_ios
— AdGuard for iOSadguard_ext_safari
— AdGuard for Safariadguard_ext_chromium
— AdGuard Browser Extension for Chrome (and chromium-based browsers, e.g. new Microsoft Edge)adguard_ext_firefox
— AdGuard Browser Extension for Firefoxadguard_ext_edge
— AdGuard Browser Extension for Edge Legacyadguard_ext_opera
— AdGuard Browser Extension for Operaadguard_ext_android_cb
— AdGuard Content Blocker for mobile Samsung and Yandex browsersext_ublock
— special case; this one is declared when a uBlock version of a filter is compiled by the FiltersRegistrycap_html_filtering
— products that support HTML filtering rules: AdGuard for Windows, AdGuard for Mac, and AdGuard for Android
!#else
— start of the block when conditions are falserules_list
,true_conditions_rules_list
,false_conditions_rules_list
— lists of rules!#endif
— end of the block
Examples
! for all AdGuard products except AdGuard for Safari
!#if (adguard && !adguard_ext_safari)
||example.org^$third-party
domain.com##div.ad
!#endif
! directives even can be combined
!#if (adguard_app_android)
!#include /androidspecific.txt
!#endif
!#if (adguard && !adguard_ext_safari)
! for all AdGuard products except AdGuard for Safari
||example.org^$third-party
domain.com##div.ad
!#else
! for AdGuard for Safari only
||subdomain.example.org^$third-party
!#endif
The !#else
directive is supported by the FiltersDownloader v1.1.20 or later.
It is already supported for filter lists compiled by the FiltersRegistry, but it still may not be supported by AdGuard products when adding a filter list with !#else
as a custom one. The following products will support it in the mentioned versions or later:
- AdGuard for Windows, Mac, and Android with CoreLibs v1.13;
- AdGuard Browser Extension v4.2.208;
- AdGuard v1.11.16 for Safari.
Safari affinity
Safari's limit for each content blocker is 150,000 active rules. But in AdGuard for Safari and AdGuard for iOS, we've split the rules into 6 content blockers, thus increasing the rule limit to 900,000.
Here is the composition of each content blocker:
- AdGuard General — Ad Blocking, Language-specific
- AdGuard Privacy — Privacy
- AdGuard Social — Social Widgets, Annoyances
- AdGuard Security — Security
- AdGuard Other — Other
- AdGuard Custom — Custom
User rules and allowlist are added to every content blocker.
The main disadvantage of using multiple content blockers is that rules from different blockers are applied independently. Blocking rules are not affected by this, but unblocking rules may cause problems. If a blocking rule is in one content blocker and an exception is in another, the exception will not work. Filter maintainers use !#safari_cb_affinity
to define Safari content blocker affinity for the rules inside of the directive block.
Syntax
!#safari_cb_affinity(content_blockers)
rules_list
!#safari_cb_affinity
where:
!#safari_cb_affinity(content_blockers)
— start of the blockcontent_blockers
— comma-separated list of content blockers. Possible values:general
— AdGuard General content blockerprivacy
— AdGuard Privacy content blockersocial
— AdGuard Social content blockersecurity
— AdGuard Security content blockerother
— AdGuard Other content blockercustom
— AdGuard Custom content blockerall
— special keyword that means that the rules must be included into all content blockers
rules_list
— list of rules!#safari_cb_affinity
— end of the block
Examples
! to unhide specific element which is hidden by AdGuard Base filter:
!#safari_cb_affinity(general)
example.org#@#.adBanner
!#safari_cb_affinity
! to allowlist basic rule from AdGuard Tracking Protection filter:
!#safari_cb_affinity(privacy)
@@||example.org^
!#safari_cb_affinity
Hints
"Hint" is a special comment, instruction to the filters compiler used on the server side (see FiltersRegistry).
Syntax
!+ HINT_NAME1(PARAMS) HINT_NAME2(PARAMS)
Multiple hints can be applied.
NOT_OPTIMIZED
hint
For each filter, AdGuard compiles two versions: full and optimized. Optimized version is much more lightweight and does not contain rules which are not used at all or used rarely.
Rules usage frequency comes from the collected filter rules statistics. But filters optimization is based on more than that — some filters have specific configuration. This is how it looks like for Base filter:
"filter": AdGuard Base filter,
"percent": 30,
"minPercent": 20,
"maxPercent": 40,
"strict": true
where:
- filter — filter identifier
- percent — expected optimization percent
~= (rules count in optimized filter) / (rules count in original filter) * 100
- minPercent — lower bound of
percent
value - maxPercent — upper bound of
percent
value - strict — if
percent < minPercent
ORpercent > maxPercent
and strict mode is on then filter compilation should fail, otherwise original rules must be used
In other words, percent
is the "compression level". For instance, for the Base filter it is configured to 40%. It means that optimization algorithm should strip 60% of rules.
Eventually, here are the two versions of the Base filter for AdGuard Browser Extension:
- full: https://filters.adtidy.org/extension/chromium/filters/2.txt
- optimized: https://filters.adtidy.org/extension/chromium/filters/2_optimized.txt
If you want to add a rule which should not be removed at optimization use the NOT_OPTIMIZED
hint:
!+ NOT_OPTIMIZED
||example.org^
And this rule will not be optimized only for AdGuard for Android:
!+ NOT_OPTIMIZED PLATFORM(android)
||example.org^
PLATFORM
and NOT_PLATFORM
hints
Used to specify the platforms to apply the rules. List of existing platforms and links to Base filter, for example, for each of them:
windows
— AdGuard for Windows — https://filters.adtidy.org/windows/filters/2.txtmac
— AdGuard for Mac — https://filters.adtidy.org/mac_v2/filters/2.txtandroid
— AdGuard for Android — https://filters.adtidy.org/android/filters/2.txtios
— AdGuard for iOS — https://filters.adtidy.org/ios/filters/2.txtext_chromium
— AdGuard Browser Extension for Chrome — https://filters.adtidy.org/extension/chromium/filters/2.txtext_chromium_mv3
— AdGuard Browser Extension for Chrome MV3 — https://filters.adtidy.org/extension/chromium-mv3/filters/2.txtext_ff
— AdGuard Browser Extension for Firefox — https://filters.adtidy.org/extension/firefox/filters/2.txtext_edge
— AdGuard Browser Extension for Edge — https://filters.adtidy.org/extension/edge/filters/2.txtext_opera
— AdGuard Browser Extension for Opera — https://filters.adtidy.org/extension/opera/filters/2.txtext_safari
— AdGuard for Safari — https://filters.adtidy.org/extension/safari/filters/2.txtext_android_cb
— AdGuard Content Blocker — https://filters.adtidy.org/extension/android-content-blocker/filters/2.txtext_ublock
— uBlock Origin — https://filters.adtidy.org/extension/ublock/filters/2.txt
Examples
This rule will be available only in AdGuard for Windows, Mac, Android:
!+ PLATFORM(windows,mac,android)
||example.org^
Except for AdGuard for Safari, AdGuard Content Blocker, and AdGuard for iOS, this rule is available on all platforms:
!+ NOT_PLATFORM(ext_safari, ext_android_cb, ios)
||example.org^
How to debug filtering rules
It may be possible to create simple filtering rules "in your head" but for anything even slightly more complicated you will need additional tools to debug and iterate them. There are tools to assist you with that. You can use DevTools in Chrome and its analogs in other browsers but most AdGuard products provide another one — Filtering log.
Filtering log
Filtering log is an advanced tool that will be helpful mostly to filter developers. It lists all web requests that pass through AdGuard, gives you exhaustive information on each of them, offers multiple sorting options, and has other useful features.
Depending on which AdGuard product you are using, Filtering log can be located in different places.
- In AdGuard for Windows, you can find it in the Ad Blocker tab or via the tray menu
- In AdGuard for Mac, it is located in Settings → Advanced → Filtering log
- In AdGuard for Android, you can find it under Statistics → Recent activity. Recent activity can also be accessed from the Assistant
- In AdGuard Browser Extension, it is accessible from the Miscellaneous settings tab or by right-clicking the extension icon. Only Chromium- and Firefox-based web browsers show applied element hiding rules (including CSS, ExtCSS) and JS rules and scriptlets in their Filtering logs
In AdGuard for iOS and AdGuard for Safari, Filtering log does not exist because of the way content blockers are implemented in Safari. AdGuard does not see the web requests and therefore cannot display them.
Selectors debugging mode
Sometimes, you might need to check the performance of a given selector or a stylesheet. In order to do it without interacting with JavaScript directly, you can use a special debug
style property. When ExtendedCss
meets this property, it enables the debugging mode either for a single selector or for all selectors, depending on the debug
value.
Open the browser console while on a web page to see the timing statistics for selector(s) that were applied there. Debugging mode displays the following stats as object where each of the debugged selectors are keys, and value is an object with such properties:
Always printed:
selectorParsed
— text of eventually parsed selectortimings
— list of DOM nodes matched by the selectorappliesCount
— total number of times that the selector has been applied on the pageappliesTimings
— time that it took to apply the selector on the page, for each of the instances that it has been applied (in milliseconds)meanTiming
— mean time that it took to apply the selector on the pagestandardDeviation
— standard deviationtimingsSum
— total time it took to apply the selector on the page across all instances
Printed only for remove pseudos:
removed
— flag to signal if elements were removed
Printed if elements are not removed:
matchedElements
— list of DOM nodes matched by the selectorstyleApplied
— parsed rule style declaration related to the selector
Examples
Debugging a single selector:
When the value of the debug
property is true
, only information about this selector will be shown in the browser console.
#$?#.banner { display: none; debug: true; }
Enabling global debug:
When the value of the debug
property is global
, the console will display information about all extended CSS selectors that have matches on the current page, for all the rules from any of the enabled filters.
#$?#.banner { display: none; debug: global; }
Testing extended selectors without AdGuard
ExtendedCss can be executed on any page without using any AdGuard product. In order to do that you should copy and execute the following code in a browser console:
!function(e,t,d){C=e.createElement(t),C.src=d,C.onload=function(){alert("ExtendedCss loaded successfully")},s=e.getElementsByTagName(t)[0],s?s.parentNode.insertBefore(C,s):(h=e.getElementsByTagName("head")[0],h.appendChild(C))}(document,"script","https://AdguardTeam.github.io/ExtendedCss/extended-css.min.js");
Alternatively, install the ExtendedCssDebugger userscript.
Now you can now use the ExtendedCss
from global scope, and run its method query()
as Document.querySelectorAll()
.
Examples
const selector = 'div.block:has=(.header:matches-css(after, content: Ads))';
// array of HTMLElements matched the `selector` is to be returned
ExtendedCss.query(selector);
Debugging scriptlets
If you are using AdGuard Browser Extension and want to debug a scriptlet or a trusted scriptlet rule, you can get additional information by opening the Filtering log. In that case, scriptlets will switch to debug mode and there will be more information in the browser console.
The following scriptlets are especially developed for debug purposes:
debug-current-inline-script
debug-on-property-read
debug-on-property-write
log-addEventListener
log-on-stack-trace
log-eval
log
The following scriptlets also may be used for debug purposes:
json-prune
prevent-fetch
prevent-requestAnimationFrame
prevent-setInterval
prevent-setTimeout
prevent-window-open
with specifiedreplacement
parameterprevent-xhr
trusted-replace-fetch-response
trusted-replace-xhr-response
Compatibility tables legend
Product shortcuts
CoreLibs apps
— AdGuard for Windows, AdGuard for Mac, and AdGuard for AndroidAdGuard for Chromium
— AdGuard Browser Extension for Chrome and other Chromium-based browsers such as Microsoft Edge, OperaAdGuard for Firefox
— AdGuard Browser Extension for FirefoxAdGuard for iOS
— AdGuard for iOS and AdGuard Pro for iOS (for mobile Safari browser)AdGuard for Safari
— AdGuard for desktop Safari browserAdGuard Content Blocker
— Content Blocker for Android mobile browsers: Samsung Internet and Yandex Browser
Compatibility shortcuts
- ✅ — fully supported
- ✅ * — supported, but reliability may vary or limitations may occur; check the modifier description for more details
- 🧩 — may already be implemented in nightly or beta versions but is not yet supported in release versions
- ⏳ — feature that has been implemented or is planned to be implemented but is not yet available in any product
- ❌ — not supported
- 👎 — deprecated; still supported but will be removed in the future
- 🚫 — removed and no longer supported