1
0
mirror of https://github.com/gorhill/uBlock.git synced 2024-09-29 22:27:12 +02:00
uBlock/src/js/traffic.js

1161 lines
38 KiB
JavaScript
Raw Normal View History

2014-06-24 00:42:43 +02:00
/*******************************************************************************
2016-03-22 15:19:41 +01:00
uBlock Origin - a browser extension to block requests.
Copyright (C) 2014-present Raymond Hill
2014-06-24 00:42:43 +02:00
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see {http://www.gnu.org/licenses/}.
Home: https://github.com/gorhill/uBlock
*/
2016-07-01 04:03:29 +02:00
'use strict';
2014-06-24 00:42:43 +02:00
/******************************************************************************/
import htmlFilteringEngine from './html-filtering.js';
import httpheaderFilteringEngine from './httpheader-filtering.js';
import logger from './logger.js';
import scriptletFilteringEngine from './scriptlet-filtering.js';
import staticNetFilteringEngine from './static-net-filtering.js';
import textEncode from './text-encode.js';
import µb from './background.js';
import {
sessionFirewall,
sessionSwitches,
sessionURLFiltering,
} from './filtering-engines.js';
import {
entityFromDomain,
isNetworkURI,
} from './uri-utils.js';
2014-06-24 00:42:43 +02:00
/******************************************************************************/
// Platform-specific behavior.
// https://github.com/uBlockOrigin/uBlock-issues/issues/42
// https://bugzilla.mozilla.org/show_bug.cgi?id=1376932
// Add proper version number detection once issue is fixed in Firefox.
let dontCacheResponseHeaders =
vAPI.webextFlavor.soup.has('firefox');
// The real actual webextFlavor value may not be set in stone, so listen
// for possible future changes.
window.addEventListener('webextFlavor', function() {
dontCacheResponseHeaders =
vAPI.webextFlavor.soup.has('firefox');
}, { once: true });
// https://github.com/uBlockOrigin/uBlock-issues/issues/1553
const supportsFloc = document.interestCohort instanceof Function;
/******************************************************************************/
const patchLocalRedirectURL = url => url.charCodeAt(0) === 0x2F /* '/' */
? vAPI.getURL(url)
: url;
/******************************************************************************/
2014-07-26 15:55:12 +02:00
// Intercept and filter web requests.
2014-07-14 17:24:59 +02:00
const onBeforeRequest = function(details) {
const fctxt = µb.filteringContext.fromWebrequestDetails(details);
2014-07-26 01:29:51 +02:00
// Special handling for root document.
2015-04-07 03:26:05 +02:00
// https://github.com/chrisaljoudi/uBlock/issues/1001
// This must be executed regardless of whether the request is
// behind-the-scene
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
if ( fctxt.itype === fctxt.MAIN_FRAME ) {
return onBeforeRootFrameRequest(fctxt);
2014-07-14 17:24:59 +02:00
}
// Special treatment: behind-the-scene requests
const tabId = details.tabId;
if ( tabId < 0 ) {
return onBeforeBehindTheSceneRequest(fctxt);
}
2014-07-26 01:29:51 +02:00
// Lookup the page store associated with this tab id.
let pageStore = µb.pageStoreFromTabId(tabId);
if ( pageStore === null ) {
const tabContext = µb.tabContextManager.mustLookup(tabId);
if ( tabContext.tabId < 0 ) {
return onBeforeBehindTheSceneRequest(fctxt);
}
vAPI.tabs.onNavigation({ tabId, frameId: 0, url: tabContext.rawURL });
2015-04-09 00:46:08 +02:00
pageStore = µb.pageStoreFromTabId(tabId);
2014-07-14 20:40:40 +02:00
}
2014-07-15 13:38:34 +02:00
const result = pageStore.filterRequest(fctxt);
pageStore.journalAddRequest(fctxt, result);
if ( logger.enabled ) {
fctxt.setRealm('network').toLogger();
}
2015-04-09 00:46:08 +02:00
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
// Redirected
if ( fctxt.redirectURL !== undefined ) {
return { redirectUrl: patchLocalRedirectURL(fctxt.redirectURL) };
2014-07-14 17:24:59 +02:00
}
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
// Not redirected
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
// Blocked
if ( result === 1 ) {
return { cancel: true };
}
2014-07-14 17:24:59 +02:00
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
// Not blocked
if (
fctxt.itype === fctxt.SUB_FRAME &&
details.parentFrameId !== -1 &&
details.aliasURL === undefined
) {
pageStore.setFrameURL(details);
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
}
if ( result === 2 ) {
return { cancel: false };
}
2014-07-14 17:24:59 +02:00
};
/******************************************************************************/
const onBeforeRootFrameRequest = function(fctxt) {
const requestURL = fctxt.url;
2015-03-26 00:28:22 +01:00
2015-03-21 21:52:35 +01:00
// Special handling for root document.
2015-04-07 03:26:05 +02:00
// https://github.com/chrisaljoudi/uBlock/issues/1001
// This must be executed regardless of whether the request is
// behind-the-scene
const requestHostname = fctxt.getHostname();
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
let result = 0;
let logData;
2015-03-26 00:28:22 +01:00
// If the site is whitelisted, disregard strict blocking
const trusted = µb.getNetFilteringSwitch(requestURL) === false;
if ( trusted ) {
result = 2;
if ( logger.enabled ) {
logData = { engine: 'u', result: 2, raw: 'whitelisted' };
}
}
2015-03-27 18:00:55 +01:00
// Permanently unrestricted?
if (
result === 0 &&
sessionSwitches.evaluateZ('no-strict-blocking', requestHostname)
) {
result = 2;
if ( logger.enabled ) {
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
logData = {
engine: 'u',
result: 2,
raw: `no-strict-blocking: ${sessionSwitches.z} true`
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
};
}
2015-03-27 18:00:55 +01:00
}
2015-03-26 00:28:22 +01:00
// Temporarily whitelisted?
if ( result === 0 && strictBlockBypasser.isBypassed(requestHostname) ) {
result = 2;
if ( logger.enabled ) {
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
logData = {
engine: 'u',
result: 2,
raw: 'no-strict-blocking: true (temporary)'
};
}
2015-03-21 21:52:35 +01:00
}
2015-03-26 00:28:22 +01:00
// Static filtering
if ( result === 0 ) {
({ result, logData } = shouldStrictBlock(fctxt, logger.enabled));
2015-03-26 00:28:22 +01:00
}
const pageStore = µb.bindTabToPageStore(fctxt.tabId, 'beforeRequest');
if ( pageStore !== null ) {
2016-10-08 16:15:31 +02:00
pageStore.journalAddRootFrame('uncommitted', requestURL);
pageStore.journalAddRequest(fctxt, result);
2015-03-26 00:28:22 +01:00
}
if ( logger.enabled ) {
fctxt.setFilter(logData);
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
}
// https://github.com/uBlockOrigin/uBlock-issues/issues/760
// Redirect non-blocked request?
if (
result !== 1 &&
trusted === false &&
pageStore !== null &&
staticNetFilteringEngine.hasQuery(fctxt)
) {
pageStore.redirectNonBlockedRequest(fctxt);
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
}
if ( logger.enabled ) {
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
fctxt.setRealm('network').toLogger();
}
// Redirected
if ( fctxt.redirectURL !== undefined ) {
return { redirectUrl: patchLocalRedirectURL(fctxt.redirectURL) };
}
2015-03-26 00:28:22 +01:00
// Not blocked
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
if ( result !== 1 ) { return; }
2015-03-26 00:28:22 +01:00
// No log data means no strict blocking (because we need to report why
// the blocking occurs.
if ( logData === undefined ) { return; }
2015-06-12 01:33:30 +02:00
2015-03-26 00:28:22 +01:00
// Blocked
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
const query = encodeURIComponent(JSON.stringify({
2015-03-26 00:28:22 +01:00
url: requestURL,
2015-03-30 19:10:29 +02:00
hn: requestHostname,
dn: fctxt.getDomain() || requestHostname,
fs: logData.raw
2015-03-26 00:28:22 +01:00
}));
2015-03-27 18:00:55 +01:00
vAPI.tabs.replace(
fctxt.tabId,
vAPI.getURL('document-blocked.html?details=') + query
);
2015-03-27 18:00:55 +01:00
return { cancel: true };
2015-03-21 21:52:35 +01:00
};
/******************************************************************************/
// Strict blocking through static filtering
//
// https://github.com/chrisaljoudi/uBlock/issues/1128
// Do not block if the match begins after the hostname,
// except when the filter is specifically of type `other`.
// https://github.com/gorhill/uBlock/issues/490
// Removing this for the time being, will need a new, dedicated type.
// https://github.com/uBlockOrigin/uBlock-issues/issues/1501
// Support explicit exception filters.
//
// Let result of match for specific `document` type be `rs`
// Let result of match for no specific type be `rg` *after* going through
// confirmation necessary for implicit matches
// Let `important` be `i`
// Let final result be logical combination of `rs` and `rg` as follow:
//
// | rs |
// +--------+--------+--------+--------|
// | 0 | 1 | 1i | 2 |
// --------+--------+--------+--------+--------+--------|
// | 0 | rg | rs | rs | rs |
// rg | 1 | rg | rs | rs | rs |
// | 1i | rg | rg | rs | rg |
// | 2 | rg | rg | rs | rs |
// --------+--------+--------+--------+--------+--------+
const shouldStrictBlock = function(fctxt, loggerEnabled) {
const snfe = staticNetFilteringEngine;
// Explicit filtering: `document` option
const rs = snfe.matchRequest(fctxt, 0b0011);
const is = rs === 1 && snfe.isBlockImportant();
let lds;
if ( rs !== 0 || loggerEnabled ) {
lds = snfe.toLogData();
}
// | rs |
// +--------+--------+--------+--------|
// | 0 | 1 | 1i | 2 |
// --------+--------+--------+--------+--------+--------|
// | 0 | rg | rs | x | rs |
// rg | 1 | rg | rs | x | rs |
// | 1i | rg | rg | x | rg |
// | 2 | rg | rg | x | rs |
// --------+--------+--------+--------+--------+--------+
if ( rs === 1 && is ) {
return { result: rs, logData: lds };
}
// Implicit filtering: no `document` option
fctxt.type = 'no_type';
let rg = snfe.matchRequest(fctxt, 0b0011);
fctxt.type = 'main_frame';
const ig = rg === 1 && snfe.isBlockImportant();
let ldg;
if ( rg !== 0 || loggerEnabled ) {
ldg = snfe.toLogData();
if ( rg === 1 && validateStrictBlock(fctxt, ldg) === false ) {
rg = 0; ldg = undefined;
}
}
// | rs |
// +--------+--------+--------+--------|
// | 0 | 1 | 1i | 2 |
// --------+--------+--------+--------+--------+--------|
// | 0 | x | rs | - | rs |
// rg | 1 | x | rs | - | rs |
// | 1i | x | x | - | x |
// | 2 | x | x | - | rs |
// --------+--------+--------+--------+--------+--------+
if ( rs === 0 || rg === 1 && ig || rg === 2 && rs !== 2 ) {
return { result: rg, logData: ldg };
}
// | rs |
// +--------+--------+--------+--------|
// | 0 | 1 | 1i | 2 |
// --------+--------+--------+--------+--------+--------|
// | 0 | - | x | - | x |
// rg | 1 | - | x | - | x |
// | 1i | - | - | - | - |
// | 2 | - | - | - | x |
// --------+--------+--------+--------+--------+--------+
return { result: rs, logData: lds };
};
/******************************************************************************/
// https://github.com/gorhill/uBlock/issues/3208
// Mind case insensitivity.
// https://github.com/uBlockOrigin/uBlock-issues/issues/1147
// Do not strict-block if the filter pattern does not contain at least one
// token character.
const validateStrictBlock = function(fctxt, logData) {
2017-11-09 21:46:25 +01:00
if ( typeof logData.regex !== 'string' ) { return false; }
if ( typeof logData.raw === 'string' && /\w/.test(logData.raw) === false ) {
return false;
}
const url = fctxt.url;
const re = new RegExp(logData.regex, 'i');
const match = re.exec(url.toLowerCase());
if ( match === null ) { return false; }
2015-03-30 23:42:12 +02:00
// https://github.com/chrisaljoudi/uBlock/issues/1128
// https://github.com/chrisaljoudi/uBlock/issues/1212
// Verify that the end of the match is anchored to the end of the
// hostname.
// https://github.com/uBlockOrigin/uAssets/issues/7619#issuecomment-653010310
// Also match FQDN.
const hostname = fctxt.getHostname();
const hnpos = url.indexOf(hostname);
const hnlen = hostname.length;
const end = match.index + match[0].length - hnpos - hnlen;
return end === 0 || end === 1 ||
end === 2 && url.charCodeAt(hnpos + hnlen) === 0x2E /* '.' */;
2015-03-30 23:42:12 +02:00
};
/******************************************************************************/
2016-10-14 16:06:34 +02:00
// Intercept and filter behind-the-scene requests.
const onBeforeBehindTheSceneRequest = function(fctxt) {
const pageStore = µb.pageStoreFromTabId(fctxt.tabId);
2018-02-26 19:59:16 +01:00
if ( pageStore === null ) { return; }
2016-10-14 16:06:34 +02:00
2017-10-19 15:35:28 +02:00
// https://github.com/gorhill/uBlock/issues/3150
// Ability to globally block CSP reports MUST also apply to
// behind-the-scene network requests.
let result = 0;
// https://github.com/uBlockOrigin/uBlock-issues/issues/339
// Need to also test against `-scheme` since tabOrigin is normalized.
// Not especially elegant but for now this accomplishes the purpose of
// not dealing with network requests fired from a synthetic scope,
// that is unless advanced user mode is enabled.
if (
fctxt.tabOrigin.endsWith('-scheme') === false &&
isNetworkURI(fctxt.tabOrigin) ||
µb.userSettings.advancedUserEnabled ||
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
fctxt.itype === fctxt.CSP_REPORT
) {
result = pageStore.filterRequest(fctxt);
// The "any-tab" scope is not whitelist-able, and in such case we must
// use the origin URL as the scope. Most such requests aren't going to
// be blocked, so we test for whitelisting and modify the result only
// when the request is being blocked.
//
// https://github.com/uBlockOrigin/uBlock-issues/issues/1478
// Also remove potential redirection when request is to be
// whitelisted.
if (
result === 1 &&
µb.getNetFilteringSwitch(fctxt.tabOrigin) === false
) {
result = 2;
fctxt.redirectURL = undefined;
fctxt.filter = { engine: 'u', result: 2, raw: 'whitelisted' };
}
}
2015-01-24 18:06:22 +01:00
// https://github.com/uBlockOrigin/uBlock-issues/issues/1204
onBeforeBehindTheSceneRequest.journalAddRequest(fctxt, result);
if ( logger.enabled ) {
fctxt.setRealm('network').toLogger();
}
2016-07-01 04:03:29 +02:00
// Redirected
if ( fctxt.redirectURL !== undefined ) {
return { redirectUrl: patchLocalRedirectURL(fctxt.redirectURL) };
}
// Blocked?
if ( result === 1 ) {
return { cancel: true };
}
2015-01-24 18:06:22 +01:00
};
// https://github.com/uBlockOrigin/uBlock-issues/issues/1204
// Report the tabless network requests to all page stores matching the
// document origin. This is an approximation, there is unfortunately no
// way to know for sure which exact page triggered a tabless network
// request.
{
let hostname = '';
let pageStores = new Set();
let pageStoresToken = 0;
let gcTimer;
const reset = function() {
hostname = '';
pageStores = new Set();
pageStoresToken = 0;
};
const gc = ( ) => {
gcTimer = undefined;
if ( pageStoresToken !== µb.pageStoresToken ) { return reset(); }
gcTimer = vAPI.setTimeout(gc, 30011);
};
onBeforeBehindTheSceneRequest.journalAddRequest = (fctxt, result) => {
const docHostname = fctxt.getDocHostname();
if (
docHostname !== hostname ||
pageStoresToken !== µb.pageStoresToken
) {
hostname = docHostname;
pageStores = new Set();
for ( const pageStore of µb.pageStores.values() ) {
if ( pageStore.tabHostname !== docHostname ) { continue; }
pageStores.add(pageStore);
}
pageStoresToken = µb.pageStoresToken;
if ( gcTimer !== undefined ) {
clearTimeout(gcTimer);
}
gcTimer = vAPI.setTimeout(gc, 30011);
}
for ( const pageStore of pageStores ) {
pageStore.journalAddRequest(fctxt, result);
}
};
}
2015-01-24 18:06:22 +01:00
/******************************************************************************/
// To handle:
// - Media elements larger than n kB
// - Scriptlet injection (requires ability to modify response body)
// - HTML filtering (requires ability to modify response body)
// - CSP injection
2014-09-24 23:38:22 +02:00
const onHeadersReceived = function(details) {
// https://github.com/uBlockOrigin/uBlock-issues/issues/610
// Process behind-the-scene requests in a special way.
if (
details.tabId < 0 &&
normalizeBehindTheSceneResponseHeaders(details) === false
) {
return;
}
2014-09-24 23:38:22 +02:00
const fctxt = µb.filteringContext.fromWebrequestDetails(details);
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
const isRootDoc = fctxt.itype === fctxt.MAIN_FRAME;
let pageStore = µb.pageStoreFromTabId(fctxt.tabId);
if ( pageStore === null ) {
if ( isRootDoc === false ) { return; }
pageStore = µb.bindTabToPageStore(fctxt.tabId, 'beforeRequest');
2014-09-24 23:38:22 +02:00
}
if ( pageStore.getNetFilteringSwitch(fctxt) === false ) { return; }
2015-04-09 00:46:08 +02:00
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
if ( fctxt.itype === fctxt.IMAGE || fctxt.itype === fctxt.MEDIA ) {
Add support for `1P`, `3P`, `header=` filter options and other changes New filter options ================== Strict partyness: `1P`, `3P` ---------------------------- The current options 1p/3p are meant to "weakly" match partyness, i.e. a network request is considered 1st-party to its context as long as both the context and the request share the same base domain. The new partyness options are meant to check for strict partyness, i.e. a network request will be considered 1st-party if and only if both the context and the request share the same hostname. For examples: - context: `www.example.org` - request: `www.example.org` - `1p`: yes, `1P`: yes - `3p`: no, `3P`: no - context: `www.example.org` - request: `subdomain.example.org` - `1p`: yes, `1P`: no - `3p`: no, `3P`: yes - context: `www.example.org` - request: `www.example.com` - `1p`: no, `1P`: no - `3p`: yes, `3P`: yes The strict partyness options will be visually emphasized in the editor so as to prevent mistakenly using `1P` or `3P` where weak partyness is meant to be used. Filter on response headers: `header=` ------------------------------------- Currently experimental and under evaluation. Disabled by default, enable by toggling `filterOnHeaders` to `true` in advanced settings. Ability to filter network requests according to whether a specific response header is present and whether it matches or does not match a specific value. For example: *$1p,3P,script,header=via:1\.1\s+google The above filter is meant to block network requests which fullfill all the following conditions: - is weakly 1st-party to the context - is not strictly 1st-party to the context - is of type `script` - has a response HTTP header named `via`, which value matches the regular expression `1\.1\s+google`. The matches are always performed in a case-insensitive manner. The header value is assumed to be a literal regular expression, except for the following special characters: - to anchor to start of string, use leading `|`, not `^` - to anchor to end of string, use trailing `|`, not `$` - to invert the test, use a leading `!` To block a network request if it merely contains a specific HTTP header is just a matter of specifying the header name without a header value: *$1p,3P,script,header=via Generic exception filters can be used to disable specific block `header=` filters, i.e. `@@*$1p,3P,script,header` will override the block `header=` filters given as example above. Dynamic filtering's `allow` rules override block `headers=` filters. Important: It is key that filter authors use as many narrowing filter options as possible when using the `header=` option, and the `header=` option should be used ONLY when other filter options are not sufficient. More documentation justifying the purpose of `header=` option will be provided eventually if ever it is decided to move it from experimental to stable status. To be decided: to restrict usage of this filter option to only uBO's own filter lists or "My filters". Changes ======= Fine tuning `queryprune=` ------------------------- The following changes have been implemented: The special value `*` (i.e. `queryprune=*`) means "remove all query parameters". If the `queryprune=` value is made only of alphanumeric characters (including `_`), the value will be internally converted to regex equivalent `^value=`. This ensures a better future compatibility with AdGuard's `removeparam=`. If the `queryprune=` value starts with `!`, the test will be inverted. This can be used to remove all query parameters EXCEPT those who match the specified value. Other ----- The legacy code to test for spurious CSP reports has been removed. This is no longer an issue ever since uBO redirects to local resources through web accessible resources. Notes ===== The following new and recently added filter options are not compatible with Chromium's manifest v3 changes: - `queryprune=` - `1P` - `3P` - `header=`
2020-11-23 14:22:43 +01:00
const result = foilLargeMediaElement(details, fctxt, pageStore);
if ( result !== undefined ) { return result; }
2015-04-09 00:46:08 +02:00
}
// Keep in mind response headers will be modified in-place if needed, so
// `details.responseHeaders` will always point to the modified response
// headers.
const { responseHeaders } = details;
if ( Array.isArray(responseHeaders) === false ) { return; }
Add support for `1P`, `3P`, `header=` filter options and other changes New filter options ================== Strict partyness: `1P`, `3P` ---------------------------- The current options 1p/3p are meant to "weakly" match partyness, i.e. a network request is considered 1st-party to its context as long as both the context and the request share the same base domain. The new partyness options are meant to check for strict partyness, i.e. a network request will be considered 1st-party if and only if both the context and the request share the same hostname. For examples: - context: `www.example.org` - request: `www.example.org` - `1p`: yes, `1P`: yes - `3p`: no, `3P`: no - context: `www.example.org` - request: `subdomain.example.org` - `1p`: yes, `1P`: no - `3p`: no, `3P`: yes - context: `www.example.org` - request: `www.example.com` - `1p`: no, `1P`: no - `3p`: yes, `3P`: yes The strict partyness options will be visually emphasized in the editor so as to prevent mistakenly using `1P` or `3P` where weak partyness is meant to be used. Filter on response headers: `header=` ------------------------------------- Currently experimental and under evaluation. Disabled by default, enable by toggling `filterOnHeaders` to `true` in advanced settings. Ability to filter network requests according to whether a specific response header is present and whether it matches or does not match a specific value. For example: *$1p,3P,script,header=via:1\.1\s+google The above filter is meant to block network requests which fullfill all the following conditions: - is weakly 1st-party to the context - is not strictly 1st-party to the context - is of type `script` - has a response HTTP header named `via`, which value matches the regular expression `1\.1\s+google`. The matches are always performed in a case-insensitive manner. The header value is assumed to be a literal regular expression, except for the following special characters: - to anchor to start of string, use leading `|`, not `^` - to anchor to end of string, use trailing `|`, not `$` - to invert the test, use a leading `!` To block a network request if it merely contains a specific HTTP header is just a matter of specifying the header name without a header value: *$1p,3P,script,header=via Generic exception filters can be used to disable specific block `header=` filters, i.e. `@@*$1p,3P,script,header` will override the block `header=` filters given as example above. Dynamic filtering's `allow` rules override block `headers=` filters. Important: It is key that filter authors use as many narrowing filter options as possible when using the `header=` option, and the `header=` option should be used ONLY when other filter options are not sufficient. More documentation justifying the purpose of `header=` option will be provided eventually if ever it is decided to move it from experimental to stable status. To be decided: to restrict usage of this filter option to only uBO's own filter lists or "My filters". Changes ======= Fine tuning `queryprune=` ------------------------- The following changes have been implemented: The special value `*` (i.e. `queryprune=*`) means "remove all query parameters". If the `queryprune=` value is made only of alphanumeric characters (including `_`), the value will be internally converted to regex equivalent `^value=`. This ensures a better future compatibility with AdGuard's `removeparam=`. If the `queryprune=` value starts with `!`, the test will be inverted. This can be used to remove all query parameters EXCEPT those who match the specified value. Other ----- The legacy code to test for spurious CSP reports has been removed. This is no longer an issue ever since uBO redirects to local resources through web accessible resources. Notes ===== The following new and recently added filter options are not compatible with Chromium's manifest v3 changes: - `queryprune=` - `1P` - `3P` - `header=`
2020-11-23 14:22:43 +01:00
if ( isRootDoc === false && µb.hiddenSettings.filterOnHeaders === true ) {
const result = pageStore.filterOnHeaders(fctxt, responseHeaders);
if ( result !== 0 ) {
if ( logger.enabled ) {
Add support for `1P`, `3P`, `header=` filter options and other changes New filter options ================== Strict partyness: `1P`, `3P` ---------------------------- The current options 1p/3p are meant to "weakly" match partyness, i.e. a network request is considered 1st-party to its context as long as both the context and the request share the same base domain. The new partyness options are meant to check for strict partyness, i.e. a network request will be considered 1st-party if and only if both the context and the request share the same hostname. For examples: - context: `www.example.org` - request: `www.example.org` - `1p`: yes, `1P`: yes - `3p`: no, `3P`: no - context: `www.example.org` - request: `subdomain.example.org` - `1p`: yes, `1P`: no - `3p`: no, `3P`: yes - context: `www.example.org` - request: `www.example.com` - `1p`: no, `1P`: no - `3p`: yes, `3P`: yes The strict partyness options will be visually emphasized in the editor so as to prevent mistakenly using `1P` or `3P` where weak partyness is meant to be used. Filter on response headers: `header=` ------------------------------------- Currently experimental and under evaluation. Disabled by default, enable by toggling `filterOnHeaders` to `true` in advanced settings. Ability to filter network requests according to whether a specific response header is present and whether it matches or does not match a specific value. For example: *$1p,3P,script,header=via:1\.1\s+google The above filter is meant to block network requests which fullfill all the following conditions: - is weakly 1st-party to the context - is not strictly 1st-party to the context - is of type `script` - has a response HTTP header named `via`, which value matches the regular expression `1\.1\s+google`. The matches are always performed in a case-insensitive manner. The header value is assumed to be a literal regular expression, except for the following special characters: - to anchor to start of string, use leading `|`, not `^` - to anchor to end of string, use trailing `|`, not `$` - to invert the test, use a leading `!` To block a network request if it merely contains a specific HTTP header is just a matter of specifying the header name without a header value: *$1p,3P,script,header=via Generic exception filters can be used to disable specific block `header=` filters, i.e. `@@*$1p,3P,script,header` will override the block `header=` filters given as example above. Dynamic filtering's `allow` rules override block `headers=` filters. Important: It is key that filter authors use as many narrowing filter options as possible when using the `header=` option, and the `header=` option should be used ONLY when other filter options are not sufficient. More documentation justifying the purpose of `header=` option will be provided eventually if ever it is decided to move it from experimental to stable status. To be decided: to restrict usage of this filter option to only uBO's own filter lists or "My filters". Changes ======= Fine tuning `queryprune=` ------------------------- The following changes have been implemented: The special value `*` (i.e. `queryprune=*`) means "remove all query parameters". If the `queryprune=` value is made only of alphanumeric characters (including `_`), the value will be internally converted to regex equivalent `^value=`. This ensures a better future compatibility with AdGuard's `removeparam=`. If the `queryprune=` value starts with `!`, the test will be inverted. This can be used to remove all query parameters EXCEPT those who match the specified value. Other ----- The legacy code to test for spurious CSP reports has been removed. This is no longer an issue ever since uBO redirects to local resources through web accessible resources. Notes ===== The following new and recently added filter options are not compatible with Chromium's manifest v3 changes: - `queryprune=` - `1P` - `3P` - `header=`
2020-11-23 14:22:43 +01:00
fctxt.setRealm('network').toLogger();
}
if ( result === 1 ) {
pageStore.journalAddRequest(fctxt, 1);
Add support for `1P`, `3P`, `header=` filter options and other changes New filter options ================== Strict partyness: `1P`, `3P` ---------------------------- The current options 1p/3p are meant to "weakly" match partyness, i.e. a network request is considered 1st-party to its context as long as both the context and the request share the same base domain. The new partyness options are meant to check for strict partyness, i.e. a network request will be considered 1st-party if and only if both the context and the request share the same hostname. For examples: - context: `www.example.org` - request: `www.example.org` - `1p`: yes, `1P`: yes - `3p`: no, `3P`: no - context: `www.example.org` - request: `subdomain.example.org` - `1p`: yes, `1P`: no - `3p`: no, `3P`: yes - context: `www.example.org` - request: `www.example.com` - `1p`: no, `1P`: no - `3p`: yes, `3P`: yes The strict partyness options will be visually emphasized in the editor so as to prevent mistakenly using `1P` or `3P` where weak partyness is meant to be used. Filter on response headers: `header=` ------------------------------------- Currently experimental and under evaluation. Disabled by default, enable by toggling `filterOnHeaders` to `true` in advanced settings. Ability to filter network requests according to whether a specific response header is present and whether it matches or does not match a specific value. For example: *$1p,3P,script,header=via:1\.1\s+google The above filter is meant to block network requests which fullfill all the following conditions: - is weakly 1st-party to the context - is not strictly 1st-party to the context - is of type `script` - has a response HTTP header named `via`, which value matches the regular expression `1\.1\s+google`. The matches are always performed in a case-insensitive manner. The header value is assumed to be a literal regular expression, except for the following special characters: - to anchor to start of string, use leading `|`, not `^` - to anchor to end of string, use trailing `|`, not `$` - to invert the test, use a leading `!` To block a network request if it merely contains a specific HTTP header is just a matter of specifying the header name without a header value: *$1p,3P,script,header=via Generic exception filters can be used to disable specific block `header=` filters, i.e. `@@*$1p,3P,script,header` will override the block `header=` filters given as example above. Dynamic filtering's `allow` rules override block `headers=` filters. Important: It is key that filter authors use as many narrowing filter options as possible when using the `header=` option, and the `header=` option should be used ONLY when other filter options are not sufficient. More documentation justifying the purpose of `header=` option will be provided eventually if ever it is decided to move it from experimental to stable status. To be decided: to restrict usage of this filter option to only uBO's own filter lists or "My filters". Changes ======= Fine tuning `queryprune=` ------------------------- The following changes have been implemented: The special value `*` (i.e. `queryprune=*`) means "remove all query parameters". If the `queryprune=` value is made only of alphanumeric characters (including `_`), the value will be internally converted to regex equivalent `^value=`. This ensures a better future compatibility with AdGuard's `removeparam=`. If the `queryprune=` value starts with `!`, the test will be inverted. This can be used to remove all query parameters EXCEPT those who match the specified value. Other ----- The legacy code to test for spurious CSP reports has been removed. This is no longer an issue ever since uBO redirects to local resources through web accessible resources. Notes ===== The following new and recently added filter options are not compatible with Chromium's manifest v3 changes: - `queryprune=` - `1P` - `3P` - `header=`
2020-11-23 14:22:43 +01:00
return { cancel: true };
}
}
}
if ( isRootDoc === false && fctxt.itype !== fctxt.SUB_FRAME ) { return; }
2017-07-22 22:58:08 +02:00
// https://github.com/gorhill/uBlock/issues/2813
// Disable the blocking of large media elements if the document is itself
// a media element: the resource was not prevented from loading so no
// point to further block large media elements for the current document.
if ( isRootDoc ) {
const contentType = headerValueFromName('content-type', responseHeaders);
if ( reMediaContentTypes.test(contentType) ) {
pageStore.allowLargeMediaElementsUntil = 0;
// Fall-through: this could be an SVG document, which supports
// script tags.
2017-07-22 22:58:08 +02:00
}
}
// At this point we have a HTML document.
const filteredHTML =
µb.canFilterResponseData && filterDocument(fctxt, details) === true;
let modifiedHeaders = false;
if ( httpheaderFilteringEngine.apply(fctxt, responseHeaders) === true ) {
modifiedHeaders = true;
}
if ( injectCSP(fctxt, pageStore, responseHeaders) === true ) {
modifiedHeaders = true;
}
if ( supportsFloc && foilFloc(fctxt, responseHeaders) ) {
modifiedHeaders = true;
}
// https://bugzilla.mozilla.org/show_bug.cgi?id=1376932
// Prevent document from being cached by the browser if we modified it,
// either through HTML filtering and/or modified response headers.
// https://github.com/uBlockOrigin/uBlock-issues/issues/229
// Use `no-cache` instead of `no-cache, no-store, must-revalidate`, this
// allows Firefox's offline mode to work as expected.
if ( (filteredHTML || modifiedHeaders) && dontCacheResponseHeaders ) {
const cacheControl = µb.hiddenSettings.cacheControlForFirefox1376932;
if ( cacheControl !== 'unset' ) {
let i = headerIndexFromName('cache-control', responseHeaders);
if ( i !== -1 ) {
responseHeaders[i].value = cacheControl;
} else {
responseHeaders.push({ name: 'Cache-Control', value: cacheControl });
}
modifiedHeaders = true;
}
}
if ( modifiedHeaders ) {
return { responseHeaders };
}
};
const reMediaContentTypes = /^(?:audio|image|video)\//;
2017-07-22 22:58:08 +02:00
/******************************************************************************/
// https://github.com/uBlockOrigin/uBlock-issues/issues/610
const normalizeBehindTheSceneResponseHeaders = function(details) {
if ( details.type !== 'xmlhttprequest' ) { return false; }
const headers = details.responseHeaders;
if ( Array.isArray(headers) === false ) { return false; }
const contentType = headerValueFromName('content-type', headers);
if ( contentType === '' ) { return false; }
if ( reMediaContentTypes.test(contentType) === false ) { return false; }
if ( contentType.startsWith('image') ) {
details.type = 'image';
} else {
details.type = 'media';
}
return true;
};
/*******************************************************************************
The response body filterer is responsible for:
- HTML filtering
In the spirit of efficiency, the response body filterer works this way:
If:
- HTML filtering: no.
Then:
No response body filtering is initiated.
If:
- HTML filtering: yes.
Then:
Assemble all response body data into a single buffer. Once all the
response data has been received, create a document from it. Then:
- Remove all DOM elements matching HTML filters.
Then serialize the resulting modified document as the new response
body.
**/
const filterDocument = (( ) => {
const filterers = new Map();
let domParser, xmlSerializer,
utf8TextDecoder, textDecoder, textEncoder;
const textDecode = function(encoding, buffer) {
2018-02-18 13:16:10 +01:00
if (
textDecoder !== undefined &&
textDecoder.encoding !== encoding
) {
textDecoder = undefined;
}
if ( textDecoder === undefined ) {
textDecoder = new TextDecoder(encoding);
}
return textDecoder.decode(buffer);
};
const reContentTypeDocument = /^(?:text\/html|application\/xhtml\+xml)/i;
const reContentTypeCharset = /charset=['"]?([^'" ]+)/i;
2018-01-03 19:59:38 +01:00
const mimeFromContentType = function(contentType) {
const match = reContentTypeDocument.exec(contentType);
2018-03-01 20:12:16 +01:00
if ( match !== null ) {
return match[0].toLowerCase();
}
};
const charsetFromContentType = function(contentType) {
const match = reContentTypeCharset.exec(contentType);
2018-01-05 00:26:52 +01:00
if ( match !== null ) {
return match[1].toLowerCase();
}
};
const charsetFromDoc = function(doc) {
let meta = doc.querySelector('meta[charset]');
2018-01-05 00:26:52 +01:00
if ( meta !== null ) {
return meta.getAttribute('charset').toLowerCase();
}
meta = doc.querySelector(
'meta[http-equiv="content-type" i][content]'
);
if ( meta !== null ) {
return charsetFromContentType(meta.getAttribute('content'));
}
};
const streamClose = function(filterer, buffer) {
if ( buffer !== undefined ) {
filterer.stream.write(buffer);
} else if ( filterer.buffer !== undefined ) {
filterer.stream.write(filterer.buffer);
}
filterer.stream.close();
};
const onStreamData = function(ev) {
const filterer = filterers.get(this);
if ( filterer === undefined ) {
this.write(ev.data);
this.disconnect();
return;
}
if (
this.status !== 'transferringdata' &&
this.status !== 'finishedtransferringdata'
) {
filterers.delete(this);
this.disconnect();
return;
}
2017-12-30 17:21:23 +01:00
// TODO:
// - Possibly improve buffer growth, if benchmarking shows it's worth
// it.
// - Also evaluate whether keeping a list of buffers and then decoding
// them in sequence using TextDecoder's "stream" option is more
// efficient. Can the data buffers be safely kept around for later
// use?
// - Informal, quick benchmarks seem to show most of the overhead is
// from calling TextDecoder.decode() and TextEncoder.encode(), and if
// confirmed, there is nothing which can be done uBO-side to reduce
// overhead.
if ( filterer.buffer === null ) {
filterer.buffer = new Uint8Array(ev.data);
return;
}
const buffer = new Uint8Array(
filterer.buffer.byteLength +
ev.data.byteLength
);
buffer.set(filterer.buffer);
buffer.set(new Uint8Array(ev.data), filterer.buffer.byteLength);
filterer.buffer = buffer;
};
const onStreamStop = function() {
const filterer = filterers.get(this);
filterers.delete(this);
if ( filterer === undefined || filterer.buffer === null ) {
this.close();
return;
}
if ( this.status !== 'finishedtransferringdata' ) { return; }
if ( domParser === undefined ) {
domParser = new DOMParser();
xmlSerializer = new XMLSerializer();
}
if ( textEncoder === undefined ) {
textEncoder = new TextEncoder();
}
let doc;
// If stream encoding is still unknnown, try to extract from document.
let charsetFound = filterer.charset,
2018-02-18 13:16:10 +01:00
charsetUsed = charsetFound;
if ( charsetFound === undefined ) {
if ( utf8TextDecoder === undefined ) {
utf8TextDecoder = new TextDecoder();
}
doc = domParser.parseFromString(
2018-02-18 13:16:10 +01:00
utf8TextDecoder.decode(filterer.buffer.slice(0, 1024)),
2018-03-01 20:12:16 +01:00
filterer.mime
);
2018-02-18 13:16:10 +01:00
charsetFound = charsetFromDoc(doc);
charsetUsed = textEncode.normalizeCharset(charsetFound);
2018-02-18 13:16:10 +01:00
if ( charsetUsed === undefined ) {
return streamClose(filterer);
}
}
doc = domParser.parseFromString(
2018-02-18 13:16:10 +01:00
textDecode(charsetUsed, filterer.buffer),
2018-03-01 20:12:16 +01:00
filterer.mime
);
2018-02-18 13:16:10 +01:00
// https://github.com/gorhill/uBlock/issues/3507
// In case of no explicit charset found, try to find one again, but
// this time with the whole document parsed.
if ( charsetFound === undefined ) {
charsetFound = textEncode.normalizeCharset(charsetFromDoc(doc));
2018-02-18 13:16:10 +01:00
if ( charsetFound !== charsetUsed ) {
if ( charsetFound === undefined ) {
return streamClose(filterer);
}
charsetUsed = charsetFound;
doc = domParser.parseFromString(
textDecode(charsetFound, filterer.buffer),
2018-03-01 20:12:16 +01:00
filterer.mime
2018-02-18 13:16:10 +01:00
);
}
}
let modified = false;
if ( filterer.selectors !== undefined ) {
if ( htmlFilteringEngine.apply(doc, filterer) ) {
modified = true;
}
}
if ( modified === false ) {
2018-02-18 13:16:10 +01:00
return streamClose(filterer);
}
// https://stackoverflow.com/questions/6088972/get-doctype-of-an-html-as-string-with-javascript/10162353#10162353
const doctypeStr = doc.doctype instanceof Object ?
xmlSerializer.serializeToString(doc.doctype) + '\n' :
'';
2018-01-03 05:06:16 +01:00
// https://github.com/gorhill/uBlock/issues/3391
let encodedStream = textEncoder.encode(
2018-01-03 05:06:16 +01:00
doctypeStr +
doc.documentElement.outerHTML
);
2018-02-18 13:16:10 +01:00
if ( charsetUsed !== 'utf-8' ) {
encodedStream = textEncode.encode(
2018-02-18 13:16:10 +01:00
charsetUsed,
2018-01-03 05:06:16 +01:00
encodedStream
);
}
streamClose(filterer, encodedStream);
};
const onStreamError = function() {
filterers.delete(this);
};
return function(fctxt, extras) {
// https://github.com/gorhill/uBlock/issues/3478
const statusCode = extras.statusCode || 0;
if ( statusCode !== 0 && (statusCode < 200 || statusCode >= 300) ) {
return;
}
const hostname = fctxt.getHostname();
if ( hostname === '' ) { return; }
const domain = fctxt.getDomain();
const request = {
stream: undefined,
tabId: fctxt.tabId,
url: fctxt.url,
hostname: hostname,
domain: domain,
entity: entityFromDomain(domain),
selectors: undefined,
buffer: null,
2018-04-03 00:40:29 +02:00
mime: 'text/html',
charset: undefined
};
request.selectors = htmlFilteringEngine.retrieve(request);
if ( request.selectors === undefined ) { return; }
const headers = extras.responseHeaders;
const contentType = headerValueFromName('content-type', headers);
if ( contentType !== '' ) {
2018-03-01 20:12:16 +01:00
request.mime = mimeFromContentType(contentType);
if ( request.mime === undefined ) { return; }
let charset = charsetFromContentType(contentType);
2018-01-05 00:26:52 +01:00
if ( charset !== undefined ) {
charset = textEncode.normalizeCharset(charset);
2018-01-03 19:59:38 +01:00
if ( charset === undefined ) { return; }
2018-01-05 00:26:52 +01:00
request.charset = charset;
}
}
// https://bugzilla.mozilla.org/show_bug.cgi?id=1426789
if ( headerValueFromName('content-disposition', headers) ) { return; }
const stream = request.stream =
browser.webRequest.filterResponseData(extras.requestId);
stream.ondata = onStreamData;
stream.onstop = onStreamStop;
stream.onerror = onStreamError;
filterers.set(stream, request);
return true;
};
})();
/******************************************************************************/
const injectCSP = function(fctxt, pageStore, responseHeaders) {
const cspSubsets = [];
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
const requestType = fctxt.type;
2015-01-24 18:06:22 +01:00
// Start collecting policies >>>>>>>>
// ======== built-in policies
const builtinDirectives = [];
2017-09-11 15:53:42 +02:00
if ( pageStore.filterScripting(fctxt, true) === 1 ) {
builtinDirectives.push(µb.cspNoScripting);
if ( logger.enabled ) {
fctxt.setRealm('network').setType('scripting').toLogger();
}
}
// https://github.com/uBlockOrigin/uBlock-issues/issues/422
// We need to derive a special context for filtering `inline-script`,
// as the embedding document for this "resource" will always be the
// frame itself, not that of the parent of the frame.
else {
const fctxt2 = fctxt.duplicate();
fctxt2.type = 'inline-script';
fctxt2.setDocOriginFromURL(fctxt.url);
const result = pageStore.filterRequest(fctxt2);
if ( result === 1 ) {
builtinDirectives.push(µb.cspNoInlineScript);
}
if ( result === 2 && logger.enabled ) {
fctxt2.setRealm('network').toLogger();
}
}
2017-09-11 15:53:42 +02:00
// https://github.com/gorhill/uBlock/issues/1539
// - Use a CSP to also forbid inline fonts if remote fonts are blocked.
fctxt.type = 'inline-font';
if ( pageStore.filterRequest(fctxt) === 1 ) {
builtinDirectives.push(µb.cspNoInlineFont);
if ( logger.enabled ) {
fctxt.setRealm('network').toLogger();
2017-09-11 15:53:42 +02:00
}
}
if ( builtinDirectives.length !== 0 ) {
cspSubsets[0] = builtinDirectives.join(', ');
2017-09-11 15:53:42 +02:00
}
// ======== filter-based policies
// Static filtering.
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
fctxt.type = requestType;
const staticDirectives =
staticNetFilteringEngine.matchAndFetchModifiers(fctxt, 'csp');
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
if ( staticDirectives !== undefined ) {
for ( const directive of staticDirectives ) {
if ( directive.result !== 1 ) { continue; }
cspSubsets.push(directive.modifier.value);
}
}
// URL filtering `allow` rules override static filtering.
if (
cspSubsets.length !== 0 &&
sessionURLFiltering.evaluateZ(
fctxt.getTabHostname(),
fctxt.url,
'csp'
) === 2
) {
if ( logger.enabled ) {
fctxt.setRealm('network')
.setType('csp')
.setFilter(sessionURLFiltering.toLogData())
.toLogger();
}
return;
}
// Dynamic filtering `allow` rules override static filtering.
if (
cspSubsets.length !== 0 &&
µb.userSettings.advancedUserEnabled &&
sessionFirewall.evaluateCellZY(
fctxt.getTabHostname(),
fctxt.getTabHostname(),
'*'
) === 2
) {
if ( logger.enabled ) {
fctxt.setRealm('network')
.setType('csp')
.setFilter(sessionFirewall.toLogData())
.toLogger();
2017-02-06 21:34:31 +01:00
}
return;
}
// <<<<<<<< All policies have been collected
// Static CSP policies will be applied.
if ( logger.enabled && staticDirectives !== undefined ) {
Add new filter option `queryprune=` Related issue: - https://github.com/uBlockOrigin/uBlock-issues/issues/760 The purpose of this new network filter option is to remove query parameters form the URL of network requests. The name `queryprune` has been picked over `querystrip` since the purpose of the option is to remove some parameters from the URL rather than all parameters. `queryprune` is a modifier option (like `csp`) in that it does not cause a network request to be blocked but rather modified before being emitted. `queryprune` must be assigned a value, which value will determine which parameters from a query string will be removed. The syntax for the value is that of regular expression *except* for the following rules: - do not wrap the regex directive between `/` - do not use regex special values `^` and `$` - do not use literal comma character in the value, though you can use hex-encoded version, `\x2c` - to match the start of a query parameter, prepend `|` - to match the end of a query parameter, append `|` `queryprune` regex-like values will be tested against each key-value parameter pair as `[key]=[value]` string. This way you can prune according to either the key, the value, or both. This commit introduces the concept of modifier filter options, which as of now are: - `csp=` - `queryprune=` They both work in similar way when used with `important` option or when used in exception filters. Modifier options can apply to any network requests, hence the logger reports the type of the network requests, and no longer use the modifier as the type, i.e. `csp` filters are no longer reported as requests of type `csp`. Though modifier options can apply to any network requests, for the time being the `csp=` modifier option still apply only to top or embedded (frame) documents, just as before. In some future we may want to apply `csp=` directives to network requests of type script, to control the behavior of service workers for example. A new built-in filter expression has been added to the logger: "modified", which allow to see all the network requests which were modified before being emitted. The translation work for this new option will be available in a future commit.
2020-10-31 15:42:53 +01:00
fctxt.setRealm('network')
.pushFilters(staticDirectives.map(a => a.logData()))
.toLogger();
}
2016-07-01 04:03:29 +02:00
if ( cspSubsets.length === 0 ) { return; }
µb.updateToolbarIcon(fctxt.tabId, 0x02);
2014-09-24 23:38:22 +02:00
// Use comma to merge CSP directives.
// Ref.: https://www.w3.org/TR/CSP2/#implementation-considerations
//
// https://github.com/gorhill/uMatrix/issues/967
// Inject a new CSP header rather than modify an existing one, except
// if the current environment does not support merging headers:
// Firefox 58/webext and less can't merge CSP headers, so we will merge
// them here.
responseHeaders.push({
name: 'Content-Security-Policy',
value: cspSubsets.join(', ')
});
return true;
2018-04-04 18:42:01 +02:00
};
/******************************************************************************/
// https://github.com/uBlockOrigin/uBlock-issues/issues/1553
// https://github.com/WICG/floc#opting-out-of-computation
const foilFloc = function(fctxt, responseHeaders) {
const hn = fctxt.getHostname();
if ( scriptletFilteringEngine.hasScriptlet(hn, 1, 'no-floc') === false ) {
return false;
}
responseHeaders.push({
name: 'Permissions-Policy',
value: 'interest-cohort=()' }
);
return true;
};
/******************************************************************************/
// https://github.com/gorhill/uBlock/issues/1163
// "Block elements by size".
// https://github.com/gorhill/uBlock/issues/1390#issuecomment-187310719
// Do not foil when the media element is fetched from the browser
// cache. This works only when the webext API supports the `fromCache`
// property (Firefox).
const foilLargeMediaElement = function(details, fctxt, pageStore) {
if ( details.fromCache === true ) { return; }
let size = 0;
if ( µb.userSettings.largeMediaSize !== 0 ) {
const headers = details.responseHeaders;
const i = headerIndexFromName('content-length', headers);
if ( i === -1 ) { return; }
size = parseInt(headers[i].value, 10) || 0;
}
2016-11-08 21:53:08 +01:00
const result = pageStore.filterLargeMediaElement(fctxt, size);
if ( result === 0 ) { return; }
if ( logger.enabled ) {
fctxt.setRealm('network').toLogger();
}
return { cancel: true };
};
/******************************************************************************/
// Caller must ensure headerName is normalized to lower case.
const headerIndexFromName = function(headerName, headers) {
let i = headers.length;
while ( i-- ) {
if ( headers[i].name.toLowerCase() === headerName ) {
return i;
}
}
return -1;
2014-09-24 23:38:22 +02:00
};
const headerValueFromName = function(headerName, headers) {
const i = headerIndexFromName(headerName, headers);
2017-07-22 22:58:08 +02:00
return i !== -1 ? headers[i].value : '';
};
2014-09-24 23:38:22 +02:00
/******************************************************************************/
const strictBlockBypasser = {
hostnameToDeadlineMap: new Map(),
cleanupTimer: undefined,
cleanup: function() {
for ( const [ hostname, deadline ] of this.hostnameToDeadlineMap ) {
if ( deadline <= Date.now() ) {
this.hostnameToDeadlineMap.delete(hostname);
}
}
},
bypass: function(hostname) {
if ( typeof hostname !== 'string' || hostname === '' ) { return; }
this.hostnameToDeadlineMap.set(
hostname,
Date.now() + µb.hiddenSettings.strictBlockingBypassDuration * 1000
);
},
isBypassed: function(hostname) {
if ( this.hostnameToDeadlineMap.size === 0 ) { return false; }
let bypassDuration =
µb.hiddenSettings.strictBlockingBypassDuration * 1000;
if ( this.cleanupTimer === undefined ) {
this.cleanupTimer = vAPI.setTimeout(
( ) => {
this.cleanupTimer = undefined;
this.cleanup();
},
bypassDuration + 10000
);
}
for (;;) {
const deadline = this.hostnameToDeadlineMap.get(hostname);
if ( deadline !== undefined ) {
if ( deadline > Date.now() ) {
this.hostnameToDeadlineMap.set(
hostname,
Date.now() + bypassDuration
);
return true;
}
this.hostnameToDeadlineMap.delete(hostname);
}
const pos = hostname.indexOf('.');
if ( pos === -1 ) { break; }
hostname = hostname.slice(pos + 1);
}
return false;
}
};
/******************************************************************************/
const webRequest = {
start: (( ) => {
vAPI.net = new vAPI.Net();
vAPI.net.suspend();
Add support for `1P`, `3P`, `header=` filter options and other changes New filter options ================== Strict partyness: `1P`, `3P` ---------------------------- The current options 1p/3p are meant to "weakly" match partyness, i.e. a network request is considered 1st-party to its context as long as both the context and the request share the same base domain. The new partyness options are meant to check for strict partyness, i.e. a network request will be considered 1st-party if and only if both the context and the request share the same hostname. For examples: - context: `www.example.org` - request: `www.example.org` - `1p`: yes, `1P`: yes - `3p`: no, `3P`: no - context: `www.example.org` - request: `subdomain.example.org` - `1p`: yes, `1P`: no - `3p`: no, `3P`: yes - context: `www.example.org` - request: `www.example.com` - `1p`: no, `1P`: no - `3p`: yes, `3P`: yes The strict partyness options will be visually emphasized in the editor so as to prevent mistakenly using `1P` or `3P` where weak partyness is meant to be used. Filter on response headers: `header=` ------------------------------------- Currently experimental and under evaluation. Disabled by default, enable by toggling `filterOnHeaders` to `true` in advanced settings. Ability to filter network requests according to whether a specific response header is present and whether it matches or does not match a specific value. For example: *$1p,3P,script,header=via:1\.1\s+google The above filter is meant to block network requests which fullfill all the following conditions: - is weakly 1st-party to the context - is not strictly 1st-party to the context - is of type `script` - has a response HTTP header named `via`, which value matches the regular expression `1\.1\s+google`. The matches are always performed in a case-insensitive manner. The header value is assumed to be a literal regular expression, except for the following special characters: - to anchor to start of string, use leading `|`, not `^` - to anchor to end of string, use trailing `|`, not `$` - to invert the test, use a leading `!` To block a network request if it merely contains a specific HTTP header is just a matter of specifying the header name without a header value: *$1p,3P,script,header=via Generic exception filters can be used to disable specific block `header=` filters, i.e. `@@*$1p,3P,script,header` will override the block `header=` filters given as example above. Dynamic filtering's `allow` rules override block `headers=` filters. Important: It is key that filter authors use as many narrowing filter options as possible when using the `header=` option, and the `header=` option should be used ONLY when other filter options are not sufficient. More documentation justifying the purpose of `header=` option will be provided eventually if ever it is decided to move it from experimental to stable status. To be decided: to restrict usage of this filter option to only uBO's own filter lists or "My filters". Changes ======= Fine tuning `queryprune=` ------------------------- The following changes have been implemented: The special value `*` (i.e. `queryprune=*`) means "remove all query parameters". If the `queryprune=` value is made only of alphanumeric characters (including `_`), the value will be internally converted to regex equivalent `^value=`. This ensures a better future compatibility with AdGuard's `removeparam=`. If the `queryprune=` value starts with `!`, the test will be inverted. This can be used to remove all query parameters EXCEPT those who match the specified value. Other ----- The legacy code to test for spurious CSP reports has been removed. This is no longer an issue ever since uBO redirects to local resources through web accessible resources. Notes ===== The following new and recently added filter options are not compatible with Chromium's manifest v3 changes: - `queryprune=` - `1P` - `3P` - `header=`
2020-11-23 14:22:43 +01:00
return ( ) => {
vAPI.net.setSuspendableListener(onBeforeRequest);
vAPI.net.addListener(
'onHeadersReceived',
onHeadersReceived,
Add support for `1P`, `3P`, `header=` filter options and other changes New filter options ================== Strict partyness: `1P`, `3P` ---------------------------- The current options 1p/3p are meant to "weakly" match partyness, i.e. a network request is considered 1st-party to its context as long as both the context and the request share the same base domain. The new partyness options are meant to check for strict partyness, i.e. a network request will be considered 1st-party if and only if both the context and the request share the same hostname. For examples: - context: `www.example.org` - request: `www.example.org` - `1p`: yes, `1P`: yes - `3p`: no, `3P`: no - context: `www.example.org` - request: `subdomain.example.org` - `1p`: yes, `1P`: no - `3p`: no, `3P`: yes - context: `www.example.org` - request: `www.example.com` - `1p`: no, `1P`: no - `3p`: yes, `3P`: yes The strict partyness options will be visually emphasized in the editor so as to prevent mistakenly using `1P` or `3P` where weak partyness is meant to be used. Filter on response headers: `header=` ------------------------------------- Currently experimental and under evaluation. Disabled by default, enable by toggling `filterOnHeaders` to `true` in advanced settings. Ability to filter network requests according to whether a specific response header is present and whether it matches or does not match a specific value. For example: *$1p,3P,script,header=via:1\.1\s+google The above filter is meant to block network requests which fullfill all the following conditions: - is weakly 1st-party to the context - is not strictly 1st-party to the context - is of type `script` - has a response HTTP header named `via`, which value matches the regular expression `1\.1\s+google`. The matches are always performed in a case-insensitive manner. The header value is assumed to be a literal regular expression, except for the following special characters: - to anchor to start of string, use leading `|`, not `^` - to anchor to end of string, use trailing `|`, not `$` - to invert the test, use a leading `!` To block a network request if it merely contains a specific HTTP header is just a matter of specifying the header name without a header value: *$1p,3P,script,header=via Generic exception filters can be used to disable specific block `header=` filters, i.e. `@@*$1p,3P,script,header` will override the block `header=` filters given as example above. Dynamic filtering's `allow` rules override block `headers=` filters. Important: It is key that filter authors use as many narrowing filter options as possible when using the `header=` option, and the `header=` option should be used ONLY when other filter options are not sufficient. More documentation justifying the purpose of `header=` option will be provided eventually if ever it is decided to move it from experimental to stable status. To be decided: to restrict usage of this filter option to only uBO's own filter lists or "My filters". Changes ======= Fine tuning `queryprune=` ------------------------- The following changes have been implemented: The special value `*` (i.e. `queryprune=*`) means "remove all query parameters". If the `queryprune=` value is made only of alphanumeric characters (including `_`), the value will be internally converted to regex equivalent `^value=`. This ensures a better future compatibility with AdGuard's `removeparam=`. If the `queryprune=` value starts with `!`, the test will be inverted. This can be used to remove all query parameters EXCEPT those who match the specified value. Other ----- The legacy code to test for spurious CSP reports has been removed. This is no longer an issue ever since uBO redirects to local resources through web accessible resources. Notes ===== The following new and recently added filter options are not compatible with Chromium's manifest v3 changes: - `queryprune=` - `1P` - `3P` - `header=`
2020-11-23 14:22:43 +01:00
{ urls: [ 'http://*/*', 'https://*/*' ] },
[ 'blocking', 'responseHeaders' ]
);
vAPI.net.unsuspend(true);
};
})(),
2015-03-26 00:28:22 +01:00
Add support for `1P`, `3P`, `header=` filter options and other changes New filter options ================== Strict partyness: `1P`, `3P` ---------------------------- The current options 1p/3p are meant to "weakly" match partyness, i.e. a network request is considered 1st-party to its context as long as both the context and the request share the same base domain. The new partyness options are meant to check for strict partyness, i.e. a network request will be considered 1st-party if and only if both the context and the request share the same hostname. For examples: - context: `www.example.org` - request: `www.example.org` - `1p`: yes, `1P`: yes - `3p`: no, `3P`: no - context: `www.example.org` - request: `subdomain.example.org` - `1p`: yes, `1P`: no - `3p`: no, `3P`: yes - context: `www.example.org` - request: `www.example.com` - `1p`: no, `1P`: no - `3p`: yes, `3P`: yes The strict partyness options will be visually emphasized in the editor so as to prevent mistakenly using `1P` or `3P` where weak partyness is meant to be used. Filter on response headers: `header=` ------------------------------------- Currently experimental and under evaluation. Disabled by default, enable by toggling `filterOnHeaders` to `true` in advanced settings. Ability to filter network requests according to whether a specific response header is present and whether it matches or does not match a specific value. For example: *$1p,3P,script,header=via:1\.1\s+google The above filter is meant to block network requests which fullfill all the following conditions: - is weakly 1st-party to the context - is not strictly 1st-party to the context - is of type `script` - has a response HTTP header named `via`, which value matches the regular expression `1\.1\s+google`. The matches are always performed in a case-insensitive manner. The header value is assumed to be a literal regular expression, except for the following special characters: - to anchor to start of string, use leading `|`, not `^` - to anchor to end of string, use trailing `|`, not `$` - to invert the test, use a leading `!` To block a network request if it merely contains a specific HTTP header is just a matter of specifying the header name without a header value: *$1p,3P,script,header=via Generic exception filters can be used to disable specific block `header=` filters, i.e. `@@*$1p,3P,script,header` will override the block `header=` filters given as example above. Dynamic filtering's `allow` rules override block `headers=` filters. Important: It is key that filter authors use as many narrowing filter options as possible when using the `header=` option, and the `header=` option should be used ONLY when other filter options are not sufficient. More documentation justifying the purpose of `header=` option will be provided eventually if ever it is decided to move it from experimental to stable status. To be decided: to restrict usage of this filter option to only uBO's own filter lists or "My filters". Changes ======= Fine tuning `queryprune=` ------------------------- The following changes have been implemented: The special value `*` (i.e. `queryprune=*`) means "remove all query parameters". If the `queryprune=` value is made only of alphanumeric characters (including `_`), the value will be internally converted to regex equivalent `^value=`. This ensures a better future compatibility with AdGuard's `removeparam=`. If the `queryprune=` value starts with `!`, the test will be inverted. This can be used to remove all query parameters EXCEPT those who match the specified value. Other ----- The legacy code to test for spurious CSP reports has been removed. This is no longer an issue ever since uBO redirects to local resources through web accessible resources. Notes ===== The following new and recently added filter options are not compatible with Chromium's manifest v3 changes: - `queryprune=` - `1P` - `3P` - `header=`
2020-11-23 14:22:43 +01:00
strictBlockBypass: hostname => {
strictBlockBypasser.bypass(hostname);
Add support for `1P`, `3P`, `header=` filter options and other changes New filter options ================== Strict partyness: `1P`, `3P` ---------------------------- The current options 1p/3p are meant to "weakly" match partyness, i.e. a network request is considered 1st-party to its context as long as both the context and the request share the same base domain. The new partyness options are meant to check for strict partyness, i.e. a network request will be considered 1st-party if and only if both the context and the request share the same hostname. For examples: - context: `www.example.org` - request: `www.example.org` - `1p`: yes, `1P`: yes - `3p`: no, `3P`: no - context: `www.example.org` - request: `subdomain.example.org` - `1p`: yes, `1P`: no - `3p`: no, `3P`: yes - context: `www.example.org` - request: `www.example.com` - `1p`: no, `1P`: no - `3p`: yes, `3P`: yes The strict partyness options will be visually emphasized in the editor so as to prevent mistakenly using `1P` or `3P` where weak partyness is meant to be used. Filter on response headers: `header=` ------------------------------------- Currently experimental and under evaluation. Disabled by default, enable by toggling `filterOnHeaders` to `true` in advanced settings. Ability to filter network requests according to whether a specific response header is present and whether it matches or does not match a specific value. For example: *$1p,3P,script,header=via:1\.1\s+google The above filter is meant to block network requests which fullfill all the following conditions: - is weakly 1st-party to the context - is not strictly 1st-party to the context - is of type `script` - has a response HTTP header named `via`, which value matches the regular expression `1\.1\s+google`. The matches are always performed in a case-insensitive manner. The header value is assumed to be a literal regular expression, except for the following special characters: - to anchor to start of string, use leading `|`, not `^` - to anchor to end of string, use trailing `|`, not `$` - to invert the test, use a leading `!` To block a network request if it merely contains a specific HTTP header is just a matter of specifying the header name without a header value: *$1p,3P,script,header=via Generic exception filters can be used to disable specific block `header=` filters, i.e. `@@*$1p,3P,script,header` will override the block `header=` filters given as example above. Dynamic filtering's `allow` rules override block `headers=` filters. Important: It is key that filter authors use as many narrowing filter options as possible when using the `header=` option, and the `header=` option should be used ONLY when other filter options are not sufficient. More documentation justifying the purpose of `header=` option will be provided eventually if ever it is decided to move it from experimental to stable status. To be decided: to restrict usage of this filter option to only uBO's own filter lists or "My filters". Changes ======= Fine tuning `queryprune=` ------------------------- The following changes have been implemented: The special value `*` (i.e. `queryprune=*`) means "remove all query parameters". If the `queryprune=` value is made only of alphanumeric characters (including `_`), the value will be internally converted to regex equivalent `^value=`. This ensures a better future compatibility with AdGuard's `removeparam=`. If the `queryprune=` value starts with `!`, the test will be inverted. This can be used to remove all query parameters EXCEPT those who match the specified value. Other ----- The legacy code to test for spurious CSP reports has been removed. This is no longer an issue ever since uBO redirects to local resources through web accessible resources. Notes ===== The following new and recently added filter options are not compatible with Chromium's manifest v3 changes: - `queryprune=` - `1P` - `3P` - `header=`
2020-11-23 14:22:43 +01:00
},
2015-03-26 00:28:22 +01:00
};
/******************************************************************************/
export { webRequest };
/******************************************************************************/