@adguard/agtree 2.1.4 → 2.3.0

package/dist/agtree.d.ts

@@ -1,5 +1,5 @@
 /*
- * AGTree v2.1.4 (build date: Mon, 25 Nov 2024 16:56:07 GMT)
+ * AGTree v2.3.0 (build date: Thu, 19 Dec 2024 15:31:22 GMT)
  * (c) 2024 Adguard Software Ltd.
  * Released under the MIT license
  * https://github.com/AdguardTeam/tsurlfilter/tree/master/packages/agtree#readme
@@ -3699,6 +3699,7 @@ declare const modifierDataSchema: zod.ZodEffects<zod.ZodTypeAny, {
     exceptionOnly: boolean;
     valueOptional: boolean;
     valueFormat: string | null;
+    valueFormatFlags: string | null;
 }, any>;
 /**
  * Type of the modifier data schema.

package/dist/agtree.js

@@ -1,5 +1,5 @@
 /*
- * AGTree v2.1.4 (build date: Mon, 25 Nov 2024 16:56:07 GMT)
+ * AGTree v2.3.0 (build date: Thu, 19 Dec 2024 15:31:22 GMT)
  * (c) 2024 Adguard Software Ltd.
  * Released under the MIT license
  * https://github.com/AdguardTeam/tsurlfilter/tree/master/packages/agtree#readme
@@ -13,7 +13,6 @@ var sprintfJs = require('sprintf-js');
 var cssTokenizer = require('@adguard/css-tokenizer');
 var tldts = require('tldts');
 var isIp = require('is-ip');
-var XRegExp = require('xregexp');
 var cloneDeep = require('clone-deep');
 var compatibilityTableData_js = require('./compatibility-table-data.js');
 var zod = require('zod');
@@ -312,6 +311,8 @@ const COMMA_DOMAIN_LIST_SEPARATOR = ',';
 const PIPE_MODIFIER_SEPARATOR = '|';
 const CSS_MEDIA_MARKER = '@media';
 const CSS_NOT_PSEUDO = 'not';
+const CSS_BLOCK_OPEN = '{';
+const CSS_BLOCK_CLOSE = '}';
 const HINT_MARKER = '!+';
 const HINT_MARKER_LEN = HINT_MARKER.length;
 const NETWORK_RULE_EXCEPTION_MARKER = '@@';
@@ -6086,14 +6087,24 @@ class CssTokenStream {
         // - end: end index of the token
         // - props: additional properties of the token, if any (we don't use it here, this is why we use underscore)
         // - balance: balance level of the token
-        tokenizeBalanced(source, (type, start, end, _, balance) => {
-            this.tokens.push({
-                type,
-                start,
-                end,
-                balance,
+        try {
+            tokenizeBalanced(source, (type, start, end, _, balance) => {
+                this.tokens.push({
+                    type,
+                    start,
+                    end,
+                    balance,
+                });
             });
-        });
+        }
+        catch (error) {
+            // If the error is an AdblockSyntaxError, adjust the error positions to the base offset
+            if (error instanceof AdblockSyntaxError) {
+                error.start += baseOffset;
+                error.end += baseOffset;
+                throw error;
+            }
+        }
         this.index = 0;
         this.baseOffset = baseOffset;
     }
@@ -7653,6 +7664,40 @@ class AdgScriptletInjectionBodyParser extends ParserBase {
     }
 }
 
+/**
+ * Represents an error that occurs when an operation is aborted.
+ */
+class AbortError extends Error {
+    constructor() {
+        super('Aborted');
+    }
+}
+// TODO: AG-38480 add a stop function to the tokenizers callback and move `hasToken` to CSS Tokenizer as well
+/**
+ * Checks if the given raw string contains any of the specified tokens.
+ * This function uses error throwing inside the abort tokenization process.
+ *
+ * @param raw - The raw string to be tokenized and checked.
+ * @param tokens - A set of token types to check for in the raw string.
+ * @returns `true` if any of the specified tokens are found in the raw string, otherwise `false`.
+ */
+const hasToken = (raw, tokens) => {
+    try {
+        cssTokenizer.tokenizeExtended(raw, (type) => {
+            if (tokens.has(type)) {
+                throw new AbortError();
+            }
+        });
+    }
+    catch (e) {
+        if (e instanceof AbortError) {
+            return true;
+        }
+        throw e;
+    }
+    return false;
+};
+
 /* eslint-disable no-param-reassign */
 /**
  * Value map for binary serialization. This helps to reduce the size of the serialized data,
@@ -8025,6 +8070,33 @@ class CosmeticRuleParser extends ParserBase {
                 body: AdgCssInjectionParser.parse(rawBody, options, baseOffset + bodyStart),
             };
         };
+        /**
+         * Parses Adb CSS injection rules
+         * eg: example.com##.foo { display: none; }
+         *
+         * @returns parsed rule
+         */
+        const parseAbpCssInjection = () => {
+            if (!options.parseAbpSpecificRules) {
+                return null;
+            }
+            // check if the rule contains both CSS block open and close characters
+            // if none of them is present we can stop parsing
+            if (rawBody.indexOf(CSS_BLOCK_OPEN) === -1 && rawBody.indexOf(CSS_BLOCK_CLOSE) === -1) {
+                return null;
+            }
+            if (!hasToken(rawBody, new Set([cssTokenizer.TokenType.OpenCurlyBracket, cssTokenizer.TokenType.CloseCurlyBracket]))) {
+                return null;
+            }
+            // try to parse the raw body as an AdGuard CSS injection rule
+            const body = AdgCssInjectionParser.parse(rawBody, options, baseOffset + bodyStart);
+            // if the parsed rule type is a 'CssInjectionRuleBody', return the parsed rule
+            return {
+                syntax: exports.AdblockSyntax.Abp,
+                type: exports.CosmeticRuleType.CssInjectionRule,
+                body,
+            };
+        };
         const parseAbpSnippetInjection = () => {
             if (!options.parseAbpSpecificRules) {
                 throw new AdblockSyntaxError(sprintfJs.sprintf(ERROR_MESSAGES$3.SYNTAX_DISABLED, exports.AdblockSyntax.Abp), baseOffset + bodyStart, baseOffset + bodyEnd);
@@ -8136,10 +8208,22 @@ class CosmeticRuleParser extends ParserBase {
         // the next function is called, and so on.
         // If all functions return null, an error should be thrown.
         const separatorMap = {
-            '##': [parseUboHtmlFiltering, parseUboScriptletInjection, parseUboCssInjection, parseElementHiding],
-            '#@#': [parseUboHtmlFiltering, parseUboScriptletInjection, parseUboCssInjection, parseElementHiding],
-            '#?#': [parseUboCssInjection, parseElementHiding],
-            '#@?#': [parseUboCssInjection, parseElementHiding],
+            '##': [
+                parseUboHtmlFiltering,
+                parseUboScriptletInjection,
+                parseUboCssInjection,
+                parseAbpCssInjection,
+                parseElementHiding,
+            ],
+            '#@#': [
+                parseUboHtmlFiltering,
+                parseUboScriptletInjection,
+                parseUboCssInjection,
+                parseAbpCssInjection,
+                parseElementHiding,
+            ],
+            '#?#': [parseUboCssInjection, parseAbpCssInjection, parseElementHiding],
+            '#@?#': [parseUboCssInjection, parseAbpCssInjection, parseElementHiding],
             '#$#': [parseAdgCssInjection, parseAbpSnippetInjection],
             '#@$#': [parseAdgCssInjection, parseAbpSnippetInjection],
             '#$?#': [parseAdgCssInjection],
@@ -8208,7 +8292,7 @@ class CosmeticRuleParser extends ParserBase {
                 result = node.body.selectorList.value;
                 break;
             case exports.CosmeticRuleType.CssInjectionRule:
-                if (node.syntax === exports.AdblockSyntax.Adg) {
+                if (node.syntax === exports.AdblockSyntax.Adg || node.syntax === exports.AdblockSyntax.Abp) {
                     result = AdgCssInjectionParser.generate(node.body);
                 }
                 else if (node.syntax === exports.AdblockSyntax.Ubo) {
@@ -10764,10 +10848,11 @@ const isCustomValueFormatValidator = (valueFormat) => {
  *
  * @param modifier Modifier AST node.
  * @param valueFormat Value format for the modifier.
+ * @param valueFormatFlags Optional; RegExp flags for the value format.
  *
  * @returns Validation result.
  */
-const validateValue = (modifier, valueFormat) => {
+const validateValue = (modifier, valueFormat, valueFormatFlags) => {
     if (isCustomValueFormatValidator(valueFormat)) {
         const validator = CUSTOM_VALUE_FORMAT_MAP[valueFormat];
         return validator(modifier);
@@ -10776,14 +10861,19 @@ const validateValue = (modifier, valueFormat) => {
     if (!modifier.value?.value) {
         return getValueRequiredValidationResult(modifierName);
     }
-    let xRegExp;
+    let regExp;
     try {
-        xRegExp = XRegExp(valueFormat);
+        if (isString(valueFormatFlags)) {
+            regExp = new RegExp(valueFormat, valueFormatFlags);
+        }
+        else {
+            regExp = new RegExp(valueFormat);
+        }
     }
     catch (e) {
         throw new Error(`${SOURCE_DATA_ERROR_PREFIX.INVALID_VALUE_FORMAT_REGEXP}: '${modifierName}'`);
     }
-    const isValid = xRegExp.test(modifier.value?.value);
+    const isValid = regExp.test(modifier.value?.value);
     if (!isValid) {
         return getInvalidValidationResult(`${VALIDATION_ERROR_PREFIX.VALUE_INVALID}: '${modifierName}'`);
     }
@@ -11367,7 +11457,7 @@ const validateForSpecificSyntax = (syntax, modifier, isException) => {
         if (!specificBlockerData.valueFormat) {
             throw new Error(`${SOURCE_DATA_ERROR_PREFIX.NO_VALUE_FORMAT_FOR_ASSIGNABLE}: '${modifierName}'`);
         }
-        return validateValue(modifier, specificBlockerData.valueFormat);
+        return validateValue(modifier, specificBlockerData.valueFormat, specificBlockerData.valueFormatFlags);
     }
     if (modifier?.value) {
         // e.g. 'third-party=true'
@@ -13064,8 +13154,14 @@ function getErrorMessage(error) {
  * Known validators that don't need to be validated as regex.
  */
 const KNOWN_VALIDATORS = new Set([
+    'csp_value',
     'domain',
+    'permissions_value',
+    'pipe_separated_apps',
+    'pipe_separated_denyallow_domains',
     'pipe_separated_domains',
+    'pipe_separated_methods',
+    'pipe_separated_stealth_options',
     'regexp',
     'url',
 ]);
@@ -13116,6 +13212,10 @@ zodToCamelCase(baseCompatibilityDataSchema.extend({
      * Its value can be a regex pattern or a known validator name (e.g. `domain`, `pipe_separated_domains`, etc.).
      */
     value_format: nonEmptyStringSchema.nullable().default(null),
+    /**
+     * Describes the flags for the `value_format` regex pattern.
+     */
+    value_format_flags: nonEmptyStringSchema.nullable().default(null),
 }).superRefine((data, ctx) => {
     // TODO: find something better, for now we can't add refine logic to the base schema:
     // https://github.com/colinhacks/zod/issues/454#issuecomment-848370721
@@ -13136,11 +13236,17 @@ zodToCamelCase(baseCompatibilityDataSchema.extend({
         const valueFormat = data.value_format.trim();
         // if it is a known validator, we don't need to validate it further
         if (KNOWN_VALIDATORS.has(valueFormat)) {
+            if (data.value_format_flags) {
+                ctx.addIssue({
+                    code: zod.ZodIssueCode.custom,
+                    message: 'value_format_flags are not allowed for known validators',
+                });
+            }
             return;
         }
         // otherwise, we need to validate it as a regex
         try {
-            XRegExp(valueFormat);
+            new RegExp(valueFormat, data.value_format_flags ?? EMPTY);
         }
         catch (error) {
             ctx.addIssue({
@@ -13149,6 +13255,12 @@ zodToCamelCase(baseCompatibilityDataSchema.extend({
             });
         }
     }
+    else if (data.value_format_flags) {
+        ctx.addIssue({
+            code: zod.ZodIssueCode.custom,
+            message: 'value_format is required for value_format_flags',
+        });
+    }
 }));
 
 /**
@@ -15699,7 +15811,7 @@ class RuleCategorizer {
     }
 }
 
-const version = "2.1.4";
+const version = "2.3.0";
 
 /**
  * @file AGTree version

package/dist/agtree.mjs

@@ -1,5 +1,5 @@
 /*
- * AGTree v2.1.4 (build date: Mon, 25 Nov 2024 16:56:07 GMT)
+ * AGTree v2.3.0 (build date: Thu, 19 Dec 2024 15:31:22 GMT)
  * (c) 2024 Adguard Software Ltd.
  * Released under the MIT license
  * https://github.com/AdguardTeam/tsurlfilter/tree/master/packages/agtree#readme
@@ -12,7 +12,6 @@ import { TokenType as TokenType$1, tokenizeExtended, getFormattedTokenName } fro
 import * as tldts from 'tldts';
 import { parse } from 'tldts';
 import isIp from 'is-ip';
-import XRegExp from 'xregexp';
 import cloneDeep from 'clone-deep';
 import { modifiersCompatibilityTableData, redirectsCompatibilityTableData, scriptletsCompatibilityTableData } from './compatibility-table-data.js';
 import zod from 'zod';
@@ -292,6 +291,8 @@ const COMMA_DOMAIN_LIST_SEPARATOR = ',';
 const PIPE_MODIFIER_SEPARATOR = '|';
 const CSS_MEDIA_MARKER = '@media';
 const CSS_NOT_PSEUDO = 'not';
+const CSS_BLOCK_OPEN = '{';
+const CSS_BLOCK_CLOSE = '}';
 const HINT_MARKER = '!+';
 const HINT_MARKER_LEN = HINT_MARKER.length;
 const NETWORK_RULE_EXCEPTION_MARKER = '@@';
@@ -6066,14 +6067,24 @@ class CssTokenStream {
         // - end: end index of the token
         // - props: additional properties of the token, if any (we don't use it here, this is why we use underscore)
         // - balance: balance level of the token
-        tokenizeBalanced(source, (type, start, end, _, balance) => {
-            this.tokens.push({
-                type,
-                start,
-                end,
-                balance,
+        try {
+            tokenizeBalanced(source, (type, start, end, _, balance) => {
+                this.tokens.push({
+                    type,
+                    start,
+                    end,
+                    balance,
+                });
             });
-        });
+        }
+        catch (error) {
+            // If the error is an AdblockSyntaxError, adjust the error positions to the base offset
+            if (error instanceof AdblockSyntaxError) {
+                error.start += baseOffset;
+                error.end += baseOffset;
+                throw error;
+            }
+        }
         this.index = 0;
         this.baseOffset = baseOffset;
     }
@@ -7633,6 +7644,40 @@ class AdgScriptletInjectionBodyParser extends ParserBase {
     }
 }
 
+/**
+ * Represents an error that occurs when an operation is aborted.
+ */
+class AbortError extends Error {
+    constructor() {
+        super('Aborted');
+    }
+}
+// TODO: AG-38480 add a stop function to the tokenizers callback and move `hasToken` to CSS Tokenizer as well
+/**
+ * Checks if the given raw string contains any of the specified tokens.
+ * This function uses error throwing inside the abort tokenization process.
+ *
+ * @param raw - The raw string to be tokenized and checked.
+ * @param tokens - A set of token types to check for in the raw string.
+ * @returns `true` if any of the specified tokens are found in the raw string, otherwise `false`.
+ */
+const hasToken = (raw, tokens) => {
+    try {
+        tokenizeExtended(raw, (type) => {
+            if (tokens.has(type)) {
+                throw new AbortError();
+            }
+        });
+    }
+    catch (e) {
+        if (e instanceof AbortError) {
+            return true;
+        }
+        throw e;
+    }
+    return false;
+};
+
 /* eslint-disable no-param-reassign */
 /**
  * Value map for binary serialization. This helps to reduce the size of the serialized data,
@@ -8005,6 +8050,33 @@ class CosmeticRuleParser extends ParserBase {
                 body: AdgCssInjectionParser.parse(rawBody, options, baseOffset + bodyStart),
             };
         };
+        /**
+         * Parses Adb CSS injection rules
+         * eg: example.com##.foo { display: none; }
+         *
+         * @returns parsed rule
+         */
+        const parseAbpCssInjection = () => {
+            if (!options.parseAbpSpecificRules) {
+                return null;
+            }
+            // check if the rule contains both CSS block open and close characters
+            // if none of them is present we can stop parsing
+            if (rawBody.indexOf(CSS_BLOCK_OPEN) === -1 && rawBody.indexOf(CSS_BLOCK_CLOSE) === -1) {
+                return null;
+            }
+            if (!hasToken(rawBody, new Set([TokenType$1.OpenCurlyBracket, TokenType$1.CloseCurlyBracket]))) {
+                return null;
+            }
+            // try to parse the raw body as an AdGuard CSS injection rule
+            const body = AdgCssInjectionParser.parse(rawBody, options, baseOffset + bodyStart);
+            // if the parsed rule type is a 'CssInjectionRuleBody', return the parsed rule
+            return {
+                syntax: AdblockSyntax.Abp,
+                type: CosmeticRuleType.CssInjectionRule,
+                body,
+            };
+        };
         const parseAbpSnippetInjection = () => {
             if (!options.parseAbpSpecificRules) {
                 throw new AdblockSyntaxError(sprintf(ERROR_MESSAGES$3.SYNTAX_DISABLED, AdblockSyntax.Abp), baseOffset + bodyStart, baseOffset + bodyEnd);
@@ -8116,10 +8188,22 @@ class CosmeticRuleParser extends ParserBase {
         // the next function is called, and so on.
         // If all functions return null, an error should be thrown.
         const separatorMap = {
-            '##': [parseUboHtmlFiltering, parseUboScriptletInjection, parseUboCssInjection, parseElementHiding],
-            '#@#': [parseUboHtmlFiltering, parseUboScriptletInjection, parseUboCssInjection, parseElementHiding],
-            '#?#': [parseUboCssInjection, parseElementHiding],
-            '#@?#': [parseUboCssInjection, parseElementHiding],
+            '##': [
+                parseUboHtmlFiltering,
+                parseUboScriptletInjection,
+                parseUboCssInjection,
+                parseAbpCssInjection,
+                parseElementHiding,
+            ],
+            '#@#': [
+                parseUboHtmlFiltering,
+                parseUboScriptletInjection,
+                parseUboCssInjection,
+                parseAbpCssInjection,
+                parseElementHiding,
+            ],
+            '#?#': [parseUboCssInjection, parseAbpCssInjection, parseElementHiding],
+            '#@?#': [parseUboCssInjection, parseAbpCssInjection, parseElementHiding],
             '#$#': [parseAdgCssInjection, parseAbpSnippetInjection],
             '#@$#': [parseAdgCssInjection, parseAbpSnippetInjection],
             '#$?#': [parseAdgCssInjection],
@@ -8188,7 +8272,7 @@ class CosmeticRuleParser extends ParserBase {
                 result = node.body.selectorList.value;
                 break;
             case CosmeticRuleType.CssInjectionRule:
-                if (node.syntax === AdblockSyntax.Adg) {
+                if (node.syntax === AdblockSyntax.Adg || node.syntax === AdblockSyntax.Abp) {
                     result = AdgCssInjectionParser.generate(node.body);
                 }
                 else if (node.syntax === AdblockSyntax.Ubo) {
@@ -10744,10 +10828,11 @@ const isCustomValueFormatValidator = (valueFormat) => {
  *
  * @param modifier Modifier AST node.
  * @param valueFormat Value format for the modifier.
+ * @param valueFormatFlags Optional; RegExp flags for the value format.
  *
  * @returns Validation result.
  */
-const validateValue = (modifier, valueFormat) => {
+const validateValue = (modifier, valueFormat, valueFormatFlags) => {
     if (isCustomValueFormatValidator(valueFormat)) {
         const validator = CUSTOM_VALUE_FORMAT_MAP[valueFormat];
         return validator(modifier);
@@ -10756,14 +10841,19 @@ const validateValue = (modifier, valueFormat) => {
     if (!modifier.value?.value) {
         return getValueRequiredValidationResult(modifierName);
     }
-    let xRegExp;
+    let regExp;
     try {
-        xRegExp = XRegExp(valueFormat);
+        if (isString(valueFormatFlags)) {
+            regExp = new RegExp(valueFormat, valueFormatFlags);
+        }
+        else {
+            regExp = new RegExp(valueFormat);
+        }
     }
     catch (e) {
         throw new Error(`${SOURCE_DATA_ERROR_PREFIX.INVALID_VALUE_FORMAT_REGEXP}: '${modifierName}'`);
     }
-    const isValid = xRegExp.test(modifier.value?.value);
+    const isValid = regExp.test(modifier.value?.value);
     if (!isValid) {
         return getInvalidValidationResult(`${VALIDATION_ERROR_PREFIX.VALUE_INVALID}: '${modifierName}'`);
     }
@@ -11347,7 +11437,7 @@ const validateForSpecificSyntax = (syntax, modifier, isException) => {
         if (!specificBlockerData.valueFormat) {
             throw new Error(`${SOURCE_DATA_ERROR_PREFIX.NO_VALUE_FORMAT_FOR_ASSIGNABLE}: '${modifierName}'`);
         }
-        return validateValue(modifier, specificBlockerData.valueFormat);
+        return validateValue(modifier, specificBlockerData.valueFormat, specificBlockerData.valueFormatFlags);
     }
     if (modifier?.value) {
         // e.g. 'third-party=true'
@@ -13044,8 +13134,14 @@ function getErrorMessage(error) {
  * Known validators that don't need to be validated as regex.
  */
 const KNOWN_VALIDATORS = new Set([
+    'csp_value',
     'domain',
+    'permissions_value',
+    'pipe_separated_apps',
+    'pipe_separated_denyallow_domains',
     'pipe_separated_domains',
+    'pipe_separated_methods',
+    'pipe_separated_stealth_options',
     'regexp',
     'url',
 ]);
@@ -13096,6 +13192,10 @@ zodToCamelCase(baseCompatibilityDataSchema.extend({
      * Its value can be a regex pattern or a known validator name (e.g. `domain`, `pipe_separated_domains`, etc.).
      */
     value_format: nonEmptyStringSchema.nullable().default(null),
+    /**
+     * Describes the flags for the `value_format` regex pattern.
+     */
+    value_format_flags: nonEmptyStringSchema.nullable().default(null),
 }).superRefine((data, ctx) => {
     // TODO: find something better, for now we can't add refine logic to the base schema:
     // https://github.com/colinhacks/zod/issues/454#issuecomment-848370721
@@ -13116,11 +13216,17 @@ zodToCamelCase(baseCompatibilityDataSchema.extend({
         const valueFormat = data.value_format.trim();
         // if it is a known validator, we don't need to validate it further
         if (KNOWN_VALIDATORS.has(valueFormat)) {
+            if (data.value_format_flags) {
+                ctx.addIssue({
+                    code: zod.ZodIssueCode.custom,
+                    message: 'value_format_flags are not allowed for known validators',
+                });
+            }
             return;
         }
         // otherwise, we need to validate it as a regex
         try {
-            XRegExp(valueFormat);
+            new RegExp(valueFormat, data.value_format_flags ?? EMPTY);
         }
         catch (error) {
             ctx.addIssue({
@@ -13129,6 +13235,12 @@ zodToCamelCase(baseCompatibilityDataSchema.extend({
             });
         }
     }
+    else if (data.value_format_flags) {
+        ctx.addIssue({
+            code: zod.ZodIssueCode.custom,
+            message: 'value_format is required for value_format_flags',
+        });
+    }
 }));
 
 /**
@@ -15679,7 +15791,7 @@ class RuleCategorizer {
     }
 }
 
-const version = "2.1.4";
+const version = "2.3.0";
 
 /**
  * @file AGTree version