@adguard/agtree 2.2.0 → 2.3.0

package/dist/agtree.d.ts

@@ -1,5 +1,5 @@
 /*
- * AGTree v2.2.0 (build date: Wed, 27 Nov 2024 16:28:27 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

package/dist/agtree.js

@@ -1,5 +1,5 @@
 /*
- * AGTree v2.2.0 (build date: Wed, 27 Nov 2024 16:28:27 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
@@ -311,6 +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 = '@@';
@@ -6085,14 +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;
     }
@@ -7652,6 +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,
@@ -8024,6 +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$3.SYNTAX_DISABLED, exports.AdblockSyntax.Abp), baseOffset + bodyStart, baseOffset + bodyEnd);
@@ -8135,10 +8208,22 @@ class CosmeticRuleParser extends ParserBase {
         // the next function is called, and so on.
         // If all functions return null, an error should be thrown.
         const separatorMap = {
-            '##': [parseUboHtmlFiltering, parseUboScriptletInjection, parseUboCssInjection, parseElementHiding],
-            '#@#': [parseUboHtmlFiltering, parseUboScriptletInjection, parseUboCssInjection, parseElementHiding],
-            '#?#': [parseUboCssInjection, parseElementHiding],
-            '#@?#': [parseUboCssInjection, parseElementHiding],
+            '##': [
+                parseUboHtmlFiltering,
+                parseUboScriptletInjection,
+                parseUboCssInjection,
+                parseAbpCssInjection,
+                parseElementHiding,
+            ],
+            '#@#': [
+                parseUboHtmlFiltering,
+                parseUboScriptletInjection,
+                parseUboCssInjection,
+                parseAbpCssInjection,
+                parseElementHiding,
+            ],
+            '#?#': [parseUboCssInjection, parseAbpCssInjection, parseElementHiding],
+            '#@?#': [parseUboCssInjection, parseAbpCssInjection, parseElementHiding],
             '#$#': [parseAdgCssInjection, parseAbpSnippetInjection],
             '#@$#': [parseAdgCssInjection, parseAbpSnippetInjection],
             '#$?#': [parseAdgCssInjection],
@@ -8207,7 +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) {
@@ -15726,7 +15811,7 @@ class RuleCategorizer {
     }
 }
 
-const version = "2.2.0";
+const version = "2.3.0";
 
 /**
  * @file AGTree version

package/dist/agtree.mjs

@@ -1,5 +1,5 @@
 /*
- * AGTree v2.2.0 (build date: Wed, 27 Nov 2024 16:28:27 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
@@ -291,6 +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 = '@@';
@@ -6065,14 +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;
     }
@@ -7632,6 +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,
@@ -8004,6 +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$3.SYNTAX_DISABLED, AdblockSyntax.Abp), baseOffset + bodyStart, baseOffset + bodyEnd);
@@ -8115,10 +8188,22 @@ class CosmeticRuleParser extends ParserBase {
         // the next function is called, and so on.
         // If all functions return null, an error should be thrown.
         const separatorMap = {
-            '##': [parseUboHtmlFiltering, parseUboScriptletInjection, parseUboCssInjection, parseElementHiding],
-            '#@#': [parseUboHtmlFiltering, parseUboScriptletInjection, parseUboCssInjection, parseElementHiding],
-            '#?#': [parseUboCssInjection, parseElementHiding],
-            '#@?#': [parseUboCssInjection, parseElementHiding],
+            '##': [
+                parseUboHtmlFiltering,
+                parseUboScriptletInjection,
+                parseUboCssInjection,
+                parseAbpCssInjection,
+                parseElementHiding,
+            ],
+            '#@#': [
+                parseUboHtmlFiltering,
+                parseUboScriptletInjection,
+                parseUboCssInjection,
+                parseAbpCssInjection,
+                parseElementHiding,
+            ],
+            '#?#': [parseUboCssInjection, parseAbpCssInjection, parseElementHiding],
+            '#@?#': [parseUboCssInjection, parseAbpCssInjection, parseElementHiding],
             '#$#': [parseAdgCssInjection, parseAbpSnippetInjection],
             '#@$#': [parseAdgCssInjection, parseAbpSnippetInjection],
             '#$?#': [parseAdgCssInjection],
@@ -8187,7 +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) {
@@ -15706,7 +15791,7 @@ class RuleCategorizer {
     }
 }
 
-const version = "2.2.0";
+const version = "2.3.0";
 
 /**
  * @file AGTree version

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@adguard/agtree",
-  "version": "2.2.0",
+  "version": "2.3.0",
   "description": "Tool set for working with adblock filter lists",
   "keywords": [
     "adblock",