entities - npm Package Compare versions

Comparing version 4.4.0 to 4.5.0

lib/decode_codepoint.d.ts

@@ -0,4 +1,19 @@
+/**
+ * Polyfill for `String.fromCodePoint`. It is used to create a string from a Unicode code point.
+ */
 export declare const fromCodePoint: (...codePoints: number[]) => string;
+/**
+ * Replace the given code point with a replacement character if it is a
+ * surrogate or is outside the valid range. Otherwise return the code
+ * point unchanged.
+ */
 export declare function replaceCodePoint(codePoint: number): number;
+/**
+ * Replace the code point if relevant, then convert it to a string.
+ *
+ * @deprecated Use `fromCodePoint(replaceCodePoint(codePoint))` instead.
+ * @param codePoint The code point to decode.
+ * @returns The decoded code point.
+ */
 export default function decodeCodePoint(codePoint: number): string;
 //# sourceMappingURL=decode_codepoint.d.ts.map

lib/decode_codepoint.js

@@ -8,2 +8,3 @@ "use strict";
     [0, 65533],
+    // C1 Unicode control character reference replacements
     [128, 8364],
@@ -37,2 +38,5 @@ [130, 8218],
 ]);
+/**
+ * Polyfill for `String.fromCodePoint`. It is used to create a string from a Unicode code point.
+ */
 exports.fromCodePoint = 
@@ -50,2 +54,7 @@ // eslint-disable-next-line @typescript-eslint/no-unnecessary-condition, node/no-unsupported-features/es-builtins
 };
+/**
+ * Replace the given code point with a replacement character if it is a
+ * surrogate or is outside the valid range. Otherwise return the code
+ * point unchanged.
+ */
 function replaceCodePoint(codePoint) {
@@ -59,2 +68,9 @@ var _a;
 exports.replaceCodePoint = replaceCodePoint;
+/**
+ * Replace the code point if relevant, then convert it to a string.
+ *
+ * @deprecated Use `fromCodePoint(replaceCodePoint(codePoint))` instead.
+ * @param codePoint The code point to decode.
+ * @returns The decoded code point.
+ */
 function decodeCodePoint(codePoint) {
@@ -61,0 +77,0 @@ return (0, exports.fromCodePoint)(replaceCodePoint(codePoint));

lib/decode.d.ts

@@ -11,12 +11,183 @@ import htmlDecodeTree from "./generated/decode-data-html.js";
 }
+export declare enum DecodingMode {
+    /** Entities in text nodes that can end with any character. */
+    Legacy = 0,
+    /** Only allow entities terminated with a semicolon. */
+    Strict = 1,
+    /** Entities in attributes have limitations on ending characters. */
+    Attribute = 2
+}
+/**
+ * Producers for character reference errors as defined in the HTML spec.
+ */
+export interface EntityErrorProducer {
+    missingSemicolonAfterCharacterReference(): void;
+    absenceOfDigitsInNumericCharacterReference(consumedCharacters: number): void;
+    validateNumericCharacterReference(code: number): void;
+}
+/**
+ * Token decoder with support of writing partial entities.
+ */
+export declare class EntityDecoder {
+    /** The tree used to decode entities. */
+    private readonly decodeTree;
+    /**
+     * The function that is called when a codepoint is decoded.
+     *
+     * For multi-byte named entities, this will be called multiple times,
+     * with the second codepoint, and the same `consumed` value.
+     *
+     * @param codepoint The decoded codepoint.
+     * @param consumed The number of bytes consumed by the decoder.
+     */
+    private readonly emitCodePoint;
+    /** An object that is used to produce errors. */
+    private readonly errors?;
+    constructor(
+    /** The tree used to decode entities. */
+    decodeTree: Uint16Array, 
+    /**
+     * The function that is called when a codepoint is decoded.
+     *
+     * For multi-byte named entities, this will be called multiple times,
+     * with the second codepoint, and the same `consumed` value.
+     *
+     * @param codepoint The decoded codepoint.
+     * @param consumed The number of bytes consumed by the decoder.
+     */
+    emitCodePoint: (cp: number, consumed: number) => void, 
+    /** An object that is used to produce errors. */
+    errors?: EntityErrorProducer | undefined);
+    /** The current state of the decoder. */
+    private state;
+    /** Characters that were consumed while parsing an entity. */
+    private consumed;
+    /**
+     * The result of the entity.
+     *
+     * Either the result index of a numeric entity, or the codepoint of a
+     * numeric entity.
+     */
+    private result;
+    /** The current index in the decode tree. */
+    private treeIndex;
+    /** The number of characters that were consumed in excess. */
+    private excess;
+    /** The mode in which the decoder is operating. */
+    private decodeMode;
+    /** Resets the instance to make it reusable. */
+    startEntity(decodeMode: DecodingMode): void;
+    /**
+     * Write an entity to the decoder. This can be called multiple times with partial entities.
+     * If the entity is incomplete, the decoder will return -1.
+     *
+     * Mirrors the implementation of `getDecoder`, but with the ability to stop decoding if the
+     * entity is incomplete, and resume when the next string is written.
+     *
+     * @param string The string containing the entity (or a continuation of the entity).
+     * @param offset The offset at which the entity begins. Should be 0 if this is not the first call.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    write(str: string, offset: number): number;
+    /**
+     * Switches between the numeric decimal and hexadecimal states.
+     *
+     * Equivalent to the `Numeric character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    private stateNumericStart;
+    private addToNumericResult;
+    /**
+     * Parses a hexadecimal numeric entity.
+     *
+     * Equivalent to the `Hexademical character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    private stateNumericHex;
+    /**
+     * Parses a decimal numeric entity.
+     *
+     * Equivalent to the `Decimal character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    private stateNumericDecimal;
+    /**
+     * Validate and emit a numeric entity.
+     *
+     * Implements the logic from the `Hexademical character reference start
+     * state` and `Numeric character reference end state` in the HTML spec.
+     *
+     * @param lastCp The last code point of the entity. Used to see if the
+     *               entity was terminated with a semicolon.
+     * @param expectedLength The minimum number of characters that should be
+     *                       consumed. Used to validate that at least one digit
+     *                       was consumed.
+     * @returns The number of characters that were consumed.
+     */
+    private emitNumericEntity;
+    /**
+     * Parses a named entity.
+     *
+     * Equivalent to the `Named character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    private stateNamedEntity;
+    /**
+     * Emit a named entity that was not terminated with a semicolon.
+     *
+     * @returns The number of characters consumed.
+     */
+    private emitNotTerminatedNamedEntity;
+    /**
+     * Emit a named entity.
+     *
+     * @param result The index of the entity in the decode tree.
+     * @param valueLength The number of bytes in the entity.
+     * @param consumed The number of characters consumed.
+     *
+     * @returns The number of characters consumed.
+     */
+    private emitNamedEntityData;
+    /**
+     * Signal to the parser that the end of the input was reached.
+     *
+     * Remaining data will be emitted and relevant errors will be produced.
+     *
+     * @returns The number of characters consumed.
+     */
+    end(): number;
+}
+/**
+ * Determines the branch of the current node that is taken given the current
+ * character. This function is used to traverse the trie.
+ *
+ * @param decodeTree The trie.
+ * @param current The current node.
+ * @param nodeIdx The index right after the current node and its value.
+ * @param char The current character.
+ * @returns The index of the next node, or -1 if no branch is taken.
+ */
 export declare function determineBranch(decodeTree: Uint16Array, current: number, nodeIdx: number, char: number): number;
 /**
- * Decodes an HTML string, allowing for entities not terminated by a semi-colon.
+ * Decodes an HTML string.
  *
  * @param str The string to decode.
+ * @param mode The decoding mode.
  * @returns The decoded string.
  */
-export declare function decodeHTML(str: string): string;
+export declare function decodeHTML(str: string, mode?: DecodingMode): string;
 /**
- * Decodes an HTML string, requiring all entities to be terminated by a semi-colon.
+ * Decodes an HTML string in an attribute.
  *
@@ -26,5 +197,12 @@ * @param str The string to decode.
  */
+export declare function decodeHTMLAttribute(str: string): string;
+/**
+ * Decodes an HTML string, requiring all entities to be terminated by a semicolon.
+ *
+ * @param str The string to decode.
+ * @returns The decoded string.
+ */
 export declare function decodeHTMLStrict(str: string): string;
 /**
- * Decodes an XML string, requiring all entities to be terminated by a semi-colon.
+ * Decodes an XML string, requiring all entities to be terminated by a semicolon.
  *
@@ -31,0 +209,0 @@ * @param str The string to decode.

lib/decode.js

 "use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
+    Object.defineProperty(o, "default", { enumerable: true, value: v });
+}) : function(o, v) {
+    o["default"] = v;
+});
+var __importStar = (this && this.__importStar) || function (mod) {
+    if (mod && mod.__esModule) return mod;
+    var result = {};
+    if (mod != null) for (var k in mod) if (k !== "default" && Object.prototype.hasOwnProperty.call(mod, k)) __createBinding(result, mod, k);
+    __setModuleDefault(result, mod);
+    return result;
+};
 var __importDefault = (this && this.__importDefault) || function (mod) {
@@ -6,3 +29,3 @@ return (mod && mod.__esModule) ? mod : { "default": mod };
 Object.defineProperty(exports, "__esModule", { value: true });
-exports.decodeXML = exports.decodeHTMLStrict = exports.decodeHTML = exports.determineBranch = exports.BinTrieFlags = exports.fromCodePoint = exports.replaceCodePoint = exports.decodeCodePoint = exports.xmlDecodeTree = exports.htmlDecodeTree = void 0;
+exports.decodeXML = exports.decodeHTMLStrict = exports.decodeHTMLAttribute = exports.decodeHTML = exports.determineBranch = exports.EntityDecoder = exports.DecodingMode = exports.BinTrieFlags = exports.fromCodePoint = exports.replaceCodePoint = exports.decodeCodePoint = exports.xmlDecodeTree = exports.htmlDecodeTree = void 0;
 var decode_data_html_js_1 = __importDefault(require("./generated/decode-data-html.js"));
@@ -12,3 +35,3 @@ exports.htmlDecodeTree = decode_data_html_js_1.default;
 exports.xmlDecodeTree = decode_data_xml_js_1.default;
-var decode_codepoint_js_1 = __importDefault(require("./decode_codepoint.js"));
+var decode_codepoint_js_1 = __importStar(require("./decode_codepoint.js"));
 exports.decodeCodePoint = decode_codepoint_js_1.default;
@@ -22,2 +45,3 @@ var decode_codepoint_js_2 = require("./decode_codepoint.js");
     CharCodes[CharCodes["SEMI"] = 59] = "SEMI";
+    CharCodes[CharCodes["EQUALS"] = 61] = "EQUALS";
     CharCodes[CharCodes["ZERO"] = 48] = "ZERO";
@@ -28,5 +52,9 @@ CharCodes[CharCodes["NINE"] = 57] = "NINE";
     CharCodes[CharCodes["LOWER_X"] = 120] = "LOWER_X";
-    /** Bit that needs to be set to convert an upper case ASCII character to lower case */
-    CharCodes[CharCodes["To_LOWER_BIT"] = 32] = "To_LOWER_BIT";
+    CharCodes[CharCodes["LOWER_Z"] = 122] = "LOWER_Z";
+    CharCodes[CharCodes["UPPER_A"] = 65] = "UPPER_A";
+    CharCodes[CharCodes["UPPER_F"] = 70] = "UPPER_F";
+    CharCodes[CharCodes["UPPER_Z"] = 90] = "UPPER_Z";
 })(CharCodes || (CharCodes = {}));
+/** Bit that needs to be set to convert an upper case ASCII character to lower case */
+var TO_LOWER_BIT = 32;
 var BinTrieFlags;
@@ -38,81 +66,398 @@ (function (BinTrieFlags) {
 })(BinTrieFlags = exports.BinTrieFlags || (exports.BinTrieFlags = {}));
-function getDecoder(decodeTree) {
-    return function decodeHTMLBinary(str, strict) {
-        var ret = "";
-        var lastIdx = 0;
-        var strIdx = 0;
-        while ((strIdx = str.indexOf("&", strIdx)) >= 0) {
-            ret += str.slice(lastIdx, strIdx);
-            lastIdx = strIdx;
-            // Skip the "&"
-            strIdx += 1;
-            // If we have a numeric entity, handle this separately.
-            if (str.charCodeAt(strIdx) === CharCodes.NUM) {
-                // Skip the leading "&#". For hex entities, also skip the leading "x".
-                var start = strIdx + 1;
-                var base = 10;
-                var cp = str.charCodeAt(start);
-                if ((cp | CharCodes.To_LOWER_BIT) === CharCodes.LOWER_X) {
-                    base = 16;
-                    strIdx += 1;
-                    start += 1;
+function isNumber(code) {
+    return code >= CharCodes.ZERO && code <= CharCodes.NINE;
+}
+function isHexadecimalCharacter(code) {
+    return ((code >= CharCodes.UPPER_A && code <= CharCodes.UPPER_F) ||
+        (code >= CharCodes.LOWER_A && code <= CharCodes.LOWER_F));
+}
+function isAsciiAlphaNumeric(code) {
+    return ((code >= CharCodes.UPPER_A && code <= CharCodes.UPPER_Z) ||
+        (code >= CharCodes.LOWER_A && code <= CharCodes.LOWER_Z) ||
+        isNumber(code));
+}
+/**
+ * Checks if the given character is a valid end character for an entity in an attribute.
+ *
+ * Attribute values that aren't terminated properly aren't parsed, and shouldn't lead to a parser error.
+ * See the example in https://html.spec.whatwg.org/multipage/parsing.html#named-character-reference-state
+ */
+function isEntityInAttributeInvalidEnd(code) {
+    return code === CharCodes.EQUALS || isAsciiAlphaNumeric(code);
+}
+var EntityDecoderState;
+(function (EntityDecoderState) {
+    EntityDecoderState[EntityDecoderState["EntityStart"] = 0] = "EntityStart";
+    EntityDecoderState[EntityDecoderState["NumericStart"] = 1] = "NumericStart";
+    EntityDecoderState[EntityDecoderState["NumericDecimal"] = 2] = "NumericDecimal";
+    EntityDecoderState[EntityDecoderState["NumericHex"] = 3] = "NumericHex";
+    EntityDecoderState[EntityDecoderState["NamedEntity"] = 4] = "NamedEntity";
+})(EntityDecoderState || (EntityDecoderState = {}));
+var DecodingMode;
+(function (DecodingMode) {
+    /** Entities in text nodes that can end with any character. */
+    DecodingMode[DecodingMode["Legacy"] = 0] = "Legacy";
+    /** Only allow entities terminated with a semicolon. */
+    DecodingMode[DecodingMode["Strict"] = 1] = "Strict";
+    /** Entities in attributes have limitations on ending characters. */
+    DecodingMode[DecodingMode["Attribute"] = 2] = "Attribute";
+})(DecodingMode = exports.DecodingMode || (exports.DecodingMode = {}));
+/**
+ * Token decoder with support of writing partial entities.
+ */
+var EntityDecoder = /** @class */ (function () {
+    function EntityDecoder(
+    /** The tree used to decode entities. */
+    decodeTree, 
+    /**
+     * The function that is called when a codepoint is decoded.
+     *
+     * For multi-byte named entities, this will be called multiple times,
+     * with the second codepoint, and the same `consumed` value.
+     *
+     * @param codepoint The decoded codepoint.
+     * @param consumed The number of bytes consumed by the decoder.
+     */
+    emitCodePoint, 
+    /** An object that is used to produce errors. */
+    errors) {
+        this.decodeTree = decodeTree;
+        this.emitCodePoint = emitCodePoint;
+        this.errors = errors;
+        /** The current state of the decoder. */
+        this.state = EntityDecoderState.EntityStart;
+        /** Characters that were consumed while parsing an entity. */
+        this.consumed = 1;
+        /**
+         * The result of the entity.
+         *
+         * Either the result index of a numeric entity, or the codepoint of a
+         * numeric entity.
+         */
+        this.result = 0;
+        /** The current index in the decode tree. */
+        this.treeIndex = 0;
+        /** The number of characters that were consumed in excess. */
+        this.excess = 1;
+        /** The mode in which the decoder is operating. */
+        this.decodeMode = DecodingMode.Strict;
+    }
+    /** Resets the instance to make it reusable. */
+    EntityDecoder.prototype.startEntity = function (decodeMode) {
+        this.decodeMode = decodeMode;
+        this.state = EntityDecoderState.EntityStart;
+        this.result = 0;
+        this.treeIndex = 0;
+        this.excess = 1;
+        this.consumed = 1;
+    };
+    /**
+     * Write an entity to the decoder. This can be called multiple times with partial entities.
+     * If the entity is incomplete, the decoder will return -1.
+     *
+     * Mirrors the implementation of `getDecoder`, but with the ability to stop decoding if the
+     * entity is incomplete, and resume when the next string is written.
+     *
+     * @param string The string containing the entity (or a continuation of the entity).
+     * @param offset The offset at which the entity begins. Should be 0 if this is not the first call.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    EntityDecoder.prototype.write = function (str, offset) {
+        switch (this.state) {
+            case EntityDecoderState.EntityStart: {
+                if (str.charCodeAt(offset) === CharCodes.NUM) {
+                    this.state = EntityDecoderState.NumericStart;
+                    this.consumed += 1;
+                    return this.stateNumericStart(str, offset + 1);
                 }
-                do
-                    cp = str.charCodeAt(++strIdx);
-                while ((cp >= CharCodes.ZERO && cp <= CharCodes.NINE) ||
-                    (base === 16 &&
-                        (cp | CharCodes.To_LOWER_BIT) >= CharCodes.LOWER_A &&
-                        (cp | CharCodes.To_LOWER_BIT) <= CharCodes.LOWER_F));
-                if (start !== strIdx) {
-                    var entity = str.substring(start, strIdx);
-                    var parsed = parseInt(entity, base);
-                    if (str.charCodeAt(strIdx) === CharCodes.SEMI) {
-                        strIdx += 1;
-                    }
-                    else if (strict) {
-                        continue;
-                    }
-                    ret += (0, decode_codepoint_js_1.default)(parsed);
-                    lastIdx = strIdx;
-                }
-                continue;
+                this.state = EntityDecoderState.NamedEntity;
+                return this.stateNamedEntity(str, offset);
             }
-            var resultIdx = 0;
-            var excess = 1;
-            var treeIdx = 0;
-            var current = decodeTree[treeIdx];
-            for (; strIdx < str.length; strIdx++, excess++) {
-                treeIdx = determineBranch(decodeTree, current, treeIdx + 1, str.charCodeAt(strIdx));
-                if (treeIdx < 0)
-                    break;
-                current = decodeTree[treeIdx];
-                var masked = current & BinTrieFlags.VALUE_LENGTH;
-                // If the branch is a value, store it and continue
-                if (masked) {
-                    // If we have a legacy entity while parsing strictly, just skip the number of bytes
-                    if (!strict || str.charCodeAt(strIdx) === CharCodes.SEMI) {
-                        resultIdx = treeIdx;
-                        excess = 0;
-                    }
-                    // The mask is the number of bytes of the value, including the current byte.
-                    var valueLength = (masked >> 14) - 1;
-                    if (valueLength === 0)
-                        break;
-                    treeIdx += valueLength;
+            case EntityDecoderState.NumericStart: {
+                return this.stateNumericStart(str, offset);
+            }
+            case EntityDecoderState.NumericDecimal: {
+                return this.stateNumericDecimal(str, offset);
+            }
+            case EntityDecoderState.NumericHex: {
+                return this.stateNumericHex(str, offset);
+            }
+            case EntityDecoderState.NamedEntity: {
+                return this.stateNamedEntity(str, offset);
+            }
+        }
+    };
+    /**
+     * Switches between the numeric decimal and hexadecimal states.
+     *
+     * Equivalent to the `Numeric character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    EntityDecoder.prototype.stateNumericStart = function (str, offset) {
+        if (offset >= str.length) {
+            return -1;
+        }
+        if ((str.charCodeAt(offset) | TO_LOWER_BIT) === CharCodes.LOWER_X) {
+            this.state = EntityDecoderState.NumericHex;
+            this.consumed += 1;
+            return this.stateNumericHex(str, offset + 1);
+        }
+        this.state = EntityDecoderState.NumericDecimal;
+        return this.stateNumericDecimal(str, offset);
+    };
+    EntityDecoder.prototype.addToNumericResult = function (str, start, end, base) {
+        if (start !== end) {
+            var digitCount = end - start;
+            this.result =
+                this.result * Math.pow(base, digitCount) +
+                    parseInt(str.substr(start, digitCount), base);
+            this.consumed += digitCount;
+        }
+    };
+    /**
+     * Parses a hexadecimal numeric entity.
+     *
+     * Equivalent to the `Hexademical character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    EntityDecoder.prototype.stateNumericHex = function (str, offset) {
+        var startIdx = offset;
+        while (offset < str.length) {
+            var char = str.charCodeAt(offset);
+            if (isNumber(char) || isHexadecimalCharacter(char)) {
+                offset += 1;
+            }
+            else {
+                this.addToNumericResult(str, startIdx, offset, 16);
+                return this.emitNumericEntity(char, 3);
+            }
+        }
+        this.addToNumericResult(str, startIdx, offset, 16);
+        return -1;
+    };
+    /**
+     * Parses a decimal numeric entity.
+     *
+     * Equivalent to the `Decimal character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    EntityDecoder.prototype.stateNumericDecimal = function (str, offset) {
+        var startIdx = offset;
+        while (offset < str.length) {
+            var char = str.charCodeAt(offset);
+            if (isNumber(char)) {
+                offset += 1;
+            }
+            else {
+                this.addToNumericResult(str, startIdx, offset, 10);
+                return this.emitNumericEntity(char, 2);
+            }
+        }
+        this.addToNumericResult(str, startIdx, offset, 10);
+        return -1;
+    };
+    /**
+     * Validate and emit a numeric entity.
+     *
+     * Implements the logic from the `Hexademical character reference start
+     * state` and `Numeric character reference end state` in the HTML spec.
+     *
+     * @param lastCp The last code point of the entity. Used to see if the
+     *               entity was terminated with a semicolon.
+     * @param expectedLength The minimum number of characters that should be
+     *                       consumed. Used to validate that at least one digit
+     *                       was consumed.
+     * @returns The number of characters that were consumed.
+     */
+    EntityDecoder.prototype.emitNumericEntity = function (lastCp, expectedLength) {
+        var _a;
+        // Ensure we consumed at least one digit.
+        if (this.consumed <= expectedLength) {
+            (_a = this.errors) === null || _a === void 0 ? void 0 : _a.absenceOfDigitsInNumericCharacterReference(this.consumed);
+            return 0;
+        }
+        // Figure out if this is a legit end of the entity
+        if (lastCp === CharCodes.SEMI) {
+            this.consumed += 1;
+        }
+        else if (this.decodeMode === DecodingMode.Strict) {
+            return 0;
+        }
+        this.emitCodePoint((0, decode_codepoint_js_1.replaceCodePoint)(this.result), this.consumed);
+        if (this.errors) {
+            if (lastCp !== CharCodes.SEMI) {
+                this.errors.missingSemicolonAfterCharacterReference();
+            }
+            this.errors.validateNumericCharacterReference(this.result);
+        }
+        return this.consumed;
+    };
+    /**
+     * Parses a named entity.
+     *
+     * Equivalent to the `Named character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    EntityDecoder.prototype.stateNamedEntity = function (str, offset) {
+        var decodeTree = this.decodeTree;
+        var current = decodeTree[this.treeIndex];
+        // The mask is the number of bytes of the value, including the current byte.
+        var valueLength = (current & BinTrieFlags.VALUE_LENGTH) >> 14;
+        for (; offset < str.length; offset++, this.excess++) {
+            var char = str.charCodeAt(offset);
+            this.treeIndex = determineBranch(decodeTree, current, this.treeIndex + Math.max(1, valueLength), char);
+            if (this.treeIndex < 0) {
+                return this.result === 0 ||
+                    // If we are parsing an attribute
+                    (this.decodeMode === DecodingMode.Attribute &&
+                        // We shouldn't have consumed any characters after the entity,
+                        (valueLength === 0 ||
+                            // And there should be no invalid characters.
+                            isEntityInAttributeInvalidEnd(char)))
+                    ? 0
+                    : this.emitNotTerminatedNamedEntity();
+            }
+            current = decodeTree[this.treeIndex];
+            valueLength = (current & BinTrieFlags.VALUE_LENGTH) >> 14;
+            // If the branch is a value, store it and continue
+            if (valueLength !== 0) {
+                // If the entity is terminated by a semicolon, we are done.
+                if (char === CharCodes.SEMI) {
+                    return this.emitNamedEntityData(this.treeIndex, valueLength, this.consumed + this.excess);
                 }
+                // If we encounter a non-terminated (legacy) entity while parsing strictly, then ignore it.
+                if (this.decodeMode !== DecodingMode.Strict) {
+                    this.result = this.treeIndex;
+                    this.consumed += this.excess;
+                    this.excess = 0;
+                }
             }
-            if (resultIdx !== 0) {
-                var valueLength = (decodeTree[resultIdx] & BinTrieFlags.VALUE_LENGTH) >> 14;
-                ret +=
-                    valueLength === 1
-                        ? String.fromCharCode(decodeTree[resultIdx] & ~BinTrieFlags.VALUE_LENGTH)
-                        : valueLength === 2
-                            ? String.fromCharCode(decodeTree[resultIdx + 1])
-                            : String.fromCharCode(decodeTree[resultIdx + 1], decodeTree[resultIdx + 2]);
-                lastIdx = strIdx - excess + 1;
+        }
+        return -1;
+    };
+    /**
+     * Emit a named entity that was not terminated with a semicolon.
+     *
+     * @returns The number of characters consumed.
+     */
+    EntityDecoder.prototype.emitNotTerminatedNamedEntity = function () {
+        var _a;
+        var _b = this, result = _b.result, decodeTree = _b.decodeTree;
+        var valueLength = (decodeTree[result] & BinTrieFlags.VALUE_LENGTH) >> 14;
+        this.emitNamedEntityData(result, valueLength, this.consumed);
+        (_a = this.errors) === null || _a === void 0 ? void 0 : _a.missingSemicolonAfterCharacterReference();
+        return this.consumed;
+    };
+    /**
+     * Emit a named entity.
+     *
+     * @param result The index of the entity in the decode tree.
+     * @param valueLength The number of bytes in the entity.
+     * @param consumed The number of characters consumed.
+     *
+     * @returns The number of characters consumed.
+     */
+    EntityDecoder.prototype.emitNamedEntityData = function (result, valueLength, consumed) {
+        var decodeTree = this.decodeTree;
+        this.emitCodePoint(valueLength === 1
+            ? decodeTree[result] & ~BinTrieFlags.VALUE_LENGTH
+            : decodeTree[result + 1], consumed);
+        if (valueLength === 3) {
+            // For multi-byte values, we need to emit the second byte.
+            this.emitCodePoint(decodeTree[result + 2], consumed);
+        }
+        return consumed;
+    };
+    /**
+     * Signal to the parser that the end of the input was reached.
+     *
+     * Remaining data will be emitted and relevant errors will be produced.
+     *
+     * @returns The number of characters consumed.
+     */
+    EntityDecoder.prototype.end = function () {
+        var _a;
+        switch (this.state) {
+            case EntityDecoderState.NamedEntity: {
+                // Emit a named entity if we have one.
+                return this.result !== 0 &&
+                    (this.decodeMode !== DecodingMode.Attribute ||
+                        this.result === this.treeIndex)
+                    ? this.emitNotTerminatedNamedEntity()
+                    : 0;
             }
+            // Otherwise, emit a numeric entity if we have one.
+            case EntityDecoderState.NumericDecimal: {
+                return this.emitNumericEntity(0, 2);
+            }
+            case EntityDecoderState.NumericHex: {
+                return this.emitNumericEntity(0, 3);
+            }
+            case EntityDecoderState.NumericStart: {
+                (_a = this.errors) === null || _a === void 0 ? void 0 : _a.absenceOfDigitsInNumericCharacterReference(this.consumed);
+                return 0;
+            }
+            case EntityDecoderState.EntityStart: {
+                // Return 0 if we have no entity.
+                return 0;
+            }
         }
-        return ret + str.slice(lastIdx);
     };
+    return EntityDecoder;
+}());
+exports.EntityDecoder = EntityDecoder;
+/**
+ * Creates a function that decodes entities in a string.
+ *
+ * @param decodeTree The decode tree.
+ * @returns A function that decodes entities in a string.
+ */
+function getDecoder(decodeTree) {
+    var ret = "";
+    var decoder = new EntityDecoder(decodeTree, function (str) { return (ret += (0, decode_codepoint_js_1.fromCodePoint)(str)); });
+    return function decodeWithTrie(str, decodeMode) {
+        var lastIndex = 0;
+        var offset = 0;
+        while ((offset = str.indexOf("&", offset)) >= 0) {
+            ret += str.slice(lastIndex, offset);
+            decoder.startEntity(decodeMode);
+            var len = decoder.write(str, 
+            // Skip the "&"
+            offset + 1);
+            if (len < 0) {
+                lastIndex = offset + decoder.end();
+                break;
+            }
+            lastIndex = offset + len;
+            // If `len` is 0, skip the current `&` and continue.
+            offset = len === 0 ? lastIndex + 1 : lastIndex;
+        }
+        var result = ret + str.slice(lastIndex);
+        // Make sure we don't keep a reference to the final string.
+        ret = "";
+        return result;
+    };
 }
+/**
+ * Determines the branch of the current node that is taken given the current
+ * character. This function is used to traverse the trie.
+ *
+ * @param decodeTree The trie.
+ * @param current The current node.
+ * @param nodeIdx The index right after the current node and its value.
+ * @param char The current character.
+ * @returns The index of the next node, or -1 if no branch is taken.
+ */
 function determineBranch(decodeTree, current, nodeIdx, char) {
@@ -155,13 +500,15 @@ var branchCount = (current & BinTrieFlags.BRANCH_LENGTH) >> 7;
 /**
- * Decodes an HTML string, allowing for entities not terminated by a semi-colon.
+ * Decodes an HTML string.
  *
  * @param str The string to decode.
+ * @param mode The decoding mode.
  * @returns The decoded string.
  */
-function decodeHTML(str) {
-    return htmlDecoder(str, false);
+function decodeHTML(str, mode) {
+    if (mode === void 0) { mode = DecodingMode.Legacy; }
+    return htmlDecoder(str, mode);
 }
 exports.decodeHTML = decodeHTML;
 /**
- * Decodes an HTML string, requiring all entities to be terminated by a semi-colon.
+ * Decodes an HTML string in an attribute.
  *
@@ -171,8 +518,18 @@ * @param str The string to decode.
  */
+function decodeHTMLAttribute(str) {
+    return htmlDecoder(str, DecodingMode.Attribute);
+}
+exports.decodeHTMLAttribute = decodeHTMLAttribute;
+/**
+ * Decodes an HTML string, requiring all entities to be terminated by a semicolon.
+ *
+ * @param str The string to decode.
+ * @returns The decoded string.
+ */
 function decodeHTMLStrict(str) {
-    return htmlDecoder(str, true);
+    return htmlDecoder(str, DecodingMode.Strict);
 }
 exports.decodeHTMLStrict = decodeHTMLStrict;
 /**
- * Decodes an XML string, requiring all entities to be terminated by a semi-colon.
+ * Decodes an XML string, requiring all entities to be terminated by a semicolon.
  *
@@ -183,5 +540,5 @@ * @param str The string to decode.
 function decodeXML(str) {
-    return xmlDecoder(str, true);
+    return xmlDecoder(str, DecodingMode.Strict);
 }
 exports.decodeXML = decodeXML;
 //# sourceMappingURL=decode.js.map

lib/encode.js

@@ -63,3 +63,3 @@ "use strict";
         }
-        // We might have a tree node without a value; skip and use a numeric entitiy.
+        // We might have a tree node without a value; skip and use a numeric entity.
         if (next !== undefined) {
@@ -66,0 +66,0 @@ ret += next;

lib/escape.js

@@ -64,2 +64,12 @@ "use strict";
 exports.escape = encodeXML;
+/**
+ * Creates a function that escapes all characters matched by the given regular
+ * expression using the given map of characters to escape to their entities.
+ *
+ * @param regex Regular expression to match characters to escape.
+ * @param map Map of characters to escape to their entities.
+ *
+ * @returns Function that escapes all characters matched by the given regular
+ * expression using the given map of characters to escape to their entities.
+ */
 function getEscaper(regex, map) {
@@ -74,3 +84,3 @@ return function escape(data) {
             }
-            // We know that this chararcter will be in the map.
+            // We know that this character will be in the map.
             result += map.get(match[0].charCodeAt(0));
@@ -77,0 +87,0 @@ // Every match will be of length 1

lib/esm/decode_codepoint.d.ts

@@ -0,4 +1,19 @@
+/**
+ * Polyfill for `String.fromCodePoint`. It is used to create a string from a Unicode code point.
+ */
 export declare const fromCodePoint: (...codePoints: number[]) => string;
+/**
+ * Replace the given code point with a replacement character if it is a
+ * surrogate or is outside the valid range. Otherwise return the code
+ * point unchanged.
+ */
 export declare function replaceCodePoint(codePoint: number): number;
+/**
+ * Replace the code point if relevant, then convert it to a string.
+ *
+ * @deprecated Use `fromCodePoint(replaceCodePoint(codePoint))` instead.
+ * @param codePoint The code point to decode.
+ * @returns The decoded code point.
+ */
 export default function decodeCodePoint(codePoint: number): string;
 //# sourceMappingURL=decode_codepoint.d.ts.map

lib/esm/decode_codepoint.js

@@ -5,2 +5,3 @@ // Adapted from https://github.com/mathiasbynens/he/blob/36afe179392226cf1b6ccdb16ebbb7a5a844d93a/src/he.js#L106-L134
     [0, 65533],
+    // C1 Unicode control character reference replacements
     [128, 8364],
@@ -34,2 +35,5 @@ [130, 8218],
 ]);
+/**
+ * Polyfill for `String.fromCodePoint`. It is used to create a string from a Unicode code point.
+ */
 export const fromCodePoint = 
@@ -47,2 +51,7 @@ // eslint-disable-next-line @typescript-eslint/no-unnecessary-condition, node/no-unsupported-features/es-builtins
 };
+/**
+ * Replace the given code point with a replacement character if it is a
+ * surrogate or is outside the valid range. Otherwise return the code
+ * point unchanged.
+ */
 export function replaceCodePoint(codePoint) {
@@ -55,2 +64,9 @@ var _a;
 }
+/**
+ * Replace the code point if relevant, then convert it to a string.
+ *
+ * @deprecated Use `fromCodePoint(replaceCodePoint(codePoint))` instead.
+ * @param codePoint The code point to decode.
+ * @returns The decoded code point.
+ */
 export default function decodeCodePoint(codePoint) {
@@ -57,0 +73,0 @@ return fromCodePoint(replaceCodePoint(codePoint));

lib/esm/decode.d.ts

@@ -11,12 +11,183 @@ import htmlDecodeTree from "./generated/decode-data-html.js";
 }
+export declare enum DecodingMode {
+    /** Entities in text nodes that can end with any character. */
+    Legacy = 0,
+    /** Only allow entities terminated with a semicolon. */
+    Strict = 1,
+    /** Entities in attributes have limitations on ending characters. */
+    Attribute = 2
+}
+/**
+ * Producers for character reference errors as defined in the HTML spec.
+ */
+export interface EntityErrorProducer {
+    missingSemicolonAfterCharacterReference(): void;
+    absenceOfDigitsInNumericCharacterReference(consumedCharacters: number): void;
+    validateNumericCharacterReference(code: number): void;
+}
+/**
+ * Token decoder with support of writing partial entities.
+ */
+export declare class EntityDecoder {
+    /** The tree used to decode entities. */
+    private readonly decodeTree;
+    /**
+     * The function that is called when a codepoint is decoded.
+     *
+     * For multi-byte named entities, this will be called multiple times,
+     * with the second codepoint, and the same `consumed` value.
+     *
+     * @param codepoint The decoded codepoint.
+     * @param consumed The number of bytes consumed by the decoder.
+     */
+    private readonly emitCodePoint;
+    /** An object that is used to produce errors. */
+    private readonly errors?;
+    constructor(
+    /** The tree used to decode entities. */
+    decodeTree: Uint16Array, 
+    /**
+     * The function that is called when a codepoint is decoded.
+     *
+     * For multi-byte named entities, this will be called multiple times,
+     * with the second codepoint, and the same `consumed` value.
+     *
+     * @param codepoint The decoded codepoint.
+     * @param consumed The number of bytes consumed by the decoder.
+     */
+    emitCodePoint: (cp: number, consumed: number) => void, 
+    /** An object that is used to produce errors. */
+    errors?: EntityErrorProducer | undefined);
+    /** The current state of the decoder. */
+    private state;
+    /** Characters that were consumed while parsing an entity. */
+    private consumed;
+    /**
+     * The result of the entity.
+     *
+     * Either the result index of a numeric entity, or the codepoint of a
+     * numeric entity.
+     */
+    private result;
+    /** The current index in the decode tree. */
+    private treeIndex;
+    /** The number of characters that were consumed in excess. */
+    private excess;
+    /** The mode in which the decoder is operating. */
+    private decodeMode;
+    /** Resets the instance to make it reusable. */
+    startEntity(decodeMode: DecodingMode): void;
+    /**
+     * Write an entity to the decoder. This can be called multiple times with partial entities.
+     * If the entity is incomplete, the decoder will return -1.
+     *
+     * Mirrors the implementation of `getDecoder`, but with the ability to stop decoding if the
+     * entity is incomplete, and resume when the next string is written.
+     *
+     * @param string The string containing the entity (or a continuation of the entity).
+     * @param offset The offset at which the entity begins. Should be 0 if this is not the first call.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    write(str: string, offset: number): number;
+    /**
+     * Switches between the numeric decimal and hexadecimal states.
+     *
+     * Equivalent to the `Numeric character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    private stateNumericStart;
+    private addToNumericResult;
+    /**
+     * Parses a hexadecimal numeric entity.
+     *
+     * Equivalent to the `Hexademical character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    private stateNumericHex;
+    /**
+     * Parses a decimal numeric entity.
+     *
+     * Equivalent to the `Decimal character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    private stateNumericDecimal;
+    /**
+     * Validate and emit a numeric entity.
+     *
+     * Implements the logic from the `Hexademical character reference start
+     * state` and `Numeric character reference end state` in the HTML spec.
+     *
+     * @param lastCp The last code point of the entity. Used to see if the
+     *               entity was terminated with a semicolon.
+     * @param expectedLength The minimum number of characters that should be
+     *                       consumed. Used to validate that at least one digit
+     *                       was consumed.
+     * @returns The number of characters that were consumed.
+     */
+    private emitNumericEntity;
+    /**
+     * Parses a named entity.
+     *
+     * Equivalent to the `Named character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    private stateNamedEntity;
+    /**
+     * Emit a named entity that was not terminated with a semicolon.
+     *
+     * @returns The number of characters consumed.
+     */
+    private emitNotTerminatedNamedEntity;
+    /**
+     * Emit a named entity.
+     *
+     * @param result The index of the entity in the decode tree.
+     * @param valueLength The number of bytes in the entity.
+     * @param consumed The number of characters consumed.
+     *
+     * @returns The number of characters consumed.
+     */
+    private emitNamedEntityData;
+    /**
+     * Signal to the parser that the end of the input was reached.
+     *
+     * Remaining data will be emitted and relevant errors will be produced.
+     *
+     * @returns The number of characters consumed.
+     */
+    end(): number;
+}
+/**
+ * Determines the branch of the current node that is taken given the current
+ * character. This function is used to traverse the trie.
+ *
+ * @param decodeTree The trie.
+ * @param current The current node.
+ * @param nodeIdx The index right after the current node and its value.
+ * @param char The current character.
+ * @returns The index of the next node, or -1 if no branch is taken.
+ */
 export declare function determineBranch(decodeTree: Uint16Array, current: number, nodeIdx: number, char: number): number;
 /**
- * Decodes an HTML string, allowing for entities not terminated by a semi-colon.
+ * Decodes an HTML string.
  *
  * @param str The string to decode.
+ * @param mode The decoding mode.
  * @returns The decoded string.
  */
-export declare function decodeHTML(str: string): string;
+export declare function decodeHTML(str: string, mode?: DecodingMode): string;
 /**
- * Decodes an HTML string, requiring all entities to be terminated by a semi-colon.
+ * Decodes an HTML string in an attribute.
  *
@@ -26,5 +197,12 @@ * @param str The string to decode.
  */
+export declare function decodeHTMLAttribute(str: string): string;
+/**
+ * Decodes an HTML string, requiring all entities to be terminated by a semicolon.
+ *
+ * @param str The string to decode.
+ * @returns The decoded string.
+ */
 export declare function decodeHTMLStrict(str: string): string;
 /**
- * Decodes an XML string, requiring all entities to be terminated by a semi-colon.
+ * Decodes an XML string, requiring all entities to be terminated by a semicolon.
  *
@@ -31,0 +209,0 @@ * @param str The string to decode.

lib/esm/decode.js

 import htmlDecodeTree from "./generated/decode-data-html.js";
 import xmlDecodeTree from "./generated/decode-data-xml.js";
-import decodeCodePoint from "./decode_codepoint.js";
+import decodeCodePoint, { replaceCodePoint, fromCodePoint, } from "./decode_codepoint.js";
 // Re-export for use by eg. htmlparser2
@@ -11,2 +11,3 @@ export { htmlDecodeTree, xmlDecodeTree, decodeCodePoint };
     CharCodes[CharCodes["SEMI"] = 59] = "SEMI";
+    CharCodes[CharCodes["EQUALS"] = 61] = "EQUALS";
     CharCodes[CharCodes["ZERO"] = 48] = "ZERO";
@@ -17,5 +18,9 @@ CharCodes[CharCodes["NINE"] = 57] = "NINE";
     CharCodes[CharCodes["LOWER_X"] = 120] = "LOWER_X";
-    /** Bit that needs to be set to convert an upper case ASCII character to lower case */
-    CharCodes[CharCodes["To_LOWER_BIT"] = 32] = "To_LOWER_BIT";
+    CharCodes[CharCodes["LOWER_Z"] = 122] = "LOWER_Z";
+    CharCodes[CharCodes["UPPER_A"] = 65] = "UPPER_A";
+    CharCodes[CharCodes["UPPER_F"] = 70] = "UPPER_F";
+    CharCodes[CharCodes["UPPER_Z"] = 90] = "UPPER_Z";
 })(CharCodes || (CharCodes = {}));
+/** Bit that needs to be set to convert an upper case ASCII character to lower case */
+const TO_LOWER_BIT = 0b100000;
 export var BinTrieFlags;
@@ -27,81 +32,396 @@ (function (BinTrieFlags) {
 })(BinTrieFlags || (BinTrieFlags = {}));
-function getDecoder(decodeTree) {
-    return function decodeHTMLBinary(str, strict) {
-        let ret = "";
-        let lastIdx = 0;
-        let strIdx = 0;
-        while ((strIdx = str.indexOf("&", strIdx)) >= 0) {
-            ret += str.slice(lastIdx, strIdx);
-            lastIdx = strIdx;
-            // Skip the "&"
-            strIdx += 1;
-            // If we have a numeric entity, handle this separately.
-            if (str.charCodeAt(strIdx) === CharCodes.NUM) {
-                // Skip the leading "&#". For hex entities, also skip the leading "x".
-                let start = strIdx + 1;
-                let base = 10;
-                let cp = str.charCodeAt(start);
-                if ((cp | CharCodes.To_LOWER_BIT) === CharCodes.LOWER_X) {
-                    base = 16;
-                    strIdx += 1;
-                    start += 1;
+function isNumber(code) {
+    return code >= CharCodes.ZERO && code <= CharCodes.NINE;
+}
+function isHexadecimalCharacter(code) {
+    return ((code >= CharCodes.UPPER_A && code <= CharCodes.UPPER_F) ||
+        (code >= CharCodes.LOWER_A && code <= CharCodes.LOWER_F));
+}
+function isAsciiAlphaNumeric(code) {
+    return ((code >= CharCodes.UPPER_A && code <= CharCodes.UPPER_Z) ||
+        (code >= CharCodes.LOWER_A && code <= CharCodes.LOWER_Z) ||
+        isNumber(code));
+}
+/**
+ * Checks if the given character is a valid end character for an entity in an attribute.
+ *
+ * Attribute values that aren't terminated properly aren't parsed, and shouldn't lead to a parser error.
+ * See the example in https://html.spec.whatwg.org/multipage/parsing.html#named-character-reference-state
+ */
+function isEntityInAttributeInvalidEnd(code) {
+    return code === CharCodes.EQUALS || isAsciiAlphaNumeric(code);
+}
+var EntityDecoderState;
+(function (EntityDecoderState) {
+    EntityDecoderState[EntityDecoderState["EntityStart"] = 0] = "EntityStart";
+    EntityDecoderState[EntityDecoderState["NumericStart"] = 1] = "NumericStart";
+    EntityDecoderState[EntityDecoderState["NumericDecimal"] = 2] = "NumericDecimal";
+    EntityDecoderState[EntityDecoderState["NumericHex"] = 3] = "NumericHex";
+    EntityDecoderState[EntityDecoderState["NamedEntity"] = 4] = "NamedEntity";
+})(EntityDecoderState || (EntityDecoderState = {}));
+export var DecodingMode;
+(function (DecodingMode) {
+    /** Entities in text nodes that can end with any character. */
+    DecodingMode[DecodingMode["Legacy"] = 0] = "Legacy";
+    /** Only allow entities terminated with a semicolon. */
+    DecodingMode[DecodingMode["Strict"] = 1] = "Strict";
+    /** Entities in attributes have limitations on ending characters. */
+    DecodingMode[DecodingMode["Attribute"] = 2] = "Attribute";
+})(DecodingMode || (DecodingMode = {}));
+/**
+ * Token decoder with support of writing partial entities.
+ */
+export class EntityDecoder {
+    constructor(
+    /** The tree used to decode entities. */
+    decodeTree, 
+    /**
+     * The function that is called when a codepoint is decoded.
+     *
+     * For multi-byte named entities, this will be called multiple times,
+     * with the second codepoint, and the same `consumed` value.
+     *
+     * @param codepoint The decoded codepoint.
+     * @param consumed The number of bytes consumed by the decoder.
+     */
+    emitCodePoint, 
+    /** An object that is used to produce errors. */
+    errors) {
+        this.decodeTree = decodeTree;
+        this.emitCodePoint = emitCodePoint;
+        this.errors = errors;
+        /** The current state of the decoder. */
+        this.state = EntityDecoderState.EntityStart;
+        /** Characters that were consumed while parsing an entity. */
+        this.consumed = 1;
+        /**
+         * The result of the entity.
+         *
+         * Either the result index of a numeric entity, or the codepoint of a
+         * numeric entity.
+         */
+        this.result = 0;
+        /** The current index in the decode tree. */
+        this.treeIndex = 0;
+        /** The number of characters that were consumed in excess. */
+        this.excess = 1;
+        /** The mode in which the decoder is operating. */
+        this.decodeMode = DecodingMode.Strict;
+    }
+    /** Resets the instance to make it reusable. */
+    startEntity(decodeMode) {
+        this.decodeMode = decodeMode;
+        this.state = EntityDecoderState.EntityStart;
+        this.result = 0;
+        this.treeIndex = 0;
+        this.excess = 1;
+        this.consumed = 1;
+    }
+    /**
+     * Write an entity to the decoder. This can be called multiple times with partial entities.
+     * If the entity is incomplete, the decoder will return -1.
+     *
+     * Mirrors the implementation of `getDecoder`, but with the ability to stop decoding if the
+     * entity is incomplete, and resume when the next string is written.
+     *
+     * @param string The string containing the entity (or a continuation of the entity).
+     * @param offset The offset at which the entity begins. Should be 0 if this is not the first call.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    write(str, offset) {
+        switch (this.state) {
+            case EntityDecoderState.EntityStart: {
+                if (str.charCodeAt(offset) === CharCodes.NUM) {
+                    this.state = EntityDecoderState.NumericStart;
+                    this.consumed += 1;
+                    return this.stateNumericStart(str, offset + 1);
                 }
-                do
-                    cp = str.charCodeAt(++strIdx);
-                while ((cp >= CharCodes.ZERO && cp <= CharCodes.NINE) ||
-                    (base === 16 &&
-                        (cp | CharCodes.To_LOWER_BIT) >= CharCodes.LOWER_A &&
-                        (cp | CharCodes.To_LOWER_BIT) <= CharCodes.LOWER_F));
-                if (start !== strIdx) {
-                    const entity = str.substring(start, strIdx);
-                    const parsed = parseInt(entity, base);
-                    if (str.charCodeAt(strIdx) === CharCodes.SEMI) {
-                        strIdx += 1;
-                    }
-                    else if (strict) {
-                        continue;
-                    }
-                    ret += decodeCodePoint(parsed);
-                    lastIdx = strIdx;
-                }
-                continue;
+                this.state = EntityDecoderState.NamedEntity;
+                return this.stateNamedEntity(str, offset);
             }
-            let resultIdx = 0;
-            let excess = 1;
-            let treeIdx = 0;
-            let current = decodeTree[treeIdx];
-            for (; strIdx < str.length; strIdx++, excess++) {
-                treeIdx = determineBranch(decodeTree, current, treeIdx + 1, str.charCodeAt(strIdx));
-                if (treeIdx < 0)
-                    break;
-                current = decodeTree[treeIdx];
-                const masked = current & BinTrieFlags.VALUE_LENGTH;
-                // If the branch is a value, store it and continue
-                if (masked) {
-                    // If we have a legacy entity while parsing strictly, just skip the number of bytes
-                    if (!strict || str.charCodeAt(strIdx) === CharCodes.SEMI) {
-                        resultIdx = treeIdx;
-                        excess = 0;
-                    }
-                    // The mask is the number of bytes of the value, including the current byte.
-                    const valueLength = (masked >> 14) - 1;
-                    if (valueLength === 0)
-                        break;
-                    treeIdx += valueLength;
+            case EntityDecoderState.NumericStart: {
+                return this.stateNumericStart(str, offset);
+            }
+            case EntityDecoderState.NumericDecimal: {
+                return this.stateNumericDecimal(str, offset);
+            }
+            case EntityDecoderState.NumericHex: {
+                return this.stateNumericHex(str, offset);
+            }
+            case EntityDecoderState.NamedEntity: {
+                return this.stateNamedEntity(str, offset);
+            }
+        }
+    }
+    /**
+     * Switches between the numeric decimal and hexadecimal states.
+     *
+     * Equivalent to the `Numeric character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    stateNumericStart(str, offset) {
+        if (offset >= str.length) {
+            return -1;
+        }
+        if ((str.charCodeAt(offset) | TO_LOWER_BIT) === CharCodes.LOWER_X) {
+            this.state = EntityDecoderState.NumericHex;
+            this.consumed += 1;
+            return this.stateNumericHex(str, offset + 1);
+        }
+        this.state = EntityDecoderState.NumericDecimal;
+        return this.stateNumericDecimal(str, offset);
+    }
+    addToNumericResult(str, start, end, base) {
+        if (start !== end) {
+            const digitCount = end - start;
+            this.result =
+                this.result * Math.pow(base, digitCount) +
+                    parseInt(str.substr(start, digitCount), base);
+            this.consumed += digitCount;
+        }
+    }
+    /**
+     * Parses a hexadecimal numeric entity.
+     *
+     * Equivalent to the `Hexademical character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    stateNumericHex(str, offset) {
+        const startIdx = offset;
+        while (offset < str.length) {
+            const char = str.charCodeAt(offset);
+            if (isNumber(char) || isHexadecimalCharacter(char)) {
+                offset += 1;
+            }
+            else {
+                this.addToNumericResult(str, startIdx, offset, 16);
+                return this.emitNumericEntity(char, 3);
+            }
+        }
+        this.addToNumericResult(str, startIdx, offset, 16);
+        return -1;
+    }
+    /**
+     * Parses a decimal numeric entity.
+     *
+     * Equivalent to the `Decimal character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    stateNumericDecimal(str, offset) {
+        const startIdx = offset;
+        while (offset < str.length) {
+            const char = str.charCodeAt(offset);
+            if (isNumber(char)) {
+                offset += 1;
+            }
+            else {
+                this.addToNumericResult(str, startIdx, offset, 10);
+                return this.emitNumericEntity(char, 2);
+            }
+        }
+        this.addToNumericResult(str, startIdx, offset, 10);
+        return -1;
+    }
+    /**
+     * Validate and emit a numeric entity.
+     *
+     * Implements the logic from the `Hexademical character reference start
+     * state` and `Numeric character reference end state` in the HTML spec.
+     *
+     * @param lastCp The last code point of the entity. Used to see if the
+     *               entity was terminated with a semicolon.
+     * @param expectedLength The minimum number of characters that should be
+     *                       consumed. Used to validate that at least one digit
+     *                       was consumed.
+     * @returns The number of characters that were consumed.
+     */
+    emitNumericEntity(lastCp, expectedLength) {
+        var _a;
+        // Ensure we consumed at least one digit.
+        if (this.consumed <= expectedLength) {
+            (_a = this.errors) === null || _a === void 0 ? void 0 : _a.absenceOfDigitsInNumericCharacterReference(this.consumed);
+            return 0;
+        }
+        // Figure out if this is a legit end of the entity
+        if (lastCp === CharCodes.SEMI) {
+            this.consumed += 1;
+        }
+        else if (this.decodeMode === DecodingMode.Strict) {
+            return 0;
+        }
+        this.emitCodePoint(replaceCodePoint(this.result), this.consumed);
+        if (this.errors) {
+            if (lastCp !== CharCodes.SEMI) {
+                this.errors.missingSemicolonAfterCharacterReference();
+            }
+            this.errors.validateNumericCharacterReference(this.result);
+        }
+        return this.consumed;
+    }
+    /**
+     * Parses a named entity.
+     *
+     * Equivalent to the `Named character reference state` in the HTML spec.
+     *
+     * @param str The string containing the entity (or a continuation of the entity).
+     * @param offset The current offset.
+     * @returns The number of characters that were consumed, or -1 if the entity is incomplete.
+     */
+    stateNamedEntity(str, offset) {
+        const { decodeTree } = this;
+        let current = decodeTree[this.treeIndex];
+        // The mask is the number of bytes of the value, including the current byte.
+        let valueLength = (current & BinTrieFlags.VALUE_LENGTH) >> 14;
+        for (; offset < str.length; offset++, this.excess++) {
+            const char = str.charCodeAt(offset);
+            this.treeIndex = determineBranch(decodeTree, current, this.treeIndex + Math.max(1, valueLength), char);
+            if (this.treeIndex < 0) {
+                return this.result === 0 ||
+                    // If we are parsing an attribute
+                    (this.decodeMode === DecodingMode.Attribute &&
+                        // We shouldn't have consumed any characters after the entity,
+                        (valueLength === 0 ||
+                            // And there should be no invalid characters.
+                            isEntityInAttributeInvalidEnd(char)))
+                    ? 0
+                    : this.emitNotTerminatedNamedEntity();
+            }
+            current = decodeTree[this.treeIndex];
+            valueLength = (current & BinTrieFlags.VALUE_LENGTH) >> 14;
+            // If the branch is a value, store it and continue
+            if (valueLength !== 0) {
+                // If the entity is terminated by a semicolon, we are done.
+                if (char === CharCodes.SEMI) {
+                    return this.emitNamedEntityData(this.treeIndex, valueLength, this.consumed + this.excess);
                 }
+                // If we encounter a non-terminated (legacy) entity while parsing strictly, then ignore it.
+                if (this.decodeMode !== DecodingMode.Strict) {
+                    this.result = this.treeIndex;
+                    this.consumed += this.excess;
+                    this.excess = 0;
+                }
             }
-            if (resultIdx !== 0) {
-                const valueLength = (decodeTree[resultIdx] & BinTrieFlags.VALUE_LENGTH) >> 14;
-                ret +=
-                    valueLength === 1
-                        ? String.fromCharCode(decodeTree[resultIdx] & ~BinTrieFlags.VALUE_LENGTH)
-                        : valueLength === 2
-                            ? String.fromCharCode(decodeTree[resultIdx + 1])
-                            : String.fromCharCode(decodeTree[resultIdx + 1], decodeTree[resultIdx + 2]);
-                lastIdx = strIdx - excess + 1;
+        }
+        return -1;
+    }
+    /**
+     * Emit a named entity that was not terminated with a semicolon.
+     *
+     * @returns The number of characters consumed.
+     */
+    emitNotTerminatedNamedEntity() {
+        var _a;
+        const { result, decodeTree } = this;
+        const valueLength = (decodeTree[result] & BinTrieFlags.VALUE_LENGTH) >> 14;
+        this.emitNamedEntityData(result, valueLength, this.consumed);
+        (_a = this.errors) === null || _a === void 0 ? void 0 : _a.missingSemicolonAfterCharacterReference();
+        return this.consumed;
+    }
+    /**
+     * Emit a named entity.
+     *
+     * @param result The index of the entity in the decode tree.
+     * @param valueLength The number of bytes in the entity.
+     * @param consumed The number of characters consumed.
+     *
+     * @returns The number of characters consumed.
+     */
+    emitNamedEntityData(result, valueLength, consumed) {
+        const { decodeTree } = this;
+        this.emitCodePoint(valueLength === 1
+            ? decodeTree[result] & ~BinTrieFlags.VALUE_LENGTH
+            : decodeTree[result + 1], consumed);
+        if (valueLength === 3) {
+            // For multi-byte values, we need to emit the second byte.
+            this.emitCodePoint(decodeTree[result + 2], consumed);
+        }
+        return consumed;
+    }
+    /**
+     * Signal to the parser that the end of the input was reached.
+     *
+     * Remaining data will be emitted and relevant errors will be produced.
+     *
+     * @returns The number of characters consumed.
+     */
+    end() {
+        var _a;
+        switch (this.state) {
+            case EntityDecoderState.NamedEntity: {
+                // Emit a named entity if we have one.
+                return this.result !== 0 &&
+                    (this.decodeMode !== DecodingMode.Attribute ||
+                        this.result === this.treeIndex)
+                    ? this.emitNotTerminatedNamedEntity()
+                    : 0;
             }
+            // Otherwise, emit a numeric entity if we have one.
+            case EntityDecoderState.NumericDecimal: {
+                return this.emitNumericEntity(0, 2);
+            }
+            case EntityDecoderState.NumericHex: {
+                return this.emitNumericEntity(0, 3);
+            }
+            case EntityDecoderState.NumericStart: {
+                (_a = this.errors) === null || _a === void 0 ? void 0 : _a.absenceOfDigitsInNumericCharacterReference(this.consumed);
+                return 0;
+            }
+            case EntityDecoderState.EntityStart: {
+                // Return 0 if we have no entity.
+                return 0;
+            }
         }
-        return ret + str.slice(lastIdx);
+    }
+}
+/**
+ * Creates a function that decodes entities in a string.
+ *
+ * @param decodeTree The decode tree.
+ * @returns A function that decodes entities in a string.
+ */
+function getDecoder(decodeTree) {
+    let ret = "";
+    const decoder = new EntityDecoder(decodeTree, (str) => (ret += fromCodePoint(str)));
+    return function decodeWithTrie(str, decodeMode) {
+        let lastIndex = 0;
+        let offset = 0;
+        while ((offset = str.indexOf("&", offset)) >= 0) {
+            ret += str.slice(lastIndex, offset);
+            decoder.startEntity(decodeMode);
+            const len = decoder.write(str, 
+            // Skip the "&"
+            offset + 1);
+            if (len < 0) {
+                lastIndex = offset + decoder.end();
+                break;
+            }
+            lastIndex = offset + len;
+            // If `len` is 0, skip the current `&` and continue.
+            offset = len === 0 ? lastIndex + 1 : lastIndex;
+        }
+        const result = ret + str.slice(lastIndex);
+        // Make sure we don't keep a reference to the final string.
+        ret = "";
+        return result;
     };
 }
+/**
+ * Determines the branch of the current node that is taken given the current
+ * character. This function is used to traverse the trie.
+ *
+ * @param decodeTree The trie.
+ * @param current The current node.
+ * @param nodeIdx The index right after the current node and its value.
+ * @param char The current character.
+ * @returns The index of the next node, or -1 if no branch is taken.
+ */
 export function determineBranch(decodeTree, current, nodeIdx, char) {
@@ -143,12 +463,13 @@ const branchCount = (current & BinTrieFlags.BRANCH_LENGTH) >> 7;
 /**
- * Decodes an HTML string, allowing for entities not terminated by a semi-colon.
+ * Decodes an HTML string.
  *
  * @param str The string to decode.
+ * @param mode The decoding mode.
  * @returns The decoded string.
  */
-export function decodeHTML(str) {
-    return htmlDecoder(str, false);
+export function decodeHTML(str, mode = DecodingMode.Legacy) {
+    return htmlDecoder(str, mode);
 }
 /**
- * Decodes an HTML string, requiring all entities to be terminated by a semi-colon.
+ * Decodes an HTML string in an attribute.
  *
@@ -158,7 +479,16 @@ * @param str The string to decode.
  */
+export function decodeHTMLAttribute(str) {
+    return htmlDecoder(str, DecodingMode.Attribute);
+}
+/**
+ * Decodes an HTML string, requiring all entities to be terminated by a semicolon.
+ *
+ * @param str The string to decode.
+ * @returns The decoded string.
+ */
 export function decodeHTMLStrict(str) {
-    return htmlDecoder(str, true);
+    return htmlDecoder(str, DecodingMode.Strict);
 }
 /**
- * Decodes an XML string, requiring all entities to be terminated by a semi-colon.
+ * Decodes an XML string, requiring all entities to be terminated by a semicolon.
  *
@@ -169,4 +499,4 @@ * @param str The string to decode.
 export function decodeXML(str) {
-    return xmlDecoder(str, true);
+    return xmlDecoder(str, DecodingMode.Strict);
 }
 //# sourceMappingURL=decode.js.map

lib/esm/encode.js

@@ -55,3 +55,3 @@ import htmlTrie from "./generated/encode-html.js";
         }
-        // We might have a tree node without a value; skip and use a numeric entitiy.
+        // We might have a tree node without a value; skip and use a numeric entity.
         if (next !== undefined) {
@@ -58,0 +58,0 @@ ret += next;

lib/esm/escape.js

@@ -58,2 +58,12 @@ export const xmlReplacer = /["&'<>$\x80-\uFFFF]/g;
 export const escape = encodeXML;
+/**
+ * Creates a function that escapes all characters matched by the given regular
+ * expression using the given map of characters to escape to their entities.
+ *
+ * @param regex Regular expression to match characters to escape.
+ * @param map Map of characters to escape to their entities.
+ *
+ * @returns Function that escapes all characters matched by the given regular
+ * expression using the given map of characters to escape to their entities.
+ */
 function getEscaper(regex, map) {
@@ -68,3 +78,3 @@ return function escape(data) {
             }
-            // We know that this chararcter will be in the map.
+            // We know that this character will be in the map.
             result += map.get(match[0].charCodeAt(0));
@@ -71,0 +81,0 @@ // Every match will be of length 1

lib/esm/generated/encode-html.d.ts

@@ -1,2 +0,2 @@
-declare type EncodeTrieNode = string | {
+type EncodeTrieNode = string | {
     v?: string;
@@ -3,0 +3,0 @@ n: number | Map<number, EncodeTrieNode>;

lib/esm/index.d.ts

@@ -0,1 +1,2 @@
+import { DecodingMode } from "./decode.js";
 /** The level of entities to support. */
@@ -8,9 +9,2 @@ export declare enum EntityLevel {
 }
-/** Determines whether some entities are allowed to be written without a trailing `;`. */
-export declare enum DecodingMode {
-    /** Support legacy HTML entities. */
-    Legacy = 0,
-    /** Do not support legacy HTML entities. */
-    Strict = 1
-}
 export declare enum EncodingMode {
@@ -61,3 +55,3 @@ /**
      */
-    mode?: DecodingMode;
+    mode?: DecodingMode | undefined;
 }
@@ -103,3 +97,3 @@ /**
 export { encodeHTML, encodeNonAsciiHTML, encodeHTML as encodeHTML4, encodeHTML as encodeHTML5, } from "./encode.js";
-export { decodeXML, decodeHTML, decodeHTMLStrict, decodeHTML as decodeHTML4, decodeHTML as decodeHTML5, decodeHTMLStrict as decodeHTML4Strict, decodeHTMLStrict as decodeHTML5Strict, decodeXML as decodeXMLStrict, } from "./decode.js";
+export { EntityDecoder, DecodingMode, decodeXML, decodeHTML, decodeHTMLStrict, decodeHTMLAttribute, decodeHTML as decodeHTML4, decodeHTML as decodeHTML5, decodeHTMLStrict as decodeHTML4Strict, decodeHTMLStrict as decodeHTML5Strict, decodeXML as decodeXMLStrict, } from "./decode.js";
 //# sourceMappingURL=index.d.ts.map

lib/esm/index.js

@@ -1,2 +0,2 @@
-import { decodeXML, decodeHTML, decodeHTMLStrict } from "./decode.js";
+import { decodeXML, decodeHTML, DecodingMode } from "./decode.js";
 import { encodeHTML, encodeNonAsciiHTML } from "./encode.js";
@@ -12,10 +12,2 @@ import { encodeXML, escapeUTF8, escapeAttribute, escapeText, } from "./escape.js";
 })(EntityLevel || (EntityLevel = {}));
-/** Determines whether some entities are allowed to be written without a trailing `;`. */
-export var DecodingMode;
-(function (DecodingMode) {
-    /** Support legacy HTML entities. */
-    DecodingMode[DecodingMode["Legacy"] = 0] = "Legacy";
-    /** Do not support legacy HTML entities. */
-    DecodingMode[DecodingMode["Strict"] = 1] = "Strict";
-})(DecodingMode || (DecodingMode = {}));
 export var EncodingMode;
@@ -57,8 +49,6 @@ (function (EncodingMode) {
 export function decode(data, options = EntityLevel.XML) {
-    const opts = typeof options === "number" ? { level: options } : options;
-    if (opts.level === EntityLevel.HTML) {
-        if (opts.mode === DecodingMode.Strict) {
-            return decodeHTMLStrict(data);
-        }
-        return decodeHTML(data);
+    const level = typeof options === "number" ? options : options.level;
+    if (level === EntityLevel.HTML) {
+        const mode = typeof options === "object" ? options.mode : undefined;
+        return decodeHTML(data, mode);
     }
@@ -75,10 +65,6 @@ return decodeXML(data);
 export function decodeStrict(data, options = EntityLevel.XML) {
+    var _a;
     const opts = typeof options === "number" ? { level: options } : options;
-    if (opts.level === EntityLevel.HTML) {
-        if (opts.mode === DecodingMode.Legacy) {
-            return decodeHTML(data);
-        }
-        return decodeHTMLStrict(data);
-    }
-    return decodeXML(data);
+    (_a = opts.mode) !== null && _a !== void 0 ? _a : (opts.mode = DecodingMode.Strict);
+    return decode(data, opts);
 }
@@ -113,5 +99,5 @@ /**
 encodeHTML as encodeHTML4, encodeHTML as encodeHTML5, } from "./encode.js";
-export { decodeXML, decodeHTML, decodeHTMLStrict, 
+export { EntityDecoder, DecodingMode, decodeXML, decodeHTML, decodeHTMLStrict, decodeHTMLAttribute, 
 // Legacy aliases (deprecated)
 decodeHTML as decodeHTML4, decodeHTML as decodeHTML5, decodeHTMLStrict as decodeHTML4Strict, decodeHTMLStrict as decodeHTML5Strict, decodeXML as decodeXMLStrict, } from "./decode.js";
 //# sourceMappingURL=index.js.map

lib/generated/encode-html.d.ts

@@ -1,2 +0,2 @@
-declare type EncodeTrieNode = string | {
+type EncodeTrieNode = string | {
     v?: string;
@@ -3,0 +3,0 @@ n: number | Map<number, EncodeTrieNode>;

lib/index.d.ts

@@ -0,1 +1,2 @@
+import { DecodingMode } from "./decode.js";
 /** The level of entities to support. */
@@ -8,9 +9,2 @@ export declare enum EntityLevel {
 }
-/** Determines whether some entities are allowed to be written without a trailing `;`. */
-export declare enum DecodingMode {
-    /** Support legacy HTML entities. */
-    Legacy = 0,
-    /** Do not support legacy HTML entities. */
-    Strict = 1
-}
 export declare enum EncodingMode {
@@ -61,3 +55,3 @@ /**
      */
-    mode?: DecodingMode;
+    mode?: DecodingMode | undefined;
 }
@@ -103,3 +97,3 @@ /**
 export { encodeHTML, encodeNonAsciiHTML, encodeHTML as encodeHTML4, encodeHTML as encodeHTML5, } from "./encode.js";
-export { decodeXML, decodeHTML, decodeHTMLStrict, decodeHTML as decodeHTML4, decodeHTML as decodeHTML5, decodeHTMLStrict as decodeHTML4Strict, decodeHTMLStrict as decodeHTML5Strict, decodeXML as decodeXMLStrict, } from "./decode.js";
+export { EntityDecoder, DecodingMode, decodeXML, decodeHTML, decodeHTMLStrict, decodeHTMLAttribute, decodeHTML as decodeHTML4, decodeHTML as decodeHTML5, decodeHTMLStrict as decodeHTML4Strict, decodeHTMLStrict as decodeHTML5Strict, decodeXML as decodeXMLStrict, } from "./decode.js";
 //# sourceMappingURL=index.d.ts.map

lib/index.js

 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
-exports.decodeXMLStrict = exports.decodeHTML5Strict = exports.decodeHTML4Strict = exports.decodeHTML5 = exports.decodeHTML4 = exports.decodeHTMLStrict = exports.decodeHTML = exports.decodeXML = exports.encodeHTML5 = exports.encodeHTML4 = exports.encodeNonAsciiHTML = exports.encodeHTML = exports.escapeText = exports.escapeAttribute = exports.escapeUTF8 = exports.escape = exports.encodeXML = exports.encode = exports.decodeStrict = exports.decode = exports.EncodingMode = exports.DecodingMode = exports.EntityLevel = void 0;
+exports.decodeXMLStrict = exports.decodeHTML5Strict = exports.decodeHTML4Strict = exports.decodeHTML5 = exports.decodeHTML4 = exports.decodeHTMLAttribute = exports.decodeHTMLStrict = exports.decodeHTML = exports.decodeXML = exports.DecodingMode = exports.EntityDecoder = exports.encodeHTML5 = exports.encodeHTML4 = exports.encodeNonAsciiHTML = exports.encodeHTML = exports.escapeText = exports.escapeAttribute = exports.escapeUTF8 = exports.escape = exports.encodeXML = exports.encode = exports.decodeStrict = exports.decode = exports.EncodingMode = exports.EntityLevel = void 0;
 var decode_js_1 = require("./decode.js");
@@ -15,10 +15,2 @@ var encode_js_1 = require("./encode.js");
 })(EntityLevel = exports.EntityLevel || (exports.EntityLevel = {}));
-/** Determines whether some entities are allowed to be written without a trailing `;`. */
-var DecodingMode;
-(function (DecodingMode) {
-    /** Support legacy HTML entities. */
-    DecodingMode[DecodingMode["Legacy"] = 0] = "Legacy";
-    /** Do not support legacy HTML entities. */
-    DecodingMode[DecodingMode["Strict"] = 1] = "Strict";
-})(DecodingMode = exports.DecodingMode || (exports.DecodingMode = {}));
 var EncodingMode;
@@ -61,8 +53,6 @@ (function (EncodingMode) {
     if (options === void 0) { options = EntityLevel.XML; }
-    var opts = typeof options === "number" ? { level: options } : options;
-    if (opts.level === EntityLevel.HTML) {
-        if (opts.mode === DecodingMode.Strict) {
-            return (0, decode_js_1.decodeHTMLStrict)(data);
-        }
-        return (0, decode_js_1.decodeHTML)(data);
+    var level = typeof options === "number" ? options : options.level;
+    if (level === EntityLevel.HTML) {
+        var mode = typeof options === "object" ? options.mode : undefined;
+        return (0, decode_js_1.decodeHTML)(data, mode);
     }
@@ -80,11 +70,7 @@ return (0, decode_js_1.decodeXML)(data);
 function decodeStrict(data, options) {
+    var _a;
     if (options === void 0) { options = EntityLevel.XML; }
     var opts = typeof options === "number" ? { level: options } : options;
-    if (opts.level === EntityLevel.HTML) {
-        if (opts.mode === DecodingMode.Legacy) {
-            return (0, decode_js_1.decodeHTML)(data);
-        }
-        return (0, decode_js_1.decodeHTMLStrict)(data);
-    }
-    return (0, decode_js_1.decodeXML)(data);
+    (_a = opts.mode) !== null && _a !== void 0 ? _a : (opts.mode = decode_js_1.DecodingMode.Strict);
+    return decode(data, opts);
 }
@@ -131,5 +117,8 @@ exports.decodeStrict = decodeStrict;
 var decode_js_2 = require("./decode.js");
+Object.defineProperty(exports, "EntityDecoder", { enumerable: true, get: function () { return decode_js_2.EntityDecoder; } });
+Object.defineProperty(exports, "DecodingMode", { enumerable: true, get: function () { return decode_js_2.DecodingMode; } });
 Object.defineProperty(exports, "decodeXML", { enumerable: true, get: function () { return decode_js_2.decodeXML; } });
 Object.defineProperty(exports, "decodeHTML", { enumerable: true, get: function () { return decode_js_2.decodeHTML; } });
 Object.defineProperty(exports, "decodeHTMLStrict", { enumerable: true, get: function () { return decode_js_2.decodeHTMLStrict; } });
+Object.defineProperty(exports, "decodeHTMLAttribute", { enumerable: true, get: function () { return decode_js_2.decodeHTMLAttribute; } });
 // Legacy aliases (deprecated)
@@ -136,0 +125,0 @@ Object.defineProperty(exports, "decodeHTML4", { enumerable: true, get: function () { return decode_js_2.decodeHTML; } });

package.json

 {
     "name": "entities",
-    "version": "4.4.0",
+    "version": "4.5.0",
     "description": "Encode & decode XML and HTML entities with ease & speed",
@@ -44,13 +44,13 @@ "author": "Felix Boehm <[e-mail]>",
         "@types/jest": "^28.1.8",
-        "@types/node": "^18.7.14",
-        "@typescript-eslint/eslint-plugin": "^5.36.1",
-        "@typescript-eslint/parser": "^5.36.1",
-        "eslint": "^8.23.0",
-        "eslint-config-prettier": "^8.5.0",
+        "@types/node": "^18.15.11",
+        "@typescript-eslint/eslint-plugin": "^5.58.0",
+        "@typescript-eslint/parser": "^5.58.0",
+        "eslint": "^8.38.0",
+        "eslint-config-prettier": "^8.8.0",
         "eslint-plugin-node": "^11.1.0",
         "jest": "^28.1.3",
-        "prettier": "^2.7.1",
+        "prettier": "^2.8.7",
         "ts-jest": "^28.0.8",
-        "typedoc": "^0.23.12",
-        "typescript": "^4.8.2"
+        "typedoc": "^0.24.1",
+        "typescript": "^5.0.4"
     },
@@ -57,0 +57,0 @@ "scripts": {

readme.md

@@ -1,2 +0,2 @@
-# entities [![NPM version](http://img.shields.io/npm/v/entities.svg)](https://npmjs.org/package/entities) [![Downloads](https://img.shields.io/npm/dm/entities.svg)](https://npmjs.org/package/entities) [![Build Status](http://img.shields.io/travis/fb55/entities.svg)](http://travis-ci.org/fb55/entities) [![Coverage](http://img.shields.io/coveralls/fb55/entities.svg)](https://coveralls.io/r/fb55/entities)
+# entities [![NPM version](https://img.shields.io/npm/v/entities.svg)](https://npmjs.org/package/entities) [![Downloads](https://img.shields.io/npm/dm/entities.svg)](https://npmjs.org/package/entities) [![Node.js CI](https://github.com/fb55/entities/actions/workflows/nodejs-test.yml/badge.svg)](https://github.com/fb55/entities/actions/workflows/nodejs-test.yml)
 
@@ -12,6 +12,6 @@ Encode & decode HTML & XML entities with ease & speed.
     process HTML entities.
--   ⚡️ Fast: `entities` is the fastes library for decoding HTML entities (as of
-    April 2022); see [performance](#performance).
+-   ⚡️ Fast: `entities` is the fastest library for decoding HTML entities (as
+    of April 2022); see [performance](#performance).
 -   🎛 Configurable: Get an output tailored for your needs. You are fine with
-    UTF8? That'll safe you some bytes. Prefer to only have ASCII characters? We
+    UTF8? That'll save you some bytes. Prefer to only have ASCII characters? We
     can do that as well!
@@ -91,3 +91,3 @@
 
-This libary wouldn't be possible without the work of these individuals. Thanks
+This library wouldn't be possible without the work of these individuals. Thanks
 to
@@ -94,0 +94,0 @@

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Total package byte prevSize

412892

increased by16.86%

Lines of code

3177

increased by52.37%

No dependency changes