package.lib.esm.decode.js Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of entities Show documentation
Show all versions of entities Show documentation
Encode & decode XML and HTML entities with ease & speed
import htmlDecodeTree from "./generated/decode-data-html.js";
import xmlDecodeTree from "./generated/decode-data-xml.js";
import decodeCodePoint, { replaceCodePoint, fromCodePoint, } from "./decode_codepoint.js";
// Re-export for use by eg. htmlparser2
export { htmlDecodeTree, xmlDecodeTree, decodeCodePoint };
export { replaceCodePoint, fromCodePoint } from "./decode_codepoint.js";
var CharCodes;
(function (CharCodes) {
CharCodes[CharCodes["NUM"] = 35] = "NUM";
CharCodes[CharCodes["SEMI"] = 59] = "SEMI";
CharCodes[CharCodes["EQUALS"] = 61] = "EQUALS";
CharCodes[CharCodes["ZERO"] = 48] = "ZERO";
CharCodes[CharCodes["NINE"] = 57] = "NINE";
CharCodes[CharCodes["LOWER_A"] = 97] = "LOWER_A";
CharCodes[CharCodes["LOWER_F"] = 102] = "LOWER_F";
CharCodes[CharCodes["LOWER_X"] = 120] = "LOWER_X";
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;
(function (BinTrieFlags) {
BinTrieFlags[BinTrieFlags["VALUE_LENGTH"] = 49152] = "VALUE_LENGTH";
BinTrieFlags[BinTrieFlags["BRANCH_LENGTH"] = 16256] = "BRANCH_LENGTH";
BinTrieFlags[BinTrieFlags["JUMP_TABLE"] = 127] = "JUMP_TABLE";
})(BinTrieFlags || (BinTrieFlags = {}));
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);
}
this.state = EntityDecoderState.NamedEntity;
return this.stateNamedEntity(str, offset);
}
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;
}
}
}
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;
}
}
}
}
/**
* 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) {
const branchCount = (current & BinTrieFlags.BRANCH_LENGTH) >> 7;
const jumpOffset = current & BinTrieFlags.JUMP_TABLE;
// Case 1: Single branch encoded in jump offset
if (branchCount === 0) {
return jumpOffset !== 0 && char === jumpOffset ? nodeIdx : -1;
}
// Case 2: Multiple branches encoded in jump table
if (jumpOffset) {
const value = char - jumpOffset;
return value < 0 || value >= branchCount
? -1
: decodeTree[nodeIdx + value] - 1;
}
// Case 3: Multiple branches encoded in dictionary
// Binary search for the character.
let lo = nodeIdx;
let hi = lo + branchCount - 1;
while (lo <= hi) {
const mid = (lo + hi) >>> 1;
const midVal = decodeTree[mid];
if (midVal < char) {
lo = mid + 1;
}
else if (midVal > char) {
hi = mid - 1;
}
else {
return decodeTree[mid + branchCount];
}
}
return -1;
}
const htmlDecoder = getDecoder(htmlDecodeTree);
const xmlDecoder = getDecoder(xmlDecodeTree);
/**
* Decodes an HTML string.
*
* @param str The string to decode.
* @param mode The decoding mode.
* @returns The decoded string.
*/
export function decodeHTML(str, mode = DecodingMode.Legacy) {
return htmlDecoder(str, mode);
}
/**
* Decodes an HTML string in an attribute.
*
* @param str The string to decode.
* @returns The decoded string.
*/
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, DecodingMode.Strict);
}
/**
* Decodes an XML string, requiring all entities to be terminated by a semicolon.
*
* @param str The string to decode.
* @returns The decoded string.
*/
export function decodeXML(str) {
return xmlDecoder(str, DecodingMode.Strict);
}
//# sourceMappingURL=decode.js.map