166 lines
No EOL
6.5 KiB
JavaScript
166 lines
No EOL
6.5 KiB
JavaScript
import htmlDecodeTree from "./generated/decode-data-html.js";
|
|
import xmlDecodeTree from "./generated/decode-data-xml.js";
|
|
import decodeCodePoint 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["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";
|
|
/** 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 = {}));
|
|
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 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;
|
|
}
|
|
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;
|
|
}
|
|
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;
|
|
}
|
|
}
|
|
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 ret + str.slice(lastIdx);
|
|
};
|
|
}
|
|
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, allowing for entities not terminated by a semi-colon.
|
|
*
|
|
* @param str The string to decode.
|
|
* @returns The decoded string.
|
|
*/
|
|
export function decodeHTML(str) {
|
|
return htmlDecoder(str, false);
|
|
}
|
|
/**
|
|
* Decodes an HTML string, requiring all entities to be terminated by a semi-colon.
|
|
*
|
|
* @param str The string to decode.
|
|
* @returns The decoded string.
|
|
*/
|
|
export function decodeHTMLStrict(str) {
|
|
return htmlDecoder(str, true);
|
|
}
|
|
/**
|
|
* Decodes an XML string, requiring all entities to be terminated by a semi-colon.
|
|
*
|
|
* @param str The string to decode.
|
|
* @returns The decoded string.
|
|
*/
|
|
export function decodeXML(str) {
|
|
return xmlDecoder(str, true);
|
|
}
|
|
//# sourceMappingURL=decode.js.map
|