parse5
Version:
HTML parser and serializer.
1,273 lines (1,272 loc) • 116 kB
JavaScript
"use strict";
Object.defineProperty(exports, "__esModule", { value: true });
exports.Tokenizer = exports.TokenizerMode = void 0;
const preprocessor_js_1 = require("./preprocessor.js");
const unicode_js_1 = require("../common/unicode.js");
const token_js_1 = require("../common/token.js");
const decode_js_1 = require("entities/lib/decode.js");
const error_codes_js_1 = require("../common/error-codes.js");
const html_js_1 = require("../common/html.js");
//C1 Unicode control character reference replacements
const C1_CONTROLS_REFERENCE_REPLACEMENTS = new Map([
[0x80, 8364],
[0x82, 8218],
[0x83, 402],
[0x84, 8222],
[0x85, 8230],
[0x86, 8224],
[0x87, 8225],
[0x88, 710],
[0x89, 8240],
[0x8a, 352],
[0x8b, 8249],
[0x8c, 338],
[0x8e, 381],
[0x91, 8216],
[0x92, 8217],
[0x93, 8220],
[0x94, 8221],
[0x95, 8226],
[0x96, 8211],
[0x97, 8212],
[0x98, 732],
[0x99, 8482],
[0x9a, 353],
[0x9b, 8250],
[0x9c, 339],
[0x9e, 382],
[0x9f, 376],
]);
//States
var State;
(function (State) {
State[State["DATA"] = 0] = "DATA";
State[State["RCDATA"] = 1] = "RCDATA";
State[State["RAWTEXT"] = 2] = "RAWTEXT";
State[State["SCRIPT_DATA"] = 3] = "SCRIPT_DATA";
State[State["PLAINTEXT"] = 4] = "PLAINTEXT";
State[State["TAG_OPEN"] = 5] = "TAG_OPEN";
State[State["END_TAG_OPEN"] = 6] = "END_TAG_OPEN";
State[State["TAG_NAME"] = 7] = "TAG_NAME";
State[State["RCDATA_LESS_THAN_SIGN"] = 8] = "RCDATA_LESS_THAN_SIGN";
State[State["RCDATA_END_TAG_OPEN"] = 9] = "RCDATA_END_TAG_OPEN";
State[State["RCDATA_END_TAG_NAME"] = 10] = "RCDATA_END_TAG_NAME";
State[State["RAWTEXT_LESS_THAN_SIGN"] = 11] = "RAWTEXT_LESS_THAN_SIGN";
State[State["RAWTEXT_END_TAG_OPEN"] = 12] = "RAWTEXT_END_TAG_OPEN";
State[State["RAWTEXT_END_TAG_NAME"] = 13] = "RAWTEXT_END_TAG_NAME";
State[State["SCRIPT_DATA_LESS_THAN_SIGN"] = 14] = "SCRIPT_DATA_LESS_THAN_SIGN";
State[State["SCRIPT_DATA_END_TAG_OPEN"] = 15] = "SCRIPT_DATA_END_TAG_OPEN";
State[State["SCRIPT_DATA_END_TAG_NAME"] = 16] = "SCRIPT_DATA_END_TAG_NAME";
State[State["SCRIPT_DATA_ESCAPE_START"] = 17] = "SCRIPT_DATA_ESCAPE_START";
State[State["SCRIPT_DATA_ESCAPE_START_DASH"] = 18] = "SCRIPT_DATA_ESCAPE_START_DASH";
State[State["SCRIPT_DATA_ESCAPED"] = 19] = "SCRIPT_DATA_ESCAPED";
State[State["SCRIPT_DATA_ESCAPED_DASH"] = 20] = "SCRIPT_DATA_ESCAPED_DASH";
State[State["SCRIPT_DATA_ESCAPED_DASH_DASH"] = 21] = "SCRIPT_DATA_ESCAPED_DASH_DASH";
State[State["SCRIPT_DATA_ESCAPED_LESS_THAN_SIGN"] = 22] = "SCRIPT_DATA_ESCAPED_LESS_THAN_SIGN";
State[State["SCRIPT_DATA_ESCAPED_END_TAG_OPEN"] = 23] = "SCRIPT_DATA_ESCAPED_END_TAG_OPEN";
State[State["SCRIPT_DATA_ESCAPED_END_TAG_NAME"] = 24] = "SCRIPT_DATA_ESCAPED_END_TAG_NAME";
State[State["SCRIPT_DATA_DOUBLE_ESCAPE_START"] = 25] = "SCRIPT_DATA_DOUBLE_ESCAPE_START";
State[State["SCRIPT_DATA_DOUBLE_ESCAPED"] = 26] = "SCRIPT_DATA_DOUBLE_ESCAPED";
State[State["SCRIPT_DATA_DOUBLE_ESCAPED_DASH"] = 27] = "SCRIPT_DATA_DOUBLE_ESCAPED_DASH";
State[State["SCRIPT_DATA_DOUBLE_ESCAPED_DASH_DASH"] = 28] = "SCRIPT_DATA_DOUBLE_ESCAPED_DASH_DASH";
State[State["SCRIPT_DATA_DOUBLE_ESCAPED_LESS_THAN_SIGN"] = 29] = "SCRIPT_DATA_DOUBLE_ESCAPED_LESS_THAN_SIGN";
State[State["SCRIPT_DATA_DOUBLE_ESCAPE_END"] = 30] = "SCRIPT_DATA_DOUBLE_ESCAPE_END";
State[State["BEFORE_ATTRIBUTE_NAME"] = 31] = "BEFORE_ATTRIBUTE_NAME";
State[State["ATTRIBUTE_NAME"] = 32] = "ATTRIBUTE_NAME";
State[State["AFTER_ATTRIBUTE_NAME"] = 33] = "AFTER_ATTRIBUTE_NAME";
State[State["BEFORE_ATTRIBUTE_VALUE"] = 34] = "BEFORE_ATTRIBUTE_VALUE";
State[State["ATTRIBUTE_VALUE_DOUBLE_QUOTED"] = 35] = "ATTRIBUTE_VALUE_DOUBLE_QUOTED";
State[State["ATTRIBUTE_VALUE_SINGLE_QUOTED"] = 36] = "ATTRIBUTE_VALUE_SINGLE_QUOTED";
State[State["ATTRIBUTE_VALUE_UNQUOTED"] = 37] = "ATTRIBUTE_VALUE_UNQUOTED";
State[State["AFTER_ATTRIBUTE_VALUE_QUOTED"] = 38] = "AFTER_ATTRIBUTE_VALUE_QUOTED";
State[State["SELF_CLOSING_START_TAG"] = 39] = "SELF_CLOSING_START_TAG";
State[State["BOGUS_COMMENT"] = 40] = "BOGUS_COMMENT";
State[State["MARKUP_DECLARATION_OPEN"] = 41] = "MARKUP_DECLARATION_OPEN";
State[State["COMMENT_START"] = 42] = "COMMENT_START";
State[State["COMMENT_START_DASH"] = 43] = "COMMENT_START_DASH";
State[State["COMMENT"] = 44] = "COMMENT";
State[State["COMMENT_LESS_THAN_SIGN"] = 45] = "COMMENT_LESS_THAN_SIGN";
State[State["COMMENT_LESS_THAN_SIGN_BANG"] = 46] = "COMMENT_LESS_THAN_SIGN_BANG";
State[State["COMMENT_LESS_THAN_SIGN_BANG_DASH"] = 47] = "COMMENT_LESS_THAN_SIGN_BANG_DASH";
State[State["COMMENT_LESS_THAN_SIGN_BANG_DASH_DASH"] = 48] = "COMMENT_LESS_THAN_SIGN_BANG_DASH_DASH";
State[State["COMMENT_END_DASH"] = 49] = "COMMENT_END_DASH";
State[State["COMMENT_END"] = 50] = "COMMENT_END";
State[State["COMMENT_END_BANG"] = 51] = "COMMENT_END_BANG";
State[State["DOCTYPE"] = 52] = "DOCTYPE";
State[State["BEFORE_DOCTYPE_NAME"] = 53] = "BEFORE_DOCTYPE_NAME";
State[State["DOCTYPE_NAME"] = 54] = "DOCTYPE_NAME";
State[State["AFTER_DOCTYPE_NAME"] = 55] = "AFTER_DOCTYPE_NAME";
State[State["AFTER_DOCTYPE_PUBLIC_KEYWORD"] = 56] = "AFTER_DOCTYPE_PUBLIC_KEYWORD";
State[State["BEFORE_DOCTYPE_PUBLIC_IDENTIFIER"] = 57] = "BEFORE_DOCTYPE_PUBLIC_IDENTIFIER";
State[State["DOCTYPE_PUBLIC_IDENTIFIER_DOUBLE_QUOTED"] = 58] = "DOCTYPE_PUBLIC_IDENTIFIER_DOUBLE_QUOTED";
State[State["DOCTYPE_PUBLIC_IDENTIFIER_SINGLE_QUOTED"] = 59] = "DOCTYPE_PUBLIC_IDENTIFIER_SINGLE_QUOTED";
State[State["AFTER_DOCTYPE_PUBLIC_IDENTIFIER"] = 60] = "AFTER_DOCTYPE_PUBLIC_IDENTIFIER";
State[State["BETWEEN_DOCTYPE_PUBLIC_AND_SYSTEM_IDENTIFIERS"] = 61] = "BETWEEN_DOCTYPE_PUBLIC_AND_SYSTEM_IDENTIFIERS";
State[State["AFTER_DOCTYPE_SYSTEM_KEYWORD"] = 62] = "AFTER_DOCTYPE_SYSTEM_KEYWORD";
State[State["BEFORE_DOCTYPE_SYSTEM_IDENTIFIER"] = 63] = "BEFORE_DOCTYPE_SYSTEM_IDENTIFIER";
State[State["DOCTYPE_SYSTEM_IDENTIFIER_DOUBLE_QUOTED"] = 64] = "DOCTYPE_SYSTEM_IDENTIFIER_DOUBLE_QUOTED";
State[State["DOCTYPE_SYSTEM_IDENTIFIER_SINGLE_QUOTED"] = 65] = "DOCTYPE_SYSTEM_IDENTIFIER_SINGLE_QUOTED";
State[State["AFTER_DOCTYPE_SYSTEM_IDENTIFIER"] = 66] = "AFTER_DOCTYPE_SYSTEM_IDENTIFIER";
State[State["BOGUS_DOCTYPE"] = 67] = "BOGUS_DOCTYPE";
State[State["CDATA_SECTION"] = 68] = "CDATA_SECTION";
State[State["CDATA_SECTION_BRACKET"] = 69] = "CDATA_SECTION_BRACKET";
State[State["CDATA_SECTION_END"] = 70] = "CDATA_SECTION_END";
State[State["CHARACTER_REFERENCE"] = 71] = "CHARACTER_REFERENCE";
State[State["NAMED_CHARACTER_REFERENCE"] = 72] = "NAMED_CHARACTER_REFERENCE";
State[State["AMBIGUOUS_AMPERSAND"] = 73] = "AMBIGUOUS_AMPERSAND";
State[State["NUMERIC_CHARACTER_REFERENCE"] = 74] = "NUMERIC_CHARACTER_REFERENCE";
State[State["HEXADEMICAL_CHARACTER_REFERENCE_START"] = 75] = "HEXADEMICAL_CHARACTER_REFERENCE_START";
State[State["HEXADEMICAL_CHARACTER_REFERENCE"] = 76] = "HEXADEMICAL_CHARACTER_REFERENCE";
State[State["DECIMAL_CHARACTER_REFERENCE"] = 77] = "DECIMAL_CHARACTER_REFERENCE";
State[State["NUMERIC_CHARACTER_REFERENCE_END"] = 78] = "NUMERIC_CHARACTER_REFERENCE_END";
})(State || (State = {}));
//Tokenizer initial states for different modes
exports.TokenizerMode = {
DATA: State.DATA,
RCDATA: State.RCDATA,
RAWTEXT: State.RAWTEXT,
SCRIPT_DATA: State.SCRIPT_DATA,
PLAINTEXT: State.PLAINTEXT,
CDATA_SECTION: State.CDATA_SECTION,
};
//Utils
//OPTIMIZATION: these utility functions should not be moved out of this module. V8 Crankshaft will not inline
//this functions if they will be situated in another module due to context switch.
//Always perform inlining check before modifying this functions ('node --trace-inlining').
function isAsciiDigit(cp) {
return cp >= unicode_js_1.CODE_POINTS.DIGIT_0 && cp <= unicode_js_1.CODE_POINTS.DIGIT_9;
}
function isAsciiUpper(cp) {
return cp >= unicode_js_1.CODE_POINTS.LATIN_CAPITAL_A && cp <= unicode_js_1.CODE_POINTS.LATIN_CAPITAL_Z;
}
function isAsciiLower(cp) {
return cp >= unicode_js_1.CODE_POINTS.LATIN_SMALL_A && cp <= unicode_js_1.CODE_POINTS.LATIN_SMALL_Z;
}
function isAsciiLetter(cp) {
return isAsciiLower(cp) || isAsciiUpper(cp);
}
function isAsciiAlphaNumeric(cp) {
return isAsciiLetter(cp) || isAsciiDigit(cp);
}
function isAsciiUpperHexDigit(cp) {
return cp >= unicode_js_1.CODE_POINTS.LATIN_CAPITAL_A && cp <= unicode_js_1.CODE_POINTS.LATIN_CAPITAL_F;
}
function isAsciiLowerHexDigit(cp) {
return cp >= unicode_js_1.CODE_POINTS.LATIN_SMALL_A && cp <= unicode_js_1.CODE_POINTS.LATIN_SMALL_F;
}
function isAsciiHexDigit(cp) {
return isAsciiDigit(cp) || isAsciiUpperHexDigit(cp) || isAsciiLowerHexDigit(cp);
}
function toAsciiLower(cp) {
return cp + 32;
}
function isWhitespace(cp) {
return cp === unicode_js_1.CODE_POINTS.SPACE || cp === unicode_js_1.CODE_POINTS.LINE_FEED || cp === unicode_js_1.CODE_POINTS.TABULATION || cp === unicode_js_1.CODE_POINTS.FORM_FEED;
}
function isEntityInAttributeInvalidEnd(nextCp) {
return nextCp === unicode_js_1.CODE_POINTS.EQUALS_SIGN || isAsciiAlphaNumeric(nextCp);
}
function isScriptDataDoubleEscapeSequenceEnd(cp) {
return isWhitespace(cp) || cp === unicode_js_1.CODE_POINTS.SOLIDUS || cp === unicode_js_1.CODE_POINTS.GREATER_THAN_SIGN;
}
//Tokenizer
class Tokenizer {
constructor(options, handler) {
this.options = options;
this.handler = handler;
this.paused = false;
/** Ensures that the parsing loop isn't run multiple times at once. */
this.inLoop = false;
/**
* Indicates that the current adjusted node exists, is not an element in the HTML namespace,
* and that it is not an integration point for either MathML or HTML.
*
* @see {@link https://html.spec.whatwg.org/multipage/parsing.html#tree-construction}
*/
this.inForeignNode = false;
this.lastStartTagName = '';
this.active = false;
this.state = State.DATA;
this.returnState = State.DATA;
this.charRefCode = -1;
this.consumedAfterSnapshot = -1;
this.currentCharacterToken = null;
this.currentToken = null;
this.currentAttr = { name: '', value: '' };
this.preprocessor = new preprocessor_js_1.Preprocessor(handler);
this.currentLocation = this.getCurrentLocation(-1);
}
//Errors
_err(code) {
var _a, _b;
(_b = (_a = this.handler).onParseError) === null || _b === void 0 ? void 0 : _b.call(_a, this.preprocessor.getError(code));
}
// NOTE: `offset` may never run across line boundaries.
getCurrentLocation(offset) {
if (!this.options.sourceCodeLocationInfo) {
return null;
}
return {
startLine: this.preprocessor.line,
startCol: this.preprocessor.col - offset,
startOffset: this.preprocessor.offset - offset,
endLine: -1,
endCol: -1,
endOffset: -1,
};
}
_runParsingLoop() {
if (this.inLoop)
return;
this.inLoop = true;
while (this.active && !this.paused) {
this.consumedAfterSnapshot = 0;
const cp = this._consume();
if (!this._ensureHibernation()) {
this._callState(cp);
}
}
this.inLoop = false;
}
//API
pause() {
this.paused = true;
}
resume(writeCallback) {
if (!this.paused) {
throw new Error('Parser was already resumed');
}
this.paused = false;
// Necessary for synchronous resume.
if (this.inLoop)
return;
this._runParsingLoop();
if (!this.paused) {
writeCallback === null || writeCallback === void 0 ? void 0 : writeCallback();
}
}
write(chunk, isLastChunk, writeCallback) {
this.active = true;
this.preprocessor.write(chunk, isLastChunk);
this._runParsingLoop();
if (!this.paused) {
writeCallback === null || writeCallback === void 0 ? void 0 : writeCallback();
}
}
insertHtmlAtCurrentPos(chunk) {
this.active = true;
this.preprocessor.insertHtmlAtCurrentPos(chunk);
this._runParsingLoop();
}
//Hibernation
_ensureHibernation() {
if (this.preprocessor.endOfChunkHit) {
this._unconsume(this.consumedAfterSnapshot);
this.active = false;
return true;
}
return false;
}
//Consumption
_consume() {
this.consumedAfterSnapshot++;
return this.preprocessor.advance();
}
_unconsume(count) {
this.consumedAfterSnapshot -= count;
this.preprocessor.retreat(count);
}
_reconsumeInState(state, cp) {
this.state = state;
this._callState(cp);
}
_advanceBy(count) {
this.consumedAfterSnapshot += count;
for (let i = 0; i < count; i++) {
this.preprocessor.advance();
}
}
_consumeSequenceIfMatch(pattern, caseSensitive) {
if (this.preprocessor.startsWith(pattern, caseSensitive)) {
// We will already have consumed one character before calling this method.
this._advanceBy(pattern.length - 1);
return true;
}
return false;
}
//Token creation
_createStartTagToken() {
this.currentToken = {
type: token_js_1.TokenType.START_TAG,
tagName: '',
tagID: html_js_1.TAG_ID.UNKNOWN,
selfClosing: false,
ackSelfClosing: false,
attrs: [],
location: this.getCurrentLocation(1),
};
}
_createEndTagToken() {
this.currentToken = {
type: token_js_1.TokenType.END_TAG,
tagName: '',
tagID: html_js_1.TAG_ID.UNKNOWN,
selfClosing: false,
ackSelfClosing: false,
attrs: [],
location: this.getCurrentLocation(2),
};
}
_createCommentToken(offset) {
this.currentToken = {
type: token_js_1.TokenType.COMMENT,
data: '',
location: this.getCurrentLocation(offset),
};
}
_createDoctypeToken(initialName) {
this.currentToken = {
type: token_js_1.TokenType.DOCTYPE,
name: initialName,
forceQuirks: false,
publicId: null,
systemId: null,
location: this.currentLocation,
};
}
_createCharacterToken(type, chars) {
this.currentCharacterToken = {
type,
chars,
location: this.currentLocation,
};
}
//Tag attributes
_createAttr(attrNameFirstCh) {
this.currentAttr = {
name: attrNameFirstCh,
value: '',
};
this.currentLocation = this.getCurrentLocation(0);
}
_leaveAttrName() {
var _a;
var _b;
const token = this.currentToken;
if ((0, token_js_1.getTokenAttr)(token, this.currentAttr.name) === null) {
token.attrs.push(this.currentAttr);
if (token.location && this.currentLocation) {
const attrLocations = ((_a = (_b = token.location).attrs) !== null && _a !== void 0 ? _a : (_b.attrs = Object.create(null)));
attrLocations[this.currentAttr.name] = this.currentLocation;
// Set end location
this._leaveAttrValue();
}
}
else {
this._err(error_codes_js_1.ERR.duplicateAttribute);
}
}
_leaveAttrValue() {
if (this.currentLocation) {
this.currentLocation.endLine = this.preprocessor.line;
this.currentLocation.endCol = this.preprocessor.col;
this.currentLocation.endOffset = this.preprocessor.offset;
}
}
//Token emission
prepareToken(ct) {
this._emitCurrentCharacterToken(ct.location);
this.currentToken = null;
if (ct.location) {
ct.location.endLine = this.preprocessor.line;
ct.location.endCol = this.preprocessor.col + 1;
ct.location.endOffset = this.preprocessor.offset + 1;
}
this.currentLocation = this.getCurrentLocation(-1);
}
emitCurrentTagToken() {
const ct = this.currentToken;
this.prepareToken(ct);
ct.tagID = (0, html_js_1.getTagID)(ct.tagName);
if (ct.type === token_js_1.TokenType.START_TAG) {
this.lastStartTagName = ct.tagName;
this.handler.onStartTag(ct);
}
else {
if (ct.attrs.length > 0) {
this._err(error_codes_js_1.ERR.endTagWithAttributes);
}
if (ct.selfClosing) {
this._err(error_codes_js_1.ERR.endTagWithTrailingSolidus);
}
this.handler.onEndTag(ct);
}
this.preprocessor.dropParsedChunk();
}
emitCurrentComment(ct) {
this.prepareToken(ct);
this.handler.onComment(ct);
this.preprocessor.dropParsedChunk();
}
emitCurrentDoctype(ct) {
this.prepareToken(ct);
this.handler.onDoctype(ct);
this.preprocessor.dropParsedChunk();
}
_emitCurrentCharacterToken(nextLocation) {
if (this.currentCharacterToken) {
//NOTE: if we have a pending character token, make it's end location equal to the
//current token's start location.
if (nextLocation && this.currentCharacterToken.location) {
this.currentCharacterToken.location.endLine = nextLocation.startLine;
this.currentCharacterToken.location.endCol = nextLocation.startCol;
this.currentCharacterToken.location.endOffset = nextLocation.startOffset;
}
switch (this.currentCharacterToken.type) {
case token_js_1.TokenType.CHARACTER: {
this.handler.onCharacter(this.currentCharacterToken);
break;
}
case token_js_1.TokenType.NULL_CHARACTER: {
this.handler.onNullCharacter(this.currentCharacterToken);
break;
}
case token_js_1.TokenType.WHITESPACE_CHARACTER: {
this.handler.onWhitespaceCharacter(this.currentCharacterToken);
break;
}
}
this.currentCharacterToken = null;
}
}
_emitEOFToken() {
const location = this.getCurrentLocation(0);
if (location) {
location.endLine = location.startLine;
location.endCol = location.startCol;
location.endOffset = location.startOffset;
}
this._emitCurrentCharacterToken(location);
this.handler.onEof({ type: token_js_1.TokenType.EOF, location });
this.active = false;
}
//Characters emission
//OPTIMIZATION: specification uses only one type of character tokens (one token per character).
//This causes a huge memory overhead and a lot of unnecessary parser loops. parse5 uses 3 groups of characters.
//If we have a sequence of characters that belong to the same group, the parser can process it
//as a single solid character token.
//So, there are 3 types of character tokens in parse5:
//1)TokenType.NULL_CHARACTER - \u0000-character sequences (e.g. '\u0000\u0000\u0000')
//2)TokenType.WHITESPACE_CHARACTER - any whitespace/new-line character sequences (e.g. '\n \r\t \f')
//3)TokenType.CHARACTER - any character sequence which don't belong to groups 1 and 2 (e.g. 'abcdef1234@@#$%^')
_appendCharToCurrentCharacterToken(type, ch) {
if (this.currentCharacterToken) {
if (this.currentCharacterToken.type !== type) {
this.currentLocation = this.getCurrentLocation(0);
this._emitCurrentCharacterToken(this.currentLocation);
this.preprocessor.dropParsedChunk();
}
else {
this.currentCharacterToken.chars += ch;
return;
}
}
this._createCharacterToken(type, ch);
}
_emitCodePoint(cp) {
const type = isWhitespace(cp)
? token_js_1.TokenType.WHITESPACE_CHARACTER
: cp === unicode_js_1.CODE_POINTS.NULL
? token_js_1.TokenType.NULL_CHARACTER
: token_js_1.TokenType.CHARACTER;
this._appendCharToCurrentCharacterToken(type, String.fromCodePoint(cp));
}
//NOTE: used when we emit characters explicitly.
//This is always for non-whitespace and non-null characters, which allows us to avoid additional checks.
_emitChars(ch) {
this._appendCharToCurrentCharacterToken(token_js_1.TokenType.CHARACTER, ch);
}
// Character reference helpers
_matchNamedCharacterReference(cp) {
let result = null;
let excess = 0;
let withoutSemicolon = false;
for (let i = 0, current = decode_js_1.htmlDecodeTree[0]; i >= 0; cp = this._consume()) {
i = (0, decode_js_1.determineBranch)(decode_js_1.htmlDecodeTree, current, i + 1, cp);
if (i < 0)
break;
excess += 1;
current = decode_js_1.htmlDecodeTree[i];
const masked = current & decode_js_1.BinTrieFlags.VALUE_LENGTH;
// If the branch is a value, store it and continue
if (masked) {
// The mask is the number of bytes of the value, including the current byte.
const valueLength = (masked >> 14) - 1;
// 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
if (cp !== unicode_js_1.CODE_POINTS.SEMICOLON &&
this._isCharacterReferenceInAttribute() &&
isEntityInAttributeInvalidEnd(this.preprocessor.peek(1))) {
//NOTE: we don't flush all consumed code points here, and instead switch back to the original state after
//emitting an ampersand. This is fine, as alphanumeric characters won't be parsed differently in attributes.
result = [unicode_js_1.CODE_POINTS.AMPERSAND];
// Skip over the value.
i += valueLength;
}
else {
// If this is a surrogate pair, consume the next two bytes.
result =
valueLength === 0
? [decode_js_1.htmlDecodeTree[i] & ~decode_js_1.BinTrieFlags.VALUE_LENGTH]
: valueLength === 1
? [decode_js_1.htmlDecodeTree[++i]]
: [decode_js_1.htmlDecodeTree[++i], decode_js_1.htmlDecodeTree[++i]];
excess = 0;
withoutSemicolon = cp !== unicode_js_1.CODE_POINTS.SEMICOLON;
}
if (valueLength === 0) {
// If the value is zero-length, we're done.
this._consume();
break;
}
}
}
this._unconsume(excess);
if (withoutSemicolon && !this.preprocessor.endOfChunkHit) {
this._err(error_codes_js_1.ERR.missingSemicolonAfterCharacterReference);
}
// We want to emit the error above on the code point after the entity.
// We always consume one code point too many in the loop, and we wait to
// unconsume it until after the error is emitted.
this._unconsume(1);
return result;
}
_isCharacterReferenceInAttribute() {
return (this.returnState === State.ATTRIBUTE_VALUE_DOUBLE_QUOTED ||
this.returnState === State.ATTRIBUTE_VALUE_SINGLE_QUOTED ||
this.returnState === State.ATTRIBUTE_VALUE_UNQUOTED);
}
_flushCodePointConsumedAsCharacterReference(cp) {
if (this._isCharacterReferenceInAttribute()) {
this.currentAttr.value += String.fromCodePoint(cp);
}
else {
this._emitCodePoint(cp);
}
}
// Calling states this way turns out to be much faster than any other approach.
_callState(cp) {
switch (this.state) {
case State.DATA: {
this._stateData(cp);
break;
}
case State.RCDATA: {
this._stateRcdata(cp);
break;
}
case State.RAWTEXT: {
this._stateRawtext(cp);
break;
}
case State.SCRIPT_DATA: {
this._stateScriptData(cp);
break;
}
case State.PLAINTEXT: {
this._statePlaintext(cp);
break;
}
case State.TAG_OPEN: {
this._stateTagOpen(cp);
break;
}
case State.END_TAG_OPEN: {
this._stateEndTagOpen(cp);
break;
}
case State.TAG_NAME: {
this._stateTagName(cp);
break;
}
case State.RCDATA_LESS_THAN_SIGN: {
this._stateRcdataLessThanSign(cp);
break;
}
case State.RCDATA_END_TAG_OPEN: {
this._stateRcdataEndTagOpen(cp);
break;
}
case State.RCDATA_END_TAG_NAME: {
this._stateRcdataEndTagName(cp);
break;
}
case State.RAWTEXT_LESS_THAN_SIGN: {
this._stateRawtextLessThanSign(cp);
break;
}
case State.RAWTEXT_END_TAG_OPEN: {
this._stateRawtextEndTagOpen(cp);
break;
}
case State.RAWTEXT_END_TAG_NAME: {
this._stateRawtextEndTagName(cp);
break;
}
case State.SCRIPT_DATA_LESS_THAN_SIGN: {
this._stateScriptDataLessThanSign(cp);
break;
}
case State.SCRIPT_DATA_END_TAG_OPEN: {
this._stateScriptDataEndTagOpen(cp);
break;
}
case State.SCRIPT_DATA_END_TAG_NAME: {
this._stateScriptDataEndTagName(cp);
break;
}
case State.SCRIPT_DATA_ESCAPE_START: {
this._stateScriptDataEscapeStart(cp);
break;
}
case State.SCRIPT_DATA_ESCAPE_START_DASH: {
this._stateScriptDataEscapeStartDash(cp);
break;
}
case State.SCRIPT_DATA_ESCAPED: {
this._stateScriptDataEscaped(cp);
break;
}
case State.SCRIPT_DATA_ESCAPED_DASH: {
this._stateScriptDataEscapedDash(cp);
break;
}
case State.SCRIPT_DATA_ESCAPED_DASH_DASH: {
this._stateScriptDataEscapedDashDash(cp);
break;
}
case State.SCRIPT_DATA_ESCAPED_LESS_THAN_SIGN: {
this._stateScriptDataEscapedLessThanSign(cp);
break;
}
case State.SCRIPT_DATA_ESCAPED_END_TAG_OPEN: {
this._stateScriptDataEscapedEndTagOpen(cp);
break;
}
case State.SCRIPT_DATA_ESCAPED_END_TAG_NAME: {
this._stateScriptDataEscapedEndTagName(cp);
break;
}
case State.SCRIPT_DATA_DOUBLE_ESCAPE_START: {
this._stateScriptDataDoubleEscapeStart(cp);
break;
}
case State.SCRIPT_DATA_DOUBLE_ESCAPED: {
this._stateScriptDataDoubleEscaped(cp);
break;
}
case State.SCRIPT_DATA_DOUBLE_ESCAPED_DASH: {
this._stateScriptDataDoubleEscapedDash(cp);
break;
}
case State.SCRIPT_DATA_DOUBLE_ESCAPED_DASH_DASH: {
this._stateScriptDataDoubleEscapedDashDash(cp);
break;
}
case State.SCRIPT_DATA_DOUBLE_ESCAPED_LESS_THAN_SIGN: {
this._stateScriptDataDoubleEscapedLessThanSign(cp);
break;
}
case State.SCRIPT_DATA_DOUBLE_ESCAPE_END: {
this._stateScriptDataDoubleEscapeEnd(cp);
break;
}
case State.BEFORE_ATTRIBUTE_NAME: {
this._stateBeforeAttributeName(cp);
break;
}
case State.ATTRIBUTE_NAME: {
this._stateAttributeName(cp);
break;
}
case State.AFTER_ATTRIBUTE_NAME: {
this._stateAfterAttributeName(cp);
break;
}
case State.BEFORE_ATTRIBUTE_VALUE: {
this._stateBeforeAttributeValue(cp);
break;
}
case State.ATTRIBUTE_VALUE_DOUBLE_QUOTED: {
this._stateAttributeValueDoubleQuoted(cp);
break;
}
case State.ATTRIBUTE_VALUE_SINGLE_QUOTED: {
this._stateAttributeValueSingleQuoted(cp);
break;
}
case State.ATTRIBUTE_VALUE_UNQUOTED: {
this._stateAttributeValueUnquoted(cp);
break;
}
case State.AFTER_ATTRIBUTE_VALUE_QUOTED: {
this._stateAfterAttributeValueQuoted(cp);
break;
}
case State.SELF_CLOSING_START_TAG: {
this._stateSelfClosingStartTag(cp);
break;
}
case State.BOGUS_COMMENT: {
this._stateBogusComment(cp);
break;
}
case State.MARKUP_DECLARATION_OPEN: {
this._stateMarkupDeclarationOpen(cp);
break;
}
case State.COMMENT_START: {
this._stateCommentStart(cp);
break;
}
case State.COMMENT_START_DASH: {
this._stateCommentStartDash(cp);
break;
}
case State.COMMENT: {
this._stateComment(cp);
break;
}
case State.COMMENT_LESS_THAN_SIGN: {
this._stateCommentLessThanSign(cp);
break;
}
case State.COMMENT_LESS_THAN_SIGN_BANG: {
this._stateCommentLessThanSignBang(cp);
break;
}
case State.COMMENT_LESS_THAN_SIGN_BANG_DASH: {
this._stateCommentLessThanSignBangDash(cp);
break;
}
case State.COMMENT_LESS_THAN_SIGN_BANG_DASH_DASH: {
this._stateCommentLessThanSignBangDashDash(cp);
break;
}
case State.COMMENT_END_DASH: {
this._stateCommentEndDash(cp);
break;
}
case State.COMMENT_END: {
this._stateCommentEnd(cp);
break;
}
case State.COMMENT_END_BANG: {
this._stateCommentEndBang(cp);
break;
}
case State.DOCTYPE: {
this._stateDoctype(cp);
break;
}
case State.BEFORE_DOCTYPE_NAME: {
this._stateBeforeDoctypeName(cp);
break;
}
case State.DOCTYPE_NAME: {
this._stateDoctypeName(cp);
break;
}
case State.AFTER_DOCTYPE_NAME: {
this._stateAfterDoctypeName(cp);
break;
}
case State.AFTER_DOCTYPE_PUBLIC_KEYWORD: {
this._stateAfterDoctypePublicKeyword(cp);
break;
}
case State.BEFORE_DOCTYPE_PUBLIC_IDENTIFIER: {
this._stateBeforeDoctypePublicIdentifier(cp);
break;
}
case State.DOCTYPE_PUBLIC_IDENTIFIER_DOUBLE_QUOTED: {
this._stateDoctypePublicIdentifierDoubleQuoted(cp);
break;
}
case State.DOCTYPE_PUBLIC_IDENTIFIER_SINGLE_QUOTED: {
this._stateDoctypePublicIdentifierSingleQuoted(cp);
break;
}
case State.AFTER_DOCTYPE_PUBLIC_IDENTIFIER: {
this._stateAfterDoctypePublicIdentifier(cp);
break;
}
case State.BETWEEN_DOCTYPE_PUBLIC_AND_SYSTEM_IDENTIFIERS: {
this._stateBetweenDoctypePublicAndSystemIdentifiers(cp);
break;
}
case State.AFTER_DOCTYPE_SYSTEM_KEYWORD: {
this._stateAfterDoctypeSystemKeyword(cp);
break;
}
case State.BEFORE_DOCTYPE_SYSTEM_IDENTIFIER: {
this._stateBeforeDoctypeSystemIdentifier(cp);
break;
}
case State.DOCTYPE_SYSTEM_IDENTIFIER_DOUBLE_QUOTED: {
this._stateDoctypeSystemIdentifierDoubleQuoted(cp);
break;
}
case State.DOCTYPE_SYSTEM_IDENTIFIER_SINGLE_QUOTED: {
this._stateDoctypeSystemIdentifierSingleQuoted(cp);
break;
}
case State.AFTER_DOCTYPE_SYSTEM_IDENTIFIER: {
this._stateAfterDoctypeSystemIdentifier(cp);
break;
}
case State.BOGUS_DOCTYPE: {
this._stateBogusDoctype(cp);
break;
}
case State.CDATA_SECTION: {
this._stateCdataSection(cp);
break;
}
case State.CDATA_SECTION_BRACKET: {
this._stateCdataSectionBracket(cp);
break;
}
case State.CDATA_SECTION_END: {
this._stateCdataSectionEnd(cp);
break;
}
case State.CHARACTER_REFERENCE: {
this._stateCharacterReference(cp);
break;
}
case State.NAMED_CHARACTER_REFERENCE: {
this._stateNamedCharacterReference(cp);
break;
}
case State.AMBIGUOUS_AMPERSAND: {
this._stateAmbiguousAmpersand(cp);
break;
}
case State.NUMERIC_CHARACTER_REFERENCE: {
this._stateNumericCharacterReference(cp);
break;
}
case State.HEXADEMICAL_CHARACTER_REFERENCE_START: {
this._stateHexademicalCharacterReferenceStart(cp);
break;
}
case State.HEXADEMICAL_CHARACTER_REFERENCE: {
this._stateHexademicalCharacterReference(cp);
break;
}
case State.DECIMAL_CHARACTER_REFERENCE: {
this._stateDecimalCharacterReference(cp);
break;
}
case State.NUMERIC_CHARACTER_REFERENCE_END: {
this._stateNumericCharacterReferenceEnd(cp);
break;
}
default: {
throw new Error('Unknown state');
}
}
}
// State machine
// Data state
//------------------------------------------------------------------
_stateData(cp) {
switch (cp) {
case unicode_js_1.CODE_POINTS.LESS_THAN_SIGN: {
this.state = State.TAG_OPEN;
break;
}
case unicode_js_1.CODE_POINTS.AMPERSAND: {
this.returnState = State.DATA;
this.state = State.CHARACTER_REFERENCE;
break;
}
case unicode_js_1.CODE_POINTS.NULL: {
this._err(error_codes_js_1.ERR.unexpectedNullCharacter);
this._emitCodePoint(cp);
break;
}
case unicode_js_1.CODE_POINTS.EOF: {
this._emitEOFToken();
break;
}
default: {
this._emitCodePoint(cp);
}
}
}
// RCDATA state
//------------------------------------------------------------------
_stateRcdata(cp) {
switch (cp) {
case unicode_js_1.CODE_POINTS.AMPERSAND: {
this.returnState = State.RCDATA;
this.state = State.CHARACTER_REFERENCE;
break;
}
case unicode_js_1.CODE_POINTS.LESS_THAN_SIGN: {
this.state = State.RCDATA_LESS_THAN_SIGN;
break;
}
case unicode_js_1.CODE_POINTS.NULL: {
this._err(error_codes_js_1.ERR.unexpectedNullCharacter);
this._emitChars(unicode_js_1.REPLACEMENT_CHARACTER);
break;
}
case unicode_js_1.CODE_POINTS.EOF: {
this._emitEOFToken();
break;
}
default: {
this._emitCodePoint(cp);
}
}
}
// RAWTEXT state
//------------------------------------------------------------------
_stateRawtext(cp) {
switch (cp) {
case unicode_js_1.CODE_POINTS.LESS_THAN_SIGN: {
this.state = State.RAWTEXT_LESS_THAN_SIGN;
break;
}
case unicode_js_1.CODE_POINTS.NULL: {
this._err(error_codes_js_1.ERR.unexpectedNullCharacter);
this._emitChars(unicode_js_1.REPLACEMENT_CHARACTER);
break;
}
case unicode_js_1.CODE_POINTS.EOF: {
this._emitEOFToken();
break;
}
default: {
this._emitCodePoint(cp);
}
}
}
// Script data state
//------------------------------------------------------------------
_stateScriptData(cp) {
switch (cp) {
case unicode_js_1.CODE_POINTS.LESS_THAN_SIGN: {
this.state = State.SCRIPT_DATA_LESS_THAN_SIGN;
break;
}
case unicode_js_1.CODE_POINTS.NULL: {
this._err(error_codes_js_1.ERR.unexpectedNullCharacter);
this._emitChars(unicode_js_1.REPLACEMENT_CHARACTER);
break;
}
case unicode_js_1.CODE_POINTS.EOF: {
this._emitEOFToken();
break;
}
default: {
this._emitCodePoint(cp);
}
}
}
// PLAINTEXT state
//------------------------------------------------------------------
_statePlaintext(cp) {
switch (cp) {
case unicode_js_1.CODE_POINTS.NULL: {
this._err(error_codes_js_1.ERR.unexpectedNullCharacter);
this._emitChars(unicode_js_1.REPLACEMENT_CHARACTER);
break;
}
case unicode_js_1.CODE_POINTS.EOF: {
this._emitEOFToken();
break;
}
default: {
this._emitCodePoint(cp);
}
}
}
// Tag open state
//------------------------------------------------------------------
_stateTagOpen(cp) {
if (isAsciiLetter(cp)) {
this._createStartTagToken();
this.state = State.TAG_NAME;
this._stateTagName(cp);
}
else
switch (cp) {
case unicode_js_1.CODE_POINTS.EXCLAMATION_MARK: {
this.state = State.MARKUP_DECLARATION_OPEN;
break;
}
case unicode_js_1.CODE_POINTS.SOLIDUS: {
this.state = State.END_TAG_OPEN;
break;
}
case unicode_js_1.CODE_POINTS.QUESTION_MARK: {
this._err(error_codes_js_1.ERR.unexpectedQuestionMarkInsteadOfTagName);
this._createCommentToken(1);
this.state = State.BOGUS_COMMENT;
this._stateBogusComment(cp);
break;
}
case unicode_js_1.CODE_POINTS.EOF: {
this._err(error_codes_js_1.ERR.eofBeforeTagName);
this._emitChars('<');
this._emitEOFToken();
break;
}
default: {
this._err(error_codes_js_1.ERR.invalidFirstCharacterOfTagName);
this._emitChars('<');
this.state = State.DATA;
this._stateData(cp);
}
}
}
// End tag open state
//------------------------------------------------------------------
_stateEndTagOpen(cp) {
if (isAsciiLetter(cp)) {
this._createEndTagToken();
this.state = State.TAG_NAME;
this._stateTagName(cp);
}
else
switch (cp) {
case unicode_js_1.CODE_POINTS.GREATER_THAN_SIGN: {
this._err(error_codes_js_1.ERR.missingEndTagName);
this.state = State.DATA;
break;
}
case unicode_js_1.CODE_POINTS.EOF: {
this._err(error_codes_js_1.ERR.eofBeforeTagName);
this._emitChars('</');
this._emitEOFToken();
break;
}
default: {
this._err(error_codes_js_1.ERR.invalidFirstCharacterOfTagName);
this._createCommentToken(2);
this.state = State.BOGUS_COMMENT;
this._stateBogusComment(cp);
}
}
}
// Tag name state
//------------------------------------------------------------------
_stateTagName(cp) {
const token = this.currentToken;
switch (cp) {
case unicode_js_1.CODE_POINTS.SPACE:
case unicode_js_1.CODE_POINTS.LINE_FEED:
case unicode_js_1.CODE_POINTS.TABULATION:
case unicode_js_1.CODE_POINTS.FORM_FEED: {
this.state = State.BEFORE_ATTRIBUTE_NAME;
break;
}
case unicode_js_1.CODE_POINTS.SOLIDUS: {
this.state = State.SELF_CLOSING_START_TAG;
break;
}
case unicode_js_1.CODE_POINTS.GREATER_THAN_SIGN: {
this.state = State.DATA;
this.emitCurrentTagToken();
break;
}
case unicode_js_1.CODE_POINTS.NULL: {
this._err(error_codes_js_1.ERR.unexpectedNullCharacter);
token.tagName += unicode_js_1.REPLACEMENT_CHARACTER;
break;
}
case unicode_js_1.CODE_POINTS.EOF: {
this._err(error_codes_js_1.ERR.eofInTag);
this._emitEOFToken();
break;
}
default: {
token.tagName += String.fromCodePoint(isAsciiUpper(cp) ? toAsciiLower(cp) : cp);
}
}
}
// RCDATA less-than sign state
//------------------------------------------------------------------
_stateRcdataLessThanSign(cp) {
if (cp === unicode_js_1.CODE_POINTS.SOLIDUS) {
this.state = State.RCDATA_END_TAG_OPEN;
}
else {
this._emitChars('<');
this.state = State.RCDATA;
this._stateRcdata(cp);
}
}
// RCDATA end tag open state
//------------------------------------------------------------------
_stateRcdataEndTagOpen(cp) {
if (isAsciiLetter(cp)) {
this.state = State.RCDATA_END_TAG_NAME;
this._stateRcdataEndTagName(cp);
}
else {
this._emitChars('</');
this.state = State.RCDATA;
this._stateRcdata(cp);
}
}
handleSpecialEndTag(_cp) {
if (!this.preprocessor.startsWith(this.lastStartTagName, false)) {
return !this._ensureHibernation();
}
this._createEndTagToken();
const token = this.currentToken;
token.tagName = this.lastStartTagName;
const cp = this.preprocessor.peek(this.lastStartTagName.length);
switch (cp) {
case unicode_js_1.CODE_POINTS.SPACE:
case unicode_js_1.CODE_POINTS.LINE_FEED:
case unicode_js_1.CODE_POINTS.TABULATION:
case unicode_js_1.CODE_POINTS.FORM_FEED: {
this._advanceBy(this.lastStartTagName.length);
this.state = State.BEFORE_ATTRIBUTE_NAME;
return false;
}
case unicode_js_1.CODE_POINTS.SOLIDUS: {
this._advanceBy(this.lastStartTagName.length);
this.state = State.SELF_CLOSING_START_TAG;
return false;
}
case unicode_js_1.CODE_POINTS.GREATER_THAN_SIGN: {
this._advanceBy(this.lastStartTagName.length);
this.emitCurrentTagToken();
this.state = State.DATA;
return false;
}
default: {
return !this._ensureHibernation();
}
}
}
// RCDATA end tag name state
//------------------------------------------------------------------
_stateRcdataEndTagName(cp) {
if (this.handleSpecialEndTag(cp)) {
this._emitChars('</');
this.state = State.RCDATA;
this._stateRcdata(cp);
}
}
// RAWTEXT less-than sign state
//------------------------------------------------------------------
_stateRawtextLessThanSign(cp) {
if (cp === unicode_js_1.CODE_POINTS.SOLIDUS) {
this.state = State.RAWTEXT_END_TAG_OPEN;
}
else {
this._emitChars('<');
this.state = State.RAWTEXT;
this._stateRawtext(cp);
}
}
// RAWTEXT end tag open state
//------------------------------------------------------------------
_stateRawtextEndTagOpen(cp) {
if (isAsciiLetter(cp)) {
this.state = State.RAWTEXT_END_TAG_NAME;
this._stateRawtextEndTagName(cp);
}
else {
this._emitChars('</');
this.state = State.RAWTEXT;
this._stateRawtext(cp);
}
}
// RAWTEXT end tag name state
//------------------------------------------------------------------
_stateRawtextEndTagName(cp) {
if (this.handleSpecialEndTag(cp)) {
this._emitChars('</');
this.state = State.RAWTEXT;
this._stateRawtext(cp);
}
}
// Script data less-than sign state
//------------------------------------------------------------------
_stateScriptDataLessThanSign(cp) {
switch (cp) {
case unicode_js_1.CODE_POINTS.SOLIDUS: {
this.state = State.SCRIPT_DATA_END_TAG_OPEN;
break;
}
case unicode_js_1.CODE_POINTS.EXCLAMATION_MARK: {
this.state = State.SCRIPT_DATA_ESCAPE_START;
this._emitChars('<!');
break;
}
default: {
this._emitChars('<');
this.state = State.SCRIPT_DATA;
this._stateScriptData(cp);
}
}
}
// Script data end tag open state
//------------------------------------------------------------------
_stateScriptDataEndTagOpen(cp) {
if (isAsciiLetter(cp)) {
this.state = State.SCRIPT_DATA_END_TAG_NAME;
this._stateScriptDataEndTagName(cp);
}
else {
this._emitChars('</');
this.state = State.SCRIPT_DATA;
this._stateScriptData(cp);
}
}
// Script data end tag name state
//------------------------------------------------------------------
_stateScriptDataEndTagName(cp) {
if (this.handleSpecialEndTag(cp)) {
this._emitChars('</');
this.state = State.SCRIPT_DATA;
this._stateScriptData(cp);
}
}
// Script data escape start state
//------------------------------------------------------------------
_stateScriptDataEscapeStart(cp) {
if (cp === unicode_js_1.CODE_POINTS.HYPHEN_MINUS) {
this.state = State.SCRIPT_DATA_ESCAPE_START_DASH;
this._emitChars('-');
}
else {
this.state = State.SCRIPT_DATA;
this._stateScriptData(cp);
}
}
// Script data escape start dash state
//------------------------------------------------------------------
_stateScriptDataEscapeStartDash(cp) {
if (cp === unicode_js_1.CODE_POINTS.HYPHEN_MINUS) {
this.state = State.SCRIPT_DATA_ESCAPED_DASH_DASH;
this._emitChars('-');
}
else {
this.state = State.SCRIPT_DATA;
this._stateScriptData(cp);
}
}
// Script data escaped state
//------------------------------------------------------------------
_stateScriptDataEscaped(cp) {
switch (cp) {