822 lines
32 KiB
JavaScript
822 lines
32 KiB
JavaScript
"use strict";
|
|
var __importDefault = (this && this.__importDefault) || function (mod) {
|
|
return (mod && mod.__esModule) ? mod : { "default": mod };
|
|
};
|
|
Object.defineProperty(exports, "__esModule", { value: true });
|
|
var decode_codepoint_1 = __importDefault(require("entities/lib/decode_codepoint"));
|
|
var decode_1 = require("entities/lib/decode");
|
|
function isWhitespace(c) {
|
|
return (c === 32 /* Space */ ||
|
|
c === 10 /* NewLine */ ||
|
|
c === 9 /* Tab */ ||
|
|
c === 12 /* FormFeed */ ||
|
|
c === 13 /* CarriageReturn */);
|
|
}
|
|
function isEndOfTagSection(c) {
|
|
return c === 47 /* Slash */ || c === 62 /* Gt */ || isWhitespace(c);
|
|
}
|
|
function isNumber(c) {
|
|
return c >= 48 /* Zero */ && c <= 57 /* Nine */;
|
|
}
|
|
function isASCIIAlpha(c) {
|
|
return ((c >= 97 /* LowerA */ && c <= 122 /* LowerZ */) ||
|
|
(c >= 65 /* UpperA */ && c <= 90 /* UpperZ */));
|
|
}
|
|
/**
|
|
* Sequences used to match longer strings.
|
|
*
|
|
* We don't have `Script`, `Style`, or `Title` here. Instead, we re-use the *End
|
|
* sequences with an increased offset.
|
|
*/
|
|
var Sequences = {
|
|
Cdata: new Uint16Array([0x43, 0x44, 0x41, 0x54, 0x41, 0x5b]),
|
|
CdataEnd: new Uint16Array([0x5d, 0x5d, 0x3e]),
|
|
CommentEnd: new Uint16Array([0x2d, 0x2d, 0x3e]),
|
|
ScriptEnd: new Uint16Array([
|
|
0x3c, 0x2f, 0x73, 0x63, 0x72, 0x69, 0x70, 0x74,
|
|
]),
|
|
StyleEnd: new Uint16Array([0x3c, 0x2f, 0x73, 0x74, 0x79, 0x6c, 0x65]),
|
|
TitleEnd: new Uint16Array([0x3c, 0x2f, 0x74, 0x69, 0x74, 0x6c, 0x65]), // `</title`
|
|
};
|
|
var Tokenizer = /** @class */ (function () {
|
|
function Tokenizer(_a, cbs) {
|
|
var _b = _a.xmlMode, xmlMode = _b === void 0 ? false : _b, _c = _a.decodeEntities, decodeEntities = _c === void 0 ? true : _c;
|
|
this.cbs = cbs;
|
|
/** The current state the tokenizer is in. */
|
|
this._state = 1 /* Text */;
|
|
/** The read buffer. */
|
|
this.buffer = "";
|
|
/** The beginning of the section that is currently being read. */
|
|
this.sectionStart = 0;
|
|
/** The index within the buffer that we are currently looking at. */
|
|
this._index = 0;
|
|
/**
|
|
* Data that has already been processed will be removed from the buffer occasionally.
|
|
* `_bufferOffset` keeps track of how many characters have been removed, to make sure position information is accurate.
|
|
*/
|
|
this.bufferOffset = 0;
|
|
/** Some behavior, eg. when decoding entities, is done while we are in another state. This keeps track of the other state type. */
|
|
this.baseState = 1 /* Text */;
|
|
/** For special parsing behavior inside of script and style tags. */
|
|
this.isSpecial = false;
|
|
/** Indicates whether the tokenizer has been paused. */
|
|
this.running = true;
|
|
/** Indicates whether the tokenizer has finished running / `.end` has been called. */
|
|
this.ended = false;
|
|
this.sequenceIndex = 0;
|
|
this.trieIndex = 0;
|
|
this.trieCurrent = 0;
|
|
this.trieResult = null;
|
|
this.entityExcess = 0;
|
|
this.xmlMode = xmlMode;
|
|
this.decodeEntities = decodeEntities;
|
|
this.entityTrie = xmlMode ? decode_1.xmlDecodeTree : decode_1.htmlDecodeTree;
|
|
}
|
|
Tokenizer.prototype.reset = function () {
|
|
this._state = 1 /* Text */;
|
|
this.buffer = "";
|
|
this.sectionStart = 0;
|
|
this._index = 0;
|
|
this.bufferOffset = 0;
|
|
this.baseState = 1 /* Text */;
|
|
this.currentSequence = undefined;
|
|
this.running = true;
|
|
this.ended = false;
|
|
};
|
|
Tokenizer.prototype.write = function (chunk) {
|
|
if (this.ended)
|
|
return this.cbs.onerror(Error(".write() after done!"));
|
|
this.buffer += chunk;
|
|
this.parse();
|
|
};
|
|
Tokenizer.prototype.end = function (chunk) {
|
|
if (this.ended)
|
|
return this.cbs.onerror(Error(".end() after done!"));
|
|
if (chunk)
|
|
this.write(chunk);
|
|
this.ended = true;
|
|
if (this.running)
|
|
this.finish();
|
|
};
|
|
Tokenizer.prototype.pause = function () {
|
|
this.running = false;
|
|
};
|
|
Tokenizer.prototype.resume = function () {
|
|
this.running = true;
|
|
if (this._index < this.buffer.length) {
|
|
this.parse();
|
|
}
|
|
if (this.ended) {
|
|
this.finish();
|
|
}
|
|
};
|
|
/**
|
|
* The start of the current section.
|
|
*/
|
|
Tokenizer.prototype.getAbsoluteSectionStart = function () {
|
|
return this.sectionStart + this.bufferOffset;
|
|
};
|
|
/**
|
|
* The current index within all of the written data.
|
|
*/
|
|
Tokenizer.prototype.getAbsoluteIndex = function () {
|
|
return this.bufferOffset + this._index;
|
|
};
|
|
Tokenizer.prototype.stateText = function (c) {
|
|
if (c === 60 /* Lt */ ||
|
|
(!this.decodeEntities && this.fastForwardTo(60 /* Lt */))) {
|
|
if (this._index > this.sectionStart) {
|
|
this.cbs.ontext(this.getSection());
|
|
}
|
|
this._state = 2 /* BeforeTagName */;
|
|
this.sectionStart = this._index;
|
|
}
|
|
else if (this.decodeEntities && c === 38 /* Amp */) {
|
|
this._state = 25 /* BeforeEntity */;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateSpecialStartSequence = function (c) {
|
|
var isEnd = this.sequenceIndex === this.currentSequence.length;
|
|
var isMatch = isEnd
|
|
? // If we are at the end of the sequence, make sure the tag name has ended
|
|
isEndOfTagSection(c)
|
|
: // Otherwise, do a case-insensitive comparison
|
|
(c | 0x20) === this.currentSequence[this.sequenceIndex];
|
|
if (!isMatch) {
|
|
this.isSpecial = false;
|
|
}
|
|
else if (!isEnd) {
|
|
this.sequenceIndex++;
|
|
return;
|
|
}
|
|
this.sequenceIndex = 0;
|
|
this._state = 3 /* InTagName */;
|
|
this.stateInTagName(c);
|
|
};
|
|
/** Look for an end tag. For <title> tags, also decode entities. */
|
|
Tokenizer.prototype.stateInSpecialTag = function (c) {
|
|
if (this.sequenceIndex === this.currentSequence.length) {
|
|
if (c === 62 /* Gt */ || isWhitespace(c)) {
|
|
var endOfText = this._index - this.currentSequence.length;
|
|
if (this.sectionStart < endOfText) {
|
|
// Spoof the index so that reported locations match up.
|
|
var actualIndex = this._index;
|
|
this._index = endOfText;
|
|
this.cbs.ontext(this.getSection());
|
|
this._index = actualIndex;
|
|
}
|
|
this.isSpecial = false;
|
|
this.sectionStart = endOfText + 2; // Skip over the `</`
|
|
this.stateInClosingTagName(c);
|
|
return; // We are done; skip the rest of the function.
|
|
}
|
|
this.sequenceIndex = 0;
|
|
}
|
|
if ((c | 0x20) === this.currentSequence[this.sequenceIndex]) {
|
|
this.sequenceIndex += 1;
|
|
}
|
|
else if (this.sequenceIndex === 0) {
|
|
if (this.currentSequence === Sequences.TitleEnd) {
|
|
// We have to parse entities in <title> tags.
|
|
if (this.decodeEntities && c === 38 /* Amp */) {
|
|
this._state = 25 /* BeforeEntity */;
|
|
}
|
|
}
|
|
else if (this.fastForwardTo(60 /* Lt */)) {
|
|
// Outside of <title> tags, we can fast-forward.
|
|
this.sequenceIndex = 1;
|
|
}
|
|
}
|
|
else {
|
|
// If we see a `<`, set the sequence index to 1; useful for eg. `<</script>`.
|
|
this.sequenceIndex = Number(c === 60 /* Lt */);
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateCDATASequence = function (c) {
|
|
if (c === Sequences.Cdata[this.sequenceIndex]) {
|
|
if (++this.sequenceIndex === Sequences.Cdata.length) {
|
|
this._state = 21 /* InCommentLike */;
|
|
this.currentSequence = Sequences.CdataEnd;
|
|
this.sequenceIndex = 0;
|
|
this.sectionStart = this._index + 1;
|
|
}
|
|
}
|
|
else {
|
|
this.sequenceIndex = 0;
|
|
this._state = 16 /* InDeclaration */;
|
|
this.stateInDeclaration(c); // Reconsume the character
|
|
}
|
|
};
|
|
/**
|
|
* When we wait for one specific character, we can speed things up
|
|
* by skipping through the buffer until we find it.
|
|
*
|
|
* @returns Whether the character was found.
|
|
*/
|
|
Tokenizer.prototype.fastForwardTo = function (c) {
|
|
while (++this._index < this.buffer.length) {
|
|
if (this.buffer.charCodeAt(this._index) === c) {
|
|
return true;
|
|
}
|
|
}
|
|
/*
|
|
* We increment the index at the end of the `parse` loop,
|
|
* so set it to `buffer.length - 1` here.
|
|
*
|
|
* TODO: Refactor `parse` to increment index before calling states.
|
|
*/
|
|
this._index = this.buffer.length - 1;
|
|
return false;
|
|
};
|
|
/**
|
|
* Comments and CDATA end with `-->` and `]]>`.
|
|
*
|
|
* Their common qualities are:
|
|
* - Their end sequences have a distinct character they start with.
|
|
* - That character is then repeated, so we have to check multiple repeats.
|
|
* - All characters but the start character of the sequence can be skipped.
|
|
*/
|
|
Tokenizer.prototype.stateInCommentLike = function (c) {
|
|
if (c === this.currentSequence[this.sequenceIndex]) {
|
|
if (++this.sequenceIndex === this.currentSequence.length) {
|
|
// Remove 2 trailing chars
|
|
var section = this.buffer.slice(this.sectionStart, this._index - 2);
|
|
if (this.currentSequence === Sequences.CdataEnd) {
|
|
this.cbs.oncdata(section);
|
|
}
|
|
else {
|
|
this.cbs.oncomment(section);
|
|
}
|
|
this.sequenceIndex = 0;
|
|
this.sectionStart = this._index + 1;
|
|
this._state = 1 /* Text */;
|
|
}
|
|
}
|
|
else if (this.sequenceIndex === 0) {
|
|
// Fast-forward to the first character of the sequence
|
|
if (this.fastForwardTo(this.currentSequence[0])) {
|
|
this.sequenceIndex = 1;
|
|
}
|
|
}
|
|
else if (c !== this.currentSequence[this.sequenceIndex - 1]) {
|
|
// Allow long sequences, eg. --->, ]]]>
|
|
this.sequenceIndex = 0;
|
|
}
|
|
};
|
|
/**
|
|
* HTML only allows ASCII alpha characters (a-z and A-Z) at the beginning of a tag name.
|
|
*
|
|
* XML allows a lot more characters here (@see https://www.w3.org/TR/REC-xml/#NT-NameStartChar).
|
|
* We allow anything that wouldn't end the tag.
|
|
*/
|
|
Tokenizer.prototype.isTagStartChar = function (c) {
|
|
return this.xmlMode ? !isEndOfTagSection(c) : isASCIIAlpha(c);
|
|
};
|
|
Tokenizer.prototype.startSpecial = function (sequence, offset) {
|
|
this.isSpecial = true;
|
|
this.currentSequence = sequence;
|
|
this.sequenceIndex = offset;
|
|
this._state = 23 /* SpecialStartSequence */;
|
|
};
|
|
Tokenizer.prototype.stateBeforeTagName = function (c) {
|
|
if (c === 33 /* ExclamationMark */) {
|
|
this._state = 15 /* BeforeDeclaration */;
|
|
this.sectionStart = this._index + 1;
|
|
}
|
|
else if (c === 63 /* Questionmark */) {
|
|
this._state = 17 /* InProcessingInstruction */;
|
|
this.sectionStart = this._index + 1;
|
|
}
|
|
else if (this.isTagStartChar(c)) {
|
|
var lower = c | 0x20;
|
|
this.sectionStart = this._index;
|
|
if (!this.xmlMode && lower === Sequences.TitleEnd[2]) {
|
|
this.startSpecial(Sequences.TitleEnd, 3);
|
|
}
|
|
else {
|
|
this._state =
|
|
!this.xmlMode && lower === Sequences.ScriptEnd[2]
|
|
? 22 /* BeforeSpecialS */
|
|
: 3 /* InTagName */;
|
|
}
|
|
}
|
|
else if (c === 47 /* Slash */) {
|
|
this._state = 5 /* BeforeClosingTagName */;
|
|
}
|
|
else {
|
|
this._state = 1 /* Text */;
|
|
this.stateText(c);
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInTagName = function (c) {
|
|
if (isEndOfTagSection(c)) {
|
|
this.cbs.onopentagname(this.getSection());
|
|
this.sectionStart = -1;
|
|
this._state = 8 /* BeforeAttributeName */;
|
|
this.stateBeforeAttributeName(c);
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateBeforeClosingTagName = function (c) {
|
|
if (isWhitespace(c)) {
|
|
// Ignore
|
|
}
|
|
else if (c === 62 /* Gt */) {
|
|
this._state = 1 /* Text */;
|
|
}
|
|
else {
|
|
this._state = this.isTagStartChar(c)
|
|
? 6 /* InClosingTagName */
|
|
: 20 /* InSpecialComment */;
|
|
this.sectionStart = this._index;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInClosingTagName = function (c) {
|
|
if (c === 62 /* Gt */ || isWhitespace(c)) {
|
|
this.cbs.onclosetag(this.getSection());
|
|
this.sectionStart = -1;
|
|
this._state = 7 /* AfterClosingTagName */;
|
|
this.stateAfterClosingTagName(c);
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateAfterClosingTagName = function (c) {
|
|
// Skip everything until ">"
|
|
if (c === 62 /* Gt */ || this.fastForwardTo(62 /* Gt */)) {
|
|
this._state = 1 /* Text */;
|
|
this.sectionStart = this._index + 1;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateBeforeAttributeName = function (c) {
|
|
if (c === 62 /* Gt */) {
|
|
this.cbs.onopentagend();
|
|
if (this.isSpecial) {
|
|
this._state = 24 /* InSpecialTag */;
|
|
this.sequenceIndex = 0;
|
|
}
|
|
else {
|
|
this._state = 1 /* Text */;
|
|
}
|
|
this.baseState = this._state;
|
|
this.sectionStart = this._index + 1;
|
|
}
|
|
else if (c === 47 /* Slash */) {
|
|
this._state = 4 /* InSelfClosingTag */;
|
|
}
|
|
else if (!isWhitespace(c)) {
|
|
this._state = 9 /* InAttributeName */;
|
|
this.sectionStart = this._index;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInSelfClosingTag = function (c) {
|
|
if (c === 62 /* Gt */) {
|
|
this.cbs.onselfclosingtag();
|
|
this._state = 1 /* Text */;
|
|
this.baseState = 1 /* Text */;
|
|
this.sectionStart = this._index + 1;
|
|
this.isSpecial = false; // Reset special state, in case of self-closing special tags
|
|
}
|
|
else if (!isWhitespace(c)) {
|
|
this._state = 8 /* BeforeAttributeName */;
|
|
this.stateBeforeAttributeName(c);
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInAttributeName = function (c) {
|
|
if (c === 61 /* Eq */ || isEndOfTagSection(c)) {
|
|
this.cbs.onattribname(this.getSection());
|
|
this.sectionStart = -1;
|
|
this._state = 10 /* AfterAttributeName */;
|
|
this.stateAfterAttributeName(c);
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateAfterAttributeName = function (c) {
|
|
if (c === 61 /* Eq */) {
|
|
this._state = 11 /* BeforeAttributeValue */;
|
|
}
|
|
else if (c === 47 /* Slash */ || c === 62 /* Gt */) {
|
|
this.cbs.onattribend(undefined);
|
|
this._state = 8 /* BeforeAttributeName */;
|
|
this.stateBeforeAttributeName(c);
|
|
}
|
|
else if (!isWhitespace(c)) {
|
|
this.cbs.onattribend(undefined);
|
|
this._state = 9 /* InAttributeName */;
|
|
this.sectionStart = this._index;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateBeforeAttributeValue = function (c) {
|
|
if (c === 34 /* DoubleQuote */) {
|
|
this._state = 12 /* InAttributeValueDq */;
|
|
this.sectionStart = this._index + 1;
|
|
}
|
|
else if (c === 39 /* SingleQuote */) {
|
|
this._state = 13 /* InAttributeValueSq */;
|
|
this.sectionStart = this._index + 1;
|
|
}
|
|
else if (!isWhitespace(c)) {
|
|
this.sectionStart = this._index;
|
|
this._state = 14 /* InAttributeValueNq */;
|
|
this.stateInAttributeValueNoQuotes(c); // Reconsume token
|
|
}
|
|
};
|
|
Tokenizer.prototype.handleInAttributeValue = function (c, quote) {
|
|
if (c === quote ||
|
|
(!this.decodeEntities && this.fastForwardTo(quote))) {
|
|
this.cbs.onattribdata(this.getSection());
|
|
this.sectionStart = -1;
|
|
this.cbs.onattribend(String.fromCharCode(quote));
|
|
this._state = 8 /* BeforeAttributeName */;
|
|
}
|
|
else if (this.decodeEntities && c === 38 /* Amp */) {
|
|
this.baseState = this._state;
|
|
this._state = 25 /* BeforeEntity */;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInAttributeValueDoubleQuotes = function (c) {
|
|
this.handleInAttributeValue(c, 34 /* DoubleQuote */);
|
|
};
|
|
Tokenizer.prototype.stateInAttributeValueSingleQuotes = function (c) {
|
|
this.handleInAttributeValue(c, 39 /* SingleQuote */);
|
|
};
|
|
Tokenizer.prototype.stateInAttributeValueNoQuotes = function (c) {
|
|
if (isWhitespace(c) || c === 62 /* Gt */) {
|
|
this.cbs.onattribdata(this.getSection());
|
|
this.sectionStart = -1;
|
|
this.cbs.onattribend(null);
|
|
this._state = 8 /* BeforeAttributeName */;
|
|
this.stateBeforeAttributeName(c);
|
|
}
|
|
else if (this.decodeEntities && c === 38 /* Amp */) {
|
|
this.baseState = this._state;
|
|
this._state = 25 /* BeforeEntity */;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateBeforeDeclaration = function (c) {
|
|
if (c === 91 /* OpeningSquareBracket */) {
|
|
this._state = 19 /* CDATASequence */;
|
|
this.sequenceIndex = 0;
|
|
}
|
|
else {
|
|
this._state =
|
|
c === 45 /* Dash */
|
|
? 18 /* BeforeComment */
|
|
: 16 /* InDeclaration */;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInDeclaration = function (c) {
|
|
if (c === 62 /* Gt */ || this.fastForwardTo(62 /* Gt */)) {
|
|
this.cbs.ondeclaration(this.getSection());
|
|
this._state = 1 /* Text */;
|
|
this.sectionStart = this._index + 1;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInProcessingInstruction = function (c) {
|
|
if (c === 62 /* Gt */ || this.fastForwardTo(62 /* Gt */)) {
|
|
this.cbs.onprocessinginstruction(this.getSection());
|
|
this._state = 1 /* Text */;
|
|
this.sectionStart = this._index + 1;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateBeforeComment = function (c) {
|
|
if (c === 45 /* Dash */) {
|
|
this._state = 21 /* InCommentLike */;
|
|
this.currentSequence = Sequences.CommentEnd;
|
|
// Allow short comments (eg. <!-->)
|
|
this.sequenceIndex = 2;
|
|
this.sectionStart = this._index + 1;
|
|
}
|
|
else {
|
|
this._state = 16 /* InDeclaration */;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInSpecialComment = function (c) {
|
|
if (c === 62 /* Gt */ || this.fastForwardTo(62 /* Gt */)) {
|
|
this.cbs.oncomment(this.getSection());
|
|
this._state = 1 /* Text */;
|
|
this.sectionStart = this._index + 1;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateBeforeSpecialS = function (c) {
|
|
var lower = c | 0x20;
|
|
if (lower === Sequences.ScriptEnd[3]) {
|
|
this.startSpecial(Sequences.ScriptEnd, 4);
|
|
}
|
|
else if (lower === Sequences.StyleEnd[3]) {
|
|
this.startSpecial(Sequences.StyleEnd, 4);
|
|
}
|
|
else {
|
|
this._state = 3 /* InTagName */;
|
|
this.stateInTagName(c); // Consume the token again
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateBeforeEntity = function (c) {
|
|
// Start excess with 1 to include the '&'
|
|
this.entityExcess = 1;
|
|
if (c === 35 /* Num */) {
|
|
this._state = 26 /* BeforeNumericEntity */;
|
|
}
|
|
else if (c === 38 /* Amp */) {
|
|
// We have two `&` characters in a row. Stay in the current state.
|
|
}
|
|
else {
|
|
this.trieIndex = 0;
|
|
this.trieCurrent = this.entityTrie[0];
|
|
this.trieResult = null;
|
|
this._state = 27 /* InNamedEntity */;
|
|
this.stateInNamedEntity(c);
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInNamedEntity = function (c) {
|
|
this.entityExcess += 1;
|
|
this.trieIndex = (0, decode_1.determineBranch)(this.entityTrie, this.trieCurrent, this.trieIndex + 1, c);
|
|
if (this.trieIndex < 0) {
|
|
this.emitNamedEntity();
|
|
this._index--;
|
|
return;
|
|
}
|
|
this.trieCurrent = this.entityTrie[this.trieIndex];
|
|
// If the branch is a value, store it and continue
|
|
if (this.trieCurrent & decode_1.BinTrieFlags.HAS_VALUE) {
|
|
// If we have a legacy entity while parsing strictly, just skip the number of bytes
|
|
if (!this.allowLegacyEntity() && c !== 59 /* Semi */) {
|
|
// No need to consider multi-byte values, as the legacy entity is always a single byte
|
|
this.trieIndex += 1;
|
|
}
|
|
else {
|
|
// Add 1 as we have already incremented the excess
|
|
var entityStart = this._index - this.entityExcess + 1;
|
|
if (entityStart > this.sectionStart) {
|
|
this.emitPartial(this.buffer.substring(this.sectionStart, entityStart));
|
|
}
|
|
// If this is a surrogate pair, combine the higher bits from the node with the next byte
|
|
this.trieResult =
|
|
this.trieCurrent & decode_1.BinTrieFlags.MULTI_BYTE
|
|
? String.fromCharCode(this.entityTrie[++this.trieIndex], this.entityTrie[++this.trieIndex])
|
|
: String.fromCharCode(this.entityTrie[++this.trieIndex]);
|
|
this.entityExcess = 0;
|
|
this.sectionStart = this._index + 1;
|
|
}
|
|
}
|
|
};
|
|
Tokenizer.prototype.emitNamedEntity = function () {
|
|
if (this.trieResult) {
|
|
this.emitPartial(this.trieResult);
|
|
}
|
|
this._state = this.baseState;
|
|
};
|
|
Tokenizer.prototype.stateBeforeNumericEntity = function (c) {
|
|
if ((c | 0x20) === 120 /* LowerX */) {
|
|
this.entityExcess++;
|
|
this._state = 29 /* InHexEntity */;
|
|
}
|
|
else {
|
|
this._state = 28 /* InNumericEntity */;
|
|
this.stateInNumericEntity(c);
|
|
}
|
|
};
|
|
Tokenizer.prototype.decodeNumericEntity = function (base, strict) {
|
|
var entityStart = this._index - this.entityExcess - 1;
|
|
var numberStart = entityStart + 2 + (base >> 4);
|
|
if (numberStart !== this._index) {
|
|
// Emit leading data if any
|
|
if (entityStart > this.sectionStart) {
|
|
this.emitPartial(this.buffer.substring(this.sectionStart, entityStart));
|
|
}
|
|
// Parse entity
|
|
var entity = this.buffer.substring(numberStart, this._index);
|
|
var parsed = parseInt(entity, base);
|
|
this.emitPartial((0, decode_codepoint_1.default)(parsed));
|
|
this.sectionStart = this._index + Number(strict);
|
|
}
|
|
this._state = this.baseState;
|
|
};
|
|
Tokenizer.prototype.stateInNumericEntity = function (c) {
|
|
if (c === 59 /* Semi */) {
|
|
this.decodeNumericEntity(10, true);
|
|
}
|
|
else if (!isNumber(c)) {
|
|
if (this.allowLegacyEntity()) {
|
|
this.decodeNumericEntity(10, false);
|
|
}
|
|
else {
|
|
this._state = this.baseState;
|
|
}
|
|
this._index--;
|
|
}
|
|
else {
|
|
this.entityExcess++;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInHexEntity = function (c) {
|
|
if (c === 59 /* Semi */) {
|
|
this.decodeNumericEntity(16, true);
|
|
}
|
|
else if ((c < 97 /* LowerA */ || c > 102 /* LowerF */) &&
|
|
(c < 65 /* UpperA */ || c > 70 /* UpperF */) &&
|
|
!isNumber(c)) {
|
|
if (this.allowLegacyEntity()) {
|
|
this.decodeNumericEntity(16, false);
|
|
}
|
|
else {
|
|
this._state = this.baseState;
|
|
}
|
|
this._index--;
|
|
}
|
|
else {
|
|
this.entityExcess++;
|
|
}
|
|
};
|
|
Tokenizer.prototype.allowLegacyEntity = function () {
|
|
return (!this.xmlMode &&
|
|
(this.baseState === 1 /* Text */ ||
|
|
this.baseState === 24 /* InSpecialTag */));
|
|
};
|
|
/**
|
|
* Remove data that has already been consumed from the buffer.
|
|
*/
|
|
Tokenizer.prototype.cleanup = function () {
|
|
// If we are inside of text, emit what we already have.
|
|
if (this.running &&
|
|
this.sectionStart !== this._index &&
|
|
(this._state === 1 /* Text */ ||
|
|
(this._state === 24 /* InSpecialTag */ &&
|
|
this.sequenceIndex === 0))) {
|
|
// TODO: We could emit attribute data here as well.
|
|
this.cbs.ontext(this.buffer.substr(this.sectionStart));
|
|
this.sectionStart = this._index;
|
|
}
|
|
var start = this.sectionStart < 0 ? this._index : this.sectionStart;
|
|
this.buffer =
|
|
start === this.buffer.length ? "" : this.buffer.substr(start);
|
|
this._index -= start;
|
|
this.bufferOffset += start;
|
|
if (this.sectionStart > 0) {
|
|
this.sectionStart = 0;
|
|
}
|
|
};
|
|
Tokenizer.prototype.shouldContinue = function () {
|
|
return this._index < this.buffer.length && this.running;
|
|
};
|
|
/**
|
|
* Iterates through the buffer, calling the function corresponding to the current state.
|
|
*
|
|
* States that are more likely to be hit are higher up, as a performance improvement.
|
|
*/
|
|
Tokenizer.prototype.parse = function () {
|
|
while (this.shouldContinue()) {
|
|
var c = this.buffer.charCodeAt(this._index);
|
|
if (this._state === 1 /* Text */) {
|
|
this.stateText(c);
|
|
}
|
|
else if (this._state === 23 /* SpecialStartSequence */) {
|
|
this.stateSpecialStartSequence(c);
|
|
}
|
|
else if (this._state === 24 /* InSpecialTag */) {
|
|
this.stateInSpecialTag(c);
|
|
}
|
|
else if (this._state === 19 /* CDATASequence */) {
|
|
this.stateCDATASequence(c);
|
|
}
|
|
else if (this._state === 12 /* InAttributeValueDq */) {
|
|
this.stateInAttributeValueDoubleQuotes(c);
|
|
}
|
|
else if (this._state === 9 /* InAttributeName */) {
|
|
this.stateInAttributeName(c);
|
|
}
|
|
else if (this._state === 21 /* InCommentLike */) {
|
|
this.stateInCommentLike(c);
|
|
}
|
|
else if (this._state === 20 /* InSpecialComment */) {
|
|
this.stateInSpecialComment(c);
|
|
}
|
|
else if (this._state === 8 /* BeforeAttributeName */) {
|
|
this.stateBeforeAttributeName(c);
|
|
}
|
|
else if (this._state === 3 /* InTagName */) {
|
|
this.stateInTagName(c);
|
|
}
|
|
else if (this._state === 6 /* InClosingTagName */) {
|
|
this.stateInClosingTagName(c);
|
|
}
|
|
else if (this._state === 2 /* BeforeTagName */) {
|
|
this.stateBeforeTagName(c);
|
|
}
|
|
else if (this._state === 10 /* AfterAttributeName */) {
|
|
this.stateAfterAttributeName(c);
|
|
}
|
|
else if (this._state === 13 /* InAttributeValueSq */) {
|
|
this.stateInAttributeValueSingleQuotes(c);
|
|
}
|
|
else if (this._state === 11 /* BeforeAttributeValue */) {
|
|
this.stateBeforeAttributeValue(c);
|
|
}
|
|
else if (this._state === 5 /* BeforeClosingTagName */) {
|
|
this.stateBeforeClosingTagName(c);
|
|
}
|
|
else if (this._state === 7 /* AfterClosingTagName */) {
|
|
this.stateAfterClosingTagName(c);
|
|
}
|
|
else if (this._state === 22 /* BeforeSpecialS */) {
|
|
this.stateBeforeSpecialS(c);
|
|
}
|
|
else if (this._state === 14 /* InAttributeValueNq */) {
|
|
this.stateInAttributeValueNoQuotes(c);
|
|
}
|
|
else if (this._state === 4 /* InSelfClosingTag */) {
|
|
this.stateInSelfClosingTag(c);
|
|
}
|
|
else if (this._state === 16 /* InDeclaration */) {
|
|
this.stateInDeclaration(c);
|
|
}
|
|
else if (this._state === 15 /* BeforeDeclaration */) {
|
|
this.stateBeforeDeclaration(c);
|
|
}
|
|
else if (this._state === 18 /* BeforeComment */) {
|
|
this.stateBeforeComment(c);
|
|
}
|
|
else if (this._state === 17 /* InProcessingInstruction */) {
|
|
this.stateInProcessingInstruction(c);
|
|
}
|
|
else if (this._state === 27 /* InNamedEntity */) {
|
|
this.stateInNamedEntity(c);
|
|
}
|
|
else if (this._state === 25 /* BeforeEntity */) {
|
|
this.stateBeforeEntity(c);
|
|
}
|
|
else if (this._state === 29 /* InHexEntity */) {
|
|
this.stateInHexEntity(c);
|
|
}
|
|
else if (this._state === 28 /* InNumericEntity */) {
|
|
this.stateInNumericEntity(c);
|
|
}
|
|
else {
|
|
// `this._state === State.BeforeNumericEntity`
|
|
this.stateBeforeNumericEntity(c);
|
|
}
|
|
this._index++;
|
|
}
|
|
this.cleanup();
|
|
};
|
|
Tokenizer.prototype.finish = function () {
|
|
if (this._state === 27 /* InNamedEntity */) {
|
|
this.emitNamedEntity();
|
|
}
|
|
// If there is remaining data, emit it in a reasonable way
|
|
if (this.sectionStart < this._index) {
|
|
this.handleTrailingData();
|
|
}
|
|
this.cbs.onend();
|
|
};
|
|
/** Handle any trailing data. */
|
|
Tokenizer.prototype.handleTrailingData = function () {
|
|
var data = this.buffer.substr(this.sectionStart);
|
|
if (this._state === 21 /* InCommentLike */) {
|
|
if (this.currentSequence === Sequences.CdataEnd) {
|
|
this.cbs.oncdata(data);
|
|
}
|
|
else {
|
|
this.cbs.oncomment(data);
|
|
}
|
|
}
|
|
else if (this._state === 28 /* InNumericEntity */ &&
|
|
this.allowLegacyEntity()) {
|
|
this.decodeNumericEntity(10, false);
|
|
// All trailing data will have been consumed
|
|
}
|
|
else if (this._state === 29 /* InHexEntity */ &&
|
|
this.allowLegacyEntity()) {
|
|
this.decodeNumericEntity(16, false);
|
|
// All trailing data will have been consumed
|
|
}
|
|
else if (this._state === 3 /* InTagName */ ||
|
|
this._state === 8 /* BeforeAttributeName */ ||
|
|
this._state === 11 /* BeforeAttributeValue */ ||
|
|
this._state === 10 /* AfterAttributeName */ ||
|
|
this._state === 9 /* InAttributeName */ ||
|
|
this._state === 13 /* InAttributeValueSq */ ||
|
|
this._state === 12 /* InAttributeValueDq */ ||
|
|
this._state === 14 /* InAttributeValueNq */ ||
|
|
this._state === 6 /* InClosingTagName */) {
|
|
/*
|
|
* If we are currently in an opening or closing tag, us not calling the
|
|
* respective callback signals that the tag should be ignored.
|
|
*/
|
|
}
|
|
else {
|
|
this.cbs.ontext(data);
|
|
}
|
|
};
|
|
Tokenizer.prototype.getSection = function () {
|
|
return this.buffer.substring(this.sectionStart, this._index);
|
|
};
|
|
Tokenizer.prototype.emitPartial = function (value) {
|
|
if (this.baseState !== 1 /* Text */ &&
|
|
this.baseState !== 24 /* InSpecialTag */) {
|
|
this.cbs.onattribdata(value);
|
|
}
|
|
else {
|
|
this.cbs.ontext(value);
|
|
}
|
|
};
|
|
return Tokenizer;
|
|
}());
|
|
exports.default = Tokenizer;
|