938 lines
36 KiB
JavaScript
938 lines
36 KiB
JavaScript
"use strict";
|
|
Object.defineProperty(exports, "__esModule", { value: true });
|
|
exports.QuoteType = void 0;
|
|
var decode_js_1 = require("entities/lib/decode.js");
|
|
var CharCodes;
|
|
(function (CharCodes) {
|
|
CharCodes[CharCodes["Tab"] = 9] = "Tab";
|
|
CharCodes[CharCodes["NewLine"] = 10] = "NewLine";
|
|
CharCodes[CharCodes["FormFeed"] = 12] = "FormFeed";
|
|
CharCodes[CharCodes["CarriageReturn"] = 13] = "CarriageReturn";
|
|
CharCodes[CharCodes["Space"] = 32] = "Space";
|
|
CharCodes[CharCodes["ExclamationMark"] = 33] = "ExclamationMark";
|
|
CharCodes[CharCodes["Number"] = 35] = "Number";
|
|
CharCodes[CharCodes["Amp"] = 38] = "Amp";
|
|
CharCodes[CharCodes["SingleQuote"] = 39] = "SingleQuote";
|
|
CharCodes[CharCodes["DoubleQuote"] = 34] = "DoubleQuote";
|
|
CharCodes[CharCodes["Dash"] = 45] = "Dash";
|
|
CharCodes[CharCodes["Slash"] = 47] = "Slash";
|
|
CharCodes[CharCodes["Zero"] = 48] = "Zero";
|
|
CharCodes[CharCodes["Nine"] = 57] = "Nine";
|
|
CharCodes[CharCodes["Semi"] = 59] = "Semi";
|
|
CharCodes[CharCodes["Lt"] = 60] = "Lt";
|
|
CharCodes[CharCodes["Eq"] = 61] = "Eq";
|
|
CharCodes[CharCodes["Gt"] = 62] = "Gt";
|
|
CharCodes[CharCodes["Questionmark"] = 63] = "Questionmark";
|
|
CharCodes[CharCodes["UpperA"] = 65] = "UpperA";
|
|
CharCodes[CharCodes["LowerA"] = 97] = "LowerA";
|
|
CharCodes[CharCodes["UpperF"] = 70] = "UpperF";
|
|
CharCodes[CharCodes["LowerF"] = 102] = "LowerF";
|
|
CharCodes[CharCodes["UpperZ"] = 90] = "UpperZ";
|
|
CharCodes[CharCodes["LowerZ"] = 122] = "LowerZ";
|
|
CharCodes[CharCodes["LowerX"] = 120] = "LowerX";
|
|
CharCodes[CharCodes["OpeningSquareBracket"] = 91] = "OpeningSquareBracket";
|
|
})(CharCodes || (CharCodes = {}));
|
|
/** All the states the tokenizer can be in. */
|
|
var State;
|
|
(function (State) {
|
|
State[State["Text"] = 1] = "Text";
|
|
State[State["BeforeTagName"] = 2] = "BeforeTagName";
|
|
State[State["InTagName"] = 3] = "InTagName";
|
|
State[State["InSelfClosingTag"] = 4] = "InSelfClosingTag";
|
|
State[State["BeforeClosingTagName"] = 5] = "BeforeClosingTagName";
|
|
State[State["InClosingTagName"] = 6] = "InClosingTagName";
|
|
State[State["AfterClosingTagName"] = 7] = "AfterClosingTagName";
|
|
// Attributes
|
|
State[State["BeforeAttributeName"] = 8] = "BeforeAttributeName";
|
|
State[State["InAttributeName"] = 9] = "InAttributeName";
|
|
State[State["AfterAttributeName"] = 10] = "AfterAttributeName";
|
|
State[State["BeforeAttributeValue"] = 11] = "BeforeAttributeValue";
|
|
State[State["InAttributeValueDq"] = 12] = "InAttributeValueDq";
|
|
State[State["InAttributeValueSq"] = 13] = "InAttributeValueSq";
|
|
State[State["InAttributeValueNq"] = 14] = "InAttributeValueNq";
|
|
// Declarations
|
|
State[State["BeforeDeclaration"] = 15] = "BeforeDeclaration";
|
|
State[State["InDeclaration"] = 16] = "InDeclaration";
|
|
// Processing instructions
|
|
State[State["InProcessingInstruction"] = 17] = "InProcessingInstruction";
|
|
// Comments & CDATA
|
|
State[State["BeforeComment"] = 18] = "BeforeComment";
|
|
State[State["CDATASequence"] = 19] = "CDATASequence";
|
|
State[State["InSpecialComment"] = 20] = "InSpecialComment";
|
|
State[State["InCommentLike"] = 21] = "InCommentLike";
|
|
// Special tags
|
|
State[State["BeforeSpecialS"] = 22] = "BeforeSpecialS";
|
|
State[State["SpecialStartSequence"] = 23] = "SpecialStartSequence";
|
|
State[State["InSpecialTag"] = 24] = "InSpecialTag";
|
|
State[State["BeforeEntity"] = 25] = "BeforeEntity";
|
|
State[State["BeforeNumericEntity"] = 26] = "BeforeNumericEntity";
|
|
State[State["InNamedEntity"] = 27] = "InNamedEntity";
|
|
State[State["InNumericEntity"] = 28] = "InNumericEntity";
|
|
State[State["InHexEntity"] = 29] = "InHexEntity";
|
|
})(State || (State = {}));
|
|
function isWhitespace(c) {
|
|
return (c === CharCodes.Space ||
|
|
c === CharCodes.NewLine ||
|
|
c === CharCodes.Tab ||
|
|
c === CharCodes.FormFeed ||
|
|
c === CharCodes.CarriageReturn);
|
|
}
|
|
function isEndOfTagSection(c) {
|
|
return c === CharCodes.Slash || c === CharCodes.Gt || isWhitespace(c);
|
|
}
|
|
function isNumber(c) {
|
|
return c >= CharCodes.Zero && c <= CharCodes.Nine;
|
|
}
|
|
function isASCIIAlpha(c) {
|
|
return ((c >= CharCodes.LowerA && c <= CharCodes.LowerZ) ||
|
|
(c >= CharCodes.UpperA && c <= CharCodes.UpperZ));
|
|
}
|
|
function isHexDigit(c) {
|
|
return ((c >= CharCodes.UpperA && c <= CharCodes.UpperF) ||
|
|
(c >= CharCodes.LowerA && c <= CharCodes.LowerF));
|
|
}
|
|
var QuoteType;
|
|
(function (QuoteType) {
|
|
QuoteType[QuoteType["NoValue"] = 0] = "NoValue";
|
|
QuoteType[QuoteType["Unquoted"] = 1] = "Unquoted";
|
|
QuoteType[QuoteType["Single"] = 2] = "Single";
|
|
QuoteType[QuoteType["Double"] = 3] = "Double";
|
|
})(QuoteType = exports.QuoteType || (exports.QuoteType = {}));
|
|
/**
|
|
* 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 Uint8Array([0x43, 0x44, 0x41, 0x54, 0x41, 0x5b]),
|
|
CdataEnd: new Uint8Array([0x5d, 0x5d, 0x3e]),
|
|
CommentEnd: new Uint8Array([0x2d, 0x2d, 0x3e]),
|
|
ScriptEnd: new Uint8Array([0x3c, 0x2f, 0x73, 0x63, 0x72, 0x69, 0x70, 0x74]),
|
|
StyleEnd: new Uint8Array([0x3c, 0x2f, 0x73, 0x74, 0x79, 0x6c, 0x65]),
|
|
TitleEnd: new Uint8Array([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 = State.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;
|
|
/** Some behavior, eg. when decoding entities, is done while we are in another state. This keeps track of the other state type. */
|
|
this.baseState = State.Text;
|
|
/** For special parsing behavior inside of script and style tags. */
|
|
this.isSpecial = false;
|
|
/** Indicates whether the tokenizer has been paused. */
|
|
this.running = true;
|
|
/** The offset of the current buffer. */
|
|
this.offset = 0;
|
|
this.currentSequence = undefined;
|
|
this.sequenceIndex = 0;
|
|
this.trieIndex = 0;
|
|
this.trieCurrent = 0;
|
|
/** For named entities, the index of the value. For numeric entities, the code point. */
|
|
this.entityResult = 0;
|
|
this.entityExcess = 0;
|
|
this.xmlMode = xmlMode;
|
|
this.decodeEntities = decodeEntities;
|
|
this.entityTrie = xmlMode ? decode_js_1.xmlDecodeTree : decode_js_1.htmlDecodeTree;
|
|
}
|
|
Tokenizer.prototype.reset = function () {
|
|
this.state = State.Text;
|
|
this.buffer = "";
|
|
this.sectionStart = 0;
|
|
this.index = 0;
|
|
this.baseState = State.Text;
|
|
this.currentSequence = undefined;
|
|
this.running = true;
|
|
this.offset = 0;
|
|
};
|
|
Tokenizer.prototype.write = function (chunk) {
|
|
this.offset += this.buffer.length;
|
|
this.buffer = chunk;
|
|
this.parse();
|
|
};
|
|
Tokenizer.prototype.end = function () {
|
|
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.offset) {
|
|
this.parse();
|
|
}
|
|
};
|
|
/**
|
|
* The current index within all of the written data.
|
|
*/
|
|
Tokenizer.prototype.getIndex = function () {
|
|
return this.index;
|
|
};
|
|
/**
|
|
* The start of the current section.
|
|
*/
|
|
Tokenizer.prototype.getSectionStart = function () {
|
|
return this.sectionStart;
|
|
};
|
|
Tokenizer.prototype.stateText = function (c) {
|
|
if (c === CharCodes.Lt ||
|
|
(!this.decodeEntities && this.fastForwardTo(CharCodes.Lt))) {
|
|
if (this.index > this.sectionStart) {
|
|
this.cbs.ontext(this.sectionStart, this.index);
|
|
}
|
|
this.state = State.BeforeTagName;
|
|
this.sectionStart = this.index;
|
|
}
|
|
else if (this.decodeEntities && c === CharCodes.Amp) {
|
|
this.state = State.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 = State.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 === CharCodes.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.sectionStart, endOfText);
|
|
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 === CharCodes.Amp) {
|
|
this.state = State.BeforeEntity;
|
|
}
|
|
}
|
|
else if (this.fastForwardTo(CharCodes.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 === CharCodes.Lt);
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateCDATASequence = function (c) {
|
|
if (c === Sequences.Cdata[this.sequenceIndex]) {
|
|
if (++this.sequenceIndex === Sequences.Cdata.length) {
|
|
this.state = State.InCommentLike;
|
|
this.currentSequence = Sequences.CdataEnd;
|
|
this.sequenceIndex = 0;
|
|
this.sectionStart = this.index + 1;
|
|
}
|
|
}
|
|
else {
|
|
this.sequenceIndex = 0;
|
|
this.state = State.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 + this.offset) {
|
|
if (this.buffer.charCodeAt(this.index - this.offset) === 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 + this.offset - 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) {
|
|
if (this.currentSequence === Sequences.CdataEnd) {
|
|
this.cbs.oncdata(this.sectionStart, this.index, 2);
|
|
}
|
|
else {
|
|
this.cbs.oncomment(this.sectionStart, this.index, 2);
|
|
}
|
|
this.sequenceIndex = 0;
|
|
this.sectionStart = this.index + 1;
|
|
this.state = State.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 = State.SpecialStartSequence;
|
|
};
|
|
Tokenizer.prototype.stateBeforeTagName = function (c) {
|
|
if (c === CharCodes.ExclamationMark) {
|
|
this.state = State.BeforeDeclaration;
|
|
this.sectionStart = this.index + 1;
|
|
}
|
|
else if (c === CharCodes.Questionmark) {
|
|
this.state = State.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]
|
|
? State.BeforeSpecialS
|
|
: State.InTagName;
|
|
}
|
|
}
|
|
else if (c === CharCodes.Slash) {
|
|
this.state = State.BeforeClosingTagName;
|
|
}
|
|
else {
|
|
this.state = State.Text;
|
|
this.stateText(c);
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInTagName = function (c) {
|
|
if (isEndOfTagSection(c)) {
|
|
this.cbs.onopentagname(this.sectionStart, this.index);
|
|
this.sectionStart = -1;
|
|
this.state = State.BeforeAttributeName;
|
|
this.stateBeforeAttributeName(c);
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateBeforeClosingTagName = function (c) {
|
|
if (isWhitespace(c)) {
|
|
// Ignore
|
|
}
|
|
else if (c === CharCodes.Gt) {
|
|
this.state = State.Text;
|
|
}
|
|
else {
|
|
this.state = this.isTagStartChar(c)
|
|
? State.InClosingTagName
|
|
: State.InSpecialComment;
|
|
this.sectionStart = this.index;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInClosingTagName = function (c) {
|
|
if (c === CharCodes.Gt || isWhitespace(c)) {
|
|
this.cbs.onclosetag(this.sectionStart, this.index);
|
|
this.sectionStart = -1;
|
|
this.state = State.AfterClosingTagName;
|
|
this.stateAfterClosingTagName(c);
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateAfterClosingTagName = function (c) {
|
|
// Skip everything until ">"
|
|
if (c === CharCodes.Gt || this.fastForwardTo(CharCodes.Gt)) {
|
|
this.state = State.Text;
|
|
this.baseState = State.Text;
|
|
this.sectionStart = this.index + 1;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateBeforeAttributeName = function (c) {
|
|
if (c === CharCodes.Gt) {
|
|
this.cbs.onopentagend(this.index);
|
|
if (this.isSpecial) {
|
|
this.state = State.InSpecialTag;
|
|
this.sequenceIndex = 0;
|
|
}
|
|
else {
|
|
this.state = State.Text;
|
|
}
|
|
this.baseState = this.state;
|
|
this.sectionStart = this.index + 1;
|
|
}
|
|
else if (c === CharCodes.Slash) {
|
|
this.state = State.InSelfClosingTag;
|
|
}
|
|
else if (!isWhitespace(c)) {
|
|
this.state = State.InAttributeName;
|
|
this.sectionStart = this.index;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInSelfClosingTag = function (c) {
|
|
if (c === CharCodes.Gt) {
|
|
this.cbs.onselfclosingtag(this.index);
|
|
this.state = State.Text;
|
|
this.baseState = State.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 = State.BeforeAttributeName;
|
|
this.stateBeforeAttributeName(c);
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInAttributeName = function (c) {
|
|
if (c === CharCodes.Eq || isEndOfTagSection(c)) {
|
|
this.cbs.onattribname(this.sectionStart, this.index);
|
|
this.sectionStart = -1;
|
|
this.state = State.AfterAttributeName;
|
|
this.stateAfterAttributeName(c);
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateAfterAttributeName = function (c) {
|
|
if (c === CharCodes.Eq) {
|
|
this.state = State.BeforeAttributeValue;
|
|
}
|
|
else if (c === CharCodes.Slash || c === CharCodes.Gt) {
|
|
this.cbs.onattribend(QuoteType.NoValue, this.index);
|
|
this.state = State.BeforeAttributeName;
|
|
this.stateBeforeAttributeName(c);
|
|
}
|
|
else if (!isWhitespace(c)) {
|
|
this.cbs.onattribend(QuoteType.NoValue, this.index);
|
|
this.state = State.InAttributeName;
|
|
this.sectionStart = this.index;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateBeforeAttributeValue = function (c) {
|
|
if (c === CharCodes.DoubleQuote) {
|
|
this.state = State.InAttributeValueDq;
|
|
this.sectionStart = this.index + 1;
|
|
}
|
|
else if (c === CharCodes.SingleQuote) {
|
|
this.state = State.InAttributeValueSq;
|
|
this.sectionStart = this.index + 1;
|
|
}
|
|
else if (!isWhitespace(c)) {
|
|
this.sectionStart = this.index;
|
|
this.state = State.InAttributeValueNq;
|
|
this.stateInAttributeValueNoQuotes(c); // Reconsume token
|
|
}
|
|
};
|
|
Tokenizer.prototype.handleInAttributeValue = function (c, quote) {
|
|
if (c === quote ||
|
|
(!this.decodeEntities && this.fastForwardTo(quote))) {
|
|
this.cbs.onattribdata(this.sectionStart, this.index);
|
|
this.sectionStart = -1;
|
|
this.cbs.onattribend(quote === CharCodes.DoubleQuote
|
|
? QuoteType.Double
|
|
: QuoteType.Single, this.index);
|
|
this.state = State.BeforeAttributeName;
|
|
}
|
|
else if (this.decodeEntities && c === CharCodes.Amp) {
|
|
this.baseState = this.state;
|
|
this.state = State.BeforeEntity;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInAttributeValueDoubleQuotes = function (c) {
|
|
this.handleInAttributeValue(c, CharCodes.DoubleQuote);
|
|
};
|
|
Tokenizer.prototype.stateInAttributeValueSingleQuotes = function (c) {
|
|
this.handleInAttributeValue(c, CharCodes.SingleQuote);
|
|
};
|
|
Tokenizer.prototype.stateInAttributeValueNoQuotes = function (c) {
|
|
if (isWhitespace(c) || c === CharCodes.Gt) {
|
|
this.cbs.onattribdata(this.sectionStart, this.index);
|
|
this.sectionStart = -1;
|
|
this.cbs.onattribend(QuoteType.Unquoted, this.index);
|
|
this.state = State.BeforeAttributeName;
|
|
this.stateBeforeAttributeName(c);
|
|
}
|
|
else if (this.decodeEntities && c === CharCodes.Amp) {
|
|
this.baseState = this.state;
|
|
this.state = State.BeforeEntity;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateBeforeDeclaration = function (c) {
|
|
if (c === CharCodes.OpeningSquareBracket) {
|
|
this.state = State.CDATASequence;
|
|
this.sequenceIndex = 0;
|
|
}
|
|
else {
|
|
this.state =
|
|
c === CharCodes.Dash
|
|
? State.BeforeComment
|
|
: State.InDeclaration;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInDeclaration = function (c) {
|
|
if (c === CharCodes.Gt || this.fastForwardTo(CharCodes.Gt)) {
|
|
this.cbs.ondeclaration(this.sectionStart, this.index);
|
|
this.state = State.Text;
|
|
this.sectionStart = this.index + 1;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInProcessingInstruction = function (c) {
|
|
if (c === CharCodes.Gt || this.fastForwardTo(CharCodes.Gt)) {
|
|
this.cbs.onprocessinginstruction(this.sectionStart, this.index);
|
|
this.state = State.Text;
|
|
this.sectionStart = this.index + 1;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateBeforeComment = function (c) {
|
|
if (c === CharCodes.Dash) {
|
|
this.state = State.InCommentLike;
|
|
this.currentSequence = Sequences.CommentEnd;
|
|
// Allow short comments (eg. <!-->)
|
|
this.sequenceIndex = 2;
|
|
this.sectionStart = this.index + 1;
|
|
}
|
|
else {
|
|
this.state = State.InDeclaration;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInSpecialComment = function (c) {
|
|
if (c === CharCodes.Gt || this.fastForwardTo(CharCodes.Gt)) {
|
|
this.cbs.oncomment(this.sectionStart, this.index, 0);
|
|
this.state = State.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 = State.InTagName;
|
|
this.stateInTagName(c); // Consume the token again
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateBeforeEntity = function (c) {
|
|
// Start excess with 1 to include the '&'
|
|
this.entityExcess = 1;
|
|
this.entityResult = 0;
|
|
if (c === CharCodes.Number) {
|
|
this.state = State.BeforeNumericEntity;
|
|
}
|
|
else if (c === CharCodes.Amp) {
|
|
// We have two `&` characters in a row. Stay in the current state.
|
|
}
|
|
else {
|
|
this.trieIndex = 0;
|
|
this.trieCurrent = this.entityTrie[0];
|
|
this.state = State.InNamedEntity;
|
|
this.stateInNamedEntity(c);
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInNamedEntity = function (c) {
|
|
this.entityExcess += 1;
|
|
this.trieIndex = (0, decode_js_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];
|
|
var masked = this.trieCurrent & 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.
|
|
var valueLength = (masked >> 14) - 1;
|
|
// If we have a legacy entity while parsing strictly, just skip the number of bytes
|
|
if (!this.allowLegacyEntity() && c !== CharCodes.Semi) {
|
|
this.trieIndex += valueLength;
|
|
}
|
|
else {
|
|
// Add 1 as we have already incremented the excess
|
|
var entityStart = this.index - this.entityExcess + 1;
|
|
if (entityStart > this.sectionStart) {
|
|
this.emitPartial(this.sectionStart, entityStart);
|
|
}
|
|
// If this is a surrogate pair, consume the next two bytes
|
|
this.entityResult = this.trieIndex;
|
|
this.trieIndex += valueLength;
|
|
this.entityExcess = 0;
|
|
this.sectionStart = this.index + 1;
|
|
if (valueLength === 0) {
|
|
this.emitNamedEntity();
|
|
}
|
|
}
|
|
}
|
|
};
|
|
Tokenizer.prototype.emitNamedEntity = function () {
|
|
this.state = this.baseState;
|
|
if (this.entityResult === 0) {
|
|
return;
|
|
}
|
|
var valueLength = (this.entityTrie[this.entityResult] & decode_js_1.BinTrieFlags.VALUE_LENGTH) >>
|
|
14;
|
|
switch (valueLength) {
|
|
case 1: {
|
|
this.emitCodePoint(this.entityTrie[this.entityResult] &
|
|
~decode_js_1.BinTrieFlags.VALUE_LENGTH);
|
|
break;
|
|
}
|
|
case 2: {
|
|
this.emitCodePoint(this.entityTrie[this.entityResult + 1]);
|
|
break;
|
|
}
|
|
case 3: {
|
|
this.emitCodePoint(this.entityTrie[this.entityResult + 1]);
|
|
this.emitCodePoint(this.entityTrie[this.entityResult + 2]);
|
|
}
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateBeforeNumericEntity = function (c) {
|
|
if ((c | 0x20) === CharCodes.LowerX) {
|
|
this.entityExcess++;
|
|
this.state = State.InHexEntity;
|
|
}
|
|
else {
|
|
this.state = State.InNumericEntity;
|
|
this.stateInNumericEntity(c);
|
|
}
|
|
};
|
|
Tokenizer.prototype.emitNumericEntity = function (strict) {
|
|
var entityStart = this.index - this.entityExcess - 1;
|
|
var numberStart = entityStart + 2 + Number(this.state === State.InHexEntity);
|
|
if (numberStart !== this.index) {
|
|
// Emit leading data if any
|
|
if (entityStart > this.sectionStart) {
|
|
this.emitPartial(this.sectionStart, entityStart);
|
|
}
|
|
this.sectionStart = this.index + Number(strict);
|
|
this.emitCodePoint((0, decode_js_1.replaceCodePoint)(this.entityResult));
|
|
}
|
|
this.state = this.baseState;
|
|
};
|
|
Tokenizer.prototype.stateInNumericEntity = function (c) {
|
|
if (c === CharCodes.Semi) {
|
|
this.emitNumericEntity(true);
|
|
}
|
|
else if (isNumber(c)) {
|
|
this.entityResult = this.entityResult * 10 + (c - CharCodes.Zero);
|
|
this.entityExcess++;
|
|
}
|
|
else {
|
|
if (this.allowLegacyEntity()) {
|
|
this.emitNumericEntity(false);
|
|
}
|
|
else {
|
|
this.state = this.baseState;
|
|
}
|
|
this.index--;
|
|
}
|
|
};
|
|
Tokenizer.prototype.stateInHexEntity = function (c) {
|
|
if (c === CharCodes.Semi) {
|
|
this.emitNumericEntity(true);
|
|
}
|
|
else if (isNumber(c)) {
|
|
this.entityResult = this.entityResult * 16 + (c - CharCodes.Zero);
|
|
this.entityExcess++;
|
|
}
|
|
else if (isHexDigit(c)) {
|
|
this.entityResult =
|
|
this.entityResult * 16 + ((c | 0x20) - CharCodes.LowerA + 10);
|
|
this.entityExcess++;
|
|
}
|
|
else {
|
|
if (this.allowLegacyEntity()) {
|
|
this.emitNumericEntity(false);
|
|
}
|
|
else {
|
|
this.state = this.baseState;
|
|
}
|
|
this.index--;
|
|
}
|
|
};
|
|
Tokenizer.prototype.allowLegacyEntity = function () {
|
|
return (!this.xmlMode &&
|
|
(this.baseState === State.Text ||
|
|
this.baseState === State.InSpecialTag));
|
|
};
|
|
/**
|
|
* Remove data that has already been consumed from the buffer.
|
|
*/
|
|
Tokenizer.prototype.cleanup = function () {
|
|
// If we are inside of text or attributes, emit what we already have.
|
|
if (this.running && this.sectionStart !== this.index) {
|
|
if (this.state === State.Text ||
|
|
(this.state === State.InSpecialTag && this.sequenceIndex === 0)) {
|
|
this.cbs.ontext(this.sectionStart, this.index);
|
|
this.sectionStart = this.index;
|
|
}
|
|
else if (this.state === State.InAttributeValueDq ||
|
|
this.state === State.InAttributeValueSq ||
|
|
this.state === State.InAttributeValueNq) {
|
|
this.cbs.onattribdata(this.sectionStart, this.index);
|
|
this.sectionStart = this.index;
|
|
}
|
|
}
|
|
};
|
|
Tokenizer.prototype.shouldContinue = function () {
|
|
return this.index < this.buffer.length + this.offset && 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 - this.offset);
|
|
switch (this.state) {
|
|
case State.Text: {
|
|
this.stateText(c);
|
|
break;
|
|
}
|
|
case State.SpecialStartSequence: {
|
|
this.stateSpecialStartSequence(c);
|
|
break;
|
|
}
|
|
case State.InSpecialTag: {
|
|
this.stateInSpecialTag(c);
|
|
break;
|
|
}
|
|
case State.CDATASequence: {
|
|
this.stateCDATASequence(c);
|
|
break;
|
|
}
|
|
case State.InAttributeValueDq: {
|
|
this.stateInAttributeValueDoubleQuotes(c);
|
|
break;
|
|
}
|
|
case State.InAttributeName: {
|
|
this.stateInAttributeName(c);
|
|
break;
|
|
}
|
|
case State.InCommentLike: {
|
|
this.stateInCommentLike(c);
|
|
break;
|
|
}
|
|
case State.InSpecialComment: {
|
|
this.stateInSpecialComment(c);
|
|
break;
|
|
}
|
|
case State.BeforeAttributeName: {
|
|
this.stateBeforeAttributeName(c);
|
|
break;
|
|
}
|
|
case State.InTagName: {
|
|
this.stateInTagName(c);
|
|
break;
|
|
}
|
|
case State.InClosingTagName: {
|
|
this.stateInClosingTagName(c);
|
|
break;
|
|
}
|
|
case State.BeforeTagName: {
|
|
this.stateBeforeTagName(c);
|
|
break;
|
|
}
|
|
case State.AfterAttributeName: {
|
|
this.stateAfterAttributeName(c);
|
|
break;
|
|
}
|
|
case State.InAttributeValueSq: {
|
|
this.stateInAttributeValueSingleQuotes(c);
|
|
break;
|
|
}
|
|
case State.BeforeAttributeValue: {
|
|
this.stateBeforeAttributeValue(c);
|
|
break;
|
|
}
|
|
case State.BeforeClosingTagName: {
|
|
this.stateBeforeClosingTagName(c);
|
|
break;
|
|
}
|
|
case State.AfterClosingTagName: {
|
|
this.stateAfterClosingTagName(c);
|
|
break;
|
|
}
|
|
case State.BeforeSpecialS: {
|
|
this.stateBeforeSpecialS(c);
|
|
break;
|
|
}
|
|
case State.InAttributeValueNq: {
|
|
this.stateInAttributeValueNoQuotes(c);
|
|
break;
|
|
}
|
|
case State.InSelfClosingTag: {
|
|
this.stateInSelfClosingTag(c);
|
|
break;
|
|
}
|
|
case State.InDeclaration: {
|
|
this.stateInDeclaration(c);
|
|
break;
|
|
}
|
|
case State.BeforeDeclaration: {
|
|
this.stateBeforeDeclaration(c);
|
|
break;
|
|
}
|
|
case State.BeforeComment: {
|
|
this.stateBeforeComment(c);
|
|
break;
|
|
}
|
|
case State.InProcessingInstruction: {
|
|
this.stateInProcessingInstruction(c);
|
|
break;
|
|
}
|
|
case State.InNamedEntity: {
|
|
this.stateInNamedEntity(c);
|
|
break;
|
|
}
|
|
case State.BeforeEntity: {
|
|
this.stateBeforeEntity(c);
|
|
break;
|
|
}
|
|
case State.InHexEntity: {
|
|
this.stateInHexEntity(c);
|
|
break;
|
|
}
|
|
case State.InNumericEntity: {
|
|
this.stateInNumericEntity(c);
|
|
break;
|
|
}
|
|
default: {
|
|
// `this._state === State.BeforeNumericEntity`
|
|
this.stateBeforeNumericEntity(c);
|
|
}
|
|
}
|
|
this.index++;
|
|
}
|
|
this.cleanup();
|
|
};
|
|
Tokenizer.prototype.finish = function () {
|
|
if (this.state === State.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 endIndex = this.buffer.length + this.offset;
|
|
if (this.state === State.InCommentLike) {
|
|
if (this.currentSequence === Sequences.CdataEnd) {
|
|
this.cbs.oncdata(this.sectionStart, endIndex, 0);
|
|
}
|
|
else {
|
|
this.cbs.oncomment(this.sectionStart, endIndex, 0);
|
|
}
|
|
}
|
|
else if (this.state === State.InNumericEntity &&
|
|
this.allowLegacyEntity()) {
|
|
this.emitNumericEntity(false);
|
|
// All trailing data will have been consumed
|
|
}
|
|
else if (this.state === State.InHexEntity &&
|
|
this.allowLegacyEntity()) {
|
|
this.emitNumericEntity(false);
|
|
// All trailing data will have been consumed
|
|
}
|
|
else if (this.state === State.InTagName ||
|
|
this.state === State.BeforeAttributeName ||
|
|
this.state === State.BeforeAttributeValue ||
|
|
this.state === State.AfterAttributeName ||
|
|
this.state === State.InAttributeName ||
|
|
this.state === State.InAttributeValueSq ||
|
|
this.state === State.InAttributeValueDq ||
|
|
this.state === State.InAttributeValueNq ||
|
|
this.state === State.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(this.sectionStart, endIndex);
|
|
}
|
|
};
|
|
Tokenizer.prototype.emitPartial = function (start, endIndex) {
|
|
if (this.baseState !== State.Text &&
|
|
this.baseState !== State.InSpecialTag) {
|
|
this.cbs.onattribdata(start, endIndex);
|
|
}
|
|
else {
|
|
this.cbs.ontext(start, endIndex);
|
|
}
|
|
};
|
|
Tokenizer.prototype.emitCodePoint = function (cp) {
|
|
if (this.baseState !== State.Text &&
|
|
this.baseState !== State.InSpecialTag) {
|
|
this.cbs.onattribentity(cp);
|
|
}
|
|
else {
|
|
this.cbs.ontextentity(cp);
|
|
}
|
|
};
|
|
return Tokenizer;
|
|
}());
|
|
exports.default = Tokenizer;
|
|
//# sourceMappingURL=Tokenizer.js.map
|