htmlparser2/lib/Parser.js

"use strict";
var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
    if (k2 === undefined) k2 = k;
    var desc = Object.getOwnPropertyDescriptor(m, k);
    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
      desc = { enumerable: true, get: function() { return m[k]; } };
    }
    Object.defineProperty(o, k2, desc);
}) : (function(o, m, k, k2) {
    if (k2 === undefined) k2 = k;
    o[k2] = m[k];
}));
var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
    Object.defineProperty(o, "default", { enumerable: true, value: v });
}) : function(o, v) {
    o["default"] = v;
});
var __importStar = (this && this.__importStar) || function (mod) {
    if (mod && mod.__esModule) return mod;
    var result = {};
    if (mod != null) for (var k in mod) if (k !== "default" && Object.prototype.hasOwnProperty.call(mod, k)) __createBinding(result, mod, k);
    __setModuleDefault(result, mod);
    return result;
};
Object.defineProperty(exports, "__esModule", { value: true });
exports.Parser = void 0;
var Tokenizer_js_1 = __importStar(require("./Tokenizer.js"));
var decode_js_1 = require("entities/lib/decode.js");
var formTags = new Set([
    "input",
    "option",
    "optgroup",
    "select",
    "button",
    "datalist",
    "textarea",
]);
var pTag = new Set(["p"]);
var tableSectionTags = new Set(["thead", "tbody"]);
var ddtTags = new Set(["dd", "dt"]);
var rtpTags = new Set(["rt", "rp"]);
var openImpliesClose = new Map([
    ["tr", new Set(["tr", "th", "td"])],
    ["th", new Set(["th"])],
    ["td", new Set(["thead", "th", "td"])],
    ["body", new Set(["head", "link", "script"])],
    ["li", new Set(["li"])],
    ["p", pTag],
    ["h1", pTag],
    ["h2", pTag],
    ["h3", pTag],
    ["h4", pTag],
    ["h5", pTag],
    ["h6", pTag],
    ["select", formTags],
    ["input", formTags],
    ["output", formTags],
    ["button", formTags],
    ["datalist", formTags],
    ["textarea", formTags],
    ["option", new Set(["option"])],
    ["optgroup", new Set(["optgroup", "option"])],
    ["dd", ddtTags],
    ["dt", ddtTags],
    ["address", pTag],
    ["article", pTag],
    ["aside", pTag],
    ["blockquote", pTag],
    ["details", pTag],
    ["div", pTag],
    ["dl", pTag],
    ["fieldset", pTag],
    ["figcaption", pTag],
    ["figure", pTag],
    ["footer", pTag],
    ["form", pTag],
    ["header", pTag],
    ["hr", pTag],
    ["main", pTag],
    ["nav", pTag],
    ["ol", pTag],
    ["pre", pTag],
    ["section", pTag],
    ["table", pTag],
    ["ul", pTag],
    ["rt", rtpTags],
    ["rp", rtpTags],
    ["tbody", tableSectionTags],
    ["tfoot", tableSectionTags],
]);
var voidElements = new Set([
    "area",
    "base",
    "basefont",
    "br",
    "col",
    "command",
    "embed",
    "frame",
    "hr",
    "img",
    "input",
    "isindex",
    "keygen",
    "link",
    "meta",
    "param",
    "source",
    "track",
    "wbr",
]);
var foreignContextElements = new Set(["math", "svg"]);
var htmlIntegrationElements = new Set([
    "mi",
    "mo",
    "mn",
    "ms",
    "mtext",
    "annotation-xml",
    "foreignobject",
    "desc",
    "title",
]);
var reNameEnd = /\s|\//;
var Parser = /** @class */ (function () {
    function Parser(cbs, options) {
        if (options === void 0) { options = {}; }
        var _a, _b, _c, _d, _e, _f;
        this.options = options;
        /** The start index of the last event. */
        this.startIndex = 0;
        /** The end index of the last event. */
        this.endIndex = 0;
        /**
         * Store the start index of the current open tag,
         * so we can update the start index for attributes.
         */
        this.openTagStart = 0;
        this.tagname = "";
        this.attribname = "";
        this.attribvalue = "";
        this.attribs = null;
        this.stack = [];
        this.buffers = [];
        this.bufferOffset = 0;
        /** The index of the last written buffer. Used when resuming after a `pause()`. */
        this.writeIndex = 0;
        /** Indicates whether the parser has finished running / `.end` has been called. */
        this.ended = false;
        this.cbs = cbs !== null && cbs !== void 0 ? cbs : {};
        this.htmlMode = !this.options.xmlMode;
        this.lowerCaseTagNames = (_a = options.lowerCaseTags) !== null && _a !== void 0 ? _a : this.htmlMode;
        this.lowerCaseAttributeNames =
            (_b = options.lowerCaseAttributeNames) !== null && _b !== void 0 ? _b : this.htmlMode;
        this.recognizeSelfClosing =
            (_c = options.recognizeSelfClosing) !== null && _c !== void 0 ? _c : !this.htmlMode;
        this.tokenizer = new ((_d = options.Tokenizer) !== null && _d !== void 0 ? _d : Tokenizer_js_1.default)(this.options, this);
        this.foreignContext = [!this.htmlMode];
        (_f = (_e = this.cbs).onparserinit) === null || _f === void 0 ? void 0 : _f.call(_e, this);
    }
    // Tokenizer event handlers
    /** @internal */
    Parser.prototype.ontext = function (start, endIndex) {
        var _a, _b;
        var data = this.getSlice(start, endIndex);
        this.endIndex = endIndex - 1;
        (_b = (_a = this.cbs).ontext) === null || _b === void 0 ? void 0 : _b.call(_a, data);
        this.startIndex = endIndex;
    };
    /** @internal */
    Parser.prototype.ontextentity = function (cp, endIndex) {
        var _a, _b;
        this.endIndex = endIndex - 1;
        (_b = (_a = this.cbs).ontext) === null || _b === void 0 ? void 0 : _b.call(_a, (0, decode_js_1.fromCodePoint)(cp));
        this.startIndex = endIndex;
    };
    /**
     * Checks if the current tag is a void element. Override this if you want
     * to specify your own additional void elements.
     */
    Parser.prototype.isVoidElement = function (name) {
        return this.htmlMode && voidElements.has(name);
    };
    /** @internal */
    Parser.prototype.onopentagname = function (start, endIndex) {
        this.endIndex = endIndex;
        var name = this.getSlice(start, endIndex);
        if (this.lowerCaseTagNames) {
            name = name.toLowerCase();
        }
        this.emitOpenTag(name);
    };
    Parser.prototype.emitOpenTag = function (name) {
        var _a, _b, _c, _d;
        this.openTagStart = this.startIndex;
        this.tagname = name;
        var impliesClose = this.htmlMode && openImpliesClose.get(name);
        if (impliesClose) {
            while (this.stack.length > 0 && impliesClose.has(this.stack[0])) {
                var element = this.stack.shift();
                (_b = (_a = this.cbs).onclosetag) === null || _b === void 0 ? void 0 : _b.call(_a, element, true);
            }
        }
        if (!this.isVoidElement(name)) {
            this.stack.unshift(name);
            if (this.htmlMode) {
                if (foreignContextElements.has(name)) {
                    this.foreignContext.unshift(true);
                }
                else if (htmlIntegrationElements.has(name)) {
                    this.foreignContext.unshift(false);
                }
            }
        }
        (_d = (_c = this.cbs).onopentagname) === null || _d === void 0 ? void 0 : _d.call(_c, name);
        if (this.cbs.onopentag)
            this.attribs = {};
    };
    Parser.prototype.endOpenTag = function (isImplied) {
        var _a, _b;
        this.startIndex = this.openTagStart;
        if (this.attribs) {
            (_b = (_a = this.cbs).onopentag) === null || _b === void 0 ? void 0 : _b.call(_a, this.tagname, this.attribs, isImplied);
            this.attribs = null;
        }
        if (this.cbs.onclosetag && this.isVoidElement(this.tagname)) {
            this.cbs.onclosetag(this.tagname, true);
        }
        this.tagname = "";
    };
    /** @internal */
    Parser.prototype.onopentagend = function (endIndex) {
        this.endIndex = endIndex;
        this.endOpenTag(false);
        // Set `startIndex` for next node
        this.startIndex = endIndex + 1;
    };
    /** @internal */
    Parser.prototype.onclosetag = function (start, endIndex) {
        var _a, _b, _c, _d, _e, _f, _g, _h;
        this.endIndex = endIndex;
        var name = this.getSlice(start, endIndex);
        if (this.lowerCaseTagNames) {
            name = name.toLowerCase();
        }
        if (this.htmlMode &&
            (foreignContextElements.has(name) ||
                htmlIntegrationElements.has(name))) {
            this.foreignContext.shift();
        }
        if (!this.isVoidElement(name)) {
            var pos = this.stack.indexOf(name);
            if (pos !== -1) {
                for (var index = 0; index <= pos; index++) {
                    var element = this.stack.shift();
                    // We know the stack has sufficient elements.
                    (_b = (_a = this.cbs).onclosetag) === null || _b === void 0 ? void 0 : _b.call(_a, element, index !== pos);
                }
            }
            else if (this.htmlMode && name === "p") {
                // Implicit open before close
                this.emitOpenTag("p");
                this.closeCurrentTag(true);
            }
        }
        else if (this.htmlMode && name === "br") {
            // We can't use `emitOpenTag` for implicit open, as `br` would be implicitly closed.
            (_d = (_c = this.cbs).onopentagname) === null || _d === void 0 ? void 0 : _d.call(_c, "br");
            (_f = (_e = this.cbs).onopentag) === null || _f === void 0 ? void 0 : _f.call(_e, "br", {}, true);
            (_h = (_g = this.cbs).onclosetag) === null || _h === void 0 ? void 0 : _h.call(_g, "br", false);
        }
        // Set `startIndex` for next node
        this.startIndex = endIndex + 1;
    };
    /** @internal */
    Parser.prototype.onselfclosingtag = function (endIndex) {
        this.endIndex = endIndex;
        if (this.recognizeSelfClosing || this.foreignContext[0]) {
            this.closeCurrentTag(false);
            // Set `startIndex` for next node
            this.startIndex = endIndex + 1;
        }
        else {
            // Ignore the fact that the tag is self-closing.
            this.onopentagend(endIndex);
        }
    };
    Parser.prototype.closeCurrentTag = function (isOpenImplied) {
        var _a, _b;
        var name = this.tagname;
        this.endOpenTag(isOpenImplied);
        // Self-closing tags will be on the top of the stack
        if (this.stack[0] === name) {
            // If the opening tag isn't implied, the closing tag has to be implied.
            (_b = (_a = this.cbs).onclosetag) === null || _b === void 0 ? void 0 : _b.call(_a, name, !isOpenImplied);
            this.stack.shift();
        }
    };
    /** @internal */
    Parser.prototype.onattribname = function (start, endIndex) {
        this.startIndex = start;
        var name = this.getSlice(start, endIndex);
        this.attribname = this.lowerCaseAttributeNames
            ? name.toLowerCase()
            : name;
    };
    /** @internal */
    Parser.prototype.onattribdata = function (start, endIndex) {
        this.attribvalue += this.getSlice(start, endIndex);
    };
    /** @internal */
    Parser.prototype.onattribentity = function (cp) {
        this.attribvalue += (0, decode_js_1.fromCodePoint)(cp);
    };
    /** @internal */
    Parser.prototype.onattribend = function (quote, endIndex) {
        var _a, _b;
        this.endIndex = endIndex;
        (_b = (_a = this.cbs).onattribute) === null || _b === void 0 ? void 0 : _b.call(_a, this.attribname, this.attribvalue, quote === Tokenizer_js_1.QuoteType.Double
            ? '"'
            : quote === Tokenizer_js_1.QuoteType.Single
                ? "'"
                : quote === Tokenizer_js_1.QuoteType.NoValue
                    ? undefined
                    : null);
        if (this.attribs &&
            !Object.prototype.hasOwnProperty.call(this.attribs, this.attribname)) {
            this.attribs[this.attribname] = this.attribvalue;
        }
        this.attribvalue = "";
    };
    Parser.prototype.getInstructionName = function (value) {
        var index = value.search(reNameEnd);
        var name = index < 0 ? value : value.substr(0, index);
        if (this.lowerCaseTagNames) {
            name = name.toLowerCase();
        }
        return name;
    };
    /** @internal */
    Parser.prototype.ondeclaration = function (start, endIndex) {
        this.endIndex = endIndex;
        var value = this.getSlice(start, endIndex);
        if (this.cbs.onprocessinginstruction) {
            var name = this.getInstructionName(value);
            this.cbs.onprocessinginstruction("!".concat(name), "!".concat(value));
        }
        // Set `startIndex` for next node
        this.startIndex = endIndex + 1;
    };
    /** @internal */
    Parser.prototype.onprocessinginstruction = function (start, endIndex) {
        this.endIndex = endIndex;
        var value = this.getSlice(start, endIndex);
        if (this.cbs.onprocessinginstruction) {
            var name = this.getInstructionName(value);
            this.cbs.onprocessinginstruction("?".concat(name), "?".concat(value));
        }
        // Set `startIndex` for next node
        this.startIndex = endIndex + 1;
    };
    /** @internal */
    Parser.prototype.oncomment = function (start, endIndex, offset) {
        var _a, _b, _c, _d;
        this.endIndex = endIndex;
        (_b = (_a = this.cbs).oncomment) === null || _b === void 0 ? void 0 : _b.call(_a, this.getSlice(start, endIndex - offset));
        (_d = (_c = this.cbs).oncommentend) === null || _d === void 0 ? void 0 : _d.call(_c);
        // Set `startIndex` for next node
        this.startIndex = endIndex + 1;
    };
    /** @internal */
    Parser.prototype.oncdata = function (start, endIndex, offset) {
        var _a, _b, _c, _d, _e, _f, _g, _h, _j, _k;
        this.endIndex = endIndex;
        var value = this.getSlice(start, endIndex - offset);
        if (!this.htmlMode || this.options.recognizeCDATA) {
            (_b = (_a = this.cbs).oncdatastart) === null || _b === void 0 ? void 0 : _b.call(_a);
            (_d = (_c = this.cbs).ontext) === null || _d === void 0 ? void 0 : _d.call(_c, value);
            (_f = (_e = this.cbs).oncdataend) === null || _f === void 0 ? void 0 : _f.call(_e);
        }
        else {
            (_h = (_g = this.cbs).oncomment) === null || _h === void 0 ? void 0 : _h.call(_g, "[CDATA[".concat(value, "]]"));
            (_k = (_j = this.cbs).oncommentend) === null || _k === void 0 ? void 0 : _k.call(_j);
        }
        // Set `startIndex` for next node
        this.startIndex = endIndex + 1;
    };
    /** @internal */
    Parser.prototype.onend = function () {
        var _a, _b;
        if (this.cbs.onclosetag) {
            // Set the end index for all remaining tags
            this.endIndex = this.startIndex;
            for (var index = 0; index < this.stack.length; index++) {
                this.cbs.onclosetag(this.stack[index], true);
            }
        }
        (_b = (_a = this.cbs).onend) === null || _b === void 0 ? void 0 : _b.call(_a);
    };
    /**
     * Resets the parser to a blank state, ready to parse a new HTML document
     */
    Parser.prototype.reset = function () {
        var _a, _b, _c, _d;
        (_b = (_a = this.cbs).onreset) === null || _b === void 0 ? void 0 : _b.call(_a);
        this.tokenizer.reset();
        this.tagname = "";
        this.attribname = "";
        this.attribs = null;
        this.stack.length = 0;
        this.startIndex = 0;
        this.endIndex = 0;
        (_d = (_c = this.cbs).onparserinit) === null || _d === void 0 ? void 0 : _d.call(_c, this);
        this.buffers.length = 0;
        this.foreignContext.length = 0;
        this.foreignContext.unshift(!this.htmlMode);
        this.bufferOffset = 0;
        this.writeIndex = 0;
        this.ended = false;
    };
    /**
     * Resets the parser, then parses a complete document and
     * pushes it to the handler.
     *
     * @param data Document to parse.
     */
    Parser.prototype.parseComplete = function (data) {
        this.reset();
        this.end(data);
    };
    Parser.prototype.getSlice = function (start, end) {
        while (start - this.bufferOffset >= this.buffers[0].length) {
            this.shiftBuffer();
        }
        var slice = this.buffers[0].slice(start - this.bufferOffset, end - this.bufferOffset);
        while (end - this.bufferOffset > this.buffers[0].length) {
            this.shiftBuffer();
            slice += this.buffers[0].slice(0, end - this.bufferOffset);
        }
        return slice;
    };
    Parser.prototype.shiftBuffer = function () {
        this.bufferOffset += this.buffers[0].length;
        this.writeIndex--;
        this.buffers.shift();
    };
    /**
     * Parses a chunk of data and calls the corresponding callbacks.
     *
     * @param chunk Chunk to parse.
     */
    Parser.prototype.write = function (chunk) {
        var _a, _b;
        if (this.ended) {
            (_b = (_a = this.cbs).onerror) === null || _b === void 0 ? void 0 : _b.call(_a, new Error(".write() after done!"));
            return;
        }
        this.buffers.push(chunk);
        if (this.tokenizer.running) {
            this.tokenizer.write(chunk);
            this.writeIndex++;
        }
    };
    /**
     * Parses the end of the buffer and clears the stack, calls onend.
     *
     * @param chunk Optional final chunk to parse.
     */
    Parser.prototype.end = function (chunk) {
        var _a, _b;
        if (this.ended) {
            (_b = (_a = this.cbs).onerror) === null || _b === void 0 ? void 0 : _b.call(_a, new Error(".end() after done!"));
            return;
        }
        if (chunk)
            this.write(chunk);
        this.ended = true;
        this.tokenizer.end();
    };
    /**
     * Pauses parsing. The parser won't emit events until `resume` is called.
     */
    Parser.prototype.pause = function () {
        this.tokenizer.pause();
    };
    /**
     * Resumes parsing after `pause` was called.
     */
    Parser.prototype.resume = function () {
        this.tokenizer.resume();
        while (this.tokenizer.running &&
            this.writeIndex < this.buffers.length) {
            this.tokenizer.write(this.buffers[this.writeIndex++]);
        }
        if (this.ended)
            this.tokenizer.end();
    };
    /**
     * Alias of `write`, for backwards compatibility.
     *
     * @param chunk Chunk to parse.
     * @deprecated
     */
    Parser.prototype.parseChunk = function (chunk) {
        this.write(chunk);
    };
    /**
     * Alias of `end`, for backwards compatibility.
     *
     * @param chunk Optional final chunk to parse.
     * @deprecated
     */
    Parser.prototype.done = function (chunk) {
        this.end(chunk);
    };
    return Parser;
}());
exports.Parser = Parser;
//# sourceMappingURL=Parser.js.map