@faubulous/mentor-rdf
Version:
A library for working with RDF vocabularies with support for basic RDFS and OWL inference.
19 lines (17 loc) • 1.88 MB
JavaScript
/*
* ATTENTION: The "eval" devtool has been used (maybe by default in mode: "development").
* This devtool is neither made for production nor for readable output files.
* It uses "eval()" calls to create a separate source file in the browser devtools.
* If you are trying to read the output file, select a different devtool (https://webpack.js.org/configuration/devtool/)
* or disable the default devtool with "devtool: false".
* If you are looking for production-ready output files, see mode: "production" (https://webpack.js.org/configuration/mode/).
*/
/******/ (() => { // webpackBootstrap
/******/ var __webpack_modules__ = ({
/***/ "./node_modules/@rubensworks/saxes/saxes.js":
/*!**************************************************!*\
!*** ./node_modules/@rubensworks/saxes/saxes.js ***!
\**************************************************/
/***/ ((__unused_webpack_module, exports, __webpack_require__) => {
"use strict";
eval("\nObject.defineProperty(exports, \"__esModule\", ({ value: true }));\nexports.SaxesParser = exports.EVENTS = void 0;\nconst ed5 = __webpack_require__(/*! xmlchars/xml/1.0/ed5 */ \"./node_modules/xmlchars/xml/1.0/ed5.js\");\nconst ed2 = __webpack_require__(/*! xmlchars/xml/1.1/ed2 */ \"./node_modules/xmlchars/xml/1.1/ed2.js\");\nconst NSed3 = __webpack_require__(/*! xmlchars/xmlns/1.0/ed3 */ \"./node_modules/xmlchars/xmlns/1.0/ed3.js\");\nvar isS = ed5.isS;\nvar isChar10 = ed5.isChar;\nvar isNameStartChar = ed5.isNameStartChar;\nvar isNameChar = ed5.isNameChar;\nvar S_LIST = ed5.S_LIST;\nvar NAME_RE = ed5.NAME_RE;\nvar isChar11 = ed2.isChar;\nvar isNCNameStartChar = NSed3.isNCNameStartChar;\nvar isNCNameChar = NSed3.isNCNameChar;\nvar NC_NAME_RE = NSed3.NC_NAME_RE;\nconst XML_NAMESPACE = \"http://www.w3.org/XML/1998/namespace\";\nconst XMLNS_NAMESPACE = \"http://www.w3.org/2000/xmlns/\";\nconst rootNS = {\n // eslint-disable-next-line @typescript-eslint/no-explicit-any, @typescript-eslint/no-unsafe-assignment\n __proto__: null,\n xml: XML_NAMESPACE,\n xmlns: XMLNS_NAMESPACE,\n};\nconst XML_ENTITIES = {\n // eslint-disable-next-line @typescript-eslint/no-explicit-any, @typescript-eslint/no-unsafe-assignment\n __proto__: null,\n amp: \"&\",\n gt: \">\",\n lt: \"<\",\n quot: \"\\\"\",\n apos: \"'\",\n};\n// EOC: end-of-chunk\nconst EOC = -1;\nconst NL_LIKE = -2;\nconst S_BEGIN = 0; // Initial state.\nconst S_BEGIN_WHITESPACE = 1; // leading whitespace\nconst S_DOCTYPE = 2; // <!DOCTYPE\nconst S_DOCTYPE_QUOTE = 3; // <!DOCTYPE \"//blah\nconst S_DTD = 4; // <!DOCTYPE \"//blah\" [ ...\nconst S_DTD_QUOTED = 5; // <!DOCTYPE \"//blah\" [ \"foo\nconst S_DTD_OPEN_WAKA = 6;\nconst S_DTD_OPEN_WAKA_BANG = 7;\nconst S_DTD_COMMENT = 8; // <!--\nconst S_DTD_COMMENT_ENDING = 9; // <!-- blah -\nconst S_DTD_COMMENT_ENDED = 10; // <!-- blah --\nconst S_DTD_PI = 11; // <?\nconst S_DTD_PI_ENDING = 12; // <?hi \"there\" ?\nconst S_TEXT = 13; // general stuff\nconst S_ENTITY = 14; // & and such\nconst S_OPEN_WAKA = 15; // <\nconst S_OPEN_WAKA_BANG = 16; // <!...\nconst S_COMMENT = 17; // <!--\nconst S_COMMENT_ENDING = 18; // <!-- blah -\nconst S_COMMENT_ENDED = 19; // <!-- blah --\nconst S_CDATA = 20; // <![CDATA[ something\nconst S_CDATA_ENDING = 21; // ]\nconst S_CDATA_ENDING_2 = 22; // ]]\nconst S_PI_FIRST_CHAR = 23; // <?hi, first char\nconst S_PI_REST = 24; // <?hi, rest of the name\nconst S_PI_BODY = 25; // <?hi there\nconst S_PI_ENDING = 26; // <?hi \"there\" ?\nconst S_XML_DECL_NAME_START = 27; // <?xml\nconst S_XML_DECL_NAME = 28; // <?xml foo\nconst S_XML_DECL_EQ = 29; // <?xml foo=\nconst S_XML_DECL_VALUE_START = 30; // <?xml foo=\nconst S_XML_DECL_VALUE = 31; // <?xml foo=\"bar\"\nconst S_XML_DECL_SEPARATOR = 32; // <?xml foo=\"bar\"\nconst S_XML_DECL_ENDING = 33; // <?xml ... ?\nconst S_OPEN_TAG = 34; // <strong\nconst S_OPEN_TAG_SLASH = 35; // <strong /\nconst S_ATTRIB = 36; // <a\nconst S_ATTRIB_NAME = 37; // <a foo\nconst S_ATTRIB_NAME_SAW_WHITE = 38; // <a foo _\nconst S_ATTRIB_VALUE = 39; // <a foo=\nconst S_ATTRIB_VALUE_QUOTED = 40; // <a foo=\"bar\nconst S_ATTRIB_VALUE_CLOSED = 41; // <a foo=\"bar\"\nconst S_ATTRIB_VALUE_UNQUOTED = 42; // <a foo=bar\nconst S_CLOSE_TAG = 43; // </a\nconst S_CLOSE_TAG_SAW_WHITE = 44; // </a >\nconst TAB = 9;\nconst NL = 0xA;\nconst CR = 0xD;\nconst SPACE = 0x20;\nconst BANG = 0x21;\nconst DQUOTE = 0x22;\nconst AMP = 0x26;\nconst SQUOTE = 0x27;\nconst MINUS = 0x2D;\nconst FORWARD_SLASH = 0x2F;\nconst SEMICOLON = 0x3B;\nconst LESS = 0x3C;\nconst EQUAL = 0x3D;\nconst GREATER = 0x3E;\nconst QUESTION = 0x3F;\nconst OPEN_BRACKET = 0x5B;\nconst CLOSE_BRACKET = 0x5D;\nconst NEL = 0x85;\nconst LS = 0x2028; // Line Separator\nconst isQuote = (c) => c === DQUOTE || c === SQUOTE;\nconst QUOTES = [DQUOTE, SQUOTE];\nconst DOCTYPE_TERMINATOR = [...QUOTES, OPEN_BRACKET, GREATER];\nconst DTD_TERMINATOR = [...QUOTES, LESS, CLOSE_BRACKET];\nconst XML_DECL_NAME_TERMINATOR = [EQUAL, QUESTION, ...S_LIST];\nconst ATTRIB_VALUE_UNQUOTED_TERMINATOR = [...S_LIST, GREATER, AMP, LESS];\nfunction nsPairCheck(parser, prefix, uri) {\n switch (prefix) {\n case \"xml\":\n if (uri !== XML_NAMESPACE) {\n parser.fail(`xml prefix must be bound to ${XML_NAMESPACE}.`);\n }\n break;\n case \"xmlns\":\n if (uri !== XMLNS_NAMESPACE) {\n parser.fail(`xmlns prefix must be bound to ${XMLNS_NAMESPACE}.`);\n }\n break;\n default:\n }\n switch (uri) {\n case XMLNS_NAMESPACE:\n parser.fail(prefix === \"\" ?\n `the default namespace may not be set to ${uri}.` :\n `may not assign a prefix (even \"xmlns\") to the URI \\\n${XMLNS_NAMESPACE}.`);\n break;\n case XML_NAMESPACE:\n switch (prefix) {\n case \"xml\":\n // Assinging the XML namespace to \"xml\" is fine.\n break;\n case \"\":\n parser.fail(`the default namespace may not be set to ${uri}.`);\n break;\n default:\n parser.fail(\"may not assign the xml namespace to another prefix.\");\n }\n break;\n default:\n }\n}\nfunction nsMappingCheck(parser, mapping) {\n for (const local of Object.keys(mapping)) {\n nsPairCheck(parser, local, mapping[local]);\n }\n}\nconst isNCName = (name) => NC_NAME_RE.test(name);\nconst isName = (name) => NAME_RE.test(name);\nconst FORBIDDEN_START = 0;\nconst FORBIDDEN_BRACKET = 1;\nconst FORBIDDEN_BRACKET_BRACKET = 2;\n/**\n * The list of supported events.\n */\nexports.EVENTS = [\n \"xmldecl\",\n \"text\",\n \"processinginstruction\",\n \"doctype\",\n \"comment\",\n \"opentagstart\",\n \"attribute\",\n \"opentag\",\n \"closetag\",\n \"cdata\",\n \"error\",\n \"end\",\n \"ready\",\n];\nconst EVENT_NAME_TO_HANDLER_NAME = {\n xmldecl: \"xmldeclHandler\",\n text: \"textHandler\",\n processinginstruction: \"piHandler\",\n doctype: \"doctypeHandler\",\n comment: \"commentHandler\",\n opentagstart: \"openTagStartHandler\",\n attribute: \"attributeHandler\",\n opentag: \"openTagHandler\",\n closetag: \"closeTagHandler\",\n cdata: \"cdataHandler\",\n error: \"errorHandler\",\n end: \"endHandler\",\n ready: \"readyHandler\",\n};\n// eslint-disable-next-line @typescript-eslint/ban-types\nclass SaxesParser {\n /**\n * Indicates whether or not the parser is closed. If ``true``, wait for\n * the ``ready`` event to write again.\n */\n get closed() {\n return this._closed;\n }\n /**\n * @param opt The parser options.\n */\n constructor(opt) {\n this.opt = opt !== null && opt !== void 0 ? opt : {};\n this.fragmentOpt = !!this.opt.fragment;\n const xmlnsOpt = this.xmlnsOpt = !!this.opt.xmlns;\n this.trackPosition = this.opt.position !== false;\n this.fileName = this.opt.fileName;\n if (xmlnsOpt) {\n // This is the function we use to perform name checks on PIs and entities.\n // When namespaces are used, colons are not allowed in PI target names or\n // entity names. So the check depends on whether namespaces are used. See:\n //\n // https://www.w3.org/XML/xml-names-19990114-errata.html\n // NE08\n //\n this.nameStartCheck = isNCNameStartChar;\n this.nameCheck = isNCNameChar;\n this.isName = isNCName;\n // eslint-disable-next-line @typescript-eslint/unbound-method\n this.processAttribs = this.processAttribsNS;\n // eslint-disable-next-line @typescript-eslint/unbound-method\n this.pushAttrib = this.pushAttribNS;\n // eslint-disable-next-line @typescript-eslint/no-explicit-any, @typescript-eslint/no-unsafe-assignment\n this.ns = Object.assign({ __proto__: null }, rootNS);\n const additional = this.opt.additionalNamespaces;\n if (additional != null) {\n nsMappingCheck(this, additional);\n Object.assign(this.ns, additional);\n }\n }\n else {\n this.nameStartCheck = isNameStartChar;\n this.nameCheck = isNameChar;\n this.isName = isName;\n // eslint-disable-next-line @typescript-eslint/unbound-method\n this.processAttribs = this.processAttribsPlain;\n // eslint-disable-next-line @typescript-eslint/unbound-method\n this.pushAttrib = this.pushAttribPlain;\n }\n //\n // The order of the members in this table needs to correspond to the state\n // numbers given to the states that correspond to the methods being recorded\n // here.\n //\n this.stateTable = [\n /* eslint-disable @typescript-eslint/unbound-method */\n this.sBegin,\n this.sBeginWhitespace,\n this.sDoctype,\n this.sDoctypeQuote,\n this.sDTD,\n this.sDTDQuoted,\n this.sDTDOpenWaka,\n this.sDTDOpenWakaBang,\n this.sDTDComment,\n this.sDTDCommentEnding,\n this.sDTDCommentEnded,\n this.sDTDPI,\n this.sDTDPIEnding,\n this.sText,\n this.sEntity,\n this.sOpenWaka,\n this.sOpenWakaBang,\n this.sComment,\n this.sCommentEnding,\n this.sCommentEnded,\n this.sCData,\n this.sCDataEnding,\n this.sCDataEnding2,\n this.sPIFirstChar,\n this.sPIRest,\n this.sPIBody,\n this.sPIEnding,\n this.sXMLDeclNameStart,\n this.sXMLDeclName,\n this.sXMLDeclEq,\n this.sXMLDeclValueStart,\n this.sXMLDeclValue,\n this.sXMLDeclSeparator,\n this.sXMLDeclEnding,\n this.sOpenTag,\n this.sOpenTagSlash,\n this.sAttrib,\n this.sAttribName,\n this.sAttribNameSawWhite,\n this.sAttribValue,\n this.sAttribValueQuoted,\n this.sAttribValueClosed,\n this.sAttribValueUnquoted,\n this.sCloseTag,\n this.sCloseTagSawWhite,\n /* eslint-enable @typescript-eslint/unbound-method */\n ];\n this._init();\n }\n _init() {\n var _a;\n this.openWakaBang = \"\";\n this.text = \"\";\n this.name = \"\";\n this.piTarget = \"\";\n this.entity = \"\";\n this.q = null;\n this.tags = [];\n this.tag = null;\n this.topNS = null;\n this.chunk = \"\";\n this.chunkPosition = 0;\n this.i = 0;\n this.prevI = 0;\n this.carriedFromPrevious = undefined;\n this.forbiddenState = FORBIDDEN_START;\n this.attribList = [];\n // The logic is organized so as to minimize the need to check\n // this.opt.fragment while parsing.\n const { fragmentOpt } = this;\n this.state = fragmentOpt ? S_TEXT : S_BEGIN;\n // We want these to be all true if we are dealing with a fragment.\n this.reportedTextBeforeRoot = this.reportedTextAfterRoot = this.closedRoot =\n this.sawRoot = fragmentOpt;\n // An XML declaration is intially possible only when parsing whole\n // documents.\n this.xmlDeclPossible = !fragmentOpt;\n this.xmlDeclExpects = [\"version\"];\n this.entityReturnState = undefined;\n let { defaultXMLVersion } = this.opt;\n if (defaultXMLVersion === undefined) {\n if (this.opt.forceXMLVersion === true) {\n throw new Error(\"forceXMLVersion set but defaultXMLVersion is not set\");\n }\n defaultXMLVersion = \"1.0\";\n }\n this.setXMLVersion(defaultXMLVersion);\n this.positionAtNewLine = 0;\n this.doctype = false;\n this._closed = false;\n this.xmlDecl = {\n version: undefined,\n encoding: undefined,\n standalone: undefined,\n };\n this.line = 1;\n this.column = 0;\n this.ENTITIES = Object.create(XML_ENTITIES);\n (_a = this.readyHandler) === null || _a === void 0 ? void 0 : _a.call(this);\n }\n /**\n * The stream position the parser is currently looking at. This field is\n * zero-based.\n *\n * This field is not based on counting Unicode characters but is to be\n * interpreted as a plain index into a JavaScript string.\n */\n get position() {\n return this.chunkPosition + this.i;\n }\n /**\n * The column number of the next character to be read by the parser. *\n * This field is zero-based. (The first column in a line is 0.)\n *\n * This field reports the index at which the next character would be in the\n * line if the line were represented as a JavaScript string. Note that this\n * *can* be different to a count based on the number of *Unicode characters*\n * due to how JavaScript handles astral plane characters.\n *\n * See [[column]] for a number that corresponds to a count of Unicode\n * characters.\n */\n get columnIndex() {\n return this.position - this.positionAtNewLine;\n }\n /**\n * Set an event listener on an event. The parser supports one handler per\n * event type. If you try to set an event handler over an existing handler,\n * the old handler is silently overwritten.\n *\n * @param name The event to listen to.\n *\n * @param handler The handler to set.\n */\n on(name, handler) {\n // eslint-disable-next-line @typescript-eslint/no-explicit-any, @typescript-eslint/no-unsafe-member-access\n this[EVENT_NAME_TO_HANDLER_NAME[name]] = handler;\n }\n /**\n * Unset an event handler.\n *\n * @parma name The event to stop listening to.\n */\n off(name) {\n // eslint-disable-next-line @typescript-eslint/no-explicit-any, @typescript-eslint/no-unsafe-member-access\n this[EVENT_NAME_TO_HANDLER_NAME[name]] = undefined;\n }\n /**\n * Make an error object. The error object will have a message that contains\n * the ``fileName`` option passed at the creation of the parser. If position\n * tracking was turned on, it will also have line and column number\n * information.\n *\n * @param message The message describing the error to report.\n *\n * @returns An error object with a properly formatted message.\n */\n makeError(message) {\n var _a;\n let msg = (_a = this.fileName) !== null && _a !== void 0 ? _a : \"\";\n if (this.trackPosition) {\n if (msg.length > 0) {\n msg += \":\";\n }\n msg += `${this.line}:${this.column}`;\n }\n if (msg.length > 0) {\n msg += \": \";\n }\n return new Error(msg + message);\n }\n /**\n * Report a parsing error. This method is made public so that client code may\n * check for issues that are outside the scope of this project and can report\n * errors.\n *\n * @param message The error to report.\n *\n * @returns this\n */\n fail(message) {\n const err = this.makeError(message);\n const handler = this.errorHandler;\n if (handler === undefined) {\n throw err;\n }\n else {\n handler(err);\n }\n return this;\n }\n /**\n * Write a XML data to the parser.\n *\n * @param chunk The XML data to write.\n *\n * @returns this\n */\n // We do need object for the type here. Yes, it often causes problems\n // but not in this case.\n write(chunk) {\n if (this.closed) {\n return this.fail(\"cannot write after close; assign an onready handler.\");\n }\n let end = false;\n if (chunk === null) {\n // We cannot return immediately because carriedFromPrevious may need\n // processing.\n end = true;\n chunk = \"\";\n }\n else if (typeof chunk === \"object\") {\n chunk = chunk.toString();\n }\n // We checked if performing a pre-decomposition of the string into an array\n // of single complete characters (``Array.from(chunk)``) would be faster\n // than the current repeated calls to ``charCodeAt``. As of August 2018, it\n // isn't. (There may be Node-specific code that would perform faster than\n // ``Array.from`` but don't want to be dependent on Node.)\n if (this.carriedFromPrevious !== undefined) {\n // The previous chunk had char we must carry over.\n chunk = `${this.carriedFromPrevious}${chunk}`;\n this.carriedFromPrevious = undefined;\n }\n let limit = chunk.length;\n const lastCode = chunk.charCodeAt(limit - 1);\n if (!end &&\n // A trailing CR or surrogate must be carried over to the next\n // chunk.\n (lastCode === CR || (lastCode >= 0xD800 && lastCode <= 0xDBFF))) {\n // The chunk ends with a character that must be carried over. We cannot\n // know how to handle it until we get the next chunk or the end of the\n // stream. So save it for later.\n this.carriedFromPrevious = chunk[limit - 1];\n limit--;\n chunk = chunk.slice(0, limit);\n }\n const { stateTable } = this;\n this.chunk = chunk;\n this.i = 0;\n while (this.i < limit) {\n // eslint-disable-next-line @typescript-eslint/no-explicit-any, @typescript-eslint/no-unsafe-argument\n stateTable[this.state].call(this);\n }\n this.chunkPosition += limit;\n return end ? this.end() : this;\n }\n /**\n * Close the current stream. Perform final well-formedness checks and reset\n * the parser tstate.\n *\n * @returns this\n */\n close() {\n return this.write(null);\n }\n /**\n * Get a single code point out of the current chunk. This updates the current\n * position if we do position tracking.\n *\n * This is the algorithm to use for XML 1.0.\n *\n * @returns The character read.\n */\n getCode10() {\n const { chunk, i } = this;\n this.prevI = i;\n // Yes, we do this instead of doing this.i++. Doing it this way, we do not\n // read this.i again, which is a bit faster.\n this.i = i + 1;\n if (i >= chunk.length) {\n return EOC;\n }\n // Using charCodeAt and handling the surrogates ourselves is faster\n // than using codePointAt.\n const code = chunk.charCodeAt(i);\n this.column++;\n if (code < 0xD800) {\n if (code >= SPACE || code === TAB) {\n return code;\n }\n switch (code) {\n case NL:\n this.line++;\n this.column = 0;\n this.positionAtNewLine = this.position;\n return NL;\n case CR:\n // We may get NaN if we read past the end of the chunk, which is fine.\n if (chunk.charCodeAt(i + 1) === NL) {\n // A \\r\\n sequence is converted to \\n so we have to skip over the\n // next character. We already know it has a size of 1 so ++ is fine\n // here.\n this.i = i + 2;\n }\n // Otherwise, a \\r is just converted to \\n, so we don't have to skip\n // ahead.\n // In either case, \\r becomes \\n.\n this.line++;\n this.column = 0;\n this.positionAtNewLine = this.position;\n return NL_LIKE;\n default:\n // If we get here, then code < SPACE and it is not NL CR or TAB.\n this.fail(\"disallowed character.\");\n return code;\n }\n }\n if (code > 0xDBFF) {\n // This is a specialized version of isChar10 that takes into account\n // that in this context code > 0xDBFF and code <= 0xFFFF. So it does not\n // test cases that don't need testing.\n if (!(code >= 0xE000 && code <= 0xFFFD)) {\n this.fail(\"disallowed character.\");\n }\n return code;\n }\n const final = 0x10000 + ((code - 0xD800) * 0x400) +\n (chunk.charCodeAt(i + 1) - 0xDC00);\n this.i = i + 2;\n // This is a specialized version of isChar10 that takes into account that in\n // this context necessarily final >= 0x10000.\n if (final > 0x10FFFF) {\n this.fail(\"disallowed character.\");\n }\n return final;\n }\n /**\n * Get a single code point out of the current chunk. This updates the current\n * position if we do position tracking.\n *\n * This is the algorithm to use for XML 1.1.\n *\n * @returns {number} The character read.\n */\n getCode11() {\n const { chunk, i } = this;\n this.prevI = i;\n // Yes, we do this instead of doing this.i++. Doing it this way, we do not\n // read this.i again, which is a bit faster.\n this.i = i + 1;\n if (i >= chunk.length) {\n return EOC;\n }\n // Using charCodeAt and handling the surrogates ourselves is faster\n // than using codePointAt.\n const code = chunk.charCodeAt(i);\n this.column++;\n if (code < 0xD800) {\n if ((code > 0x1F && code < 0x7F) || (code > 0x9F && code !== LS) ||\n code === TAB) {\n return code;\n }\n switch (code) {\n case NL: // 0xA\n this.line++;\n this.column = 0;\n this.positionAtNewLine = this.position;\n return NL;\n case CR: { // 0xD\n // We may get NaN if we read past the end of the chunk, which is\n // fine.\n const next = chunk.charCodeAt(i + 1);\n if (next === NL || next === NEL) {\n // A CR NL or CR NEL sequence is converted to NL so we have to skip\n // over the next character. We already know it has a size of 1.\n this.i = i + 2;\n }\n // Otherwise, a CR is just converted to NL, no skip.\n }\n /* yes, fall through */\n case NEL: // 0x85\n case LS: // Ox2028\n this.line++;\n this.column = 0;\n this.positionAtNewLine = this.position;\n return NL_LIKE;\n default:\n this.fail(\"disallowed character.\");\n return code;\n }\n }\n if (code > 0xDBFF) {\n // This is a specialized version of isCharAndNotRestricted that takes into\n // account that in this context code > 0xDBFF and code <= 0xFFFF. So it\n // does not test cases that don't need testing.\n if (!(code >= 0xE000 && code <= 0xFFFD)) {\n this.fail(\"disallowed character.\");\n }\n return code;\n }\n const final = 0x10000 + ((code - 0xD800) * 0x400) +\n (chunk.charCodeAt(i + 1) - 0xDC00);\n this.i = i + 2;\n // This is a specialized version of isCharAndNotRestricted that takes into\n // account that in this context necessarily final >= 0x10000.\n if (final > 0x10FFFF) {\n this.fail(\"disallowed character.\");\n }\n return final;\n }\n /**\n * Like ``getCode`` but with the return value normalized so that ``NL`` is\n * returned for ``NL_LIKE``.\n */\n getCodeNorm() {\n const c = this.getCode();\n return c === NL_LIKE ? NL : c;\n }\n unget() {\n this.i = this.prevI;\n this.column--;\n }\n /**\n * Capture characters into a buffer until encountering one of a set of\n * characters.\n *\n * @param chars An array of codepoints. Encountering a character in the array\n * ends the capture. (``chars`` may safely contain ``NL``.)\n *\n * @return The character code that made the capture end, or ``EOC`` if we hit\n * the end of the chunk. The return value cannot be NL_LIKE: NL is returned\n * instead.\n */\n captureTo(chars) {\n let { i: start } = this;\n const { chunk } = this;\n // eslint-disable-next-line no-constant-condition\n while (true) {\n const c = this.getCode();\n const isNLLike = c === NL_LIKE;\n const final = isNLLike ? NL : c;\n if (final === EOC || chars.includes(final)) {\n this.text += chunk.slice(start, this.prevI);\n return final;\n }\n if (isNLLike) {\n this.text += `${chunk.slice(start, this.prevI)}\\n`;\n start = this.i;\n }\n }\n }\n /**\n * Capture characters into a buffer until encountering a character.\n *\n * @param char The codepoint that ends the capture. **NOTE ``char`` MAY NOT\n * CONTAIN ``NL``.** Passing ``NL`` will result in buggy behavior.\n *\n * @return ``true`` if we ran into the character. Otherwise, we ran into the\n * end of the current chunk.\n */\n captureToChar(char) {\n let { i: start } = this;\n const { chunk } = this;\n // eslint-disable-next-line no-constant-condition\n while (true) {\n let c = this.getCode();\n switch (c) {\n case NL_LIKE:\n this.text += `${chunk.slice(start, this.prevI)}\\n`;\n start = this.i;\n c = NL;\n break;\n case EOC:\n this.text += chunk.slice(start);\n return false;\n default:\n }\n if (c === char) {\n this.text += chunk.slice(start, this.prevI);\n return true;\n }\n }\n }\n /**\n * Capture characters that satisfy ``isNameChar`` into the ``name`` field of\n * this parser.\n *\n * @return The character code that made the test fail, or ``EOC`` if we hit\n * the end of the chunk. The return value cannot be NL_LIKE: NL is returned\n * instead.\n */\n captureNameChars() {\n const { chunk, i: start } = this;\n // eslint-disable-next-line no-constant-condition\n while (true) {\n const c = this.getCode();\n if (c === EOC) {\n this.name += chunk.slice(start);\n return EOC;\n }\n // NL is not a name char so we don't have to test specifically for it.\n if (!isNameChar(c)) {\n this.name += chunk.slice(start, this.prevI);\n return c === NL_LIKE ? NL : c;\n }\n }\n }\n /**\n * Skip white spaces.\n *\n * @return The character that ended the skip, or ``EOC`` if we hit\n * the end of the chunk. The return value cannot be NL_LIKE: NL is returned\n * instead.\n */\n skipSpaces() {\n // eslint-disable-next-line no-constant-condition\n while (true) {\n const c = this.getCodeNorm();\n if (c === EOC || !isS(c)) {\n return c;\n }\n }\n }\n setXMLVersion(version) {\n this.currentXMLVersion = version;\n /* eslint-disable @typescript-eslint/unbound-method */\n if (version === \"1.0\") {\n this.isChar = isChar10;\n this.getCode = this.getCode10;\n }\n else {\n this.isChar = isChar11;\n this.getCode = this.getCode11;\n }\n /* eslint-enable @typescript-eslint/unbound-method */\n }\n // STATE ENGINE METHODS\n // This needs to be a state separate from S_BEGIN_WHITESPACE because we want\n // to be sure never to come back to this state later.\n sBegin() {\n // We are essentially peeking at the first character of the chunk. Since\n // S_BEGIN can be in effect only when we start working on the first chunk,\n // the index at which we must look is necessarily 0. Note also that the\n // following test does not depend on decoding surrogates.\n // If the initial character is 0xFEFF, ignore it.\n if (this.chunk.charCodeAt(0) === 0xFEFF) {\n this.i++;\n this.column++;\n }\n this.state = S_BEGIN_WHITESPACE;\n }\n sBeginWhitespace() {\n // We need to know whether we've encountered spaces or not because as soon\n // as we run into a space, an XML declaration is no longer possible. Rather\n // than slow down skipSpaces even in places where we don't care whether it\n // skipped anything or not, we check whether prevI is equal to the value of\n // i from before we skip spaces.\n const iBefore = this.i;\n const c = this.skipSpaces();\n if (this.prevI !== iBefore) {\n this.xmlDeclPossible = false;\n }\n switch (c) {\n case LESS:\n this.state = S_OPEN_WAKA;\n // We could naively call closeText but in this state, it is not normal\n // to have text be filled with any data.\n if (this.text.length !== 0) {\n throw new Error(\"no-empty text at start\");\n }\n break;\n case EOC:\n break;\n default:\n this.unget();\n this.state = S_TEXT;\n this.xmlDeclPossible = false;\n }\n }\n sDoctype() {\n var _a;\n const c = this.captureTo(DOCTYPE_TERMINATOR);\n switch (c) {\n case GREATER: {\n (_a = this.doctypeHandler) === null || _a === void 0 ? void 0 : _a.call(this, this.text);\n this.text = \"\";\n this.state = S_TEXT;\n this.doctype = true; // just remember that we saw it.\n break;\n }\n case EOC:\n break;\n default:\n this.text += String.fromCodePoint(c);\n if (c === OPEN_BRACKET) {\n this.state = S_DTD;\n }\n else if (isQuote(c)) {\n this.state = S_DOCTYPE_QUOTE;\n this.q = c;\n }\n }\n }\n sDoctypeQuote() {\n const q = this.q;\n if (this.captureToChar(q)) {\n this.text += String.fromCodePoint(q);\n this.q = null;\n this.state = S_DOCTYPE;\n }\n }\n sDTD() {\n const c = this.captureTo(DTD_TERMINATOR);\n if (c === EOC) {\n return;\n }\n this.text += String.fromCodePoint(c);\n if (c === CLOSE_BRACKET) {\n this.state = S_DOCTYPE;\n }\n else if (c === LESS) {\n this.state = S_DTD_OPEN_WAKA;\n }\n else if (isQuote(c)) {\n this.state = S_DTD_QUOTED;\n this.q = c;\n }\n }\n sDTDQuoted() {\n const q = this.q;\n if (this.captureToChar(q)) {\n this.text += String.fromCodePoint(q);\n this.state = S_DTD;\n this.q = null;\n }\n }\n sDTDOpenWaka() {\n const c = this.getCodeNorm();\n this.text += String.fromCodePoint(c);\n switch (c) {\n case BANG:\n this.state = S_DTD_OPEN_WAKA_BANG;\n this.openWakaBang = \"\";\n break;\n case QUESTION:\n this.state = S_DTD_PI;\n break;\n default:\n this.state = S_DTD;\n }\n }\n sDTDOpenWakaBang() {\n const char = String.fromCodePoint(this.getCodeNorm());\n const owb = this.openWakaBang += char;\n this.text += char;\n if (owb !== \"-\") {\n this.state = owb === \"--\" ? S_DTD_COMMENT : S_DTD;\n this.openWakaBang = \"\";\n }\n }\n sDTDComment() {\n if (this.captureToChar(MINUS)) {\n this.text += \"-\";\n this.state = S_DTD_COMMENT_ENDING;\n }\n }\n sDTDCommentEnding() {\n const c = this.getCodeNorm();\n this.text += String.fromCodePoint(c);\n this.state = c === MINUS ? S_DTD_COMMENT_ENDED : S_DTD_COMMENT;\n }\n sDTDCommentEnded() {\n const c = this.getCodeNorm();\n this.text += String.fromCodePoint(c);\n if (c === GREATER) {\n this.state = S_DTD;\n }\n else {\n this.fail(\"malformed comment.\");\n // <!-- blah -- bloo --> will be recorded as\n // a comment of \" blah -- bloo \"\n this.state = S_DTD_COMMENT;\n }\n }\n sDTDPI() {\n if (this.captureToChar(QUESTION)) {\n this.text += \"?\";\n this.state = S_DTD_PI_ENDING;\n }\n }\n sDTDPIEnding() {\n const c = this.getCodeNorm();\n this.text += String.fromCodePoint(c);\n if (c === GREATER) {\n this.state = S_DTD;\n }\n }\n sText() {\n //\n // We did try a version of saxes where the S_TEXT state was split in two\n // states: one for text inside the root element, and one for text\n // outside. This was avoiding having to test this.tags.length to decide\n // what implementation to actually use.\n //\n // Peformance testing on gigabyte-size files did not show any advantage to\n // using the two states solution instead of the current one. Conversely, it\n // made the code a bit more complicated elsewhere. For instance, a comment\n // can appear before the root element so when a comment ended it was\n // necessary to determine whether to return to the S_TEXT state or to the\n // new text-outside-root state.\n //\n if (this.tags.length !== 0) {\n this.handleTextInRoot();\n }\n else {\n this.handleTextOutsideRoot();\n }\n }\n sEntity() {\n // This is essentially a specialized version of captureToChar(SEMICOLON...)\n let { i: start } = this;\n const { chunk } = this;\n // eslint-disable-next-line no-labels, no-restricted-syntax\n loop: \n // eslint-disable-next-line no-constant-condition\n while (true) {\n switch (this.getCode()) {\n case NL_LIKE:\n this.entity += `${chunk.slice(start, this.prevI)}\\n`;\n start = this.i;\n break;\n case SEMICOLON: {\n const { entityReturnState } = this;\n const entity = this.entity + chunk.slice(start, this.prevI);\n this.state = entityReturnState;\n let parsed;\n if (entity === \"\") {\n this.fail(\"empty entity name.\");\n parsed = \"&;\";\n }\n else {\n parsed = this.parseEntity(entity);\n this.entity = \"\";\n }\n if (entityReturnState !== S_TEXT || this.textHandler !== undefined) {\n this.text += parsed;\n }\n // eslint-disable-next-line no-labels\n break loop;\n }\n case EOC:\n this.entity += chunk.slice(start);\n // eslint-disable-next-line no-labels\n break loop;\n default:\n }\n }\n }\n sOpenWaka() {\n // Reminder: a state handler is called with at least one character\n // available in the current chunk. So the first call to get code inside of\n // a state handler cannot return ``EOC``. That's why we don't test\n // for it.\n const c = this.getCode();\n // either a /, ?, !, or text is coming next.\n if (isNameStartChar(c)) {\n this.state = S_OPEN_TAG;\n this.unget();\n this.xmlDeclPossible = false;\n }\n else {\n switch (c) {\n case FORWARD_SLASH:\n this.state = S_CLOSE_TAG;\n this.xmlDeclPossible = false;\n break;\n case BANG:\n this.state = S_OPEN_WAKA_BANG;\n this.openWakaBang = \"\";\n this.xmlDeclPossible = false;\n break;\n case QUESTION:\n this.state = S_PI_FIRST_CHAR;\n break;\n default:\n this.fail(\"disallowed character in tag name\");\n this.state = S_TEXT;\n this.xmlDeclPossible = false;\n }\n }\n }\n sOpenWakaBang() {\n this.openWakaBang += String.fromCodePoint(this.getCodeNorm());\n switch (this.openWakaBang) {\n case \"[CDATA[\":\n if (!this.sawRoot && !this.reportedTextBeforeRoot) {\n this.fail(\"text data outside of root node.\");\n this.reportedTextBeforeRoot = true;\n }\n if (this.closedRoot && !this.reportedTextAfterRoot) {\n this.fail(\"text data outside of root node.\");\n this.reportedTextAfterRoot = true;\n }\n this.state = S_CDATA;\n this.openWakaBang = \"\";\n break;\n case \"--\":\n this.state = S_COMMENT;\n this.openWakaBang = \"\";\n break;\n case \"DOCTYPE\":\n this.state = S_DOCTYPE;\n if (this.doctype || this.sawRoot) {\n this.fail(\"inappropriately located doctype declaration.\");\n }\n this.openWakaBang = \"\";\n break;\n default:\n // 7 happens to be the maximum length of the string that can possibly\n // match one of the cases above.\n if (this.openWakaBang.length >= 7) {\n this.fail(\"incorrect syntax.\");\n }\n }\n }\n sComment() {\n if (this.captureToChar(MINUS)) {\n this.state = S_COMMENT_ENDING;\n }\n }\n sCommentEnding() {\n var _a;\n const c = this.getCodeNorm();\n if (c === MINUS) {\n this.state = S_COMMENT_ENDED;\n (_a = this.commentHandler) === null || _a === void 0 ? void 0 : _a.call(this, this.text);\n this.text = \"\";\n }\n else {\n this.text += `-${String.fromCodePoint(c)}`;\n this.state = S_COMMENT;\n }\n }\n sCommentEnded() {\n const c = this.getCodeNorm();\n if (c !== GREATER) {\n this.fail(\"malformed comment.\");\n // <!-- blah -- bloo --> will be recorded as\n // a comment of \" blah -- bloo \"\n this.text += `--${String.fromCodePoint(c)}`;\n this.state = S_COMMENT;\n }\n else {\n this.state = S_TEXT;\n }\n }\n sCData() {\n if (this.captureToChar(CLOSE_BRACKET)) {\n this.state = S_CDATA_ENDING;\n }\n }\n sCDataEnding() {\n const c = this.getCodeNorm();\n if (c === CLOSE_BRACKET) {\n this.state = S_CDATA_ENDING_2;\n }\n else {\n this.text += `]${String.fromCodePoint(c)}`;\n this.state = S_CDATA;\n }\n }\n sCDataEnding2() {\n var _a;\n const c = this.getCodeNorm();\n switch (c) {\n case GREATER: {\n (_a = this.cdataHandler) === null || _a === void 0 ? void 0 : _a.call(this, this.text);\n this.text = \"\";\n this.state = S_TEXT;\n break;\n }\n case CLOSE_BRACKET:\n this.text += \"]\";\n break;\n default:\n this.text += `]]${String.fromCodePoint(c)}`;\n this.state = S_CDATA;\n }\n }\n // We need this separate state to check the first character fo the pi target\n // with this.nameStartCheck which allows less characters than this.nameCheck.\n sPIFirstChar() {\n const c = this.getCodeNorm();\n // This is first because in the case where the file is well-formed this is\n // the branch taken. We optimize for well-formedness.\n if (this.nameStartCheck(c)) {\n this.piTarget += String.fromCodePoint(c);\n this.state = S_PI_REST;\n }\n else if (c === QUESTION || isS(c)) {\n this.fail(\"processing instruction without a target.\");\n this.state = c === QUESTION ? S_PI_ENDING : S_PI_BODY;\n }\n else {\n this.fail(\"disallowed character in processing instruction name.\");\n this.piTarget += String.fromCodePoint(c);\n this.state = S_PI_REST;\n }\n }\n sPIRest() {\n // Capture characters into a piTarget while ``this.nameCheck`` run on the\n // character read returns true.\n const { chunk, i: start } = this;\n // eslint-disable-next-line no-constant-condition\n while (true) {\n const c = this.getCodeNorm();\n if (c === EOC) {\n this.piTarget += chunk.slice(start);\n return;\n }\n // NL cannot satisfy this.nameCheck so we don't have to test specifically\n // for it.\n if (!this.nameCheck(c)) {\n this.piTarget += chunk.slice(start, this.prevI);\n const isQuestion = c === QUESTION;\n if (isQuestion || isS(c)) {\n if (this.piTarget === \"xml\") {\n if (!this.xmlDeclPossible) {\n this.fail(\"an XML declaration must be at the start of the document.\");\n }\n this.state = isQuestion ? S_XML_DECL_ENDING : S_XML_DECL_NAME_START;\n }\n else {\n this.state = isQuestion ? S_PI_ENDING : S_PI_BODY;\n }\n }\n else {\n this.fail(\"disallowed character in processing instruction name.\");\n this.piTarget += String.fromCodePoint(c);\n }\n break;\n }\n }\n }\n sPIBody() {\n if (this.text.length === 0) {\n const c = this.getCodeNorm();\n if (c === QUESTION) {\n this.state = S_PI_ENDING;\n }\n else if (!isS(c)) {\n this.text = String.fromCodePoint(c);\n }\n }\n // The question mark character is not valid inside any of the XML\n // declaration name/value pairs.\n else if (this.captureToChar(QUESTION)) {\n this.state = S_PI_ENDING;\n }\n }\n sPIEnding() {\n var _a;\n const c = this.getCodeNorm();\n if (c === GREATER) {\n const { piTarget } = this;\n if (piTarget.toLowerCase() === \"xml\") {\n this.fail(\"the XML declaration must appear at the start of the document.\");\n }\n (_a = this.piHandler) === null || _a === void 0 ? void 0 : _a.call(this, {\n target: piTarget,\n body: this.text,\n });\n this.piTarget = this.text = \"\";\n this.state = S_TEXT;\n }\n else if (c === QUESTION) {\n // We ran into ?? as part of a processing instruction. We initially took\n // the first ? as a sign that the PI was ending, but it is not. So we have\n // to add it to the body but we take the new ? as a sign that the PI is\n // ending.\n this.text += \"?\";\n }\n else {\n this.text += `?${String.fromCodePoint(c)}`;\n this.state = S_PI_BODY;\n }\n this.xmlDeclPossible = false;\n }\n sXMLDeclNameStart() {\n const c = this.skipSpaces();\n // The question mark character is not valid inside any of the XML\n // declaration name/value pairs.\n if (c === QUESTION) {\n // It is valid to go to S_XML_DECL_ENDING from this state.\n this.state = S_XML_DECL_ENDING;\n return;\n }\n if (c !== EOC) {\n this.state = S_XML_DECL_NAME;\n this.name = String.fromCodePoint(c);\n }\n }\n sXMLDeclName() {\n const c = this.captureTo(XML_DECL_NAME_TERMINATOR);\n // The question mark character is not valid inside any of the XML\n // declaration name/value pairs.\n if (c === QUESTION) {\n this.state = S_XML_DECL_ENDING;\n this.name += this.text;\n this.text = \"\";\n this.fail(\"XML declaration is incomplete.\");\n return;\n }\n if (!(isS(c) || c === EQUAL)) {\n return;\n }\n this.name += this.text;\n this.text = \"\";\n if (!this.xmlDeclExpects.includes(this.name)) {\n switch (this.name.length) {\n case 0:\n this.fail(\"did not expect any more name/value pairs.\");\n break;\n case 1:\n this.fail(`expected the name ${this.xmlDeclExpects[0]}.`);\n break;\n default:\n this.fail(`expected one of ${this.xmlDeclExpects.join(\", \")}`);\n }\n }\n this.state = c === EQUAL ? S_XML_DECL_VALUE_START : S_XML_DECL_EQ;\n }\n sXMLDeclEq() {\n const c = this.getCodeNorm();\n // The question mark character is not valid inside any of the XML\n // declaration name/value pairs.\n if (c === QUESTION) {\n this.state = S_XML_DECL_ENDING;\n this.fail(\"XML declaration is incomplete.\");\n return;\n }\n if (isS(c)) {\n return;\n }\n if (c !== EQUAL) {\n this.fail(\"value required.\");\n }\n this.state = S_XML_DECL_VALUE_START;\n }\n sXMLDeclValueStart() {\n const c = this.getCodeNorm();\n // The question mark character is not valid inside any of the XML\n // declaration name/value pairs.\n if (c === QUESTION) {\n this.state = S_XML_DECL_ENDING;\n this.fail(\"XML declaration is incomplete.\");\n return;\n }\n if (isS(c)) {\n return;\n }\n if (!isQuote(c)) {\n this.fail(\"value must be quoted.\");\n this.q = SPACE;\n }\n else {\n this.q = c;\n }\n this.state = S_XML_DECL_VALUE;\n }\n sXMLDeclValue() {\n const c = this.captureTo([this.q, QUESTION]);\n // The question mark character is not valid inside any of the XML\n // declaration name/value pairs.\n if (c === QUESTION) {\n this.state = S_XML_DECL_ENDING;\n this.text = \"\";\n this.fail(\"XML declaration is incomplete.\");\n return;\n }\n if (c === EOC) {\n return;\n }\n const value = this.text;\n this.text = \"\";\n switch (this.name) {\n case \"version\": {\n this.xmlDeclExpects = [\"encoding\", \"standalone\"];\n const version = value;\n this.xmlDecl.version = ve