wikiparser-node 0.3.0 → 0.4.0

package/parser/commentAndExt.js

@@ -1,26 +1,28 @@
 'use strict';
 
-const /** @type {Parser} */ Parser = require('..');
+const Parser = require('..');
 
 /**
- * @param {string} text
+ * 解析HTML注释和扩展标签
+ * @param {string} text wikitext
  * @param {accum} accum
+ * @param {boolean} includeOnly 是否嵌入
  */
 const parseCommentAndExt = (text, config = Parser.getConfig(), accum = [], includeOnly = false) => {
-	const onlyinclude = /<onlyinclude>(.*?)<\/onlyinclude>/gs;
-	if (includeOnly && onlyinclude.test(text)) { // `<onlyinclude>`拥有最高优先级
-		return text.replace(onlyinclude, /** @param {string} inner */ (_, inner) => {
-			const str = `\x00${accum.length}e\x7f`,
-				OnlyincludeToken = require('../src/onlyinclude');
+	const onlyinclude = /<onlyinclude>(.*?)<\/onlyinclude>/gsu;
+	if (includeOnly && text.search(onlyinclude) !== -1) { // `<onlyinclude>`拥有最高优先级
+		return text.replaceAll(onlyinclude, /** @param {string} inner */ (_, inner) => {
+			const str = `\0${accum.length}e\x7F`;
+			const OnlyincludeToken = require('../src/onlyinclude');
 			new OnlyincludeToken(inner, config, accum);
 			return str;
-		}).replace(/(?<=^|\x00\d+e\x7f).*?(?=$|\x00\d+e\x7f)/gs, substr => {
+		}).replaceAll(/(?<=^|\0\d+e\x7F).*?(?=$|\0\d+e\x7F)/gsu, substr => {
 			if (substr === '') {
 				return '';
 			}
 			const NoincludeToken = require('../src/nowiki/noinclude');
 			new NoincludeToken(substr, config, accum);
-			return `\x00${accum.length - 1}c\x7f`;
+			return `\0${accum.length - 1}c\x7F`;
 		});
 	}
 	const ext = config.ext.join('|'),
@@ -28,22 +30,22 @@ const parseCommentAndExt = (text, config = Parser.getConfig(), accum = [], inclu
 		noincludeRegex = includeOnly ? 'noinclude' : 'includeonly',
 		regex = new RegExp(
 			'<!--.*?(?:-->|$)|' // comment
-			+ `<${includeRegex}(?:\\s.*?)?>|</${includeRegex}\\s*>|` // <includeonly>
-			+ `<(${ext})(\\s.*?)?(?:/>|>(.*?)</(\\1\\s*)>)|` // 扩展标签
-			+ `<(${noincludeRegex})(\\s.*?)?(?:/>|>(.*?)(?:</(\\5\\s*)>|$))`, // <noinclude>
-			'gis',
+			+ `<${includeRegex}(?:\\s[^>]*?)?>|</${includeRegex}\\s*>|` // <includeonly>
+			+ `<(${ext})(\\s[^>]*?)?(?:/>|>(.*?)</(\\1\\s*)>)|` // 扩展标签
+			+ `<(${noincludeRegex})(\\s[^>]*?)?(?:/>|>(.*?)(?:</(\\5\\s*)>|$))`, // <noinclude>
+			'gisu',
 		);
 	return text.replace(
 		regex,
 		/** @type {function(...string): string} */
 		(substr, name, attr, inner, closing, include, includeAttr, includeInner, includeClosing) => {
-			const str = `\x00${accum.length}${name ? 'e' : 'c'}\x7f`;
+			const str = `\0${accum.length}${name ? 'e' : 'c'}\x7F`;
 			if (name) {
 				const ExtToken = require('../src/tagPair/ext');
 				new ExtToken(name, attr, inner, closing, config, accum);
 			} else if (substr.startsWith('<!--')) {
-				const CommentToken = require('../src/nowiki/comment'),
-					closed = substr.endsWith('-->');
+				const CommentToken = require('../src/nowiki/comment');
+				const closed = substr.endsWith('-->');
 				new CommentToken(substr.slice(4, closed ? -3 : undefined), closed, config, accum);
 			} else if (include) {
 				const IncludeToken = require('../src/tagPair/include');

package/parser/converter.js

@@ -1,32 +1,33 @@
 'use strict';
 
-const /** @type {Parser} */ Parser = require('..');
+const Parser = require('..');
 
 /**
- * @param {string} firstChild
+ * 解析语言变体转换
+ * @param {string} wikitext wikitext
  * @param {accum} accum
  */
-const parseConverter = (firstChild, config = Parser.getConfig(), accum = []) => {
-	const ConverterToken = require('../src/converter'),
-		regex1 = /-{/g,
-		regex2 = /-{|}-/g,
+const parseConverter = (wikitext, config = Parser.getConfig(), accum = []) => {
+	const ConverterToken = require('../src/converter');
+	const regex1 = /-\{/gu,
+		regex2 = /-\{|\}-/gu,
 		/** @type {RegExpExecArray[]} */ stack = [];
 	let regex = regex1,
-		mt = regex.exec(firstChild);
+		mt = regex.exec(wikitext);
 	while (mt) {
 		const {0: syntax, index} = mt;
 		if (syntax === '}-') {
 			const top = stack.pop(),
 				{length} = accum,
-				str = firstChild.slice(top.index + 2, index),
+				str = wikitext.slice(top.index + 2, index),
 				i = str.indexOf('|'),
 				[flags, text] = i === -1 ? [[], str] : [str.slice(0, i).split(';'), str.slice(i + 1)],
-				temp = text.replace(/(?<=&[#a-z0-9]+);/i, '\x01'),
+				temp = text.replace(/(&[#a-z\d]+);/iu, '$1\x01'),
 				variants = `(?:${config.variants.join('|')})`,
-				rules = temp.split(new RegExp(`;(?=\\s*(?:${variants}|[^;]*?=>\\s*${variants})\\s*:)`))
+				rules = temp.split(new RegExp(`;(?=\\s*(?:${variants}|[^;]*?=>\\s*${variants})\\s*:)`, 'u'))
 					.map(rule => rule.replaceAll('\x01', ';'));
 			new ConverterToken(flags, rules, config, accum);
-			firstChild = `${firstChild.slice(0, top.index)}\x00${length}v\x7f${firstChild.slice(index + 2)}`;
+			wikitext = `${wikitext.slice(0, top.index)}\0${length}v\x7F${wikitext.slice(index + 2)}`;
 			if (stack.length === 0) {
 				regex = regex1;
 			}
@@ -35,9 +36,9 @@ const parseConverter = (firstChild, config = Parser.getConfig(), accum = []) =>
 			stack.push(mt);
 			regex = regex2;
 		}
-		mt = regex.exec(firstChild);
+		mt = regex.exec(wikitext);
 	}
-	return firstChild;
+	return wikitext;
 };
 
 Parser.parsers.parseConverter = __filename;

package/parser/externalLinks.js

@@ -1,28 +1,29 @@
 'use strict';
 
 const {extUrlChar} = require('../util/string'),
-	/** @type {Parser} */ Parser = require('..');
+	Parser = require('..');
 
 /**
- * @param {string} firstChild
+ * 解析外部链接
+ * @param {string} wikitext wikitext
  * @param {accum} accum
  */
-const parseExternalLinks = (firstChild, config = Parser.getConfig(), accum = []) => {
-	const ExtLinkToken = require('../src/extLink'),
-		regex = new RegExp(
-			`\\[((?:${config.protocol}|//)${extUrlChar})(\\p{Zs}*)([^\\]\x01-\x08\x0a-\x1f\ufffd]*)\\]`,
-			'gui',
-		);
-	return firstChild.replace(regex, /** @type {function(...string): string} */ (_, url, space, text) => {
+const parseExternalLinks = (wikitext, config = Parser.getConfig(), accum = []) => {
+	const ExtLinkToken = require('../src/extLink');
+	const regex = new RegExp(
+		`\\[((?:${config.protocol}|//)${extUrlChar})(\\p{Zs}*)([^\\]\x01-\x08\x0A-\x1F\uFFFD]*)\\]`,
+		'giu',
+	);
+	return wikitext.replace(regex, /** @type {function(...string): string} */ (_, url, space, text) => {
 		const {length} = accum,
-			mt = url.match(/&[lg]t;/);
+			mt = /&[lg]t;/u.exec(url);
 		if (mt) {
 			url = url.slice(0, mt.index);
 			space = '';
 			text = `${url.slice(mt.index)}${space}${text}`;
 		}
 		new ExtLinkToken(url, space, text, config, accum);
-		return `\x00${length}w\x7f`;
+		return `\0${length}w\x7F`;
 	});
 };
 

package/parser/hrAndDoubleUnderscore.js

@@ -1,25 +1,34 @@
 'use strict';
 
-const /** @type {Parser} */ Parser = require('..');
+const Parser = require('..'),
+	AstText = require('../lib/text');
 
 /**
- * @param {string} firstChild
+ * 解析\<hr\>和状态开关
+ * @param {{firstChild: AstText, type: string}} root 根节点
  * @param {accum} accum
  */
-const parseHrAndDoubleUnderscore = (firstChild, config = Parser.getConfig(), accum = []) => {
+const parseHrAndDoubleUnderscore = ({firstChild: {data}, type}, config = Parser.getConfig(), accum = []) => {
 	const HrToken = require('../src/nowiki/hr'),
-		DoubleUnderscoreToken = require('../src/nowiki/doubleUnderscore'),
-		{doubleUnderscore} = config;
-	return firstChild.replace(/(?<=^(?:\x00\d+c\x7f)*)-{4,}/mg, m => {
+		DoubleUnderscoreToken = require('../src/nowiki/doubleUnderscore');
+	const {doubleUnderscore} = config;
+	if (type !== 'root') {
+		data = `\0${data}`;
+	}
+	data = data.replaceAll(/^((?:\0\d+c\x7F)*)(-{4,})/gmu, (_, lead, m) => {
 		new HrToken(m.length, config, accum);
-		return `\x00${accum.length - 1}r\x7f`;
-	}).replace(new RegExp(`__(${doubleUnderscore.flat().join('|')})__`, 'ig'), /** @param {string} p1 */(m, p1) => {
-		if (doubleUnderscore[0].includes(p1.toLowerCase()) || doubleUnderscore[1].includes(p1)) {
-			new DoubleUnderscoreToken(p1, config, accum);
-			return `\x00${accum.length - 1}u\x7f`;
-		}
-		return m;
-	});
+		return `${lead}\0${accum.length - 1}r\x7F`;
+	}).replaceAll(
+		new RegExp(`__(${doubleUnderscore.flat().join('|')})__`, 'giu'),
+		/** @param {string} p1 */ (m, p1) => {
+			if (doubleUnderscore[0].includes(p1.toLowerCase()) || doubleUnderscore[1].includes(p1)) {
+				new DoubleUnderscoreToken(p1, config, accum);
+				return `\0${accum.length - 1}u\x7F`;
+			}
+			return m;
+		},
+	);
+	return type === 'root' ? data : data.slice(1);
 };
 
 Parser.parsers.parseHrAndDoubleUnderscore = __filename;

package/parser/html.js

@@ -1,26 +1,27 @@
 'use strict';
 
-const /** @type {Parser} */ Parser = require('..');
+const Parser = require('..');
 
 /**
- * @param {string} firstChild
+ * 解析HTML标签
+ * @param {string} wikitext wikitext
  * @param {accum} accum
  */
-const parseHtml = (firstChild, config = Parser.getConfig(), accum = []) => {
-	const regex = /^(\/?)([a-z][^\s/>]*)([^>]*?)(\/?>)([^<]*)$/i,
+const parseHtml = (wikitext, config = Parser.getConfig(), accum = []) => {
+	const regex = /^(\/?)([a-z][^\s/>]*)(\s[^>]*?)?(\/?>)([^<]*)$/iu,
 		elements = config.html.flat(),
-		bits = firstChild.split('<');
+		bits = wikitext.split('<');
 	let text = bits.shift();
 	for (const x of bits) {
-		const mt = x.match(regex),
+		const mt = regex.exec(x),
 			t = mt?.[2],
 			name = t?.toLowerCase();
 		if (!mt || !elements.includes(name)) {
 			text += `<${x}`;
 			continue;
 		}
-		const [, slash,, params, brace, rest] = mt,
-			AttributeToken = require('../src/attribute'),
+		const AttributeToken = require('../src/attribute');
+		const [, slash,, params = '', brace, rest] = mt,
 			attr = new AttributeToken(params, 'html-attr', name, config, accum),
 			itemprop = attr.getAttr('itemprop');
 		if (name === 'meta' && (itemprop === undefined || attr.getAttr('content') === undefined)
@@ -30,7 +31,7 @@ const parseHtml = (firstChild, config = Parser.getConfig(), accum = []) => {
 			accum.pop();
 			continue;
 		}
-		text += `\x00${accum.length}x\x7f${rest}`;
+		text += `\0${accum.length}x\x7F${rest}`;
 		const HtmlToken = require('../src/html');
 		new HtmlToken(t, attr, slash === '/', brace === '/>', config, accum);
 	}

package/parser/links.js

@@ -1,37 +1,38 @@
 'use strict';
 
-const /** @type {Parser} */ Parser = require('..'),
+const Parser = require('..'),
 	Token = require('../src');
 
 /**
- * @param {string} firstChild
+ * 解析内部链接
+ * @param {string} wikitext wikitext
  * @param {accum} accum
  */
-const parseLinks = (firstChild, config = Parser.getConfig(), accum = []) => {
-	const parseQuotes = require('./quotes.js'),
-		regex = /^([^\n<>[\]{}|]+)(?:\|(.+?))?]](.*)/s,
-		regexImg = /^([^\n<>[\]{}|]+)\|(.*)/s,
-		regexExt = new RegExp(`^\\s*(?:${config.protocol})`, 'i'),
-		bits = firstChild.split('[[');
+const parseLinks = (wikitext, config = Parser.getConfig(), accum = []) => {
+	const parseQuotes = require('./quotes.js');
+	const regex = /^([^\n<>[\]{}|]+)(?:\|(.*?[^\]]))?\]\](.*)$/su,
+		regexImg = /^([^\n<>[\]{}|]+)\|(.*)$/su,
+		regexExt = new RegExp(`^\\s*(?:${config.protocol})`, 'iu'),
+		bits = wikitext.split('[[');
 	let s = bits.shift();
 	for (let i = 0; i < bits.length; i++) {
 		let mightBeImg, link, text, after;
 		const x = bits[i],
-			m = x.match(regex);
+			m = regex.exec(x);
 		if (m) {
 			[, link, text, after] = m;
-			if (after.startsWith(']') && text?.includes('[')) {
+			if (after[0] === ']' && text?.includes('[')) {
 				text += ']';
 				after = after.slice(1);
 			}
 		} else {
-			const m2 = x.match(regexImg);
+			const m2 = regexImg.exec(x);
 			if (m2) {
 				mightBeImg = true;
 				[, link, text] = m2;
 			}
 		}
-		if (link === undefined || regexExt.test(link) || /\x00\d+[exhbru]\x7f/.test(link)) {
+		if (link === undefined || regexExt.test(link) || /\0\d+[exhbru]\x7F/u.test(link)) {
 			s += `[[${x}`;
 			continue;
 		}
@@ -41,7 +42,7 @@ const parseLinks = (firstChild, config = Parser.getConfig(), accum = []) => {
 				page = decodeURIComponent(link);
 			} catch {}
 		}
-		const force = link.trim().startsWith(':');
+		const force = link.trim()[0] === ':';
 		if (force && mightBeImg) {
 			s += `[[${x}`;
 			continue;
@@ -79,7 +80,7 @@ const parseLinks = (firstChild, config = Parser.getConfig(), accum = []) => {
 			}
 		}
 		text &&= parseQuotes(text, config, accum);
-		s += `\x00${accum.length}l\x7f${after}`;
+		s += `\0${accum.length}l\x7F${after}`;
 		let LinkToken = require('../src/link');
 		if (!force) {
 			if (!interwiki && ns === 6) {

package/parser/list.js

@@ -1,26 +1,27 @@
 'use strict';
 
-const /** @type {Parser} */ Parser = require('..');
+const Parser = require('..');
 
 /**
- * @param {string} text
+ * 解析列表
+ * @param {string} text wikitext
  * @param {accum} accum
  */
 const parseList = (text, config = Parser.getConfig(), accum = []) => {
-	const mt = text.match(/^((?:\x00\d+c\x7f)*)([;:*#]+)/);
+	const mt = /^((?:\0\d+c\x7F)*)([;:*#]+)/u.exec(text);
 	if (!mt) {
 		return text;
 	}
-	const ListToken = require('../src/nowiki/list'),
-		[total, comment, prefix] = mt;
-	text = `${comment}\x00${accum.length}d\x7f${text.slice(total.length)}`;
+	const ListToken = require('../src/nowiki/list');
+	const [total, comment, prefix] = mt;
+	text = `${comment}\0${accum.length}d\x7F${text.slice(total.length)}`;
 	new ListToken(prefix, config, accum);
 	let dt = prefix.split(';').length - 1;
 	if (!dt) {
 		return text;
 	}
 	const DdToken = require('../src/nowiki/dd');
-	let regex = /:+|-{/g,
+	let regex = /:+|-\{/gu,
 		ex = regex.exec(text),
 		lc = 0;
 	while (ex && dt) {
@@ -28,16 +29,16 @@ const parseList = (text, config = Parser.getConfig(), accum = []) => {
 		if (syntax[0] === ':') {
 			if (syntax.length >= dt) {
 				new DdToken(':'.repeat(dt), config, accum);
-				return `${text.slice(0, index)}\x00${accum.length - 1}d\x7f${text.slice(index + dt)}`;
+				return `${text.slice(0, index)}\0${accum.length - 1}d\x7F${text.slice(index + dt)}`;
 			}
-			text = `${text.slice(0, index)}\x00${accum.length}d\x7f${text.slice(regex.lastIndex)}`;
+			text = `${text.slice(0, index)}\0${accum.length}d\x7F${text.slice(regex.lastIndex)}`;
 			dt -= syntax.length;
 			regex.lastIndex = index + 4 + String(accum.length).length;
 			new DdToken(syntax, config, accum);
 		} else if (syntax === '-{') {
 			if (!lc) {
 				const {lastIndex} = regex;
-				regex = /-{|}-/g;
+				regex = /-\{|\}-/gu;
 				regex.lastIndex = lastIndex;
 			}
 			lc++;
@@ -45,7 +46,7 @@ const parseList = (text, config = Parser.getConfig(), accum = []) => {
 			lc--;
 			if (!lc) {
 				const {lastIndex} = regex;
-				regex = /:+|-{/g;
+				regex = /:+|-\{/gu;
 				regex.lastIndex = lastIndex;
 			}
 		}

package/parser/magicLinks.js

@@ -1,28 +1,29 @@
 'use strict';
 
 const {extUrlChar} = require('../util/string'),
-	/** @type {Parser} */ Parser = require('..');
+	Parser = require('..');
 
 /**
- * @param {string} firstChild
+ * 解析自由外链
+ * @param {string} wikitext wikitext
  * @param {accum} accum
  */
-const parseMagicLinks = (firstChild, config = Parser.getConfig(), accum = []) => {
-	const MagicLinkToken = require('../src/magicLink'),
-		regex = new RegExp(`\\b(?:${config.protocol})(${extUrlChar})`, 'gui');
-	return firstChild.replace(regex, /** @param {string} p1 */ (m, p1) => {
+const parseMagicLinks = (wikitext, config = Parser.getConfig(), accum = []) => {
+	const MagicLinkToken = require('../src/magicLink');
+	const regex = new RegExp(`\\b(?:${config.protocol})(${extUrlChar})`, 'giu');
+	return wikitext.replace(regex, /** @param {string} p1 */ (m, p1) => {
 		let trail = '',
 			url = m;
-		const m2 = url.match(/&(?:lt|gt|nbsp|#x0*(?:3[ce]|a0)|#0*(?:6[02]|160));/i);
+		const m2 = /&(?:lt|gt|nbsp|#x0*(?:3[ce]|a0)|#0*(?:6[02]|160));/iu.exec(url);
 		if (m2) {
 			trail = url.slice(m2.index);
 			url = url.slice(0, m2.index);
 		}
-		const sep = new RegExp(`[,;.:!?${url.includes('(') ? '' : ')'}]+$`),
-			sepChars = url.match(sep);
+		const sep = new RegExp(`[,;.:!?${url.includes('(') ? '' : ')'}]+$`, 'u'),
+			sepChars = sep.exec(url);
 		if (sepChars) {
 			let correction = 0;
-			if (sepChars[0].startsWith(';') && /&(?:[a-z]+|#x[\da-f]+|#\d+)$/i.test(url.slice(0, sepChars.index))) {
+			if (sepChars[0][0] === ';' && /&(?:[a-z]+|#x[\da-f]+|#\d+)$/iu.test(url.slice(0, sepChars.index))) {
 				correction = 1;
 			}
 			trail = `${url.slice(sepChars.index + correction)}${trail}`;
@@ -32,7 +33,7 @@ const parseMagicLinks = (firstChild, config = Parser.getConfig(), accum = []) =>
 			return m;
 		}
 		new MagicLinkToken(url, false, config, accum);
-		return `\x00${accum.length - 1}w\x7f${trail}`;
+		return `\0${accum.length - 1}w\x7F${trail}`;
 	});
 };
 

package/parser/quotes.js

@@ -1,13 +1,14 @@
 'use strict';
 
-const /** @type {Parser} */ Parser = require('..');
+const Parser = require('..');
 
 /**
- * @param {string} text
+ * 解析单引号
+ * @param {string} text wikitext
  * @param {accum} accum
  */
 const parseQuotes = (text, config = Parser.getConfig(), accum = []) => {
-	const arr = text.split(/('{2,})/),
+	const arr = text.split(/('{2,})/u),
 		{length} = arr;
 	if (length === 1) {
 		return text;
@@ -16,7 +17,7 @@ const parseQuotes = (text, config = Parser.getConfig(), accum = []) => {
 		nItalic = 0,
 		firstSingle, firstMulti, firstSpace;
 	for (let i = 1; i < length; i += 2) {
-		const len = arr[i].length;
+		const {length: len} = arr[i];
 		switch (len) {
 			case 2:
 				nItalic++;
@@ -54,7 +55,7 @@ const parseQuotes = (text, config = Parser.getConfig(), accum = []) => {
 	const QuoteToken = require('../src/nowiki/quote');
 	for (let i = 1; i < length; i += 2) {
 		new QuoteToken(arr[i].length, config, accum);
-		arr[i] = `\x00${accum.length - 1}q\x7f`;
+		arr[i] = `\0${accum.length - 1}q\x7F`;
 	}
 	return arr.join('');
 };

package/parser/selector.js

@@ -0,0 +1,175 @@
+'use strict';
+
+const Parser = require('..');
+
+const /** @type {pseudo[]} */ simplePseudos = [
+		'root',
+		'first-child',
+		'first-of-type',
+		'last-child',
+		'last-of-type',
+		'only-child',
+		'only-of-type',
+		'empty',
+		'parent',
+		'header',
+		'hidden',
+		'visible',
+		'only-whitespace',
+		'local-link',
+		'read-only',
+		'read-write',
+		'invalid',
+		'required',
+		'optional',
+	],
+	/** @type {pseudo[]} */ complexPseudos = [
+		'is',
+		'not',
+		'nth-child',
+		'nth-of-type',
+		'nth-last-child',
+		'nth-last-of-type',
+		'contains',
+		'has',
+		'lang',
+	],
+	specialChars = [
+		['[', '['],
+		[']', ']'],
+		['(', '('],
+		[')', ')'],
+		['"', '"'],
+		["'", '''],
+		[':', ':'],
+		['\\', '\'],
+		['&', '&'],
+	],
+	pseudoRegex = new RegExp(`:(${complexPseudos.join('|')})$`, 'u'),
+	regularRegex = /[[(,>+~]|\s+/u, // eslint-disable-line regexp/no-super-linear-move
+	attributeRegex = /^\s*(\w+)\s*(?:([~|^$*!]?=)\s*("[^"]*"|'[^']*'|[^\s[\]]+)(?:\s+(i))?\s*)?\]/u,
+	functionRegex = /^(\s*"[^"]*"\s*|\s*'[^']*'\s*|[^()]*)\)/u;
+
+/**
+ * 清理转义符号
+ * @param {string} selector
+ */
+const sanitize = selector => {
+	for (const [c, escaped] of specialChars) {
+		selector = selector.replaceAll(`\\${c}`, escaped);
+	}
+	return selector;
+};
+
+/**
+ * 还原转义符号
+ * @param {string|undefined} selector
+ */
+const desanitize = selector => {
+	if (selector === undefined) {
+		return undefined;
+	}
+	for (const [c, escaped] of specialChars) {
+		selector = selector.replaceAll(escaped, c);
+	}
+	return selector.trim();
+};
+
+/**
+ * 去除首尾的引号
+ * @param {string|undefined} val 属性值或伪选择器函数的参数
+ */
+const deQuote = val => {
+	if (val === undefined) {
+		return undefined;
+	}
+	const quotes = /^(["']).*\1$/u.exec(val)?.[1];
+	return quotes ? val.slice(1, -1) : val;
+};
+
+/**
+ * 解析简单伪选择器
+ * @param {SelectorArray} step 当前顶部
+ * @param {string} str 不含属性和复杂伪选择器的语句
+ * @throws `SyntaxError` 非法的选择器
+ */
+const pushSimple = (step, str) => {
+	const pieces = str.trim().split(':'),
+		// eslint-disable-next-line unicorn/explicit-length-check
+		i = pieces.slice(1).findIndex(pseudo => simplePseudos.includes(pseudo)) + 1 || pieces.length;
+	if (pieces.slice(i).some(pseudo => !simplePseudos.includes(pseudo))) {
+		throw new SyntaxError(`非法的选择器！\n${str}\n可能需要将':'转义为'\\:'。`);
+	}
+	step.push(desanitize(pieces.slice(0, i).join(':')), ...pieces.slice(i).map(piece => `:${piece}`));
+};
+
+/**
+ * 解析选择器
+ * @param {string} selector
+ * @throws `SyntaxError` 非法的选择器
+ */
+const parseSelector = selector => {
+	selector = selector.trim();
+	const /** @type {SelectorArray[][]} */ stack = [[[]]];
+	let sanitized = sanitize(selector),
+		regex = regularRegex,
+		mt = regex.exec(sanitized),
+		[condition] = stack,
+		[step] = condition;
+	while (mt) {
+		let {0: syntax, index} = mt;
+		if (syntax.trim() === '') {
+			index += syntax.length;
+			const char = sanitized[index];
+			syntax = [',', '>', '+', '~'].includes(char) ? char : '';
+		}
+		if (syntax === ',') { // 情形1：并列
+			pushSimple(step, sanitized.slice(0, index));
+			condition = [[]];
+			[step] = condition;
+			stack.push(condition);
+		} else if (['>', '+', '~', ''].includes(syntax)) { // 情形2：关系
+			pushSimple(step, sanitized.slice(0, index));
+			if (!step.some(Boolean)) {
+				throw new SyntaxError(`非法的选择器！\n${selector}\n可能需要通用选择器'*'。`);
+			}
+			step.relation = syntax;
+			step = [];
+			condition.push(step);
+		} else if (syntax === '[') { // 情形3：属性开启
+			pushSimple(step, sanitized.slice(0, index));
+			regex = attributeRegex;
+		} else if (syntax.at(-1) === ']') { // 情形4：属性闭合
+			mt[3] = desanitize(deQuote(mt[3]));
+			step.push(mt.slice(1));
+			regex = regularRegex;
+		} else if (syntax === '(') { // 情形5：伪选择器开启
+			const pseudoExec = pseudoRegex.exec(sanitized.slice(0, index));
+			if (!pseudoExec) {
+				throw new SyntaxError(`非法的选择器！\n${desanitize(sanitized)}\n请检查伪选择器是否存在。`);
+			}
+			pushSimple(step, sanitized.slice(0, pseudoExec.index));
+			step.push(pseudoExec[1]); // 临时存放复杂伪选择器
+			regex = functionRegex;
+		} else { // 情形6：伪选择器闭合
+			const /** @type {pseudo} */ pseudo = step.pop();
+			mt.push(pseudo);
+			mt[1] = deQuote(mt[1]);
+			step.push(mt.slice(1));
+			regex = regularRegex;
+		}
+		sanitized = sanitized.slice(index + syntax.length);
+		if ([',', '>', '+', '~'].includes(syntax)) {
+			sanitized = sanitized.trim();
+		}
+		mt = regex.exec(sanitized);
+	}
+	if (regex !== regularRegex) {
+		throw new SyntaxError(`非法的选择器！\n${selector}\n检测到未闭合的'${regex === attributeRegex ? '[' : '('}'`);
+	}
+	pushSimple(step, sanitized);
+	return stack;
+};
+
+Parser.parsers.parseSelector = __filename;
+module.exports = parseSelector;