wikiparser-node 0.8.1-m → 0.9.0

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "wikiparser-node",
-  "version": "0.8.1-m",
+  "version": "0.9.0",
   "description": "A Node.js parser for MediaWiki markup with AST",
   "keywords": [
     "mediawiki",
@@ -15,22 +15,23 @@
   "author": "Bhsd",
   "files": [
     "/index.js",
-    "/parser/",
-    "/util/",
+    "/config/",
+    "/i18n/",
     "/lib/",
+    "/mixin/",
+    "/parser/",
     "/src/",
-    "/config/",
-    "/mixin/"
+    "/tool/",
+    "/util/"
   ],
   "repository": {
     "type": "git",
     "url": "git+https://github.com/bhsd-harry/wikiparser-node.git"
   },
   "scripts": {
-    "test": "eslint .",
+    "test": "eslint . && node test/test.js",
     "real": "node test/real.js",
-    "single": "node test/single.js",
-    "publish": "npm publish --tag mini"
+    "single": "node --prof test/single.js && node --prof-process isolate-0x*-v8.log > test/processed.txt && rm isolate-0x*-v8.log"
   },
   "devDependencies": {
     "@types/node": "^17.0.23",

package/parser/brackets.js

@@ -14,6 +14,7 @@ const {removeComment} = require('../util/string'),
  */
 const parseBrackets = (text, config = Parser.getConfig(), accum = []) => {
 	const source = `${config.excludes.includes('heading') ? '' : '^(\0\\d+c\x7F)*={1,6}|'}\\[\\[|\\{{2,}|-\\{(?!\\{)`,
+		{parserFunction: [,,, subst]} = config,
 		/** @type {BracketExecArray[]} */ stack = [],
 		closes = {'=': '\n', '{': '\\}{2,}|\\|', '-': '\\}-', '[': '\\]\\]'},
 		/** @type {Record<string, string>} */ marks = {'!': '!', '!!': '+', '(!': '{', '!)': '}', '!-': '-', '=': '~'};
@@ -62,12 +63,17 @@ const parseBrackets = (text, config = Parser.getConfig(), accum = []) => {
 			let skip = false,
 				ch = 't';
 			if (close.length === 3) {
-				new ArgToken(parts.map(part => part.join('=')), config, accum);
+				const argParts = parts.map(part => part.join('=')),
+					str = argParts.length > 1 && removeComment(argParts[1]).trim();
+				new ArgToken(argParts, config, accum);
+				if (str && str.endsWith(':') && subst.includes(str.slice(0, -1).toLowerCase())) {
+					ch = 's';
+				}
 			} else {
 				const name = removeComment(parts[0][0]).trim();
 				if (name in marks) {
 					ch = marks[name]; // 标记{{!}}等
-				} else if (/^(?:fullurl|canonicalurl|filepath):.|^server$/iu.test(name)) {
+				} else if (/^(?:filepath|(?:full|canonical)urle?):.|^server$/iu.test(name)) {
 					ch = 'm';
 				} else if (/^#vardefine:./iu.test(name)) {
 					ch = 'c';
@@ -116,4 +122,5 @@ const parseBrackets = (text, config = Parser.getConfig(), accum = []) => {
 	return text;
 };
 
+Parser.parsers.parseBrackets = __filename;
 module.exports = parseBrackets;

package/parser/commentAndExt.js

@@ -20,12 +20,9 @@ const parseCommentAndExt = (text, config = Parser.getConfig(), accum = [], inclu
 			const str = `\0${accum.length}e\x7F`;
 			new OnlyincludeToken(inner, config, accum);
 			return str;
-		}).replace(/(^|\0\d+e\x7F)(.*?)(?=$|\0\d+e\x7F)/gsu, (_, lead, substr) => {
-			if (substr === '') {
-				return lead;
-			}
+		}).replace(/(?<=^|\0\d+e\x7F)[^\0]+(?=$|\0\d+e\x7F)/gu, substr => {
 			new NoincludeToken(substr, config, accum);
-			return `${lead}\0${accum.length - 1}c\x7F`;
+			return `\0${accum.length - 1}c\x7F`;
 		});
 	}
 	const ext = config.ext.join('|'),
@@ -58,4 +55,5 @@ const parseCommentAndExt = (text, config = Parser.getConfig(), accum = [], inclu
 	);
 };
 
+Parser.parsers.parseCommentAndExt = __filename;
 module.exports = parseCommentAndExt;

package/parser/converter.js

@@ -42,4 +42,5 @@ const parseConverter = (wikitext, config = Parser.getConfig(), accum = []) => {
 	return wikitext;
 };
 
+Parser.parsers.parseConverter = __filename;
 module.exports = parseConverter;

package/parser/externalLinks.js

@@ -29,4 +29,5 @@ const parseExternalLinks = (wikitext, config = Parser.getConfig(), accum = []) =
 	});
 };
 
+Parser.parsers.parseExternalLinks = __filename;
 module.exports = parseExternalLinks;

package/parser/hrAndDoubleUnderscore.js

@@ -34,4 +34,5 @@ const parseHrAndDoubleUnderscore = ({firstChild: {data}, type, name}, config = P
 	return type === 'root' || type === 'ext-inner' && name === 'poem' ? data : data.slice(1);
 };
 
+Parser.parsers.parseHrAndDoubleUnderscore = __filename;
 module.exports = parseHrAndDoubleUnderscore;

package/parser/html.js

@@ -38,4 +38,5 @@ const parseHtml = (wikitext, config = Parser.getConfig(), accum = []) => {
 	return text;
 };
 
+Parser.parsers.parseHtml = __filename;
 module.exports = parseHtml;

package/parser/links.js

@@ -44,12 +44,12 @@ const parseLinks = (wikitext, config = Parser.getConfig(), accum = []) => {
 			continue;
 		}
 		const title = Parser.normalizeTitle(link, 0, false, config, true, true, true),
-			{ns, valid} = title;
+			{ns, interwiki, valid} = title;
 		if (!valid) {
 			s += `[[${x}`;
 			continue;
 		} else if (mightBeImg) {
-			if (ns !== 6) {
+			if (interwiki || ns !== 6) {
 				s += `[[${x}`;
 				continue;
 			}
@@ -79,15 +79,16 @@ const parseLinks = (wikitext, config = Parser.getConfig(), accum = []) => {
 		s += `\0${accum.length}l\x7F${after}`;
 		let SomeLinkToken = LinkToken;
 		if (!force) {
-			if (ns === 6) {
+			if (!interwiki && ns === 6) {
 				SomeLinkToken = FileToken;
-			} else if (ns === 14) {
+			} else if (!interwiki && ns === 14) {
 				SomeLinkToken = CategoryToken;
 			}
 		}
-		new SomeLinkToken(link, text, title, config, accum, delimiter);
+		new SomeLinkToken(link, text, config, accum, delimiter);
 	}
 	return s;
 };
 
+Parser.parsers.parseLinks = __filename;
 module.exports = parseLinks;

package/parser/list.js

@@ -55,4 +55,5 @@ const parseList = (text, config = Parser.getConfig(), accum = []) => {
 	return text;
 };
 
+Parser.parsers.parseList = __filename;
 module.exports = parseList;

package/parser/magicLinks.js

@@ -10,10 +10,10 @@ const {extUrlChar, extUrlCharFirst} = require('../util/string'),
  * @param {accum} accum
  */
 const parseMagicLinks = (wikitext, config = Parser.getConfig(), accum = []) => {
-	const regex = new RegExp(`(^|[^\\p{L}\\d_])(?:${config.protocol})(${extUrlCharFirst}${extUrlChar})`, 'giu');
-	return wikitext.replace(regex, /** @param {string} p1 */ (m, lead, p1) => {
+	const regex = new RegExp(`(?<![\\p{L}\\d_])(?:${config.protocol})(${extUrlCharFirst}${extUrlChar})`, 'giu');
+	return wikitext.replace(regex, /** @param {string} p1 */ (m, p1) => {
 		let trail = '',
-			url = lead ? m.slice(1) : m;
+			url = m;
 		const m2 = /&(?:lt|gt|nbsp|#x0*(?:3[ce]|a0)|#0*(?:6[02]|160));/iu.exec(url);
 		if (m2) {
 			trail = url.slice(m2.index);
@@ -33,8 +33,9 @@ const parseMagicLinks = (wikitext, config = Parser.getConfig(), accum = []) => {
 			return m;
 		}
 		new MagicLinkToken(url, false, config, accum);
-		return `${lead}\0${accum.length - 1}w\x7F${trail}`;
+		return `\0${accum.length - 1}w\x7F${trail}`;
 	});
 };
 
+Parser.parsers.parseMagicLinks = __filename;
 module.exports = parseMagicLinks;

package/parser/quotes.js

@@ -60,4 +60,5 @@ const parseQuotes = (text, config = Parser.getConfig(), accum = []) => {
 	return arr.join('');
 };
 
+Parser.parsers.parseQuotes = __filename;
 module.exports = parseQuotes;

package/parser/selector.js

@@ -0,0 +1,177 @@
+'use strict';
+
+const Parser = require('..');
+
+const /** @type {Set<pseudo>} */ simplePseudos = new Set([
+		'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 = [
+		['[', '&lbrack;'],
+		[']', '&rbrack;'],
+		['(', '&lpar;'],
+		[')', '&rpar;'],
+		['"', '&quot;'],
+		[`'`, '&apos;'],
+		[':', '&colon;'],
+		['\\', '&bsol;'],
+		['&', '&amp;'],
+	],
+	pseudoRegex = new RegExp(`:(${complexPseudos.join('|')})$`, 'u'),
+	regularRegex = /[[(,>+~]|\s+/u,
+	attributeRegex = /^\s*(\w+)\s*(?:([~|^$*!]?=)\s*("[^"]*"|'[^']*'|[^\s[\]]+)(?:\s+(i))?\s*)?\]/u,
+	functionRegex = /^(\s*"[^"]*"\s*|\s*'[^']*'\s*|[^()]*)\)/u,
+	grouping = new Set([',', '>', '+', '~']),
+	combinator = new Set(['>', '+', '~', '']);
+
+/**
+ * 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$/u.exec(val)?.[1];
+	return quotes ? val.slice(1, -1) : val;
+};
+
+/**
+ * 解析简单伪选择器
+ * @param {SelectorArray} step 当前顶部
+ * @param {string} str 不含属性和复杂伪选择器的语句
+ * @throws `SyntaxError` 非法的选择器
+ */
+const pushSimple = (step, str) => {
+	const pieces = str.trim().split(':'),
+		// eslint-disable-next-line unicorn/explicit-length-check
+		i = pieces.slice(1).findIndex(pseudo => simplePseudos.has(pseudo)) + 1 || pieces.length;
+	if (pieces.slice(i).some(pseudo => !simplePseudos.has(pseudo))) {
+		throw new SyntaxError(`非法的选择器！\n${str}\n可能需要将':'转义为'\\:'。`);
+	}
+	step.push(desanitize(pieces.slice(0, i).join(':')), ...pieces.slice(i).map(piece => `:${piece}`));
+};
+
+/**
+ * 解析选择器
+ * @param {string} selector
+ * @throws `SyntaxError` 非法的选择器
+ */
+const parseSelector = selector => {
+	selector = selector.trim();
+	const /** @type {SelectorArray[][]} */ stack = [[[]]];
+	let sanitized = sanitize(selector),
+		regex = regularRegex,
+		mt = regex.exec(sanitized),
+		[condition] = stack,
+		[step] = condition;
+	while (mt) {
+		let {0: syntax, index} = mt;
+		if (syntax.trim() === '') {
+			index += syntax.length;
+			const char = sanitized[index];
+			syntax = grouping.has(char) ? char : '';
+		}
+		if (syntax === ',') { // 情形1：并列
+			pushSimple(step, sanitized.slice(0, index));
+			condition = [[]];
+			[step] = condition;
+			stack.push(condition);
+		} else if (combinator.has(syntax)) { // 情形2：关系
+			pushSimple(step, sanitized.slice(0, index));
+			if (!step.some(Boolean)) {
+				throw new SyntaxError(`非法的选择器！\n${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.endsWith(']')) { // 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${desanitize(sanitized)}\n请检查伪选择器是否存在。`);
+			}
+			pushSimple(step, sanitized.slice(0, pseudoExec.index));
+			step.push(pseudoExec[1]); // 临时存放复杂伪选择器
+			regex = functionRegex;
+		} else { // 情形6：伪选择器闭合
+			const /** @type {pseudo} */ pseudo = step.pop();
+			mt.push(pseudo);
+			mt[1] = deQuote(mt[1]);
+			step.push(mt.slice(1));
+			regex = regularRegex;
+		}
+		sanitized = sanitized.slice(index + syntax.length);
+		if (grouping.has(syntax)) {
+			sanitized = sanitized.trim();
+		}
+		mt = regex.exec(sanitized);
+	}
+	if (regex === regularRegex) {
+		pushSimple(step, sanitized);
+		return stack;
+	}
+	throw new SyntaxError(`非法的选择器！\n${selector}\n检测到未闭合的'${regex === attributeRegex ? '[' : '('}'`);
+};
+
+Parser.parsers.parseSelector = __filename;
+module.exports = parseSelector;

package/parser/table.js

@@ -110,4 +110,5 @@ const parseTable = ({firstChild: {data}, type, name}, config = Parser.getConfig(
 	return out.slice(1);
 };
 
+Parser.parsers.parseTable = __filename;
 module.exports = parseTable;

package/src/arg.js

@@ -1,6 +1,6 @@
 'use strict';
 
-const {text} = require('../util/string'),
+const {text, noWrap} = require('../util/string'),
 	{generateForSelf, generateForChild} = require('../util/lint'),
 	Parser = require('..'),
 	Token = require('.');
@@ -24,11 +24,13 @@ class ArgToken extends Token {
 	 */
 	constructor(parts, config = Parser.getConfig(), accum = []) {
 		super(undefined, config, true, accum, {
+			AtomToken: 0, Token: 1, HiddenToken: '2:',
 		});
 		for (let i = 0; i < parts.length; i++) {
 			if (i === 0 || i > 1) {
 				const AtomToken = i === 0 ? require('./atom') : require('./atom/hidden');
 				const token = new AtomToken(parts[i], i === 0 ? 'arg-name' : undefined, config, accum, {
+					'Stage-2': ':', '!HeadingToken': '',
 				});
 				super.insertAt(token);
 			} else {
@@ -37,13 +39,15 @@ class ArgToken extends Token {
 				super.insertAt(token.setAttribute('stage', 2));
 			}
 		}
+		this.getAttribute('protectChildren')(0);
 	}
 
 	/**
 	 * @override
+	 * @param {string} selector
 	 */
 	toString(selector) {
-		return `{{{${super.toString(selector, '|')}}}}`;
+		return selector && this.matches(selector) ? '' : `{{{${super.toString(selector, '|')}}}}`;
 	}
 
 	/**
@@ -64,14 +68,19 @@ class ArgToken extends Token {
 		return 1;
 	}
 
+	/** @override */
+	print() {
+		return super.print({pre: '{{{', post: '}}}', sep: '|'});
+	}
+
 	/**
 	 * @override
 	 * @param {number} start 起始位置
 	 * @returns {LintError[]}
 	 */
-	lint(start = 0) {
+	lint(start = this.getAbsoluteIndex()) {
 		if (!this.getAttribute('include')) {
-			return [generateForSelf(this, {start}, '未预期的模板参数')];
+			return [generateForSelf(this, {start}, 'unexpected template argument')];
 		}
 		const {childNodes: [argName, argDefault, ...rest]} = this,
 			errors = argName.lint(start + 3);
@@ -80,10 +89,119 @@ class ArgToken extends Token {
 		}
 		if (rest.length > 0) {
 			const rect = {start, ...this.getRootNode().posFromIndex(start)};
-			errors.push(...rest.map(child => generateForChild(child, rect, '三重括号内的不可见部分')));
+			errors.push(...rest.map(child => {
+				const error = generateForChild(child, rect, 'invisible content inside triple brackets'),
+					{startIndex, startCol, excerpt} = error;
+				return {...error, startIndex: startIndex - 1, startCol: startCol - 1, excerpt: `|${excerpt}`};
+			}));
 		}
 		return errors;
 	}
+
+	/** @override */
+	cloneNode() {
+		const [name, ...cloned] = this.cloneChildNodes();
+		return Parser.run(() => {
+			const token = new ArgToken([''], this.getAttribute('config'));
+			token.firstChild.safeReplaceWith(name);
+			token.append(...cloned);
+			token.afterBuild();
+			return token;
+		});
+	}
+
+	/** @override */
+	afterBuild() {
+		this.setAttribute('name', this.firstChild.text().trim());
+		const /** @type {AstListener} */ argListener = ({prevTarget}) => {
+			if (prevTarget === this.firstChild) {
+				this.setAttribute('name', prevTarget.text().trim());
+			}
+		};
+		this.addEventListener(['remove', 'insert', 'replace', 'text'], argListener);
+	}
+
+	/**
+	 * 移除无效部分
+	 * @complexity `n`
+	 */
+	removeRedundant() {
+		Parser.run(() => {
+			for (let i = this.length - 1; i > 1; i--) {
+				super.removeAt(i);
+			}
+		});
+	}
+
+	/**
+	 * 移除子节点，且在移除`arg-default`子节点时自动移除全部多余子节点
+	 * @param {number} i 移除位置
+	 * @returns {Token}
+	 */
+	removeAt(i) {
+		if (i === 1) {
+			this.removeRedundant();
+		}
+		return super.removeAt(i);
+	}
+
+	/**
+	 * @override
+	 * @param {Token} token 待插入的子节点
+	 * @param {number} i 插入位置
+	 * @throws `RangeError` 不可插入多余子节点
+	 */
+	insertAt(token, i = this.length) {
+		const j = i < 0 ? i + this.length : i;
+		if (j > 1) {
+			throw new RangeError(`${this.constructor.name} 不可插入多余的子节点！`);
+		}
+		super.insertAt(token, i);
+		if (j === 1) {
+			token.type = 'arg-default';
+		}
+		return token;
+	}
+
+	/**
+	 * 设置参数名
+	 * @param {string} name 新参数名
+	 * @throws `SyntaxError` 非法的参数名
+	 */
+	setName(name) {
+		name = String(name);
+		const root = Parser.parse(`{{{${name}}}}`, this.getAttribute('include'), 2, this.getAttribute('config')),
+			{length, firstChild: arg} = root;
+		if (length !== 1 || arg.type !== 'arg' || arg.length !== 1) {
+			throw new SyntaxError(`非法的参数名称：${noWrap(name)}`);
+		}
+		const {firstChild} = arg;
+		arg.destroy(true);
+		this.firstChild.safeReplaceWith(firstChild);
+	}
+
+	/**
+	 * 设置预设值
+	 * @param {string} value 预设值
+	 * @throws `SyntaxError` 非法的参数预设值
+	 */
+	setDefault(value) {
+		value = String(value);
+		const root = Parser.parse(`{{{|${value}}}}`, this.getAttribute('include'), 2, this.getAttribute('config')),
+			{length, firstChild: arg} = root;
+		if (length !== 1 || arg.type !== 'arg' || arg.length !== 2) {
+			throw new SyntaxError(`非法的参数预设值：${noWrap(value)}`);
+		}
+		const {childNodes: [, oldDefault]} = this,
+			{lastChild} = arg;
+		arg.destroy(true);
+		if (oldDefault) {
+			oldDefault.safeReplaceWith(lastChild);
+		} else {
+			this.insertAt(lastChild);
+		}
+	}
 }
 
+Parser.classes.ArgToken = __filename;
 module.exports = ArgToken;

package/src/atom/hidden.js

@@ -1,6 +1,7 @@
 'use strict';
 
 const hidden = require('../../mixin/hidden'),
+	Parser = require('../..'),
 	AtomToken = require('.');
 
 /** 不可见的节点 */
@@ -8,4 +9,5 @@ class HiddenToken extends hidden(AtomToken) {
 	type = 'hidden';
 }
 
+Parser.classes.HiddenToken = __filename;
 module.exports = HiddenToken;

package/src/atom/index.js

@@ -14,6 +14,7 @@ class AtomToken extends Token {
 	 * @param {string} wikitext wikitext
 	 * @param {string|undefined} type Token.type
 	 * @param {accum} accum
+	 * @param {acceptable} acceptable 可接受的子节点设置
 	 */
 	constructor(wikitext, type, config = Parser.getConfig(), accum = [], acceptable = undefined) {
 		super(wikitext, config, true, accum, acceptable);
@@ -21,6 +22,22 @@ class AtomToken extends Token {
 			this.type = type;
 		}
 	}
+
+	/**
+	 * @override
+	 * @this {AtomToken & {constructor: typeof AtomToken}}
+	 */
+	cloneNode() {
+		const cloned = this.cloneChildNodes(),
+			config = this.getAttribute('config'),
+			acceptable = this.getAttribute('acceptable');
+		return Parser.run(() => {
+			const token = new this.constructor(undefined, this.type, config, [], acceptable);
+			token.append(...cloned);
+			return token;
+		});
+	}
 }
 
+Parser.classes.AtomToken = __filename;
 module.exports = AtomToken;