wikiparser-node 0.3.1 → 0.5.0

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,
+	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${desanitize(sanitized)}\n请检查伪选择器是否存在。`);
+			}
+			pushSimple(step, sanitized.slice(0, pseudoExec.index));
+			step.push(pseudoExec[1]); // 临时存放复杂伪选择器
+			regex = functionRegex;
+		} else { // 情形6：伪选择器闭合
+			const /** @type {pseudo} */ pseudo = step.pop();
+			mt.push(pseudo);
+			mt[1] = deQuote(mt[1]);
+			step.push(mt.slice(1));
+			regex = regularRegex;
+		}
+		sanitized = sanitized.slice(index + syntax.length);
+		if ([',', '>', '+', '~'].includes(syntax)) {
+			sanitized = sanitized.trim();
+		}
+		mt = regex.exec(sanitized);
+	}
+	if (regex === regularRegex) {
+		pushSimple(step, sanitized);
+		return stack;
+	}
+	throw new SyntaxError(`非法的选择器！\n${selector}\n检测到未闭合的'${regex === attributeRegex ? '[' : '('}'`);
+};
+
+Parser.parsers.parseSelector = __filename;
+module.exports = parseSelector;

package/parser/table.js

@@ -1,29 +1,36 @@
 'use strict';
 
-const /** @type {Parser} */ Parser = require('..');
+const Parser = require('..'),
+	AstText = require('../lib/text'),
+	Token = require('../src'),
+	TableToken = require('../src/table'),
+	TrToken = require('../src/table/tr'),
+	TdToken = require('../src/table/td'),
+	DdToken = require('../src/nowiki/dd');
 
 /**
- * `tr`和`td`包含开头的换行
- * @param {{firstChild: string, type: string}}
+ * 解析表格，注意`tr`和`td`包含开头的换行
+ * @param {Token & {firstChild: AstText}} root 根节点
  * @param {accum} accum
  */
-const parseTable = ({firstChild, type}, config = Parser.getConfig(), accum = []) => {
-	const Token = require('../src'),
-		TableToken = require('../src/table'),
-		TrToken = require('../src/table/tr'),
-		TdToken = require('../src/table/td'),
-		DdToken = require('../src/nowiki/dd'),
-		/** @type {TrToken[]} */ stack = [],
-		lines = firstChild.split('\n');
-	let out = type === 'root' ? '' : `\n${lines.shift()}`;
-	const /** @type {(str: string, top: TrToken & {firstChild: string}) => void} */ push = (str, top) => {
+const parseTable = ({firstChild: {data}, type, name}, config = Parser.getConfig(), accum = []) => {
+	const /** @type {TrToken[]} */ stack = [],
+		lines = data.split('\n');
+	let out = type === 'root' || type === 'ext-inner' && name === 'poem' ? '' : `\n${lines.shift()}`;
+
+	/**
+	 * 向表格中插入纯文本
+	 * @param {string} str 待插入的文本
+	 * @param {TrToken} top 当前解析的表格或表格行
+	 */
+	const push = (str, top) => {
 		if (!top) {
 			out += str;
 			return;
 		}
-		const {lastElementChild} = top;
-		if (lastElementChild.isPlain()) {
-			lastElementChild.setText(lastElementChild.firstChild + str, 0);
+		const /** @type {Token & {firstChild: AstText}} */ {lastChild} = top;
+		if (lastChild.isPlain()) {
+			lastChild.firstChild.appendData(str);
 		} else {
 			const token = new Token(str, config, true, accum);
 			token.type = 'table-inter';
@@ -32,9 +39,9 @@ const parseTable = ({firstChild, type}, config = Parser.getConfig(), accum = [])
 	};
 	for (const outLine of lines) {
 		let top = stack.pop();
-		const [spaces] = /^(?:\s|\0\d+c\x7f)*/.exec(outLine);
-		const line = outLine.slice(spaces.length),
-			matchesStart = /^(:*)((?:\s|\0\d+c\x7f)*)(\{\||\{\0\d+!\x7f|\0\d+\{\x7f)(.*)$/.exec(line);
+		const [spaces] = /^(?:\s|\0\d+c\x7F)*/u.exec(outLine),
+			line = outLine.slice(spaces.length),
+			matchesStart = /^(:*)((?:\s|\0\d+c\x7F)*)(\{\||\{\0\d+!\x7F|\0\d+\{\x7F)(.*)$/u.exec(line);
 		if (matchesStart) {
 			while (top && top.type !== 'td') {
 				top = stack.pop();
@@ -43,7 +50,7 @@ const parseTable = ({firstChild, type}, config = Parser.getConfig(), accum = [])
 			if (indent) {
 				new DdToken(indent, config, accum);
 			}
-			push(`\n${spaces}${indent && `\0${accum.length - 1}d\x7f`}${moreSpaces}\0${accum.length}b\x7f`, top);
+			push(`\n${spaces}${indent && `\0${accum.length - 1}d\x7F`}${moreSpaces}\0${accum.length}b\x7F`, top);
 			const table = new TableToken(tableSyntax, attr, config, accum);
 			stack.push(...top ? [top] : [], table);
 			continue;
@@ -51,8 +58,8 @@ const parseTable = ({firstChild, type}, config = Parser.getConfig(), accum = [])
 			out += `\n${outLine}`;
 			continue;
 		}
-		const matches
-			= /^(?:(\|\}|\0\d+!\x7f\}|\0\d+\}\x7f)|(\|-+|\0\d+!\x7f-+|\0\d+-\x7f-*)(?!-)|(!|(?:\||\0\d+!\x7f)\+?))(.*)$/
+		const matches = // eslint-disable-line operator-linebreak
+			/^(?:(\|\}|\0\d+!\x7F\}|\0\d+\}\x7F)|(\|-+|\0\d+!\x7F-+|\0\d+-\x7F-*)(?!-)|(!|(?:\||\0\d+!\x7F)\+?))(.*)$/u
 				.exec(line);
 		if (!matches) {
 			push(`\n${outLine}`, top);
@@ -81,8 +88,8 @@ const parseTable = ({firstChild, type}, config = Parser.getConfig(), accum = [])
 				top = stack.pop();
 			}
 			const regex = cell === '!'
-				? /!!|(?:\||\0\d+!\x7f){2}|\0\d+\+\x7f/g
-				: /(?:\||\0\d+!\x7f){2}|\0\d+\+\x7f/g;
+				? /!!|(?:\||\0\d+!\x7F){2}|\0\d+\+\x7F/gu
+				: /(?:\||\0\d+!\x7F){2}|\0\d+\+\x7F/gu;
 			let mt = regex.exec(attr),
 				lastIndex = 0,
 				lastSyntax = `\n${spaces}${cell}`;

package/src/arg.js

@@ -1,7 +1,8 @@
 'use strict';
 
 const {text, noWrap} = require('../util/string'),
-	/** @type {Parser} */ Parser = require('..'),
+	{generateForChild} = require('../util/lint'),
+	Parser = require('..'),
 	Token = require('.');
 
 /**
@@ -12,68 +13,86 @@ class ArgToken extends Token {
 	type = 'arg';
 
 	/**
-	 * @param {string[]} parts
+	 * @param {string[]} parts 以'|'分隔的各部分
 	 * @param {accum} accum
 	 * @complexity `n`
 	 */
 	constructor(parts, config = Parser.getConfig(), accum = []) {
 		super(undefined, config, true, accum, {AtomToken: 0, Token: 1, HiddenToken: '2:'});
-		for (const [i, part] of parts.entries()) {
+		for (let i = 0; i < parts.length; i++) {
 			if (i === 0 || i > 1) {
-				const AtomToken = i === 0 ? require('./atom') : require('./atom/hidden'),
-					token = new AtomToken(part, `arg-${i === 0 ? 'name' : 'redundant'}`, config, accum, {
-						'Stage-2': ':', '!HeadingToken': '',
-					});
+				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': '',
+				});
 				this.appendChild(token);
 			} else {
-				const token = new Token(part, config, true, accum);
+				const token = new Token(parts[i], config, true, accum);
 				token.type = 'arg-default';
 				this.appendChild(token.setAttribute('stage', 2));
 			}
 		}
-		this.protectChildren(0);
+		this.getAttribute('protectChildren')(0);
 	}
 
+	/** @override */
 	cloneNode() {
-		const [name, ...cloned] = this.cloneChildren();
+		const [name, ...cloned] = this.cloneChildNodes();
 		return Parser.run(() => {
 			const token = new ArgToken([''], this.getAttribute('config'));
-			token.firstElementChild.safeReplaceWith(name);
+			token.firstChild.safeReplaceWith(name);
 			token.append(...cloned);
 			return token.afterBuild();
 		});
 	}
 
+	/** @override */
 	afterBuild() {
-		this.setAttribute('name', this.firstElementChild.text().trim());
-		const that = this,
-			/** @type {AstListener} */ argListener = ({prevTarget}) => {
-				if (prevTarget === that.firstElementChild) {
-					that.setAttribute('name', prevTarget.text().trim());
-				}
-			};
+		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);
 		return this;
 	}
 
-	toString() {
-		return `{{{${super.toString('|')}}}}`;
+	/**
+	 * @override
+	 * @param {string} selector
+	 */
+	toString(selector) {
+		return selector && this.matches(selector) ? '' : `{{{${super.toString(selector, '|')}}}}`;
 	}
 
+	/** @override */
 	getPadding() {
 		return 3;
 	}
 
+	/** @override */
 	getGaps() {
 		return 1;
 	}
 
-	/** @returns {string} */
+	/** @override */
+	print() {
+		return super.print({pre: '{{{', post: '}}}', sep: '|'});
+	}
+
+	/**
+	 * @override
+	 * @returns {string}
+	 */
 	text() {
-		return `{{{${text(this.children.slice(0, 2), '|')}}}}`;
+		return `{{{${text(this.childNodes.slice(0, 2), '|')}}}}`;
 	}
 
-	/** @complexity `n` */
+	/**
+	 * 移除无效部分
+	 * @complexity `n`
+	 */
 	removeRedundant() {
 		Parser.run(() => {
 			for (let i = this.childNodes.length - 1; i > 1; i--) {
@@ -83,8 +102,26 @@ class ArgToken extends Token {
 	}
 
 	/**
-	 * 删除`arg-default`子节点时自动删除全部`arg-redundant`子节点
-	 * @param {number} i
+	 * @override
+	 * @param {number} start 起始位置
+	 * @returns {LintError[]}
+	 */
+	lint(start = 0) {
+		const {childNodes: [argName, argDefault, ...rest]} = this,
+			errors = argName.lint(start + 3);
+		if (argDefault) {
+			errors.push(...argDefault.lint(start + 4 + String(argName).length));
+		}
+		if (rest.length > 0) {
+			const rect = this.getRootNode().posFromIndex(start);
+			errors.push(...rest.map(child => generateForChild(child, rect, '三重括号内的不可见部分')));
+		}
+		return errors;
+	}
+
+	/**
+	 * 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) {
@@ -94,11 +131,16 @@ class ArgToken extends Token {
 		return super.removeAt(i);
 	}
 
-	/** @param {Token} token */
+	/**
+	 * @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.childNodes.length) {
 		const j = i < 0 ? i + this.childNodes.length : i;
 		if (j > 1 && !Parser.running) {
-			throw new RangeError(`${this.constructor.name} 不可插入 arg-redundant 子节点！`);
+			throw new RangeError(`${this.constructor.name} 不可插入多余的子节点！`);
 		}
 		super.insertAt(token, i);
 		if (j === 1) {
@@ -107,36 +149,42 @@ class ArgToken extends Token {
 		return token;
 	}
 
-	/** @param {string} name */
+	/**
+	 * 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(`{{{${name}}}}`, this.getAttribute('include'), 2, this.getAttribute('config')),
-			{childNodes: {length}, firstElementChild} = root;
-		if (length !== 1 || firstElementChild?.type !== 'arg' || firstElementChild.childNodes.length !== 1) {
+			{childNodes: {length}, firstChild: arg} = root;
+		if (length !== 1 || arg.type !== 'arg' || arg.childNodes.length !== 1) {
 			throw new SyntaxError(`非法的参数名称：${noWrap(name)}`);
 		}
-		const newName = firstElementChild.firstElementChild;
-		root.destroy();
-		firstElementChild.destroy();
-		this.firstElementChild.safeReplaceWith(newName);
+		const {firstChild} = arg;
+		arg.destroy(true);
+		this.firstChild.safeReplaceWith(firstChild);
 	}
 
-	/** @param {string} value */
+	/**
+	 * 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(`{{{|${value}}}}`, this.getAttribute('include'), 2, this.getAttribute('config')),
-			{childNodes: {length}, firstElementChild} = root;
-		if (length !== 1 || firstElementChild?.type !== 'arg' || firstElementChild.childNodes.length !== 2) {
+			{childNodes: {length}, firstChild: arg} = root;
+		if (length !== 1 || arg.type !== 'arg' || arg.childNodes.length !== 2) {
 			throw new SyntaxError(`非法的参数预设值：${noWrap(value)}`);
 		}
-		const [, oldDefault] = this.children,
-			newDefault = firstElementChild.lastElementChild;
-		root.destroy();
-		firstElementChild.destroy();
+		const {childNodes: [, oldDefault]} = this,
+			{lastChild} = arg;
+		arg.destroy(true);
 		if (oldDefault) {
-			oldDefault.safeReplaceWith(newDefault);
+			oldDefault.safeReplaceWith(lastChild);
 		} else {
-			this.appendChild(newDefault);
+			this.appendChild(lastChild);
 		}
 	}
 }

package/src/atom/hidden.js

@@ -1,10 +1,13 @@
 'use strict';
 
 const hidden = require('../../mixin/hidden'),
-	/** @type {Parser} */ Parser = require('../..'),
+	Parser = require('../..'),
 	AtomToken = require('.');
 
-class HiddenToken extends hidden(AtomToken) {}
+/** 不可见的节点 */
+class HiddenToken extends hidden(AtomToken) {
+	type = 'hidden';
+}
 
 Parser.classes.HiddenToken = __filename;
 module.exports = HiddenToken;

package/src/atom/index.js

@@ -1,29 +1,37 @@
 'use strict';
 
-const /** @type {Parser} */ Parser = require('../..'),
+const Parser = require('../..'),
 	Token = require('..');
 
 /**
  * 不会被继续解析的plain Token
- * @classdesc `{childNodes: (string|Token)[]}`
+ * @classdesc `{childNodes: (AstText|Token)[]}`
  */
 class AtomToken extends Token {
+	type = 'plain';
+
 	/**
-	 * @param {?string} wikitext
+	 * @param {string} wikitext wikitext
+	 * @param {string|undefined} type Token.type
 	 * @param {accum} accum
-	 * @param {acceptable} acceptable
+	 * @param {acceptable} acceptable 可接受的子节点设置
 	 */
-	constructor(wikitext, type = 'plain', config = Parser.getConfig(), accum = [], acceptable = null) {
+	constructor(wikitext, type, config = Parser.getConfig(), accum = [], acceptable = null) {
 		super(wikitext, config, true, accum, acceptable);
-		this.type = type;
+		if (type) {
+			this.type = type;
+		}
 	}
 
+	/**
+	 * @override
+	 * @this {AtomToken & {constructor: typeof AtomToken}}
+	 */
 	cloneNode() {
-		const cloned = this.cloneChildren(),
-			/** @type {typeof AtomToken} */ Constructor = this.constructor,
+		const cloned = this.cloneChildNodes(),
 			config = this.getAttribute('config'),
 			acceptable = this.getAttribute('acceptable'),
-			token = Parser.run(() => new Constructor(undefined, this.type, config, [], acceptable));
+			token = Parser.run(() => new this.constructor(undefined, this.type, config, [], acceptable));
 		token.append(...cloned);
 		return token;
 	}