wikiparser-node 0.8.1 → 0.9.0

package/src/transclude.js

@@ -1,6 +1,6 @@
 'use strict';
 
-const {removeComment, escapeRegExp, text, noWrap, print} = require('../util/string'),
+const {removeComment, escapeRegExp, text, noWrap, print, decodeHtml} = require('../util/string'),
 	{externalUse} = require('../util/debug'),
 	{generateForChild} = require('../util/lint'),
 	Parser = require('..'),
@@ -18,8 +18,9 @@ class TranscludeToken extends Token {
 	modifier = '';
 	/** @type {Record<string, Set<ParameterToken>>} */ #args = {};
 	/** @type {Set<string>} */ #keys = new Set();
-	#fragment = '';
+	#fragment;
 	#valid = true;
+	#raw = false;
 
 	/** 是否存在重复参数 */
 	get duplication() {
@@ -36,14 +37,18 @@ class TranscludeToken extends Token {
 			this.typeError('setModifier', 'String');
 		}
 		const {parserFunction: [,, raw, subst]} = this.getAttribute('config'),
-			lcModifier = modifier.trimStart().toLowerCase(),
-			isRaw = raw.includes(lcModifier),
-			isSubst = subst.includes(lcModifier),
-			wasRaw = raw.includes(this.modifier.trimStart().toLowerCase());
-		if (wasRaw && isRaw || !wasRaw && (isSubst || modifier === '')
+			lcModifier = removeComment(modifier).trim();
+		if (modifier && !lcModifier.endsWith(':')) {
+			return false;
+		}
+		const magicWord = lcModifier.slice(0, -1).toLowerCase(),
+			isRaw = raw.includes(magicWord),
+			isSubst = subst.includes(magicWord);
+		if (this.#raw && isRaw || !this.#raw && (isSubst || modifier === '')
 			|| (Parser.running || this.length > 1) && (isRaw || isSubst || modifier === '')
 		) {
 			this.setAttribute('modifier', modifier);
+			this.#raw = isRaw;
 			return Boolean(modifier);
 		}
 		return false;
@@ -60,15 +65,20 @@ class TranscludeToken extends Token {
 		super(undefined, config, true, accum, {
 			AtomToken: 0, SyntaxToken: 0, ParameterToken: '1:',
 		});
-		const {parserFunction: [insensitive, sensitive, raw]} = config;
 		this.seal('modifier');
-		if (title.includes(':')) {
-			const [modifier, ...arg] = title.split(':');
-			if (this.setModifier(modifier)) {
-				title = arg.join(':');
+		const {parserFunction: [insensitive, sensitive]} = config,
+			argSubst = /^(?:\s|\0\d+c\x7F)*\0\d+s\x7F/u.exec(title)?.[0];
+		if (argSubst) {
+			this.setAttribute('modifier', argSubst);
+			title = title.slice(argSubst.length);
+		} else if (title.includes(':')) {
+			const [modifier, ...arg] = title.split(':'),
+				[mt] = /^(?:\s|\0\d+c\x7F)*/u.exec(arg[0] ?? '');
+			if (this.setModifier(`${modifier}:${mt}`)) {
+				title = arg.join(':').slice(mt.length);
 			}
 		}
-		if (title.includes(':') || parts.length === 0 && !raw.includes(this.modifier.toLowerCase())) {
+		if (title.includes(':') || parts.length === 0 && !this.#raw) {
 			const [magicWord, ...arg] = title.split(':'),
 				cleaned = removeComment(magicWord),
 				name = cleaned[arg.length > 0 ? 'trimStart' : 'trim'](),
@@ -103,7 +113,7 @@ class TranscludeToken extends Token {
 			}
 		}
 		if (this.type === 'template') {
-			const name = removeComment(title).split('#')[0].trim();
+			const name = removeComment(decodeHtml(title)).split('#')[0].trim();
 			if (!name || /\0\d+[eh!+-]\x7F|[<>[\]{}\n]|%[\da-f]{2}/u.test(name)) {
 				accum.pop();
 				throw new SyntaxError(`非法的模板名称：${noWrap(name)}`);
@@ -132,6 +142,9 @@ class TranscludeToken extends Token {
 
 	/** @override */
 	afterBuild() {
+		if (this.modifier.includes('\0')) {
+			this.setAttribute('modifier', this.getAttribute('buildFromStr')(this.modifier, 'string'));
+		}
 		if (this.isTemplate()) {
 			const isTemplate = this.type === 'template',
 				titleObj = this.normalizeTitle(this.childNodes[isTemplate ? 0 : 1].text(), isTemplate ? 10 : 828);
@@ -182,7 +195,7 @@ class TranscludeToken extends Token {
 			return '';
 		}
 		const {childNodes, firstChild, modifier} = this;
-		return `{{${modifier}${modifier && ':'}${
+		return `{{${modifier}${
 			this.type === 'magic-word'
 				? `${String(firstChild)}${childNodes.length > 1 ? ':' : ''}${childNodes.slice(1).map(String).join('|')}`
 				: super.toString(selector, '|')
@@ -198,7 +211,7 @@ class TranscludeToken extends Token {
 		const {childNodes, firstChild, modifier, type, name} = this;
 		return type === 'magic-word' && name === 'vardefine'
 			? ''
-			: `{{${modifier}${modifier && ':'}${
+			: `{{${modifier}${
 				this.type === 'magic-word'
 					? `${firstChild.text()}${childNodes.length > 1 ? ':' : ''}${text(childNodes.slice(1), '|')}`
 					: super.text('|')
@@ -207,7 +220,7 @@ class TranscludeToken extends Token {
 
 	/** @override */
 	getPadding() {
-		return this.modifier ? this.modifier.length + 3 : 2;
+		return this.modifier.length + 2;
 	}
 
 	/** @override */
@@ -218,7 +231,7 @@ class TranscludeToken extends Token {
 	/** @override */
 	print() {
 		const {childNodes, firstChild, modifier} = this;
-		return `<span class="wpb-${this.type}">{{${modifier}${modifier && ':'}${
+		return `<span class="wpb-${this.type}">{{${modifier}${
 			this.type === 'magic-word'
 				? `${firstChild.print()}${childNodes.length > 1 ? ':' : ''}${print(childNodes.slice(1), {sep: '|'})}`
 				: print(childNodes, {sep: '|'})
@@ -229,25 +242,25 @@ class TranscludeToken extends Token {
 	 * @override
 	 * @param {number} start 起始位置
 	 */
-	lint(start = 0) {
+	lint(start = this.getAbsoluteIndex()) {
 		const errors = super.lint(start),
 			{type, childNodes} = this;
 		let rect;
 		if (!this.isTemplate()) {
 			return errors;
-		} else if (this.#fragment) {
+		} else if (this.#fragment !== undefined) {
 			rect = {start, ...this.getRootNode().posFromIndex(start)};
-			errors.push(generateForChild(childNodes[type === 'template' ? 0 : 1], rect, '多余的fragment'));
+			errors.push(generateForChild(childNodes[type === 'template' ? 0 : 1], rect, 'useless fragment'));
 		}
 		if (!this.#valid) {
 			rect = {start, ...this.getRootNode().posFromIndex(start)};
-			errors.push(generateForChild(childNodes[1], rect, '非法的模块名称'));
+			errors.push(generateForChild(childNodes[1], rect, 'illegal module name'));
 		}
 		const duplicatedArgs = this.getDuplicatedArgs();
 		if (duplicatedArgs.length > 0) {
 			rect ||= {start, ...this.getRootNode().posFromIndex(start)};
 			errors.push(...duplicatedArgs.flatMap(([, args]) => args).map(
-				arg => generateForChild(arg, rect, '重复参数'),
+				arg => generateForChild(arg, rect, 'duplicated parameter'),
 			));
 		}
 		return errors;
@@ -413,7 +426,11 @@ class TranscludeToken extends Token {
 			config = this.getAttribute('config');
 		return Parser.run(() => {
 			const token = new TranscludeToken(this.type === 'template' ? '' : first.text(), [], config);
-			token.setModifier(this.modifier);
+			if (this.#raw) {
+				token.setModifier(this.modifier);
+			} else {
+				token.setAttribute('modifier', this.modifier);
+			}
 			token.firstChild.safeReplaceWith(first);
 			token.afterBuild();
 			token.append(...cloned);
@@ -423,12 +440,12 @@ class TranscludeToken extends Token {
 
 	/** 替换引用 */
 	subst() {
-		this.setModifier('subst');
+		this.setModifier('subst:');
 	}
 
 	/** 安全的替换引用 */
 	safesubst() {
-		this.setModifier('safesubst');
+		this.setModifier('safesubst:');
 	}
 
 	/**

package/util/lint.js

@@ -1,16 +1,17 @@
 'use strict';
 
-const Token = require('../src');
+const Parser = require('..'),
+	Token = require('../src');
 
 /**
  * 生成对于子节点的LintError对象
  * @param {Token} child 子节点
  * @param {{top: number, left: number, start: number}} boundingRect 父节点的绝对定位
- * @param {string} message 错误信息
+ * @param {string} msg 错误信息
  * @param {'error'|'warning'} severity 严重程度
  * @returns {LintError}
  */
-const generateForChild = (child, boundingRect, message, severity = 'error') => {
+const generateForChild = (child, boundingRect, msg, severity = 'error') => {
 	const index = child.getRelativeIndex(),
 		{offsetHeight, offsetWidth, parentNode, length} = child,
 		{top: offsetTop, left: offsetLeft} = parentNode.posFromIndex(index),
@@ -23,23 +24,23 @@ const generateForChild = (child, boundingRect, message, severity = 'error') => {
 		endLine = startLine + offsetHeight - 1,
 		startCol = offsetTop ? offsetLeft : left + offsetLeft,
 		endCol = offsetHeight > 1 ? offsetWidth : startCol + offsetWidth;
-	return {message, severity, startIndex, endIndex, startLine, endLine, startCol, endCol, excerpt};
+	return {message: Parser.msg(msg), severity, startIndex, endIndex, startLine, endLine, startCol, endCol, excerpt};
 };
 
 /**
  * 生成对于自己的LintError对象
  * @param {Token} token 节点
  * @param {{top: number, left: number, start: number}} boundingRect 绝对定位
- * @param {string} message 错误信息
+ * @param {string} msg 错误信息
  * @param {'error'|'warning'} severity 严重程度
  * @returns {LintError}
  */
-const generateForSelf = (token, boundingRect, message, severity = 'error') => {
+const generateForSelf = (token, boundingRect, msg, severity = 'error') => {
 	const {start} = boundingRect,
 		{offsetHeight, offsetWidth, length} = token,
 		{top, left} = 'top' in boundingRect ? boundingRect : token.getRootNode().posFromIndex(start);
 	return {
-		message,
+		message: Parser.msg(msg),
 		severity,
 		startIndex: start,
 		endIndex: start + length,

package/util/string.js

@@ -72,6 +72,15 @@ const text = (childNodes, separator = '') => {
 	return childNodes.map(child => typeof child === 'string' ? child : child.text()).join(separator);
 };
 
+/**
+ * decode HTML entities
+ * @param {string} str 原字符串
+ */
+const decodeHtml = str => str?.replace(
+	/&#(\d+|x[\da-f]+);/giu,
+	(_, code) => String.fromCodePoint(`${code[0].toLowerCase() === 'x' ? '0' : ''}${code}`),
+);
+
 /**
  * optionally convert to lower cases
  * @param {string} val 属性值
@@ -103,5 +112,15 @@ const normalizeSpace = token => {
 };
 
 module.exports = {
-	extUrlCharFirst, extUrlChar, removeComment, print, escapeRegExp, explode, text, toCase, noWrap, normalizeSpace,
+	extUrlCharFirst,
+	extUrlChar,
+	removeComment,
+	print,
+	escapeRegExp,
+	explode,
+	text,
+	decodeHtml,
+	toCase,
+	noWrap,
+	normalizeSpace,
 };