securemark 0.294.5 → 0.294.7

package/CHANGELOG.md

@@ -1,5 +1,13 @@
 # Changelog
 
+## 0.294.7
+
+- Refactoring.
+
+## 0.294.6
+
+- Refactoring.
+
 ## 0.294.5
 
 - Refactoring.

package/dist/index.js

@@ -1,4 +1,4 @@
-/*! securemark v0.294.5 https://github.com/falsandtru/securemark | (c) 2017, falsandtru | UNLICENSED License */
+/*! securemark v0.294.7 https://github.com/falsandtru/securemark | (c) 2017, falsandtru | UNLICENSED License */
 (function webpackUniversalModuleDefinition(root, factory) {
 	if(typeof exports === 'object' && typeof module === 'object')
 		module.exports = factory(require("Prism"), require("DOMPurify"));
@@ -2894,9 +2894,7 @@ function indent(opener, parser = false, separation = false) {
     } = context;
     context.position = source.length;
     return new parser_1.List([new parser_1.Data(source.slice(position))]);
-  }))), ([indent]) => indent.length * 2 + -(indent[0] === ' '), [], 2 ** 4 - 1)), separation), (lines, context) => {
-    return parser((0, parser_1.subinput)(trimBlockEnd(lines.foldl((acc, node) => acc + node.value, '')), context));
-  }));
+  }))), ([indent]) => indent.length * 2 + -(indent[0] === ' '), [], 2 ** 4 - 1)), separation), (lines, context) => parser((0, parser_1.subinput)(trimBlockEnd(lines.foldl((acc, node) => acc + node.value, '')), context))));
 }
 exports.indent = indent;
 function trimBlockEnd(block) {
@@ -3094,6 +3092,11 @@ function surround(opener, parser, closer, optional = false, f, g, backtracks = [
     case 'object':
       opener = (0, combinator_1.clear)((0, combinator_1.matcher)(opener, true));
   }
+  switch (typeof parser) {
+    case 'string':
+    case 'object':
+      parser = (0, combinator_1.clear)((0, combinator_1.matcher)(parser, true));
+  }
   switch (typeof closer) {
     case 'string':
     case 'object':
@@ -3124,16 +3127,14 @@ function surround(opener, parser, closer, optional = false, f, g, backtracks = [
     if (!nodesM && !optional) {
       setBacktrack(context, backtracks, position);
       const result = g?.([nodesO, nodesM], context);
-      revert(context, linebreak);
-      return result;
+      return result || void revert(context, linebreak);
     }
     const nodesC = nodesM || optional ? closer(input) : undefined;
     context.range = context.position - position;
     if (!nodesC) {
       setBacktrack(context, backtracks, position);
       const result = g?.([nodesO, nodesM], context);
-      revert(context, linebreak);
-      return result;
+      return result || void revert(context, linebreak);
     }
     if (context.position === position) {
       return void revert(context, linebreak);
@@ -3142,10 +3143,8 @@ function surround(opener, parser, closer, optional = false, f, g, backtracks = [
     const result = f ? f([nodesO, nodesM, nodesC], context) : nodesO.import(nodesM ?? new parser_1.List()).import(nodesC);
     if (result) {
       context.linebreak ||= linebreak;
-    } else {
-      revert(context, linebreak);
     }
-    return result;
+    return result || void revert(context, linebreak);
   });
 }
 exports.surround = surround;
@@ -3183,7 +3182,7 @@ function isBacktrack(context, backtracks, position = context.position, length =
 }
 exports.isBacktrack = isBacktrack;
 function setBacktrack(context, backtracks, position, length = 1) {
-  // 以降バックトラックの可能性がなく記録不要の場合もあるが判別が面倒なので省略
+  // バックトラックの可能性がなく記録不要の場合もあるが判別が面倒なので省略
   const {
     source
   } = context;
@@ -4396,15 +4395,16 @@ exports.normalize = normalize;
 function format(source) {
   return source.replace(/\r\n?/g, '\n');
 }
+const invalid = new RegExp([/(?![\t\r\n])[\x00-\x1F\x7F]/g.source, /(?!\u200D)[\u2006\u200B-\u200F\u202A-\u202F\u2060\uFEFF]|(?<![\u1820\u1821])\u180E/g.source, /[\uD800-\uDBFF](?![\uDC00-\uDFFF])|(?<![\uD800-\uDBFF])[\uDC00-\uDFFF]/g.source].join('|'), 'g');
 function sanitize(source) {
-  return source.replace(/[\x00-\x08\x0B\x0C\x0E-\x1F\x7F]|[\u2006\u200B-\u200F\u202A-\u202F\u2060\uFEFF]|(?<![\u1820\u1821])\u180E/g, UNICODE_REPLACEMENT_CHARACTER).replace(/[\uD800-\uDBFF][\uDC00-\uDFFF]?|[\uDC00-\uDFFF]/g, char => char.length === 1 ? UNICODE_REPLACEMENT_CHARACTER : char);
+  return source.replace(invalid, UNICODE_REPLACEMENT_CHARACTER);
 }
 // https://dev.w3.org/html5/html-author/charref
 // https://en.wikipedia.org/wiki/Whitespace_character
 exports.invisibleHTMLEntityNames = ['Tab', 'NewLine', 'NonBreakingSpace', 'nbsp', 'shy', 'ensp', 'emsp', 'emsp13', 'emsp14', 'numsp', 'puncsp', 'ThinSpace', 'thinsp', 'VeryThinSpace', 'hairsp', 'ZeroWidthSpace', 'NegativeVeryThinSpace', 'NegativeThinSpace', 'NegativeMediumSpace', 'NegativeThickSpace', 'zwj', 'zwnj', 'lrm', 'rlm', 'MediumSpace', 'NoBreak', 'ApplyFunction', 'af', 'InvisibleTimes', 'it', 'InvisibleComma', 'ic'];
-const unreadableHTMLEntityNames = exports.invisibleHTMLEntityNames.slice(2);
-const unreadableEscapableCharacters = unreadableHTMLEntityNames.map(name => (0, htmlentity_1.unsafehtmlentity)((0, parser_1.input)(`&${name};`, {})).head.value);
-const unreadableEscapableCharacter = new RegExp(`[${unreadableEscapableCharacters.join('')}]`, 'g');
+const unreadableEscapeHTMLEntityNames = exports.invisibleHTMLEntityNames.filter(name => !['Tab', 'NewLine', 'NonBreakingSpace', 'nbsp', 'zwj', 'zwnj'].includes(name));
+const unreadableEscapeCharacters = unreadableEscapeHTMLEntityNames.map(name => (0, htmlentity_1.unsafehtmlentity)((0, parser_1.input)(`&${name};`, {})).head.value);
+const unreadableEscapeCharacter = new RegExp(`[${unreadableEscapeCharacters.join('')}]`, 'g');
 // https://www.pandanoir.info/entry/2018/03/11/193000
 // http://anti.rosx.net/etc/memo/002_space.html
 // http://nicowiki.com/%E7%A9%BA%E7%99%BD%E3%83%BB%E7%89%B9%E6%AE%8A%E8%A8%98%E5%8F%B7.html
@@ -4416,7 +4416,7 @@ const unreadableSpecialCharacters = (/* unused pure expression or super */ null
 // ZERO WIDTH NON-JOINER
 '\u200C',
 // ZERO WIDTH JOINER
-'\u200D',
+//'\u200D',
 // LEFT-TO-RIGHT MARK
 '\u200E',
 // RIGHT-TO-LEFT MARK
@@ -4439,7 +4439,7 @@ const unreadableSpecialCharacters = (/* unused pure expression or super */ null
 '\uFEFF']));
 // 特殊不可視文字はエディタおよびソースビューアでは等幅および強調表示により可視化する
 function escape(source) {
-  return source.replace(unreadableEscapableCharacter, char => `&${unreadableHTMLEntityNames[unreadableEscapableCharacters.indexOf(char)]};`);
+  return source.replace(unreadableEscapeCharacter, char => `&${unreadableEscapeHTMLEntityNames[unreadableEscapeCharacters.indexOf(char)]};`);
 }
 exports.escape = escape;
 
@@ -6041,6 +6041,7 @@ const reference_1 = __webpack_require__(9047);
 const template_1 = __webpack_require__(4510);
 const remark_1 = __webpack_require__(8948);
 const extension_1 = __webpack_require__(2743);
+const label_1 = __webpack_require__(2178);
 const link_1 = __webpack_require__(3628);
 const ruby_1 = __webpack_require__(7304);
 const html_1 = __webpack_require__(5013);
@@ -6096,7 +6097,7 @@ exports.inline = (0, combinator_1.lazy)(() => (0, combinator_1.union)([input =>
       return (0, html_1.html)(input);
     case '$':
       if (source[position + 1] === '{') return (0, math_1.math)(input);
-      return (0, extension_1.extension)(input) || (0, math_1.math)(input);
+      return (0, label_1.label)(input) || (0, math_1.math)(input);
     case '+':
       if (source[position + 1] === '+') return (0, insertion_1.insertion)(input);
       break;
@@ -6208,16 +6209,45 @@ const account_1 = __webpack_require__(4107);
 const hashtag_1 = __webpack_require__(5764);
 const hashnum_1 = __webpack_require__(8684);
 const anchor_1 = __webpack_require__(8535);
-exports.autolink = (0, combinator_1.lazy)(() => (0, combinator_1.validate)(new RegExp([/(?<![0-9a-z])@/yi.source, /(?<![^\p{C}\p{S}\p{P}\s]|emoji)#/yiu.source, /(?<![0-9a-z])>>/yi.source, /(?<![0-9a-z][.+-]?|[@#])!?[0-9a-z]/yi.source].join('|').replace(/emoji/g, hashtag_1.emoji), 'yiu'), (0, combinator_1.state)(~1 /* State.autolink */, (0, combinator_1.union)([url_1.lineurl, url_1.url, email_1.email,
-// Escape unmatched email-like strings.
-//str(/[0-9a-z]+(?:[_.+-][0-9a-z]+[:@]?|:|@(?=@))*/yi),
-channel_1.channel, account_1.account,
-// Escape unmatched account-like strings.
-//str(/@+(?:[0-9a-z]+(?:[_.+-][0-9a-z]+)*)?/yi),
-hashtag_1.hashtag, hashnum_1.hashnum,
-// Escape unmatched hashtag-like strings.
-//str(new RegExp(/#+(?:(?:[^\p{C}\p{S}\p{P}\s]|emoji)+(?:['_.+-](?:[^\p{C}\p{S}\p{P}\s]|emoji)+)*)?/yu.source.replace(/emoji/g, emoji), 'yu')),
-anchor_1.anchor]))));
+const text_1 = __webpack_require__(5655);
+exports.autolink = (0, combinator_1.lazy)(() => (0, combinator_1.state)(~1 /* State.autolink */, input => {
+  const {
+    context: {
+      source,
+      position
+    }
+  } = input;
+  if (position === source.length) return;
+  const fst = source[position];
+  switch (fst) {
+    case '@':
+      return (0, channel_1.channel)(input) || (0, account_1.account)(input);
+    case '#':
+      return (0, hashtag_1.hashtag)(input) || (0, hashnum_1.hashnum)(input);
+    case '>':
+      return (0, anchor_1.anchor)(input);
+    case '!':
+      if (!source.startsWith('http', position + 1)) break;
+      if (position === 0) return (0, url_1.lineurl)(input);
+      switch (source[position - 1]) {
+        case '\r':
+        case '\n':
+          return (0, url_1.lineurl)(input);
+      }
+      break;
+    case 'h':
+      if (!source.startsWith('http', position)) return;
+      if (position === 0) return (0, url_1.lineurl)(input) || (0, url_1.url)(input) || (0, email_1.email)(input);
+      switch (source[position - 1]) {
+        case '\r':
+        case '\n':
+          return (0, url_1.lineurl)(input) || (0, url_1.url)(input) || (0, email_1.email)(input);
+      }
+      return (0, url_1.url)(input) || (0, email_1.email)(input);
+    default:
+      if ((0, text_1.isAlphanumeric)(fst)) return (0, email_1.email)(input);
+  }
+}));
 
 /***/ },
 
@@ -6234,10 +6264,9 @@ exports.account = void 0;
 const parser_1 = __webpack_require__(605);
 const combinator_1 = __webpack_require__(3484);
 const link_1 = __webpack_require__(3628);
-const source_1 = __webpack_require__(8745);
 const dom_1 = __webpack_require__(394);
 // https://example/@user must be a user page or a redirect page going there.
-exports.account = (0, combinator_1.lazy)(() => (0, combinator_1.rewrite)((0, combinator_1.surround)(/(?<![0-9a-z])@/yi, (0, source_1.str)(/[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?(?:\.[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?)*\//yi), (0, source_1.str)(/[a-z][0-9a-z]*(?:[-.][0-9a-z]+)*(?![-.]?[0-9a-z@#]|>>|:\S)/yi), true, undefined, undefined, [3 | 0 /* Backtrack.autolink */]), (0, combinator_1.constraint)(1 /* State.autolink */, (0, combinator_1.state)(1 /* State.autolink */, (0, combinator_1.fmap)((0, combinator_1.convert)(source => `[${source}]{ ${source.includes('/') ? `https://${source.slice(1).replace('/', '/@')}` : `/${source}`} }`, (0, combinator_1.union)([link_1.unsafelink]), false), ([{
+exports.account = (0, combinator_1.lazy)(() => (0, combinator_1.rewrite)((0, combinator_1.surround)(/(?<![0-9a-z])@/yi, /[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?(?:\.[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?)*\//yi, /[a-z][0-9a-z]*(?:[-.][0-9a-z]+)*(?![-.]?[0-9a-z@#]|>>|:\S)/yi, true, undefined, undefined, [3 | 0 /* Backtrack.autolink */]), (0, combinator_1.constraint)(1 /* State.autolink */, (0, combinator_1.state)(1 /* State.autolink */, (0, combinator_1.fmap)((0, combinator_1.convert)(source => `[${source}]{ ${source.includes('/') ? `https://${source.slice(1).replace('/', '/@')}` : `/${source}`} }`, (0, combinator_1.union)([link_1.unsafelink]), false), ([{
   value
 }]) => new parser_1.List([new parser_1.Data((0, dom_1.define)(value, {
   class: 'account'
@@ -6258,7 +6287,6 @@ exports.anchor = void 0;
 const parser_1 = __webpack_require__(605);
 const combinator_1 = __webpack_require__(3484);
 const link_1 = __webpack_require__(3628);
-const source_1 = __webpack_require__(8745);
 const dom_1 = __webpack_require__(394);
 // Timeline(pseudonym): user/tid
 // Thread(anonymous): cid
@@ -6267,7 +6295,7 @@ const dom_1 = __webpack_require__(394);
 // cid: YYYY-MMDD-HHMM-SSmmm
 // 内部表現はUnixTimeに統一する(時系列順)
 // 外部表現は投稿ごとに投稿者の投稿時のタイムゾーンに統一する(非時系列順)
-exports.anchor = (0, combinator_1.lazy)(() => (0, combinator_1.rewrite)((0, combinator_1.open)(/(?<![0-9a-z])>>/yi, (0, source_1.str)(/(?:[a-z][0-9a-z]*(?:-[0-9a-z]+)*\/)?[0-9a-z]+(?:-[0-9a-z]+)*(?!-?[0-9a-z@#]|>>|:\S)/yi), false, [3 | 0 /* Backtrack.autolink */]), (0, combinator_1.constraint)(1 /* State.autolink */, (0, combinator_1.state)(1 /* State.autolink */, (0, combinator_1.fmap)((0, combinator_1.convert)(source => `[${source}]{ ${source.includes('/') ? `/@${source.slice(2).replace('/', '/timeline?at=')}` : `?at=${source.slice(2)}`} }`, (0, combinator_1.union)([link_1.unsafelink]), false), ([{
+exports.anchor = (0, combinator_1.lazy)(() => (0, combinator_1.rewrite)((0, combinator_1.open)(/(?<![0-9a-z])>>/yi, /(?:[a-z][0-9a-z]*(?:-[0-9a-z]+)*\/)?[0-9a-z]+(?:-[0-9a-z]+)*(?!-?[0-9a-z@#]|>>|:\S)/yi, false, [3 | 0 /* Backtrack.autolink */]), (0, combinator_1.constraint)(1 /* State.autolink */, (0, combinator_1.state)(1 /* State.autolink */, (0, combinator_1.fmap)((0, combinator_1.convert)(source => `[${source}]{ ${source.includes('/') ? `/@${source.slice(2).replace('/', '/timeline?at=')}` : `?at=${source.slice(2)}`} }`, (0, combinator_1.union)([link_1.unsafelink]), false), ([{
   value
 }]) => new parser_1.List([new parser_1.Data((0, dom_1.define)(value, {
   class: 'anchor'
@@ -6292,7 +6320,7 @@ const hashtag_1 = __webpack_require__(5764);
 const source_1 = __webpack_require__(8745);
 const dom_1 = __webpack_require__(394);
 // https://example/@user?ch=a+b must be a user channel page or a redirect page going there.
-exports.channel = (0, combinator_1.lazy)(() => (0, combinator_1.rewrite)((0, combinator_1.sequence)([(0, combinator_1.surround)(/(?<![0-9a-z])@/yi, (0, source_1.str)(/[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?(?:\.[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?)*\//yi), (0, source_1.str)(/[a-z][0-9a-z]*(?:[-.][0-9a-z]+)*(?![-.]?[0-9a-z@]|>>|:\S)/yi), true, undefined, undefined, [3 | 0 /* Backtrack.autolink */]), (0, combinator_1.some)((0, combinator_1.verify)((0, combinator_1.surround)('#', (0, source_1.str)(new RegExp([/(?!['_])(?:[^\p{C}\p{S}\p{P}\s]|emoji|'(?=[0-9A-Za-z])|_(?=[^\p{C}\p{S}\p{P}\s]|emoji))+/yu.source].join('').replace(/emoji/g, hashtag_1.emoji), 'yu')), (0, source_1.str)(new RegExp([/(?![0-9a-z@]|>>|:\S|[^\p{C}\p{S}\p{P}\s]|emoji)/yu.source].join('').replace(/emoji/g, hashtag_1.emoji), 'yu')), false, undefined, undefined, [3 | 0 /* Backtrack.autolink */]), ([{
+exports.channel = (0, combinator_1.lazy)(() => (0, combinator_1.rewrite)((0, combinator_1.sequence)([(0, combinator_1.surround)(/(?<![0-9a-z])@/yi, /[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?(?:\.[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?)*\//yi, /[a-z][0-9a-z]*(?:[-.][0-9a-z]+)*(?![-.]?[0-9a-z@]|>>|:\S)/yi, true, undefined, undefined, [3 | 0 /* Backtrack.autolink */]), (0, combinator_1.some)((0, combinator_1.verify)((0, combinator_1.surround)('#', (0, source_1.str)(new RegExp([/(?!['_])(?:[^\p{C}\p{S}\p{P}\s]|emoji|'(?=[0-9A-Za-z])|_(?=[^\p{C}\p{S}\p{P}\s]|emoji))+/yu.source].join('|').replace(/emoji/g, hashtag_1.emoji.source), 'yu')), new RegExp([/(?![0-9a-z@]|>>|:\S|[^\p{C}\p{S}\p{P}\s]|emoji)/yu.source].join('|').replace(/emoji/g, hashtag_1.emoji.source), 'yu'), false, undefined, undefined, [3 | 0 /* Backtrack.autolink */]), ([{
   value
 }]) => !/^[0-9]{1,4}$|^[0-9]{5}/.test(value)))]), (0, combinator_1.constraint)(1 /* State.autolink */, (0, combinator_1.state)(1 /* State.autolink */, (0, combinator_1.fmap)((0, combinator_1.convert)(source => `[${source}]{ ${source.includes('/') ? `https://${source.slice(1, source.indexOf('#')).replace('/', '/@')}` : `/${source.slice(0, source.indexOf('#'))}`} }`, (0, combinator_1.union)([link_1.unsafelink]), false), ([{
   value: el
@@ -6353,9 +6381,8 @@ const parser_1 = __webpack_require__(605);
 const combinator_1 = __webpack_require__(3484);
 const link_1 = __webpack_require__(3628);
 const hashtag_1 = __webpack_require__(5764);
-const source_1 = __webpack_require__(8745);
 const dom_1 = __webpack_require__(394);
-exports.hashnum = (0, combinator_1.lazy)(() => (0, combinator_1.rewrite)((0, combinator_1.open)(new RegExp([/(?<![^\p{C}\p{S}\p{P}\s]|emoji)#/yiu.source].join('').replace(/emoji/g, hashtag_1.emoji), 'yu'), (0, source_1.str)(new RegExp([/[0-9]{1,9}(?![0-9a-z@#]|>>|:\S|[^\p{C}\p{S}\p{P}\s]|emoji)/yu.source].join('').replace(/emoji/g, hashtag_1.emoji), 'yu')), false, [1 | 0 /* Backtrack.autolink */]), (0, combinator_1.constraint)(1 /* State.autolink */, (0, combinator_1.state)(1 /* State.autolink */, (0, combinator_1.fmap)((0, combinator_1.convert)(source => `[${source}]{ ${source.slice(1)} }`, (0, combinator_1.union)([link_1.unsafelink]), false), ([{
+exports.hashnum = (0, combinator_1.lazy)(() => (0, combinator_1.rewrite)((0, combinator_1.open)(new RegExp([/(?<![^\p{C}\p{S}\p{P}\s]|emoji)#/yu.source].join('|').replace(/emoji/g, hashtag_1.emoji.source), 'yu'), new RegExp([/[0-9]{1,9}(?![0-9a-z@#]|>>|:\S|[^\p{C}\p{S}\p{P}\s]|emoji)/yu.source].join('|').replace(/emoji/g, hashtag_1.emoji.source), 'yu'), false, [1 | 0 /* Backtrack.autolink */]), (0, combinator_1.constraint)(1 /* State.autolink */, (0, combinator_1.state)(1 /* State.autolink */, (0, combinator_1.fmap)((0, combinator_1.convert)(source => `[${source}]{ ${source.slice(1)} }`, (0, combinator_1.union)([link_1.unsafelink]), false), ([{
   value
 }]) => new parser_1.List([new parser_1.Data((0, dom_1.define)(value, {
   class: 'hashnum',
@@ -6381,8 +6408,8 @@ const source_1 = __webpack_require__(8745);
 const dom_1 = __webpack_require__(394);
 // https://example/hashtags/a must be a hashtag page or a redirect page going there.
 // https://github.com/tc39/proposal-regexp-unicode-property-escapes#matching-emoji
-exports.emoji = String.raw`\p{Emoji_Modifier_Base}\p{Emoji_Modifier}?|\p{Emoji_Presentation}|\p{Emoji}\uFE0F`;
-exports.hashtag = (0, combinator_1.lazy)(() => (0, combinator_1.rewrite)((0, combinator_1.verify)((0, combinator_1.surround)(new RegExp([/(?<![^\p{C}\p{S}\p{P}\s]|emoji)#/yiu.source].join('').replace(/emoji/g, exports.emoji), 'yu'), (0, source_1.str)(new RegExp([/(?!['_])(?:[^\p{C}\p{S}\p{P}\s]|emoji|'(?=[0-9A-Za-z])|_(?=[^\p{C}\p{S}\p{P}\s]|emoji))+/yu.source].join('').replace(/emoji/g, exports.emoji), 'yu')), (0, source_1.str)(new RegExp([/(?![0-9a-z@#]|>>|:\S|[^\p{C}\p{S}\p{P}\s]|emoji)/yu.source].join('').replace(/emoji/g, exports.emoji), 'yu')), false, undefined, undefined, [3 | 0 /* Backtrack.autolink */]), ([{
+exports.emoji = /\p{Emoji_Modifier_Base}\p{Emoji_Modifier}?|\p{Emoji_Presentation}|\p{Emoji}\uFE0F|\u200D/u;
+exports.hashtag = (0, combinator_1.lazy)(() => (0, combinator_1.rewrite)((0, combinator_1.verify)((0, combinator_1.surround)(new RegExp([/(?<![^\p{C}\p{S}\p{P}\s]|emoji)#/yu.source].join('|').replace(/emoji/g, exports.emoji.source), 'yu'), (0, source_1.str)(new RegExp([/(?!['_])(?:[^\p{C}\p{S}\p{P}\s]|emoji|'(?=[0-9A-Za-z])|_(?=[^\p{C}\p{S}\p{P}\s]|emoji))+/yu.source].join('|').replace(/emoji/g, exports.emoji.source), 'yu')), new RegExp([/(?![0-9a-z@#]|>>|:\S|[^\p{C}\p{S}\p{P}\s]|emoji)/yu.source].join('|').replace(/emoji/g, exports.emoji.source), 'yu'), false, undefined, undefined, [3 | 0 /* Backtrack.autolink */]), ([{
   value
 }]) => !/^[0-9]{1,4}$|^[0-9]{5}/.test(value)), (0, combinator_1.constraint)(1 /* State.autolink */, (0, combinator_1.state)(1 /* State.autolink */, (0, combinator_1.fmap)((0, combinator_1.convert)(source => `[${source}]{ ${`/hashtags/${source.slice(1)}`} }`, (0, combinator_1.union)([link_1.unsafelink]), false), ([{
   value
@@ -7065,7 +7092,7 @@ Object.setPrototypeOf(attrspecs, null);
 Object.values(attrspecs).forEach(o => Object.setPrototypeOf(o, null));
 exports.html = (0, combinator_1.lazy)(() => (0, combinator_1.validate)(/<[a-z]+(?=[ >])/yi, (0, combinator_1.union)([(0, combinator_1.surround)(
 // https://html.spec.whatwg.org/multipage/syntax.html#void-elements
-(0, source_1.str)(/<(?:area|base|br|col|embed|hr|img|input|link|meta|source|track|wbr)(?=[ >])/yi), (0, combinator_1.some)((0, combinator_1.union)([exports.attribute])), (0, combinator_1.open)((0, source_1.str)(/ ?/y), (0, source_1.str)('>'), true), true, ([as, bs = new parser_1.List(), cs], context) => new parser_1.List([new parser_1.Data(elem(as.head.value.slice(1), false, [...(0, util_1.unwrap)(as.import(bs).import(cs))], new parser_1.List(), new parser_1.List(), context))]), ([as, bs = new parser_1.List()], context) => new parser_1.List([new parser_1.Data(elem(as.head.value.slice(1), false, [...(0, util_1.unwrap)(as.import(bs))], new parser_1.List(), new parser_1.List(), context))])), (0, combinator_1.match)(new RegExp(String.raw`<(${TAGS.join('|')})(?=[ >])`, 'y'), (0, memoize_1.memoize)(([, tag]) => (0, combinator_1.surround)((0, combinator_1.surround)((0, source_1.str)(`<${tag}`), (0, combinator_1.some)(exports.attribute), (0, combinator_1.open)((0, source_1.str)(/ ?/y), (0, source_1.str)('>'), true), true, ([as, bs = new parser_1.List(), cs]) => as.import(bs).import(cs), ([as, bs = new parser_1.List()]) => as.import(bs)),
+(0, source_1.str)(/<(?:area|base|br|col|embed|hr|img|input|link|meta|source|track|wbr)(?=[ >])/y), (0, combinator_1.some)((0, combinator_1.union)([exports.attribute])), (0, combinator_1.open)((0, source_1.str)(/ ?/y), (0, source_1.str)('>'), true), true, ([as, bs = new parser_1.List(), cs], context) => new parser_1.List([new parser_1.Data(elem(as.head.value.slice(1), false, [...(0, util_1.unwrap)(as.import(bs).import(cs))], new parser_1.List(), new parser_1.List(), context))]), ([as, bs = new parser_1.List()], context) => new parser_1.List([new parser_1.Data(elem(as.head.value.slice(1), false, [...(0, util_1.unwrap)(as.import(bs))], new parser_1.List(), new parser_1.List(), context))])), (0, combinator_1.match)(new RegExp(String.raw`<(${TAGS.join('|')})(?=[ >])`, 'y'), (0, memoize_1.memoize)(([, tag]) => (0, combinator_1.surround)((0, combinator_1.surround)((0, source_1.str)(`<${tag}`), (0, combinator_1.some)(exports.attribute), (0, combinator_1.open)((0, source_1.str)(/ ?/y), (0, source_1.str)('>'), true), true, ([as, bs = new parser_1.List(), cs]) => as.import(bs).import(cs), ([as, bs = new parser_1.List()]) => as.import(bs)),
 // 不可視のHTML構造が可視構造を変化させるべきでない。
 // 可視のHTMLは優先度変更を検討する。
 // このため<>は将来的に共通構造を変化させる可能性があり
@@ -7140,19 +7167,10 @@ Object.defineProperty(exports, "__esModule", ({
 exports.htmlentity = exports.unsafehtmlentity = void 0;
 const parser_1 = __webpack_require__(605);
 const combinator_1 = __webpack_require__(3484);
+const source_1 = __webpack_require__(8745);
 const util_1 = __webpack_require__(4992);
 const dom_1 = __webpack_require__(394);
-exports.unsafehtmlentity = (0, combinator_1.focus)(/&(?:[0-9A-Za-z]+;?)?/y,
-//({ source }) => [[parser(source) ?? `${Command.Error}${source}`], '']));
-({
-  context
-}) => {
-  const {
-    source
-  } = context;
-  context.position += source.length;
-  return source.length > 1 && source.at(-1) === ';' ? new parser_1.List([new parser_1.Data(parser(source) ?? source)]) : new parser_1.List([new parser_1.Data(source)]);
-});
+exports.unsafehtmlentity = (0, combinator_1.surround)((0, source_1.str)('&'), (0, source_1.str)(/[0-9A-Za-z]+/y), (0, source_1.str)(';'), false, ([as, bs, cs]) => new parser_1.List([new parser_1.Data(parser(as.head.value + bs.head.value + cs.head.value))]), ([as, bs]) => new parser_1.List([new parser_1.Data(as.head.value + (bs?.head?.value ?? ''))]), [3 | 64 /* Backtrack.bracket */]);
 exports.htmlentity = (0, combinator_1.fmap)((0, combinator_1.union)([exports.unsafehtmlentity]), ([{
   value
 }]) => new parser_1.List([length === 1 || value.at(-1) !== ';' ? new parser_1.Data(value) : new parser_1.Data((0, dom_1.html)('span', {
@@ -7162,8 +7180,7 @@ exports.htmlentity = (0, combinator_1.fmap)((0, combinator_1.union)([exports.uns
 const parser = (el => entity => {
   if (entity === '&NewLine;') return ' ';
   el.innerHTML = entity;
-  const text = el.textContent;
-  return entity === text ? undefined : text;
+  return el.textContent;
 })((0, dom_1.html)('span'));
 
 /***/ },
@@ -7772,6 +7789,7 @@ exports.ruby = (0, combinator_1.lazy)(() => (0, combinator_1.bind)((0, combinato
       }, acc) => value + ' ' + acc, '').trim())), new parser_1.Data((0, dom_1.html)('rp', ')'))])))))]);
   }
 }));
+const delimiter = /[$"`\[\](){}<>（）［］｛｝]|\\?\n/y;
 const text = input => {
   const {
     context
@@ -7785,11 +7803,12 @@ const text = input => {
   for (let {
     position
   } = context; position < source.length; position = context.position) {
-    if (/[$"`\[\](){}<>（）［］｛｝]|\\?\n/yi.test(source.slice(position, position + 2))) break;
+    delimiter.lastIndex = position;
+    if (delimiter.test(source)) break;
     switch (source[position]) {
       case '&':
         {
-          const result = (0, htmlentity_1.unsafehtmlentity)(input) ?? (0, source_1.txt)(input);
+          const result = source[position + 1] !== ' ' ? (0, htmlentity_1.unsafehtmlentity)(input) ?? (0, source_1.txt)(input) : (0, source_1.txt)(input);
           acc.last.value += result.head.value;
           continue;
         }
@@ -8517,7 +8536,7 @@ exports.strs = strs;
 Object.defineProperty(exports, "__esModule", ({
   value: true
 }));
-exports.backToEmailHead = exports.backToUrlHead = exports.backToWhitespace = exports.next = exports.canSkip = exports.linebreak = exports.txt = exports.text = exports.nonWhitespace = void 0;
+exports.isAlphanumeric = exports.backToEmailHead = exports.backToUrlHead = exports.backToWhitespace = exports.next = exports.canSkip = exports.linebreak = exports.txt = exports.text = exports.nonWhitespace = void 0;
 const parser_1 = __webpack_require__(605);
 const combinator_1 = __webpack_require__(3484);
 const dom_1 = __webpack_require__(394);
@@ -8678,8 +8697,9 @@ function backToEmailHead(source, position, index) {
 exports.backToEmailHead = backToEmailHead;
 function isAlphanumeric(char) {
   if (char < '0' || '\x7F' < char) return false;
-  return '0' <= char && char <= '9' || 'a' <= char && char <= 'z' || 'A' <= char && char <= 'Z';
+  return '0' <= char && char <= '9' || 'A' <= char && char <= 'Z' || 'a' <= char && char <= 'z';
 }
+exports.isAlphanumeric = isAlphanumeric;
 //const dict = new class {
 //  constructor() {
 //    [
@@ -8724,7 +8744,6 @@ function seek(source, position) {
       case '@':
       case '#':
       case '$':
-      case '&':
       case '"':
       case '`':
       case '[':
@@ -8760,6 +8779,9 @@ function seek(source, position) {
       case ':':
         if (source[i + 1] === '/' && source[i + 2] === '/') return i;
         continue;
+      case '&':
+        if (source[i + 1] !== ' ') return i;
+        continue;
       case ' ':
       case '\t':
       case '　':

package/markdown.d.ts

@@ -1057,7 +1057,9 @@ export namespace MarkdownParser {
     export interface UnsafeHTMLEntityParser extends
       // ©
       Inline<'unsafehtmlentity'>,
-      Parser<string, Context, []> {
+      Parser<string, Context, [
+        SourceParser.StrParser,
+      ]> {
     }
     export interface ShortMediaParser extends
       // !https://host

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "securemark",
-  "version": "0.294.5",
+  "version": "0.294.7",
   "description": "Secure markdown renderer working on browsers for user input data.",
   "private": false,
   "homepage": "https://github.com/falsandtru/securemark",

package/src/combinator/control/manipulation/indent.ts

@@ -16,7 +16,7 @@ export function indent<N>(opener: RegExp | Parser<N>, parser: Parser<N> | boolea
     opener = / {1,4}|\t{1,2}/y;
   }
   assert(!opener.flags.match(/[gm]/) && opener.sticky && !opener.source.startsWith('^'));
-  assert(parser);
+  assert(parser = parser as Parser<N>);
   return failsafe(bind(block(match(
     opener,
     memoize(
@@ -27,10 +27,8 @@ export function indent<N>(opener: RegExp | Parser<N>, parser: Parser<N> | boolea
         return new List([new Data(source.slice(position))]);
       }))),
     ([indent]) => indent.length * 2 + -(indent[0] === ' '), [], 2 ** 4 - 1)), separation),
-    (lines, context) => {
-      assert(parser = parser as Parser<N>);
-      return parser(subinput(trimBlockEnd(lines.foldl((acc, node) => acc + node.value, '')), context));
-    }));
+    (lines, context) =>
+      parser(subinput(trimBlockEnd(lines.foldl((acc, node) => acc + node.value, '')), context))));
 }
 
 function trimBlockEnd(block: string): string {

package/src/combinator/control/manipulation/surround.ts

@@ -29,8 +29,22 @@ export function surround<P extends Parser<unknown>, S = string>(
   g?: (rss: [List<Data<S>>, List<Data<Node<P>>> | undefined], context: Context<P>) => Result<Node<P>, Context<P>, SubParsers<P>>,
   backtracks?: readonly number[],
 ): P;
+export function surround<P extends Parser<string>, S = string>(
+  opener: string | RegExp | Parser<S, Context<P>>, parser: string | RegExp | P, closer: string | RegExp | Parser<S, Context<P>>,
+  optional?: false,
+  f?: (rss: [List<Data<S>>, List<Data<Node<P>>>, List<Data<S>>], context: Context<P>) => Result<Node<P>, Context<P>, SubParsers<P>>,
+  g?: (rss: [List<Data<S>>, List<Data<Node<P>>> | undefined], context: Context<P>) => Result<Node<P>, Context<P>, SubParsers<P>>,
+  backtracks?: readonly number[],
+): P;
+export function surround<P extends Parser<string>, S = string>(
+  opener: string | RegExp | Parser<S, Context<P>>, parser: string | RegExp | P, closer: string | RegExp | Parser<S, Context<P>>,
+  optional?: boolean,
+  f?: (rss: [List<Data<S>>, List<Data<Node<P>>> | undefined, List<Data<S>>], context: Context<P>) => Result<Node<P>, Context<P>, SubParsers<P>>,
+  g?: (rss: [List<Data<S>>, List<Data<Node<P>>> | undefined], context: Context<P>) => Result<Node<P>, Context<P>, SubParsers<P>>,
+  backtracks?: readonly number[],
+): P;
 export function surround<N>(
-  opener: string | RegExp | Parser<N>, parser: Parser<N>, closer: string | RegExp | Parser<N>,
+  opener: string | RegExp | Parser<N>, parser: string | RegExp | Parser<N>, closer: string | RegExp | Parser<N>,
   optional: boolean = false,
   f?: (rss: [List<Data<N>>, List<Data<N>>, List<Data<N>>], context: Ctx) => Result<N>,
   g?: (rss: [List<Data<N>>, List<Data<N>> | undefined], context: Ctx) => Result<N>,
@@ -41,11 +55,19 @@ export function surround<N>(
     case 'object':
       opener = clear(matcher(opener, true));
   }
+  assert(opener);
+  switch (typeof parser) {
+    case 'string':
+    case 'object':
+      parser = clear(matcher(parser, true));
+  }
+  assert(parser);
   switch (typeof closer) {
     case 'string':
     case 'object':
       closer = clear(matcher(closer, true));
   }
+  assert(closer);
   return failsafe(input => {
     const { context } = input;
     const { source, position } = context;
@@ -66,8 +88,7 @@ export function surround<N>(
     if (!nodesM && !optional) {
       setBacktrack(context, backtracks, position);
       const result = g?.([nodesO, nodesM], context);
-      revert(context, linebreak);
-      return result;
+      return result || void revert(context, linebreak);
     }
     const nodesC = nodesM || optional ? closer(input) : undefined;
     assert(context.position >= position);
@@ -75,8 +96,7 @@ export function surround<N>(
     if (!nodesC) {
       setBacktrack(context, backtracks, position);
       const result = g?.([nodesO, nodesM], context);
-      revert(context, linebreak);
-      return result;
+      return result || void revert(context, linebreak);
     }
     if (context.position === position) {
       return void revert(context, linebreak);
@@ -88,10 +108,7 @@ export function surround<N>(
     if (result) {
       context.linebreak ||= linebreak;
     }
-    else {
-      revert(context, linebreak);
-    }
-    return result;
+    return result || void revert(context, linebreak);
   });
 }
 export function open<P extends Parser<unknown>>(
@@ -100,13 +117,19 @@ export function open<P extends Parser<unknown>>(
   optional?: boolean,
   backtracks?: readonly number[],
 ): P;
+export function open<P extends Parser<string>>(
+  opener: string | RegExp | Parser<Node<P>, Context<P>>,
+  parser: string | RegExp | P,
+  optional?: boolean,
+  backtracks?: readonly number[],
+): P;
 export function open<N>(
   opener: string | RegExp | Parser<N, Ctx>,
-  parser: Parser<N>,
+  parser: string | RegExp | Parser<N>,
   optional?: boolean,
   backtracks?: readonly number[],
 ): Parser<N> {
-  return surround(opener, parser, '', optional, undefined, undefined, backtracks);
+  return surround(opener, parser as Parser<N>, '', optional, undefined, undefined, backtracks);
 }
 export function close<P extends Parser<unknown>>(
   parser: P,
@@ -114,13 +137,19 @@ export function close<P extends Parser<unknown>>(
   optional?: boolean,
   backtracks?: readonly number[],
 ): P;
+export function close<P extends Parser<string>>(
+  parser: string | RegExp | P,
+  closer: string | RegExp | Parser<Node<P>, Context<P>>,
+  optional?: boolean,
+  backtracks?: readonly number[],
+): P;
 export function close<N>(
-  parser: Parser<N>,
+  parser: string | RegExp | Parser<N>,
   closer: string | RegExp | Parser<N, Ctx>,
   optional?: boolean,
   backtracks?: readonly number[],
 ): Parser<N> {
-  return surround('', parser, closer, optional, undefined, undefined, backtracks);
+  return surround('', parser as Parser<N>, closer, optional, undefined, undefined, backtracks);
 }
 
 const statesize = 2;
@@ -153,7 +182,7 @@ export function setBacktrack(
   position: number,
   length: number = 1,
 ): void {
-  // 以降バックトラックの可能性がなく記録不要の場合もあるが判別が面倒なので省略
+  // バックトラックの可能性がなく記録不要の場合もあるが判別が面倒なので省略
   const { source } = context;
   if (position === source.length) return;
   if (length === 0) return;

package/src/combinator/data/parser.ts

@@ -39,8 +39,11 @@ export interface CtxOptions {
   // 区間別テーブルは固定サイズであるためプールして再使用できる。
   // 従って分割時のデータ構造は区間ごとに探索木を動的に生成しデータ数に応じてテーブルに移行するのが最も効率的である。
   // これにより最悪時間計算量線形化に要する最悪空間計算量が+1nに局限される。
+  // またはテーブルの参照が高速なら変換せず併用してもよい。
   // 木とテーブルいずれにおいてもバックトラックデータとオーバーヘッドを合わせた追加データサイズの最大値は
   // セグメントサイズに制約されるため入力サイズに対する最大追加データサイズの平均比率はかなり小さくなる。
+  // 必要なテーブルの最大サイズは最大セグメントサイズであるため最大追加データサイズは入力サイズにかかわらず
+  // 10KB*並列数に留まり最大数百文字以下の短文ならば数百byte*並列数となる。
   //
   // 1. データ数が規定数を超えたら区間テーブルを生成しデータを振り分ける。
   //   - 子ノードのポインタだけ保持するとしても1ノード複数データ保持で圧縮できるかは微妙。

package/src/parser/api/normalize.test.ts

@@ -67,12 +67,20 @@ describe('Unit: parser/normalize', () => {
       assert(normalize('\x01---\na: b\x01\n---\n\n!> \x01---\na: b\x01\n---') === '\uFFFD---\na: b\uFFFD\n---\n\n!> \uFFFD---\na: b\uFFFD\n---');
     });
 
+    it('emoji', () => {
+      assert(normalize('😀') === '😀');
+      assert(normalize('🤚🏽') === '🤚🏽');
+      assert(normalize('👨‍👩‍👧') === '👨‍👩‍👧');
+      assert(normalize('🇺🇳') === '🇺🇳');
+      assert(normalize('#️⃣*️⃣0️⃣1️⃣2️⃣3️⃣4️⃣5️⃣6️⃣7️⃣8️⃣9️⃣') === '#️⃣*️⃣0️⃣1️⃣2️⃣3️⃣4️⃣5️⃣6️⃣7️⃣8️⃣9️⃣');
+    });
+
   });
 
   describe('escape', () => {
     it('', () => {
       assert(escape('\u200B') === '​');
-      assert(escape('\u200D') === '‍');
+      assert(escape('\u200F') === '‏');
     });
 
   });

package/src/parser/api/normalize.ts

@@ -9,17 +9,16 @@ export function normalize(source: string): string {
 }
 
 function format(source: string): string {
-  return source
-    .replace(/\r\n?/g, '\n');
+  return source.replace(/\r\n?/g, '\n');
 }
 
+const invalid = new RegExp([
+  /(?![\t\r\n])[\x00-\x1F\x7F]/g.source,
+  /(?!\u200D)[\u2006\u200B-\u200F\u202A-\u202F\u2060\uFEFF]|(?<![\u1820\u1821])\u180E/g.source,
+  /[\uD800-\uDBFF](?![\uDC00-\uDFFF])|(?<![\uD800-\uDBFF])[\uDC00-\uDFFF]/g.source,
+].join('|'), 'g');
 function sanitize(source: string): string {
-  return source
-    .replace(/[\x00-\x08\x0B\x0C\x0E-\x1F\x7F]|[\u2006\u200B-\u200F\u202A-\u202F\u2060\uFEFF]|(?<![\u1820\u1821])\u180E/g, UNICODE_REPLACEMENT_CHARACTER)
-    .replace(/[\uD800-\uDBFF][\uDC00-\uDFFF]?|[\uDC00-\uDFFF]/g, char =>
-      char.length === 1
-        ? UNICODE_REPLACEMENT_CHARACTER
-        : char);
+  return source.replace(invalid, UNICODE_REPLACEMENT_CHARACTER);
 }
 
 // https://dev.w3.org/html5/html-author/charref
@@ -58,13 +57,20 @@ export const invisibleHTMLEntityNames = [
   'InvisibleComma',
   'ic',
 ] as const;
-const unreadableHTMLEntityNames: readonly string[] = invisibleHTMLEntityNames.slice(2);
-const unreadableEscapableCharacters = unreadableHTMLEntityNames
+const unreadableEscapeHTMLEntityNames = invisibleHTMLEntityNames.filter(name => ![
+  'Tab',
+  'NewLine',
+  'NonBreakingSpace',
+  'nbsp',
+  'zwj',
+  'zwnj',
+].includes(name));
+const unreadableEscapeCharacters = unreadableEscapeHTMLEntityNames
   .map(name => unsafehtmlentity(input(`&${name};`, {}))!.head!.value);
-assert(unreadableEscapableCharacters.length === unreadableHTMLEntityNames.length);
-assert(unreadableEscapableCharacters.every(c => c.length === 1));
-const unreadableEscapableCharacter = new RegExp(`[${unreadableEscapableCharacters.join('')}]`, 'g');
-assert(!unreadableEscapableCharacter.source.includes('&'));
+assert(unreadableEscapeCharacters.length === unreadableEscapeHTMLEntityNames.length);
+assert(unreadableEscapeCharacters.every(c => c.length === 1));
+const unreadableEscapeCharacter = new RegExp(`[${unreadableEscapeCharacters.join('')}]`, 'g');
+assert(!unreadableEscapeCharacter.source.includes('&'));
 
 // https://www.pandanoir.info/entry/2018/03/11/193000
 // http://anti.rosx.net/etc/memo/002_space.html
@@ -77,7 +83,7 @@ const unreadableSpecialCharacters = [
   // ZERO WIDTH NON-JOINER
   '\u200C',
   // ZERO WIDTH JOINER
-  '\u200D',
+  //'\u200D',
   // LEFT-TO-RIGHT MARK
   '\u200E',
   // RIGHT-TO-LEFT MARK
@@ -103,7 +109,6 @@ assert(unreadableSpecialCharacters.every(c => sanitize(c) === UNICODE_REPLACEMEN
 
 // 特殊不可視文字はエディタおよびソースビューアでは等幅および強調表示により可視化する
 export function escape(source: string): string {
-  return source
-    .replace(unreadableEscapableCharacter, char =>
-      `&${unreadableHTMLEntityNames[unreadableEscapableCharacters.indexOf(char)]};`);
+  return source.replace(unreadableEscapeCharacter, char =>
+    `&${unreadableEscapeHTMLEntityNames[unreadableEscapeCharacters.indexOf(char)]};`);
 }

package/src/parser/api/parse.test.ts

@@ -361,9 +361,9 @@ describe('Unit: parser/api/parse', () => {
 
     it('backtrack', function () {
       this.timeout(5000);
-      // 最悪計算量での実行速度はCommonMarkの公式JS実装の32nより速い。
+      // 最悪計算量での実行速度はCommonMarkの公式JS実装の32nに対して3倍遅い程度。
       // 5n = annotation/reference + link + url/math + ruby + text
-      const source = `((([[[[#$[${'.'.repeat(19997)}`;
+      const source = `((([[[[#$[${'.'.repeat(19998)}`;
       assert.deepStrictEqual(
         [...parse(source, {}, { resources: { clock: 100000, recursions: [100] } }).children]
           .map(el => el.tagName),
@@ -372,7 +372,7 @@ describe('Unit: parser/api/parse', () => {
 
     it('backtrack error', function () {
       this.timeout(5000);
-      const source = `((([[[[#$[${'.'.repeat(19997 + 1)}`;
+      const source = `((([[[[#$[${'.'.repeat(19998 + 1)}`;
       assert.deepStrictEqual(
         [...parse(source, {}, { resources: { clock: 100000, recursions: [100] } }).children]
           .map(el => el.tagName),

package/src/parser/inline/autolink/account.ts

@@ -3,7 +3,6 @@ import { State, Backtrack } from '../../context';
 import { List, Data } from '../../../combinator/data/parser';
 import { union, state, constraint, rewrite, surround, convert, fmap, lazy } from '../../../combinator';
 import { unsafelink } from '../link';
-import { str } from '../../source';
 import { define } from 'typed-dom/dom';
 
 // https://example/@user must be a user page or a redirect page going there.
@@ -11,8 +10,8 @@ import { define } from 'typed-dom/dom';
 export const account: AutolinkParser.AccountParser = lazy(() => rewrite(
   surround(
     /(?<![0-9a-z])@/yi,
-    str(/[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?(?:\.[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?)*\//yi),
-    str(/[a-z][0-9a-z]*(?:[-.][0-9a-z]+)*(?![-.]?[0-9a-z@#]|>>|:\S)/yi),
+    /[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?(?:\.[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?)*\//yi,
+    /[a-z][0-9a-z]*(?:[-.][0-9a-z]+)*(?![-.]?[0-9a-z@#]|>>|:\S)/yi,
     true, undefined, undefined,
     [3 | Backtrack.autolink]),
   constraint(State.autolink, state(State.autolink, fmap(convert(

package/src/parser/inline/autolink/anchor.ts

@@ -3,7 +3,6 @@ import { State, Backtrack } from '../../context';
 import { List, Data } from '../../../combinator/data/parser';
 import { union, state, constraint, rewrite, open, convert, fmap, lazy } from '../../../combinator';
 import { unsafelink } from '../link';
-import { str } from '../../source';
 import { define } from 'typed-dom/dom';
 
 // Timeline(pseudonym): user/tid
@@ -19,7 +18,7 @@ import { define } from 'typed-dom/dom';
 export const anchor: AutolinkParser.AnchorParser = lazy(() => rewrite(
   open(
     /(?<![0-9a-z])>>/yi,
-    str(/(?:[a-z][0-9a-z]*(?:-[0-9a-z]+)*\/)?[0-9a-z]+(?:-[0-9a-z]+)*(?!-?[0-9a-z@#]|>>|:\S)/yi),
+    /(?:[a-z][0-9a-z]*(?:-[0-9a-z]+)*\/)?[0-9a-z]+(?:-[0-9a-z]+)*(?!-?[0-9a-z@#]|>>|:\S)/yi,
     false,
     [3 | Backtrack.autolink]),
   constraint(State.autolink, state(State.autolink, fmap(convert(

package/src/parser/inline/autolink/channel.ts

@@ -13,18 +13,18 @@ export const channel: AutolinkParser.ChannelParser = lazy(() => rewrite(
   sequence([
     surround(
       /(?<![0-9a-z])@/yi,
-      str(/[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?(?:\.[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?)*\//yi),
-      str(/[a-z][0-9a-z]*(?:[-.][0-9a-z]+)*(?![-.]?[0-9a-z@]|>>|:\S)/yi),
+      /[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?(?:\.[0-9a-z](?:(?:[0-9a-z]|-(?=[0-9a-z])){0,61}[0-9a-z])?)*\//yi,
+      /[a-z][0-9a-z]*(?:[-.][0-9a-z]+)*(?![-.]?[0-9a-z@]|>>|:\S)/yi,
       true, undefined, undefined,
       [3 | Backtrack.autolink]),
     some(verify(surround(
       '#',
       str(new RegExp([
         /(?!['_])(?:[^\p{C}\p{S}\p{P}\s]|emoji|'(?=[0-9A-Za-z])|_(?=[^\p{C}\p{S}\p{P}\s]|emoji))+/yu.source,
-      ].join('').replace(/emoji/g, emoji), 'yu')),
-      str(new RegExp([
+      ].join('|').replace(/emoji/g, emoji.source), 'yu')),
+      new RegExp([
         /(?![0-9a-z@]|>>|:\S|[^\p{C}\p{S}\p{P}\s]|emoji)/yu.source,
-      ].join('').replace(/emoji/g, emoji), 'yu')),
+      ].join('|').replace(/emoji/g, emoji.source), 'yu'),
       false, undefined, undefined,
       [3 | Backtrack.autolink]),
       ([{ value }]) => !/^[0-9]{1,4}$|^[0-9]{5}/.test(value as string))),

package/src/parser/inline/autolink/hashnum.ts

@@ -4,17 +4,16 @@ import { List, Data } from '../../../combinator/data/parser';
 import { union, state, constraint, rewrite, open, convert, fmap, lazy } from '../../../combinator';
 import { unsafelink } from '../link';
 import { emoji } from './hashtag';
-import { str } from '../../source';
 import { define } from 'typed-dom/dom';
 
 export const hashnum: AutolinkParser.HashnumParser = lazy(() => rewrite(
   open(
     new RegExp([
-      /(?<![^\p{C}\p{S}\p{P}\s]|emoji)#/yiu.source,
-    ].join('').replace(/emoji/g, emoji), 'yu'),
-    str(new RegExp([
+      /(?<![^\p{C}\p{S}\p{P}\s]|emoji)#/yu.source,
+    ].join('|').replace(/emoji/g, emoji.source), 'yu'),
+    new RegExp([
       /[0-9]{1,9}(?![0-9a-z@#]|>>|:\S|[^\p{C}\p{S}\p{P}\s]|emoji)/yu.source,
-    ].join('').replace(/emoji/g, emoji), 'yu')),
+    ].join('|').replace(/emoji/g, emoji.source), 'yu'),
     false,
     [1 | Backtrack.autolink]),
   constraint(State.autolink, state(State.autolink, fmap(convert(

package/src/parser/inline/autolink/hashtag.test.ts

@@ -54,10 +54,14 @@ describe('Unit: parser/inline/autolink/hashtag', () => {
       assert.deepStrictEqual(inspect(parser('#a_b'), ctx), [['<a class="hashtag" href="/hashtags/a_b">#a_b</a>'], '']);
       assert.deepStrictEqual(inspect(parser('#a__b'), ctx), [['<a class="hashtag" href="/hashtags/a">#a</a>'], '__b']);
       assert.deepStrictEqual(inspect(parser('#あ'), ctx), [['<a class="hashtag" href="/hashtags/あ">#あ</a>'], '']);
-      assert.deepStrictEqual(inspect(parser('#👩'), ctx), [['<a class="hashtag" href="/hashtags/👩">#👩</a>'], '']);
+      assert.deepStrictEqual(inspect(parser('#😀'), ctx), [['<a class="hashtag" href="/hashtags/😀">#😀</a>'], '']);
+      assert.deepStrictEqual(inspect(parser('#🤚🏽'), ctx), [['<a class="hashtag" href="/hashtags/🤚🏽">#🤚🏽</a>'], '']);
+      assert.deepStrictEqual(inspect(parser('#👨‍👩‍👧'), ctx), [['<a class="hashtag" href="/hashtags/👨‍👩‍👧">#👨‍👩‍👧</a>'], '']);
+      assert.deepStrictEqual(inspect(parser('#🇺🇳'), ctx), [['<a class="hashtag" href="/hashtags/🇺🇳">#🇺🇳</a>'], '']);
+      assert.deepStrictEqual(inspect(parser('##️⃣*️⃣0️⃣1️⃣2️⃣3️⃣4️⃣5️⃣6️⃣7️⃣8️⃣9️⃣'), ctx), [['<a class="hashtag" href="/hashtags/#️⃣*️⃣0️⃣1️⃣2️⃣3️⃣4️⃣5️⃣6️⃣7️⃣8️⃣9️⃣">##️⃣*️⃣0️⃣1️⃣2️⃣3️⃣4️⃣5️⃣6️⃣7️⃣8️⃣9️⃣</a>'], '']);
       assert.deepStrictEqual(inspect(parser('#1a'), ctx), [['<a class="hashtag" href="/hashtags/1a">#1a</a>'], '']);
       assert.deepStrictEqual(inspect(parser('#1あ'), ctx), [['<a class="hashtag" href="/hashtags/1あ">#1あ</a>'], '']);
-      assert.deepStrictEqual(inspect(parser('#1👩'), ctx), [['<a class="hashtag" href="/hashtags/1👩">#1👩</a>'], '']);
+      assert.deepStrictEqual(inspect(parser('#1😀'), ctx), [['<a class="hashtag" href="/hashtags/1😀">#1😀</a>'], '']);
       assert.deepStrictEqual(inspect(parser(`#a'`), ctx), [[`<a class="hashtag" href="/hashtags/a">#a</a>`], `'`]);
       assert.deepStrictEqual(inspect(parser(`#a''`), ctx), [[`<a class="hashtag" href="/hashtags/a">#a</a>`], `''`]);
       assert.deepStrictEqual(inspect(parser(`#a'b`), ctx), [[`<a class="hashtag" href="/hashtags/a'b">#a'b</a>`], '']);

package/src/parser/inline/autolink/hashtag.ts

@@ -9,19 +9,19 @@ import { define } from 'typed-dom/dom';
 // https://example/hashtags/a must be a hashtag page or a redirect page going there.
 
 // https://github.com/tc39/proposal-regexp-unicode-property-escapes#matching-emoji
-export const emoji = String.raw`\p{Emoji_Modifier_Base}\p{Emoji_Modifier}?|\p{Emoji_Presentation}|\p{Emoji}\uFE0F`;
+export const emoji = /\p{Emoji_Modifier_Base}\p{Emoji_Modifier}?|\p{Emoji_Presentation}|\p{Emoji}\uFE0F|\u200D/u;
 
 export const hashtag: AutolinkParser.HashtagParser = lazy(() => rewrite(
   verify(surround(
     new RegExp([
-      /(?<![^\p{C}\p{S}\p{P}\s]|emoji)#/yiu.source,
-    ].join('').replace(/emoji/g, emoji), 'yu'),
+      /(?<![^\p{C}\p{S}\p{P}\s]|emoji)#/yu.source,
+    ].join('|').replace(/emoji/g, emoji.source), 'yu'),
     str(new RegExp([
       /(?!['_])(?:[^\p{C}\p{S}\p{P}\s]|emoji|'(?=[0-9A-Za-z])|_(?=[^\p{C}\p{S}\p{P}\s]|emoji))+/yu.source,
-    ].join('').replace(/emoji/g, emoji), 'yu')),
-    str(new RegExp([
+    ].join('|').replace(/emoji/g, emoji.source), 'yu')),
+    new RegExp([
       /(?![0-9a-z@#]|>>|:\S|[^\p{C}\p{S}\p{P}\s]|emoji)/yu.source,
-    ].join('').replace(/emoji/g, emoji), 'yu')),
+    ].join('|').replace(/emoji/g, emoji.source), 'yu'),
     false, undefined, undefined,
     [3 | Backtrack.autolink]),
     ([{ value }]) => !/^[0-9]{1,4}$|^[0-9]{5}/.test(value)),

package/src/parser/inline/autolink.ts

@@ -1,35 +1,47 @@
 import { AutolinkParser } from '../inline';
 import { State } from '../context';
-import { union, state, validate, lazy } from '../../combinator';
+import { state, lazy } from '../../combinator';
 import { url, lineurl } from './autolink/url';
 import { email } from './autolink/email';
 import { channel } from './autolink/channel';
 import { account } from './autolink/account';
-import { hashtag, emoji } from './autolink/hashtag';
+import { hashtag } from './autolink/hashtag';
 import { hashnum } from './autolink/hashnum';
 import { anchor } from './autolink/anchor';
+import { isAlphanumeric } from '../source/text';
 
 export const autolink: AutolinkParser = lazy(() =>
-  validate(new RegExp([
-    /(?<![0-9a-z])@/yi.source,
-    /(?<![^\p{C}\p{S}\p{P}\s]|emoji)#/yiu.source,
-    /(?<![0-9a-z])>>/yi.source,
-    /(?<![0-9a-z][.+-]?|[@#])!?[0-9a-z]/yi.source,
-  ].join('|').replace(/emoji/g, emoji), 'yiu'),
   state(~State.autolink,
-  union([
-    lineurl,
-    url,
-    email,
-    // Escape unmatched email-like strings.
-    //str(/[0-9a-z]+(?:[_.+-][0-9a-z]+[:@]?|:|@(?=@))*/yi),
-    channel,
-    account,
-    // Escape unmatched account-like strings.
-    //str(/@+(?:[0-9a-z]+(?:[_.+-][0-9a-z]+)*)?/yi),
-    hashtag,
-    hashnum,
-    // Escape unmatched hashtag-like strings.
-    //str(new RegExp(/#+(?:(?:[^\p{C}\p{S}\p{P}\s]|emoji)+(?:['_.+-](?:[^\p{C}\p{S}\p{P}\s]|emoji)+)*)?/yu.source.replace(/emoji/g, emoji), 'yu')),
-    anchor,
-  ]))));
+  input => {
+    const { context: { source, position } } = input;
+    if (position === source.length) return;
+    const fst = source[position];
+    switch (fst) {
+      case '@':
+        return channel(input) || account(input);
+      case '#':
+        return hashtag(input) || hashnum(input);
+      case '>':
+        return anchor(input);
+      case '!':
+        if (!source.startsWith('http', position + 1)) break;
+        if (position === 0) return lineurl(input);
+        switch (source[position - 1]) {
+          case '\r':
+          case '\n':
+            return lineurl(input);
+        }
+        break;
+      case 'h':
+        if (!source.startsWith('http', position)) return;
+        if (position === 0) return lineurl(input) || url(input) || email(input);
+        switch (source[position - 1]) {
+          case '\r':
+          case '\n':
+            return lineurl(input) || url(input) || email(input);
+        }
+        return url(input) || email(input);
+      default:
+        if (isAlphanumeric(fst)) return email(input);
+    }
+  }));

package/src/parser/inline/html.ts

@@ -23,7 +23,7 @@ export const html: HTMLParser = lazy(() => validate(/<[a-z]+(?=[ >])/yi,
   union([
     surround(
       // https://html.spec.whatwg.org/multipage/syntax.html#void-elements
-      str(/<(?:area|base|br|col|embed|hr|img|input|link|meta|source|track|wbr)(?=[ >])/yi),
+      str(/<(?:area|base|br|col|embed|hr|img|input|link|meta|source|track|wbr)(?=[ >])/y),
       some(union([attribute])),
       open(str(/ ?/y), str('>'), true),
       true,

package/src/parser/inline/htmlentity.ts

@@ -1,19 +1,19 @@
 import { HTMLEntityParser, UnsafeHTMLEntityParser } from '../inline';
+import { Backtrack } from '../context';
 import { List, Data } from '../../combinator/data/parser';
-import { union, focus, fmap } from '../../combinator';
+import { union, surround, fmap } from '../../combinator';
+import { str } from '../source';
 import { invalid } from '../util';
 import { html } from 'typed-dom/dom';
 
-export const unsafehtmlentity: UnsafeHTMLEntityParser = focus(
-  /&(?:[0-9A-Za-z]+;?)?/y,
-  //({ source }) => [[parser(source) ?? `${Command.Error}${source}`], '']));
-  ({ context }) => {
-    const { source } = context;
-    context.position += source.length;
-    return source.length > 1 && source.at(-1) === ';'
-      ? new List([new Data(parser(source) ?? source)])
-      : new List([new Data(source)]);
-  });
+export const unsafehtmlentity: UnsafeHTMLEntityParser = surround(
+  str('&'), str(/[0-9A-Za-z]+/y), str(';'),
+  false,
+  ([as, bs, cs]) =>
+    new List([new Data(parser(as.head!.value + bs.head!.value + cs.head!.value))]),
+  ([as, bs]) =>
+    new List([new Data(as.head!.value + (bs?.head?.value ?? ''))]),
+  [3 | Backtrack.bracket]);
 
 export const htmlentity: HTMLEntityParser = fmap(
   union([unsafehtmlentity]),
@@ -23,14 +23,11 @@ export const htmlentity: HTMLEntityParser = fmap(
       : new Data(html('span', {
           class: 'invalid',
           ...invalid('htmlentity', 'syntax', 'Invalid HTML entity'),
-      }, value))
+        }, value))
   ]));
 
-const parser = (el => (entity: string): string | undefined => {
+const parser = (el => (entity: string): string => {
   if (entity === '
') return ' ';
   el.innerHTML = entity;
-  const text = el.textContent!;
-  return entity === text
-    ? undefined
-    : text;
+  return el.textContent!;
 })(html('span'));

package/src/parser/inline/ruby.ts

@@ -63,6 +63,8 @@ export const ruby: RubyParser = lazy(() => bind(
     }
   }));
 
+const delimiter = /[$"`\[\](){}<>（）［］｛｝]|\\?\n/y;
+
 const text: RubyParser.TextParser = input => {
   const { context } = input;
   const { source } = context;
@@ -70,11 +72,14 @@ const text: RubyParser.TextParser = input => {
   let state = false;
   context.sequential = true;
   for (let { position } = context; position < source.length; position = context.position) {
-    if (/[$"`\[\](){}<>（）［］｛｝]|\\?\n/yi.test(source.slice(position, position + 2))) break;
+    delimiter.lastIndex = position;
+    if (delimiter.test(source)) break;
     assert(source[position] !== '\n');
     switch (source[position]) {
       case '&': {
-        const result = unsafehtmlentity(input) ?? txt(input)!;
+        const result = source[position + 1] !== ' '
+          ? unsafehtmlentity(input) ?? txt(input)!
+          : txt(input)!;
         assert(result);
         acc.last!.value += result.head!.value;
         continue;

package/src/parser/inline.ts

@@ -5,6 +5,7 @@ import { reference } from './inline/reference';
 import { template } from './inline/template';
 import { remark } from './inline/remark';
 import { extension } from './inline/extension';
+import { label } from './inline/extension/label';
 import { textlink } from './inline/link';
 import { ruby } from './inline/ruby';
 import { html } from './inline/html';
@@ -94,7 +95,7 @@ export const inline: InlineParser = lazy(() => union([
         return html(input);
       case '$':
         if (source[position + 1] === '{') return math(input);
-        return extension(input)
+        return label(input)
             || math(input);
       case '+':
         if (source[position + 1] === '+') return insertion(input);

package/src/parser/source/text.ts

@@ -185,12 +185,12 @@ export function backToEmailHead(source: string, position: number, index: number)
     ? delim
     : index;
 }
-function isAlphanumeric(char: string): boolean {
+export function isAlphanumeric(char: string): boolean {
   assert(char.length === 1);
   if (char < '0' || '\x7F' < char) return false;
   return '0' <= char && char <= '9'
-      || 'a' <= char && char <= 'z'
-      || 'A' <= char && char <= 'Z';
+      || 'A' <= char && char <= 'Z'
+      || 'a' <= char && char <= 'z';
 }
 
 //const dict = new class {
@@ -238,7 +238,6 @@ function seek(source: string, position: number): number {
       case '@':
       case '#':
       case '$':
-      case '&':
       case '"':
       case '`':
       case '[':
@@ -274,6 +273,9 @@ function seek(source: string, position: number): number {
       case ':':
         if (source[i + 1] === '/' && source[i + 2] === '/') return i;
         continue;
+      case '&':
+        if (source[i + 1] !== ' ') return i;
+        continue;
       case ' ':
       case '\t':
       case '　':