Fixed punctuation check in emphasis

lib/common/utils.js

@@ -41,7 +41,7 @@ function arrayReplaceAt(src, pos, newElements) {
 
 ////////////////////////////////////////////////////////////////////////////////
 
-var UNESCAPE_MD_RE = /\\([\\!"#$%&'()*+,.\/:;<=>?@[\]^_`{|}~-])/g;
+var UNESCAPE_MD_RE = /\\([!"#$%&'()*+,\-.\/:;<=>?@[\\\]^_`{|}~])/g;
 
 function unescapeMd(str) {
   if (str.indexOf('\\') < 0) { return str; }
@@ -223,6 +223,54 @@ function isPunctChar(char) {
   return BMP_PUNCT_RE.test(char);
 }
 
+
+// Markdown ASCII punctuation characters.
+//
+// !, ", #, $, %, &, ', (, ), *, +, ,, -, ., /, :, ;, <, =, >, ?, @, [, \, ], ^, _, `, {, |, }, or ~
+// http://spec.commonmark.org/0.15/#ascii-punctuation-character
+//
+// Don't confuse with unicode punctuation !!! It lacks some chars in ascii range.
+//
+function isMdAsciiPunct(ch) {
+  switch (ch) {
+    case 0x21/* ! */:
+    case 0x22/* " */:
+    case 0x23/* # */:
+    case 0x24/* $ */:
+    case 0x25/* % */:
+    case 0x26/* & */:
+    case 0x27/* ' */:
+    case 0x28/* ( */:
+    case 0x29/* ) */:
+    case 0x2A/* * */:
+    case 0x2B/* + */:
+    case 0x2C/* , */:
+    case 0x2D/* - */:
+    case 0x2E/* . */:
+    case 0x2F/* / */:
+    case 0x3A/* : */:
+    case 0x3B/* ; */:
+    case 0x3C/* < */:
+    case 0x3D/* = */:
+    case 0x3E/* > */:
+    case 0x3F/* ? */:
+    case 0x40/* @ */:
+    case 0x5B/* [ */:
+    case 0x5C/* \ */:
+    case 0x5D/* ] */:
+    case 0x5E/* ^ */:
+    case 0x5F/* _ */:
+    case 0x60/* ` */:
+    case 0x7B/* { */:
+    case 0x7C/* | */:
+    case 0x7D/* } */:
+    case 0x7E/* ~ */:
+      return true;
+    default:
+      return false;
+  }
+}
+
 ////////////////////////////////////////////////////////////////////////////////
 
 exports.assign            = assign;
@@ -236,5 +284,6 @@ exports.escapeHtml        = escapeHtml;
 exports.arrayReplaceAt    = arrayReplaceAt;
 exports.normalizeLink     = normalizeLink;
 exports.isWhiteSpace      = isWhiteSpace;
+exports.isMdAsciiPunct    = isMdAsciiPunct;
 exports.isPunctChar       = isPunctChar;
 exports.escapeRE          = escapeRE;

lib/rules_inline/emphasis.js

@@ -4,6 +4,7 @@
 
 var isWhiteSpace   = require('../common/utils').isWhiteSpace;
 var isPunctChar    = require('../common/utils').isPunctChar;
+var isMdAsciiPunct = require('../common/utils').isMdAsciiPunct;
 
 
 function isAlphaNum(code) {
@@ -31,8 +32,10 @@ function scanDelims(state, start) {
 
   nextChar = pos < max ? state.src.charCodeAt(pos) : -1;
 
-  isLastPunctChar = lastChar >= 0 && isPunctChar(String.fromCharCode(lastChar));
-  isNextPunctChar = nextChar >= 0 && isPunctChar(String.fromCharCode(nextChar));
+  isLastPunctChar = lastChar >= 0 &&
+    (isMdAsciiPunct(lastChar) || isPunctChar(String.fromCharCode(lastChar)));
+  isNextPunctChar = nextChar >= 0 &&
+    (isMdAsciiPunct(nextChar) || isPunctChar(String.fromCharCode(nextChar)));
   isLastWhiteSpace = lastChar >= 0 && isWhiteSpace(lastChar);
   isNextWhiteSpace = nextChar >= 0 && isWhiteSpace(nextChar);
 

lib/rules_inline/text.js

@@ -7,30 +7,35 @@
 // Rule to skip pure text
 // '{}$%@~+=:' reserved for extentions
 
+// !, ", #, $, %, &, ', (, ), *, +, ,, -, ., /, :, ;, <, =, >, ?, @, [, \, ], ^, _, `, {, |, }, or ~
+
+// !!!! Don't confuse with "Markdown ASCII Punctuation" chars
+// http://spec.commonmark.org/0.15/#ascii-punctuation-character
 function isTerminatorChar(ch) {
   switch (ch) {
     case 0x0A/* \n */:
-    case 0x5C/* \ */:
-    case 0x60/* ` */:
-    case 0x2A/* * */:
-    case 0x5F/* _ */:
-    case 0x5E/* ^ */:
-    case 0x5B/* [ */:
-    case 0x5D/* ] */:
     case 0x21/* ! */:
     case 0x23/* # */:
+    case 0x24/* $ */:
+    case 0x25/* % */:
     case 0x26/* & */:
+    case 0x2A/* * */:
+    case 0x2B/* + */:
+    case 0x2D/* - */:
+    case 0x3A/* : */:
     case 0x3C/* < */:
+    case 0x3D/* = */:
     case 0x3E/* > */:
+    case 0x40/* @ */:
+    case 0x5B/* [ */:
+    case 0x5C/* \ */:
+    case 0x5D/* ] */:
+    case 0x5E/* ^ */:
+    case 0x5F/* _ */:
+    case 0x60/* ` */:
     case 0x7B/* { */:
     case 0x7D/* } */:
-    case 0x24/* $ */:
-    case 0x25/* % */:
-    case 0x40/* @ */:
     case 0x7E/* ~ */:
-    case 0x2B/* + */:
-    case 0x3D/* = */:
-    case 0x3A/* : */:
       return true;
     default:
       return false;

test/fixtures/markdown-it/commonmark_extras.txt

@@ -52,6 +52,14 @@ Issue #55:
 .
 
 
+Issue #35. `<` shoud work as punctuation
+.
+an **(:**<br>
+.
+<p>an <strong>(:</strong><br></p>
+.
+
+
 Should unescape only needed things in link destinations/titles:
 .
 [test](<\f\o\o\>\\>)