Ignore underscores in the middle of words.

lib/src/inline_parser.dart

Index: lib/src/inline_parser.dart
diff --git a/lib/src/inline_parser.dart b/lib/src/inline_parser.dart
index 3b1d9faa327e37eda63a04ed6e5ffc8c713de919..9f0db08c5d04e1748a96dce4f9a0d52a58a4683f 100644
--- a/lib/src/inline_parser.dart
+++ b/lib/src/inline_parser.dart
@@ -43,13 +43,11 @@ class InlineParser {
     // Parse "**strong**" tags.
     new TagSyntax(r'\*\*', tag: 'strong'),
     // Parse "__strong__" tags.
-    new TagSyntax(r'__', tag: 'strong'),
+    new TagSyntax(r'\b__', tag: 'strong', end: r'__\b'),

[nweiz, 2015/09/14 21:52:13]

Using \b here isn't very non-English-friendly; it'll only take ASCII
alphanumeric characters into account
(http://www.ecma-international.org/ecma-262/5.1/#sec-15.10.2.6), so something
like á_ḇ_č will still get formatted wrong.

[Bob Nystrom, 2015/09/15 00:19:16]

Hmm, good point. I think this is sort of a stopgap solution until a more
comprehensive implementation that follows the full CommonMark spec can be done.

     // Parse "*emphasis*" tags.
     new TagSyntax(r'\*', tag: 'em'),
     // Parse "_emphasis_" tags.
-    // TODO(rnystrom): Underscores in the middle of a word should not be
-    // parsed as emphasis like_in_this.
-    new TagSyntax(r'_', tag: 'em'),
+    new TagSyntax(r'\b_', tag: 'em', end: r'_\b'),
     new CodeSyntax(),
     // We will add the LinkSyntax once we know about the specific link resolver.
   ];