No longer allow reference labels that only contain whitespaces.

Knagis · Jun 25, 2015 · 7a8c040 · 7a8c040
1 parent 3286b03
commit 7a8c040
Show file tree

Hide file tree

Showing 4 changed files with 33 additions and 2 deletions.
diff --git a/CommonMark/Parser/InlineMethods.cs b/CommonMark/Parser/InlineMethods.cs
@@ -1171,6 +1171,9 @@ public static int ParseReference(Subject subj)
             if (lab == null || lab.Value.Length > Reference.MaximumReferenceLabelLength)
                 goto INVALID;
 
+            if (!Scanner.HasNonWhitespace(lab.Value))
+                goto INVALID;
+
             // colon:
             if (peek_char(subj) == ':')
                 subj.Position++;

diff --git a/CommonMark/Parser/Scanner.cs b/CommonMark/Parser/Scanner.cs
@@ -275,7 +275,7 @@ public static int scan_spacechars(string s, int pos, int sourceLength)
 
             for (var i = pos; i < sourceLength; i++)
             {
-                if (s[i] != ' ' && s[i] != '\n')
+                if (!Utilities.IsWhitespace(s[i]))
                     return i - pos;
             }
 
@@ -611,5 +611,25 @@ public static int scan_entity(string s, int pos, int length, out string namedEnt
 
             return 0;
         }
+
+        /// <summary>
+        /// Determines if the given string has non-whitespace characters in it
+        /// </summary>
+        public static bool HasNonWhitespace(Syntax.StringPart part)
+        {
+            var s = part.Source;
+            var i = part.StartIndex;
+            var l = i + part.Length;
+
+            while (i < l)
+            {
+                if (!Utilities.IsWhitespace(s[i]))
+                    return true;
+
+                i++;
+            }
+
+            return false;
+        }
     }
 }
diff --git a/CommonMark/Parser/ScannerCharacterMatcher.cs b/CommonMark/Parser/ScannerCharacterMatcher.cs
@@ -14,7 +14,7 @@ internal static class ScannerCharacterMatcher
         internal static bool MatchWhitespaces(string data, ref char currentCharacter, ref int currentPosition, int lastPosition)
         {
             var matched = false;
-            while ((currentCharacter == ' ' || currentCharacter == '\n') && currentPosition < lastPosition)
+            while (Utilities.IsWhitespace(currentCharacter) && currentPosition < lastPosition)
             {
                 currentCharacter = data[++currentPosition];
                 matched = true;

diff --git a/CommonMark/Utilities.cs b/CommonMark/Utilities.cs
@@ -41,6 +41,14 @@ public static bool IsAsciiLetterOrDigit(char c)
             return (c >= 'a' && c <= 'z') || (c >= '0' && c <= '9') || (c >= 'A' && c <= 'Z');
         }
 
+#if OptimizeFor45
+        [System.Runtime.CompilerServices.MethodImpl(System.Runtime.CompilerServices.MethodImplOptions.AggressiveInlining)]
+#endif
+        public static bool IsWhitespace(char c)
+        {
+            return c == ' ' || c == '\t' || c == '\r' || c == '\n' || c == '\f';
+        }
+
         /// <summary>
         /// Checks if the given character is an Unicode space or punctuation character.
         /// </summary>