diff --git a/example b/example
index 9a10fc8..41968b7 100644
--- a/example
+++ b/example
@@ -1,7 +1,7 @@
 takentaal v0.1.0
 #  {5000} Full work plan
 
-This is the description of the entire work plan.
+This is the (draft) description of the entire work plan.
 
 ## {1000} First task
 
@@ -12,9 +12,9 @@ This description has # ' " [] symbols and ü © Ð Latin-1 chars.
 -   {500} First subtask
 /   {500} Second subtask
 
-## Second task {1000}
+## {1000} Second task
 
 This is the description of the second task.
 
 *   {500} First subtask
-- {500} Second subtask
+- {500} \2nd subtask
diff --git a/takentaal.g4 b/takentaal.g4
index 5679125..f8c8d83 100644
--- a/takentaal.g4
+++ b/takentaal.g4
@@ -1,3 +1,17 @@
+/**
+ * This file defines the grammar for takentaal.
+ * It is divided into parser rules (lowercase) and lexer rules (uppercase).
+ * The parser splits an input into tokens accoring to the lexer rules.
+ * At any point, all lexer rules are considered. If multiple rules match,
+ * a lexer rule is chosen as follows:
+ *  - the rule that matches the longest input is chosen
+ *  - any implicit rule, e.g. 'a', is chosen
+ *  - the first defined rule is chosen.
+ * Since this grammar has to match unquoted texts and text are usually longer
+ * than other token matches, the TEXT rule disallows many characters as the
+ * first character to start with.
+ */
+
 grammar takentaal;
 
 takentaal
@@ -73,14 +87,14 @@ WS
     ;
 
 EOL
-    : '\n'+
+    : ' '* '\n'+
     ;
 
 INT
     : DIGIT+
     ;
 
-DIGIT
+fragment DIGIT
     : [0-9]
     ;
 
@@ -92,14 +106,22 @@ END_AMOUNT
     : '}'
     ;
 
-STARTCHAR
-    : [!-"$-/:-\u007A\u007C\u007E]
+// all special characters, including ' ' and digits are subtracted from the printable character range
+// '!' '#' '-' '/' '*'
+fragment STARTCHAR
+    : ["$-)+-,.:-z|~\u00A0-\u33FF]
     ;
 
-CHAR
-    : [ -\u007E\u00A0-\u33FF] // ASCII and UNICODE
+// A text should not end with a space, so the ENDHAR omits the space
+fragment ENDCHAR
+    : ["-~\u00A0-\u33FF]
     ;
 
+fragment CHAR
+    : [ -~\u00A0-\u33FF] // ASCII and UNICODE
+    ;
+
+// A text cannot start with a special character or has to be placed in quotes
 TEXT
-    : STARTCHAR CHAR*
+    : STARTCHAR (CHAR* ENDCHAR)?
     ;