Merge pull request #83 from Dinh-Hung-Tu/82

Issue #82 solution

Author: kirsle

rivescript/parser.py

@@ -535,12 +535,12 @@ def check_syntax(self, cmd, line):
             if parts[0] == "begin" and len(parts) > 1:
                 return "The 'begin' label takes no additional arguments, should be verbatim '> begin'"
             elif parts[0] == "topic":
-                match = re.match(RE.name_syntax, line)
-                if match:
+                search = re.search(RE.name_syntax, line)
+                if search:
                     return "Topics should be lowercased and contain only numbers and letters"
             elif parts[0] == "object":
-                match = re.match(RE.name_syntax, line)
-                if match:
+                search = re.search(RE.obj_syntax, line) # Upper case is allowed
+                if search:
                     return "Objects can only contain numbers and letters"
         elif cmd == '+' or cmd == '%' or cmd == '@':
             # + Trigger, % Previous, @ Redirect
@@ -585,12 +585,12 @@ def check_syntax(self, cmd, line):
 
             # In UTF-8 mode, most symbols are allowed.
             if self.utf8:
-                match = re.match(RE.utf8_trig, line)
-                if match:
+                search = re.search(RE.utf8_trig, line)
+                if search:
                     return "Triggers can't contain uppercase letters, backslashes or dots in UTF-8 mode."
             else:
-                match = re.match(RE.trig_syntax, line)
-                if match:
+                search = re.search(RE.trig_syntax, line)
+                if search:
                     return "Triggers may only contain lowercase letters, numbers, and these symbols: ( | ) [ ] * _ # @ { } < > ="
         elif cmd == '-' or cmd == '^' or cmd == '/':
             # - Trigger, ^ Continue, / Comment

rivescript/regexp.py

@@ -24,6 +24,7 @@ class RE(object):
     array       = re.compile(r'\@(.+?)\b')
     def_syntax  = re.compile(r'^.+(?:\s+.+|)\s*=\s*.+?$')
     name_syntax = re.compile(r'[^a-z0-9_\-\s]')
+    obj_syntax  = re.compile(r'[^A-Za-z0-9_\-\s]')
     utf8_trig   = re.compile(r'[A-Z\\.]')
     trig_syntax = re.compile(r'[^a-z0-9(\|)\[\]*_#@{}<>=\s]')
     cond_syntax = re.compile(r'^.+?\s*(?:==|eq|!=|ne|<>|<|<=|>|>=)\s*.+?=>.+?$')

tests/test_format_message.py

@@ -28,4 +28,32 @@ def test_format_triggers(self):
 
         """)
         self.reply("hi there", "hi there")
-        self.reply("hi  here", "hi here")
+        self.reply("hi  here", "hi here")
+
+    def test_invalid_character_raise_exception(self):
+        self.assertRaises(Exception, self.new, """
+            + $hello
+            - hi
+        """)  # This test passes with `match`, which only check at the beginning
+        self.assertRaises(Exception, self.new, """
+            + hello$
+            - hi
+        """)   # This test does not pass because the beginning is good, no $
+        self.assertRaises(Exception, self.new, """
+            > topic Greetings
+                + hello
+                - hi
+            <topics
+        """)
+        self.assertRaises(Exception, self.new, """
+            > object hash %perl
+                my ($rs, $args) = @_;
+                my $method = shift @{$args};
+            <object
+        """)  # Test for character violation in object, no %
+        self.new("""
+            > object hash Perl
+                my ($rs, $args) = @_;
+                my $method = shift @{$args};
+            <object
+        """) # No exception raised for uppercase character in object