Disallow Unicode literals in character classes when Unicode is disabled.

When Unicode mode is disabled, we also disable the use of Unicode literals
in the regular expression, since it can lead to unintuitive behavior. In
this case, Unicode literals in character classes were not disallowed, and
subsequent code filtered them out, which resulted in an empty character
class. The compiler assumes that empty character classes are not allowed,
and so this causes an assert to trigger.

Fixes #250.

regex-syntax/src/parser.rs

@@ -610,6 +610,11 @@ impl Parser {
     fn parse_class_range(&mut self, class: &mut CharClass, start: char)
                         -> Result<()> {
         if !self.bump_if('-') {
+            // Make sure we haven't parsed Unicode literals when we shouldn't have.
+            if !self.flags.unicode {
+                let _ = try!(self.codepoint_to_one_byte(start));
+            }
+
             // Not a range, so just push a singleton range.
             class.ranges.push(ClassRange::one(start));
             return Ok(());
@@ -651,6 +656,11 @@ impl Parser {
                 end: end,
             }));
         }
+        // Make sure we haven't parsed Unicode literals when we shouldn't have.
+        if !self.flags.unicode {
+            let _ = try!(self.codepoint_to_one_byte(start));
+            let _ = try!(self.codepoint_to_one_byte(end));
+        }
         class.ranges.push(ClassRange::new(start, end));
         Ok(())
     }
@@ -2405,6 +2415,13 @@ mod tests {
         test_err!(r"☃(?-u:\pL)", 9, ErrorKind::UnicodeNotAllowed, flags);
     }
 
+    #[test]
+    fn unicode_class_literal_not_allowed() {
+        let flags = Flags { allow_bytes: true, .. Flags::default() };
+        test_err!(r"(?-u)[☃]", 7, ErrorKind::UnicodeNotAllowed, flags);
+        test_err!(r"(?-u)[☃-☃]", 9, ErrorKind::UnicodeNotAllowed, flags);
+    }
+
     #[test]
     fn unicode_hex_not_allowed() {
         let flags = Flags { allow_bytes: true, .. Flags::default() };