Somewhat better index checking in the tokenizer

src/Sapl/SaplTokenizer.icl

@@ -18,11 +18,16 @@ not_eol c = not (c == '\n')
 
 find_first_string :: !String !Int (Char String Int -> Bool) -> Int
 find_first_string line start f
-	| start == size line
-		= size line
-	| f line.[start-1] line start
-		= find_first_string line (start + 1) f 
-		= start
+    # lineSz = size line
+	| start == lineSz
+		= lineSz
+    # newPos = start - 1
+    | newPos >= lineSz
+        = abort "find_first_string newPos >= lineSz"
+	| f line.[newPos] line start
+		= find_first_string line (start + 1) f
+	| otherwise
+    	= start
 
 read_string_lit :: !Char !Int !String -> (!Int, !UString)
 read_string_lit qc start str 
@@ -30,27 +35,29 @@ read_string_lit qc start str
 		= (nextbase, reverse l)
 where
 	decode p cs
-		| eof p			 	= (p, cs)	
-		| str.[p] == qc 	= (p+1, cs)
-		| str.[p] == '\\' 	= let (n,c) = decodeBSChar (p+1) in decode n [c:cs]
-							= decode (p+1) [fromChar str.[p]:cs]
+		| eof p			= (p, cs)
+        # strP = str.[p]
+		| strP == qc 	= (p+1, cs)
+		| strP == '\\' 	= let (n,c) = decodeBSChar (p+1) in decode n [c:cs]
+		| otherwise		= decode (p+1) [fromChar strP : cs]
 	
 	decodeBSChar p
-		| eof p 			= (p  ,fromChar '\\')
-		| str.[p] == '0'	= (p+1,fromInt 0)
-		| str.[p] == 'a'	= (p+1,fromInt 7)
-		| str.[p] == 'b'	= (p+1,fromChar '\b')
-		| str.[p] == 'f'	= (p+1,fromChar '\f')		
-		| str.[p] == 'n'	= (p+1,fromChar '\n')
-		| str.[p] == 'r'	= (p+1,fromChar '\r')		
-		| str.[p] == 't'	= (p+1,fromChar '\t')
-		| str.[p] == 'v'	= (p+1,fromChar '\v')
-		| str.[p] == '\''	= (p+1,fromChar '\'')
-		| str.[p] == '"'	= (p+1,fromChar '"')
-		| str.[p] == 'x'	= decodeHex 2 2 0
-		| str.[p] == 'u'	= decodeHex 4 4 0					
-		| str.[p] == 'U'	= decodeHex 8 8 0
-							= (p+1,fromChar str.[p])	// skip the backslash otherwise
+		| eof p 		= (p  ,fromChar '\\')
+        # strP = str.[p]
+		| strP == '0'	= (p+1,fromInt 0)
+		| strP == 'a'	= (p+1,fromInt 7)
+		| strP == 'b'	= (p+1,fromChar '\b')
+		| strP == 'f'	= (p+1,fromChar '\f')
+		| strP == 'n'	= (p+1,fromChar '\n')
+		| strP == 'r'	= (p+1,fromChar '\r')
+		| strP == 't'	= (p+1,fromChar '\t')
+		| strP == 'v'	= (p+1,fromChar '\v')
+		| strP == '\''	= (p+1,fromChar '\'')
+		| strP == '"'	= (p+1,fromChar '"')
+		| strP == 'x'	= decodeHex 2 2 0
+		| strP == 'u'	= decodeHex 4 4 0
+		| strP == 'U'	= decodeHex 8 8 0
+		| otherwise		= (p+1,fromChar strP) // skip the backslash otherwise
 	where
 		decodeHex :: !Int !Int !Int -> (!Int, !UChar)
 		decodeHex len s acc
@@ -140,12 +147,13 @@ read_token base line
 			"select" = return (TSelectKeyword, stop)
 			"let"	 = return (TLetKeyword, stop)
 			"in"	 = return (TInKeyword, stop)
-			str		 = if (str.[0] == '!') 
-							(return (TStrictIdentifier (str % (1, size str)), stop))
-							(return (TIdentifier str, stop))
+			str
+              # strSz = size str
+              | strSz > 0 && str.[0] == '!' = return (TStrictIdentifier (str % (1, strSz)), stop)
+              | otherwise                   = return (TIdentifier str, stop)
 where
 	tstr stop = line % (start, stop - 1)
-	start = (skipChars line base is_space)
+	start = skipChars line base is_space
 	
 	rnoarg t length = (start, start + length, t)	
 	return (a, newbase) = (start, newbase, a)