diff --git a/titan-core/src/main/java/com/thinkaurelius/titan/core/attribute/Text.java b/titan-core/src/main/java/com/thinkaurelius/titan/core/attribute/Text.java index 4b247d84e6..04779151ef 100644 --- a/titan-core/src/main/java/com/thinkaurelius/titan/core/attribute/Text.java +++ b/titan-core/src/main/java/com/thinkaurelius/titan/core/attribute/Text.java @@ -166,12 +166,12 @@ public static List tokenize(String str) { ArrayList tokens = new ArrayList(); int previous = 0; for (int p = 0; p < str.length(); p++) { - if (!Character.isLetterOrDigit(str.charAt(p))) { + if (!Character.isLetterOrDigit(str.codePointAt(p))) { if (p > previous + MIN_TOKEN_LENGTH) tokens.add(str.substring(previous, p)); previous = p + 1; } } - if (previous + MIN_TOKEN_LENGTH < str.length()) tokens.add(str.substring(previous, str.length())); + if (previous + MIN_TOKEN_LENGTH <= str.length()) tokens.add(str.substring(previous, str.length())); return tokens; }