rewrite garbage char to scala, improve form validation

modules/common/src/main/Form.scala

@@ -66,17 +66,20 @@ object Form {
   def trim(m: Mapping[String]) = m.transform[String](_.trim, identity)
 
   // trims and removes garbage chars before validation
-  val cleanTextFormatter: Formatter[String] = new Formatter[String] {
+  private val cleanTextFormatter: Formatter[String] = new Formatter[String] {
     def bind(key: String, data: Map[String, String]) =
       data
         .get(key)
         .map(_.trim)
-        .map(StringUtils.removeGarbageChars)
+        .map(String.normalize)
         .toRight(Seq(FormError(key, "error.required", Nil)))
-    def unbind(key: String, value: String) = Map(key -> StringUtils.removeGarbageChars(value.trim))
+    def unbind(key: String, value: String) = Map(key -> String.normalize(value.trim))
   }
 
-  val cleanText: Mapping[String] = of(cleanTextFormatter)
+  val cleanText: Mapping[String] = of(cleanTextFormatter).verifying(
+    "The text contains invalid chars",
+    s => !String.hasGarbageChars(s)
+  )
   def cleanText(minLength: Int = 0, maxLength: Int = Int.MaxValue): Mapping[String] =
     (minLength, maxLength) match {
       case (min, Int.MaxValue) => cleanText.verifying(Constraints.minLength(min))

modules/common/src/main/String.scala

@@ -24,6 +24,21 @@ object String {
 
   def urlencode(str: String): String = java.net.URLEncoder.encode(str, "US-ASCII")
 
+  def hasGarbageChars(str: String) = str.chars().anyMatch(isGarbageChar)
+
+  def isGarbageChar(c: Int) =
+    // invisible chars https://www.compart.com/en/unicode/block/U+2000
+    (c >= '\u2000' && c <= '\u200F') ||
+      // weird stuff https://www.compart.com/en/unicode/block/U+2000
+      (c >= '\u2028' && c <= '\u202F') ||
+      // bunch of probably useless blocks https://www.compart.com/en/unicode/block/U+2100
+      (c >= '\u2100' && c <= '\u2C5F') ||
+      // decorative chars ꧁ ꧂
+      (c == '\ua9c1' || c == '\ua9c2')
+
+  // convert weird chars into letters when possible
+  def normalize(str: String): String = Normalizer.normalize(str, Normalizer.Form.NFKC)
+
   def decodeUriPath(input: String): Option[String] = {
     try {
       play.utils.UriEncoding.decodePath(input, "UTF-8").some

modules/common/src/main/java/StringUtils.java

@@ -78,26 +78,6 @@ public class StringUtils {
         sb.append(sArr, start, end - start);
     }
 
-    public static String removeGarbageChars(String s) {
-        s = Normalizer.normalize(s, Normalizer.Form.NFKC);
-        final char[] sArr = s.toCharArray();
-        final int size = sArr.length;
-        final StringBuilder sb = new StringBuilder(size);
-        for (int i = 0; i < size; i++) {
-            final char c = sArr[i];
-            // invisible chars https://www.compart.com/en/unicode/block/U+2000
-            if (c >= '\u2000' && c <= '\u200F') continue;
-            // weird stuff https://www.compart.com/en/unicode/block/U+2000
-            if (c >= '\u2028' && c <= '\u202F') continue;
-            // bunch of probably useless blocks https://www.compart.com/en/unicode/block/U+2100
-            if (c >= '\u2100' && c <= '\u2C5F') continue;
-            // decorative chars ꧁ ꧂
-            if (c == '\ua9c1' || c == '\ua9c2') continue;
-            sb.append(c);
-        }
-        return sb.toString();
-    }
-
     /**
      * https://github.com/tdebatty/java-string-similarity/blob/master/src/main/java/info/debatty/java/stringsimilarity/Levenshtein.java
      *