Switch from RubyArray addAll to concat method to preserve the encoding and avoid implicit deconding in addAll iterator

Author: andsel

logstash-core/src/main/java/org/logstash/common/BufferedTokenizerExt.java

@@ -85,22 +85,22 @@ public RubyArray extract(final ThreadContext context, IRubyObject data) {
         final RubyArray entities = data.convertToString().split(delimiter, -1);
         if (!bufferFullErrorNotified) {
             input.clear();
-            input.addAll(entities);
+            input.concat(entities);
         } else {
             // after a full buffer signal
             if (input.isEmpty()) {
                 // after a buffer full error, the remaining part of the line, till next delimiter,
                 // has to be consumed, unless the input buffer doesn't still contain fragments of
                 // subsequent tokens.
                 entities.shift(context);
-                input.addAll(entities);
+                input.concat(entities);
             } else {
                 // merge last of the input with first of incoming data segment
                 if (!entities.isEmpty()) {
                     RubyString last = ((RubyString) input.pop(context));
                     RubyString nextFirst = ((RubyString) entities.shift(context));
                     entities.unshift(last.concat(nextFirst));
-                    input.addAll(entities);
+                    input.concat(entities);
                 }
             }
         }

logstash-core/src/test/java/org/logstash/common/BufferedTokenizerExtTest.java

@@ -102,6 +102,6 @@ public void shouldNotChangeEncodingOfTokensAfterPartitioning() {
 
         // verify encoding "ISO8859-1" is preserved in the Java to Ruby String conversion
         RubyEncoding encoding = (RubyEncoding) firstToken.callMethod(context, "encoding");
-        assertEquals("ISO8859-1", encoding.toString());
+        assertEquals("ISO-8859-1", encoding.toString());
     }
 }