test maappings

Author: ahorek

core/src/main/java/org/jruby/RubyString.java

@@ -282,6 +282,9 @@ final boolean singleByteOptimizable() {
     final boolean singleByteOptimizable(Encoding enc) {
         return StringSupport.isSingleByteOptimizable(this, enc);
     }
+    final boolean singleByteOptimizable(int flags, Encoding enc) {
+        return StringSupport.isSingleByteOptimizable(flags, this, enc);
+    }
 
     @SuppressWarnings("ReferenceEquality")
     final Encoding isCompatibleWith(EncodingCapable other) {
@@ -1926,8 +1929,7 @@ public IRubyObject upcase_bang(ThreadContext context, IRubyObject arg0, IRubyObj
     private IRubyObject upcase_bang(ThreadContext context, int flags) {
         modifyAndKeepCodeRange();
         Encoding enc = checkDummyEncoding();
-        if (((flags & Config.CASE_ASCII_ONLY) != 0 && (enc.isUTF8() || enc.maxLength() == 1)) ||
-                (flags & Config.CASE_FOLD_TURKISH_AZERI) == 0 && getCodeRange() == CR_7BIT) {
+        if (singleByteOptimizable(flags, enc)) {
             int s = value.getBegin();
             int end = s + value.getRealSize();
             byte[]bytes = value.getUnsafeBytes();
@@ -2000,8 +2002,7 @@ public IRubyObject downcase_bang(ThreadContext context, IRubyObject arg0, IRubyO
     private IRubyObject downcase_bang(ThreadContext context, int flags) {
         modifyAndKeepCodeRange();
         Encoding enc = checkDummyEncoding();
-        if (((flags & Config.CASE_ASCII_ONLY) != 0 && (enc.isUTF8() || enc.maxLength() == 1)) ||
-                (flags & Config.CASE_FOLD_TURKISH_AZERI) == 0 && getCodeRange() == CR_7BIT) {
+        if (singleByteOptimizable(flags, enc)) {
             int s = value.getBegin();
             int end = s + value.getRealSize();
             byte[]bytes = value.getUnsafeBytes();

core/src/main/java/org/jruby/util/StringSupport.java

@@ -53,6 +53,7 @@
 import org.jruby.runtime.Helpers;
 import org.jruby.runtime.ThreadContext;
 import org.jruby.runtime.builtin.IRubyObject;
+import org.jruby.util.StringSupport.TrTables;
 import org.jruby.util.collections.IntHashMap;
 import org.jruby.util.io.EncodingUtils;
 import sun.misc.Unsafe;
@@ -1700,7 +1701,30 @@ private static NeighborChar predChar(Encoding enc, byte[]bytes, int p, int len)
     }
 
     public static boolean isSingleByteOptimizable(CodeRangeable string, Encoding encoding) {
-        return string.getCodeRange() == CR_7BIT || encoding.maxLength() == 1;
+        if (encoding instanceof ASCIIEncoding || encoding instanceof USASCIIEncoding) {
+            return true;
+        }
+        else if (encoding.isUTF8()) {
+            // For UTF-8 it's worth scanning the string coderange when unknown.
+            return string.scanForCodeRange() == CR_7BIT;
+        }
+        /* Conservative.  It may be ENC_CODERANGE_UNKNOWN. */
+        if (string.getCodeRange() == CR_7BIT) {
+            return true;
+        }
+    
+        if (encoding.maxLength() == 1) {
+            return true;
+        }
+    
+        /* Conservative.  Possibly single byte.
+         * "\xa1" in Shift_JIS for example. */
+        return false;
+    }
+
+    public static boolean isSingleByteOptimizable(int flags, CodeRangeable string, Encoding encoding) {
+        return (((flags & Config.CASE_ASCII_ONLY) != 0 && (encoding.isUTF8() || encoding.maxLength() == 1)) ||
+                !((flags & Config.CASE_FOLD_TURKISH_AZERI) == 0 && string.getCodeRange() == CR_7BIT));
     }
 
     /**
@@ -1852,7 +1876,16 @@ public static boolean isAsciiOnly(CodeRangeable string) {
     }
 
     private static boolean isAsciiOnly(Encoding encoding, final int codeRange) {
-        return encoding.isAsciiCompatible() && codeRange == CR_7BIT;
+        switch(codeRange) {
+        case CR_UNKNOWN:
+          return codeRange == CR_7BIT && encoding.isAsciiCompatible();
+        case CR_7BIT:
+          return true;
+        default:
+          return false;
+        }
+
+        // return encoding.isAsciiCompatible() && codeRange == CR_7BIT;
     }
 
     /**