Avoid charset provider lookups for UTF-8 QR encoding (#2077)

* Avoid charset provider lookups for UTF-8 QR encoding

* Simplify charset lookup handling

2 files changed, 39 insertions, 21 deletions

diff --git a/core/src/main/java/com/google/zxing/common/CharacterSetECI.java b/core/src/main/java/com/google/zxing/common/CharacterSetECI.java
index 6ea51f1e6..ae89e9bd1 100644
--- a/core/src/main/java/com/google/zxing/common/CharacterSetECI.java
+++ b/core/src/main/java/com/google/zxing/common/CharacterSetECI.java
@@ -19,7 +19,7 @@ package com.google.zxing.common;
 import com.google.zxing.FormatException;
 
 import java.nio.charset.Charset;
-
+import java.nio.charset.StandardCharsets;
 import java.util.HashMap;
 import java.util.Map;
 
@@ -33,7 +33,7 @@ public enum CharacterSetECI {
 
   // Enum name is a Java encoding valid for java.lang and java.io
   Cp437(new int[]{0,2}),
-  ISO8859_1(new int[]{1,3}, "ISO-8859-1"),
+  ISO8859_1(new int[]{1,3}, StandardCharsets.ISO_8859_1, "ISO-8859-1"),
   ISO8859_2(4, "ISO-8859-2"),
   ISO8859_3(5, "ISO-8859-3"),
   ISO8859_4(6, "ISO-8859-4"),
@@ -53,31 +53,29 @@ public enum CharacterSetECI {
   Cp1251(22, "windows-1251"),
   Cp1252(23, "windows-1252"),
   Cp1256(24, "windows-1256"),
-  UnicodeBigUnmarked(25, "UTF-16BE", "UnicodeBig"),
-  UTF8(26, "UTF-8"),
-  ASCII(new int[] {27, 170}, "US-ASCII"),
+  UnicodeBigUnmarked(25, StandardCharsets.UTF_16BE, "UTF-16BE", "UnicodeBig"),
+  UTF8(26, StandardCharsets.UTF_8, "UTF-8"),
+  ASCII(new int[] {27, 170}, StandardCharsets.US_ASCII, "US-ASCII"),
   Big5(28),
   GB18030(29, "GB2312", "EUC_CN", "GBK"),
   EUC_KR(30, "EUC-KR");
 
-  // only character sets supported by the current JVM are registered here
   private static final Map<Integer,CharacterSetECI> VALUE_TO_ECI = new HashMap<>();
   private static final Map<String,CharacterSetECI> NAME_TO_ECI = new HashMap<>();
   static {
     for (CharacterSetECI eci : values()) {
-      if (Charset.isSupported(eci.name())) {
-        for (int value : eci.values) {
-          VALUE_TO_ECI.put(value, eci);
-        }
-        NAME_TO_ECI.put(eci.name(), eci);
-        for (String name : eci.otherEncodingNames) {
-          NAME_TO_ECI.put(name, eci);
-        }
+      for (int value : eci.values) {
+        VALUE_TO_ECI.put(value, eci);
+      }
+      NAME_TO_ECI.put(eci.name(), eci);
+      for (String name : eci.otherEncodingNames) {
+        NAME_TO_ECI.put(name, eci);
       }
     }
   }
 
   private final int[] values;
+  private final Charset standardCharset;
   private final String[] otherEncodingNames;
 
   CharacterSetECI(int value) {
@@ -85,12 +83,20 @@ public enum CharacterSetECI {
   }
 
   CharacterSetECI(int value, String... otherEncodingNames) {
-    this.values = new int[] {value};
-    this.otherEncodingNames = otherEncodingNames;
+    this(new int[] {value}, null, otherEncodingNames);
+  }
+
+  CharacterSetECI(int value, Charset standardCharset, String... otherEncodingNames) {
+    this(new int[] {value}, standardCharset, otherEncodingNames);
   }
 
   CharacterSetECI(int[] values, String... otherEncodingNames) {
+    this(values, null, otherEncodingNames);
+  }
+
+  CharacterSetECI(int[] values, Charset standardCharset, String... otherEncodingNames) {
     this.values = values;
+    this.standardCharset = standardCharset;
     this.otherEncodingNames = otherEncodingNames;
   }
 
@@ -99,16 +105,24 @@ public enum CharacterSetECI {
   }
 
   public Charset getCharset() {
+    if (standardCharset != null) {
+      return standardCharset;
+    }
     return Charset.forName(name());
   }
 
+  private boolean isCharsetSupported() {
+    return standardCharset != null || Charset.isSupported(name());
+  }
+
   /**
    * @param charset Java character set object
    * @return CharacterSetECI representing ECI for character encoding, or null if it is legal
    *   but unsupported
    */
   public static CharacterSetECI getCharacterSetECI(Charset charset) {
-    return NAME_TO_ECI.get(charset.name());
+    CharacterSetECI eci = NAME_TO_ECI.get(charset.name());
+    return eci == null || eci.isCharsetSupported() ? eci : null;
   }
 
   /**
@@ -121,7 +135,8 @@ public enum CharacterSetECI {
     if (value < 0 || value >= 900) {
       throw FormatException.getFormatInstance();
     }
-    return VALUE_TO_ECI.get(value);
+    CharacterSetECI eci = VALUE_TO_ECI.get(value);
+    return eci == null || eci.isCharsetSupported() ? eci : null;
   }
 
   /**
@@ -130,7 +145,8 @@ public enum CharacterSetECI {
    *   but unsupported
    */
   public static CharacterSetECI getCharacterSetECIByName(String name) {
-    return NAME_TO_ECI.get(name);
+    CharacterSetECI eci = NAME_TO_ECI.get(name);
+    return eci == null || eci.isCharsetSupported() ? eci : null;
   }
 
 }
diff --git a/core/src/main/java/com/google/zxing/qrcode/encoder/Encoder.java b/core/src/main/java/com/google/zxing/qrcode/encoder/Encoder.java
index 97f316ddd..ad02a5829 100644
--- a/core/src/main/java/com/google/zxing/qrcode/encoder/Encoder.java
+++ b/core/src/main/java/com/google/zxing/qrcode/encoder/Encoder.java
@@ -19,8 +19,8 @@ package com.google.zxing.qrcode.encoder;
 import com.google.zxing.EncodeHintType;
 import com.google.zxing.WriterException;
 import com.google.zxing.common.BitArray;
-import com.google.zxing.common.StringUtils;
 import com.google.zxing.common.CharacterSetECI;
+import com.google.zxing.common.StringUtils;
 import com.google.zxing.common.reedsolomon.GenericGF;
 import com.google.zxing.common.reedsolomon.ReedSolomonEncoder;
 import com.google.zxing.qrcode.decoder.ErrorCorrectionLevel;
@@ -249,7 +249,9 @@ public final class Encoder {
    * if it is Shift_JIS, and the input is only double-byte Kanji, then we return {@link Mode#KANJI}.
    */
   private static Mode chooseMode(String content, Charset encoding) {
-    if (StringUtils.SHIFT_JIS_CHARSET != null &&
+    if (encoding != null &&
+        ("Shift_JIS".equalsIgnoreCase(encoding.name()) || "SJIS".equalsIgnoreCase(encoding.name())) &&
+        StringUtils.SHIFT_JIS_CHARSET != null &&
         StringUtils.SHIFT_JIS_CHARSET.equals(encoding) &&
         isOnlyDoubleByteKanji(content)) {
       // Choose Kanji mode if all input are double-byte characters