From: srowen <srowen@59b500cc-1b3d-0410-9834-0bbf25fbcc57>
Date: Fri, 15 Feb 2008 16:21:55 +0000 (+0000)
Subject: Fixed UTF-8 detection, again, and added unit test
X-Git-Url: http://git.rot13.org/?a=commitdiff_plain;h=4618d7794696e813e530b5ffc1df447176aebcf3;hp=55937e92102b86ece5f367d98e44907aa70bd6f1;p=zxing.git

Fixed UTF-8 detection, again, and added unit test

git-svn-id: http://zxing.googlecode.com/svn/trunk@204 59b500cc-1b3d-0410-9834-0bbf25fbcc57
---

diff --git a/core/src/com/google/zxing/qrcode/decoder/DecodedBitStreamParser.java b/core/src/com/google/zxing/qrcode/decoder/DecodedBitStreamParser.java
index e69ca8d5..4de854f9 100644
--- a/core/src/com/google/zxing/qrcode/decoder/DecodedBitStreamParser.java
+++ b/core/src/com/google/zxing/qrcode/decoder/DecodedBitStreamParser.java
@@ -208,28 +208,31 @@ final class DecodedBitStreamParser {
     // If we see something else in that second byte, we'll make the risky guess
     // that it's UTF-8.
     int length = bytes.length;
+    boolean canBeISO88591 = true;
     for (int i = 0; i < length; i++) {
       int value = bytes[i] & 0xFF;
       if (value >= 0x80 && value <= 0x9F && i < length - 1) {
+        canBeISO88591 = false;
         // ISO-8859-1 shouldn't use this, but before we decide it is Shift_JIS,
         // just double check that it is followed by a byte that's valid in
         // the Shift_JIS encoding
         int nextValue = bytes[i + 1] & 0xFF;
         if ((value & 0x1) == 0) {
-          // if even,
-          if (nextValue >= 0x9F && nextValue <= 0xFC) {
-            return SHIFT_JIS;
+          // if even, next value should be in [0x9F,0xFC]
+          // if not, we'll guess UTF-8
+          if (nextValue < 0x9F || nextValue > 0xFC) {
+            return UTF8;
           }
         } else {
-          if (nextValue >= 0x40 && nextValue <= 0x9E) {
-            return SHIFT_JIS;
+          // if odd, next value should be in [0x40,0x9E]
+          // if not, we'll guess UTF-8
+          if (nextValue < 0x40 || nextValue > 0x9E) {
+            return UTF8;
           }
         }
-        // otherwise we're going to take a guess that it's UTF-8
-        return UTF8;
       }
     }
-    return ISO88591;
+    return canBeISO88591 ? ISO88591 : SHIFT_JIS;
   }
 
 }
diff --git a/core/test/data/blackbox/qrcode-1/21.png b/core/test/data/blackbox/qrcode-1/21.png
new file mode 100644
index 00000000..b924a35a
Binary files /dev/null and b/core/test/data/blackbox/qrcode-1/21.png differ
diff --git a/core/test/data/blackbox/qrcode-1/21.txt b/core/test/data/blackbox/qrcode-1/21.txt
new file mode 100644
index 00000000..6bc3cf2e
--- /dev/null
+++ b/core/test/data/blackbox/qrcode-1/21.txt
@@ -0,0 +1,2 @@
+ÁRVÍZTŰRŐ TÜKÖRFÚRÓGÉP
+árvíztűrő tükörfúrógép
\ No newline at end of file