From ed4a0639ba50b9c498e13a8f9ecd665d113f2df4 Mon Sep 17 00:00:00 2001
From: Eric Mader <emader@svn.icu-project.org>
Date: Fri, 20 May 2005 22:33:10 +0000
Subject: [PATCH] ICU-4060 Implement CharsetMatch.getReader(),
 CharsetMatch.getString(), CharsetDetector.getReader(),
 CharsetDetector.getString(), relationship between ISO-8859-x and
 Windows-125x.

X-SVN-Rev: 17637
---
 .../src/com/ibm/icu/text/CharsetDetector.java | 31 +++++++++++++++++--
 icu4j/src/com/ibm/icu/text/CharsetMatch.java  | 31 +++++++++++++++++--
 .../com/ibm/icu/text/CharsetRecog_sbcs.java   | 17 +++++-----
 3 files changed, 64 insertions(+), 15 deletions(-)

diff --git a/icu4j/src/com/ibm/icu/text/CharsetDetector.java b/icu4j/src/com/ibm/icu/text/CharsetDetector.java
index ae034ba9ec3..3a145d439e6 100644
--- a/icu4j/src/com/ibm/icu/text/CharsetDetector.java
+++ b/icu4j/src/com/ibm/icu/text/CharsetDetector.java
@@ -212,7 +212,15 @@ public class CharsetDetector {
      *           or null or an empty string if none is available.
      */
     public Reader getReader(InputStream in, String declaredEncoding) {
-        return null;
+        fDeclaredEncoding = declaredEncoding;
+        
+        try {
+            setText(in);
+            
+            return detect().getReader();
+        } catch (IOException e) {
+            return null;
+        }
     }
 
     /**
@@ -231,7 +239,14 @@ public class CharsetDetector {
      */
     public String getString(byte[] in, String declaredEncoding)
     {
-        return null;
+        fDeclaredEncoding = declaredEncoding;
+       
+        try {
+            setText(in);
+            return detect().getString(-1);
+        } catch (IOException e) {
+            return null;
+        }
     }
 
  
@@ -343,7 +358,14 @@ public class CharsetDetector {
         for (srci=0; srci<fInputLen; srci++) {
             int val = fInputBytes[srci] & 0x00ff;
             fByteStats[val]++;
-        }        
+        }
+        
+        for (int i = 0x80; i <= 0x9F; i += 1) {
+            if (fByteStats[i] != 0) {
+                fC1Bytes = true;
+                break;
+            }
+        }
      }
 
     /**
@@ -359,6 +381,9 @@ public class CharsetDetector {
                    new short[256];  //   Value is percent, not absolute.
                                     //   Value is rounded up, so zero really means zero occurences.
     
+    boolean     fC1Bytes =          // True if any bytes in the range 0x80 - 0x9F are in the input;
+                   false;
+    
     String      fDeclaredEncoding;
     
     
diff --git a/icu4j/src/com/ibm/icu/text/CharsetMatch.java b/icu4j/src/com/ibm/icu/text/CharsetMatch.java
index af54a89f758..0aa992dca29 100644
--- a/icu4j/src/com/ibm/icu/text/CharsetMatch.java
+++ b/icu4j/src/com/ibm/icu/text/CharsetMatch.java
@@ -6,7 +6,10 @@
 */
 package com.ibm.icu.text;
 
+import java.io.ByteArrayInputStream;
+import java.io.IOException;
 import java.io.InputStream;
+import java.io.InputStreamReader;
 import java.io.Reader;
 
 
@@ -36,7 +39,18 @@ public class CharsetMatch implements Comparable {
      * @return the Reader for the Unicode character data.
      */
     public Reader getReader() {
-        return null;
+        InputStream inputStream = fInputStream;
+        
+        if (inputStream == null) {
+            inputStream = new ByteArrayInputStream(fRawInput, 0, fRawLength);
+        }
+        
+        try {
+            fInputStream.reset();
+            return new InputStreamReader(fInputStream, getName());
+        } catch (IOException e) {
+            return null;
+        }
     }
     
     
@@ -66,7 +80,20 @@ public class CharsetMatch implements Comparable {
     public String getString(int maxLength) throws java.io.IOException {
         String result = null;
         if (fInputStream != null) {
-            // TODO:  read the stream in somehow.
+            StringBuffer sb = new StringBuffer();
+            char[] buffer = new char[1024];
+            Reader reader = getReader();
+            int max = maxLength < 0? Integer.MAX_VALUE : maxLength;
+            int bytesRead = 0;
+            
+            while ((bytesRead = reader.read(buffer, 0, Math.min(max, 1024))) >= 0) {
+                sb.append(buffer, 0, bytesRead);
+                max -= bytesRead;
+            }
+            
+            reader.close();
+            
+            return sb.toString();
         } else {
             result = new String(fRawInput, getName());            
         }
diff --git a/icu4j/src/com/ibm/icu/text/CharsetRecog_sbcs.java b/icu4j/src/com/ibm/icu/text/CharsetRecog_sbcs.java
index 69d42b80540..b7fe00f428c 100644
--- a/icu4j/src/com/ibm/icu/text/CharsetRecog_sbcs.java
+++ b/icu4j/src/com/ibm/icu/text/CharsetRecog_sbcs.java
@@ -148,17 +148,14 @@ public abstract class CharsetRecog_sbcs extends CharsetRecognizer {
         }
     }
     
-    private int byteIndex;
-    
-    void reset()
-    {
-        byteIndex = 0;
-    }
+    protected boolean haveC1Bytes = false;
     
     int match(CharsetDetector det, int[] ngrams,  byte[] byteMap)
     {
         NGramParser parser = new NGramParser(ngrams, byteMap);
         
+        haveC1Bytes = det.fC1Bytes;
+        
         return parser.parse(det);
     }
     
@@ -201,7 +198,7 @@ public abstract class CharsetRecog_sbcs extends CharsetRecognizer {
 
         public String getName()
         {
-            return "ISO-8859-1";
+            return haveC1Bytes? "windows-1252" : "ISO-8859-1";
         }
     }
 
@@ -444,7 +441,7 @@ public abstract class CharsetRecog_sbcs extends CharsetRecognizer {
 
         public String getName()
         {
-            return "ISO-8859-2";
+            return haveC1Bytes? "windows-1250" : "ISO-8859-2";
         }
     }
     
@@ -630,7 +627,7 @@ public abstract class CharsetRecog_sbcs extends CharsetRecognizer {
 
         public String getName()
         {
-            return "ISO-8859-7";
+            return haveC1Bytes? "windows-1253" : "ISO-8859-7";
         }
     }
     
@@ -693,7 +690,7 @@ public abstract class CharsetRecog_sbcs extends CharsetRecognizer {
 
         public String getName()
         {
-            return "ISO-8859-9";
+            return haveC1Bytes? "windows-1254" : "ISO-8859-9";
         }
     }