LUCENE-1466: added chainable CharFilter stage before Tokenizer to allow mapping of characters before tokenization

git-svn-id: https://svn.apache.org/repos/asf/lucene/java/trunk@787795 13f79535-47bb-0310-9956-ffa450edef68

CHANGES.txt

@@ -236,7 +236,7 @@ API Changes
     NumericRangeQuery and its new indexing format for numeric or
     date values.  (Uwe Schindler)
     
-23. LUCENE-1630: Deprecate Weight in favor of QueryWeight, which adds
+24. LUCENE-1630: Deprecate Weight in favor of QueryWeight, which adds
     a scorer(IndexReader, boolean /* scoreDocsInOrder */, boolean /*
     topScorer */) method instead of scorer(IndexReader) (now
     deprecated). The new method is used by IndexSearcher to mate
@@ -254,6 +254,11 @@ API Changes
     out of order when used with a Collector that can accept docs out
     of order. (Shai Erera via Mike McCandless)
  	
+25. LUCENE-1466: Changed Tokenizer.input to be a CharStream; added
+    CharFilter and MappingCharFilter, which allows chaining & mapping
+    of characters before tokenizers run.  (Koji Sekiguchi via Mike
+    McCandless)
+
 Bug fixes
 
 1. LUCENE-1415: MultiPhraseQuery has incorrect hashCode() and equals()

common-build.xml

@@ -42,7 +42,7 @@
   <property name="Name" value="Lucene"/>
   <property name="dev.version" value="2.9-dev"/>
   <property name="version" value="${dev.version}"/>
-  <property name="compatibility.tag" value="lucene_2_4_back_compat_tests_20090623"/>
+  <property name="compatibility.tag" value="lucene_2_4_back_compat_tests_20090624"/>
   <property name="spec.version" value="${version}"/>	
   <property name="year" value="2000-${current.year}"/>
   <property name="final.name" value="lucene-${name}-${version}"/>

contrib/analyzers/src/java/org/apache/lucene/analysis/cjk/CJKTokenizer.java

@@ -96,7 +96,7 @@ public final class CJKTokenizer extends Tokenizer {
      * @param in I/O reader
      */
     public CJKTokenizer(Reader in) {
-        input = in;
+      super(in);
     }
 
     //~ Methods ----------------------------------------------------------------
@@ -253,7 +253,7 @@ public final class CJKTokenizer extends Tokenizer {
       
         if (length > 0) {
             return reusableToken.reinit
-                (buffer, 0, length, start, start+length, TOKEN_TYPE_NAMES[tokenType]);
+                (buffer, 0, length, input.correctOffset(start), input.correctOffset(start+length), TOKEN_TYPE_NAMES[tokenType]);
         } else if (dataLen == -1) {
           return null;
         }

contrib/analyzers/src/java/org/apache/lucene/analysis/cn/ChineseTokenizer.java

@@ -55,7 +55,7 @@ public final class ChineseTokenizer extends Tokenizer {
 
 
     public ChineseTokenizer(Reader in) {
-        input = in;
+      super(in);
     }
 
     private int offset = 0, bufferIndex=0, dataLen=0;
@@ -81,7 +81,7 @@ public final class ChineseTokenizer extends Tokenizer {
         if (length>0) {
             //System.out.println(new String(buffer, 0,
             //length));
-          return token.reinit(buffer, 0, length, start, start+length);
+          return token.reinit(buffer, 0, length, input.correctOffset(start), input.correctOffset(start+length));
         }
         else
             return null;

contrib/analyzers/src/java/org/apache/lucene/analysis/cn/smart/SentenceTokenizer.java

@@ -45,6 +45,7 @@ public class SentenceTokenizer extends Tokenizer {
   private Token t = new Token();
 
   public SentenceTokenizer(Reader reader) {
+    super(reader);
     bufferInput = new BufferedReader(reader, 2048);
   }
 
@@ -91,7 +92,7 @@ public class SentenceTokenizer extends Tokenizer {
       return null;
     else {
       t.clear();
-      t.reinit(buffer.toString(), tokenStart, tokenEnd, "sentence");
+      t.reinit(buffer.toString(), input.correctOffset(tokenStart), input.correctOffset(tokenEnd), "sentence");
       return t;
     }
   }

contrib/analyzers/src/java/org/apache/lucene/analysis/ngram/EdgeNGramTokenizer.java

@@ -140,8 +140,8 @@ public class EdgeNGramTokenizer extends Tokenizer {
     int start = side == Side.FRONT ? 0 : inLen - gramSize;
     int end = start + gramSize;
     reusableToken.setTermBuffer(inStr, start, gramSize);
-    reusableToken.setStartOffset(start);
-    reusableToken.setEndOffset(end);
+    reusableToken.setStartOffset(input.correctOffset(start));
+    reusableToken.setEndOffset(input.correctOffset(end));
     gramSize++;
     return reusableToken;
   }

contrib/analyzers/src/java/org/apache/lucene/analysis/ngram/NGramTokenizer.java

@@ -85,6 +85,6 @@ public class NGramTokenizer extends Tokenizer {
 
     int oldPos = pos;
     pos++;
-    return reusableToken.reinit(inStr, oldPos, gramSize, oldPos, oldPos+gramSize);
+    return reusableToken.reinit(inStr, oldPos, gramSize, input.correctOffset(oldPos), input.correctOffset(oldPos+gramSize));
   }
 }

contrib/wikipedia/src/java/org/apache/lucene/wikipedia/analysis/WikipediaTokenizer.java

@@ -17,6 +17,7 @@
 
 package org.apache.lucene.wikipedia.analysis;
 
+import org.apache.lucene.analysis.CharReader;
 import org.apache.lucene.analysis.Token;
 import org.apache.lucene.analysis.Tokenizer;
 
@@ -107,7 +108,7 @@ public class WikipediaTokenizer extends Tokenizer {
   private Iterator tokens = null;
 
   void setInput(Reader reader) {
-    this.input = reader;
+    this.input = CharReader.get(reader);
   }
 
   /**
@@ -190,8 +191,8 @@ public class WikipediaTokenizer extends Tokenizer {
     //trim the buffer
     String s = buffer.toString().trim();
     reusableToken.setTermBuffer(s.toCharArray(), 0, s.length());
-    reusableToken.setStartOffset(theStart);
-    reusableToken.setEndOffset(theStart + s.length());
+    reusableToken.setStartOffset(input.correctOffset(theStart));
+    reusableToken.setEndOffset(input.correctOffset(theStart + s.length()));
     reusableToken.setFlags(UNTOKENIZED_TOKEN_FLAG);
     //The way the loop is written, we will have proceeded to the next token.  We need to pushback the scanner to lastPos
     if (tmpTokType != WikipediaTokenizerImpl.YYEOF){
@@ -229,8 +230,8 @@ public class WikipediaTokenizer extends Tokenizer {
     //trim the buffer
     String s = buffer.toString().trim();
     reusableToken.setTermBuffer(s.toCharArray(), 0, s.length());
-    reusableToken.setStartOffset(theStart);
-    reusableToken.setEndOffset(theStart + s.length());
+    reusableToken.setStartOffset(input.correctOffset(theStart));
+    reusableToken.setEndOffset(input.correctOffset(theStart + s.length()));
     reusableToken.setFlags(UNTOKENIZED_TOKEN_FLAG);
     //The way the loop is written, we will have proceeded to the next token.  We need to pushback the scanner to lastPos
     if (tmpTokType != WikipediaTokenizerImpl.YYEOF){
@@ -243,8 +244,8 @@ public class WikipediaTokenizer extends Tokenizer {
   private void setupToken(final Token reusableToken) {
     scanner.getText(reusableToken);
     final int start = scanner.yychar();
-    reusableToken.setStartOffset(start);
-    reusableToken.setEndOffset(start + reusableToken.termLength());
+    reusableToken.setStartOffset(input.correctOffset(start));
+    reusableToken.setEndOffset(input.correctOffset(start + reusableToken.termLength()));
   }
 
   /*
@@ -258,7 +259,7 @@ public class WikipediaTokenizer extends Tokenizer {
   }
 
   public void reset(Reader reader) throws IOException {
-    input = reader;
+    setInput(reader);
     reset();
   }
 

src/java/org/apache/lucene/analysis/BaseCharFilter.java

@@ -0,0 +1,90 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.lucene.analysis;
+
+import java.util.ArrayList;
+import java.util.List;
+
+/**
+ * Base utility class for implementing a {@link
+ * CharFilter}.  You record mappings by calling {@link
+ * #addOffCorrectMap}, and then invoke the correct method.
+ * This class is not particularly efficient, eg a new class
+ * instance is created for every call to {@link
+ * #addOffCorrectMap}, which is appended to a private list.
+ * When retrieving a mapping, that list is linearly
+ * checked.
+ * @version $Id$
+ */
+public abstract class BaseCharFilter extends CharFilter {
+
+  //private List<OffCorrectMap> pcmList;
+  private List pcmList;
+  
+  public BaseCharFilter(CharStream in) {
+    super(in);
+  }
+
+  /** Retrieve the corrected offset.  Note that this method
+   *  is slow if you correct positions far before the most
+   *  recently added position. */
+  protected int correct(int currentOff) {
+    if (pcmList == null || pcmList.isEmpty()) {
+      return currentOff;
+    }
+    for (int i = pcmList.size() - 1; i >= 0; i--) {
+      if (currentOff >= ((OffCorrectMap) pcmList.get(i)).off) {
+        return currentOff + ((OffCorrectMap) pcmList.get(i)).cumulativeDiff;
+      }
+    }
+    return currentOff;
+  }
+  
+  protected int getLastCumulativeDiff() {
+    return pcmList == null || pcmList.isEmpty() ?
+      0 : ((OffCorrectMap)pcmList.get(pcmList.size() - 1)).cumulativeDiff;
+  }
+
+  protected void addOffCorrectMap(int off, int cumulativeDiff) {
+    if (pcmList == null) {
+      pcmList = new ArrayList();
+    }
+    pcmList.add(new OffCorrectMap(off, cumulativeDiff));
+  }
+
+  static class OffCorrectMap {
+
+    int off;
+    int cumulativeDiff;
+
+    OffCorrectMap(int off, int cumulativeDiff) {
+      this.off = off;
+      this.cumulativeDiff = cumulativeDiff;
+    }
+
+    public String toString() {
+      StringBuffer sb = new StringBuffer();
+      sb.append('(');
+      sb.append(off);
+      sb.append(',');
+      sb.append(cumulativeDiff);
+      sb.append(')');
+      return sb.toString();
+    }
+  }
+}

src/java/org/apache/lucene/analysis/CharFilter.java

@@ -0,0 +1,61 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.lucene.analysis;
+
+import java.io.IOException;
+
+/**
+ * Subclasses of CharFilter can be chained to filter CharStream.
+ *
+ * @version $Id$
+ *
+ */
+public abstract class CharFilter extends CharStream {
+
+  protected CharStream input;
+
+  protected CharFilter(CharStream in) {
+    input = in;
+  }
+
+  /**
+   * Subclass may want to override to correct the current offset.
+   *
+   * @param currentOff current offset
+   * @return corrected offset
+   */
+  protected int correct(int currentOff) {
+    return currentOff;
+  }
+
+  /**
+   * Chains the corrected offset through the input
+   * CharFilter.
+   */
+  public final int correctOffset(int currentOff) {
+    return input.correctOffset(correct(currentOff));
+  }
+
+  public void close() throws IOException {
+    input.close();
+  }
+
+  public int read(char[] cbuf, int off, int len) throws IOException {
+    return input.read(cbuf, off, len);
+  }
+}

src/java/org/apache/lucene/analysis/CharReader.java

@@ -0,0 +1,53 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.lucene.analysis;
+
+import java.io.IOException;
+import java.io.Reader;
+
+/**
+ * CharReader is a Reader wrapper. It reads chars from Reader and outputs CharStream.
+ *
+ * @version $Id$
+ *
+ */
+public final class CharReader extends CharStream {
+
+  protected Reader input;
+  
+  public static CharStream get(Reader input) {
+    return input instanceof CharStream ?
+      (CharStream)input : new CharReader(input);
+  }
+
+  private CharReader(Reader in) {
+    input = in;
+  }
+
+  public int correctOffset(int currentOff) {
+    return currentOff;
+  }
+
+  public void close() throws IOException {
+    input.close();
+  }
+
+  public int read(char[] cbuf, int off, int len) throws IOException {
+    return input.read(cbuf, off, len);
+  }
+}

src/java/org/apache/lucene/analysis/CharStream.java

@@ -0,0 +1,36 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.lucene.analysis;
+
+import java.io.Reader;
+
+/**
+ * CharStream adds <a href="#correctOffset(int)">correctOffset</a> functionality over Reader.
+ *
+ * @version $Id$
+ */
+public abstract class CharStream extends Reader {
+
+  /**
+   * Called by CharFilter(s) and Tokenizer to correct token offset.
+   *
+   * @param currentOff current offset
+   * @return corrected token offset
+   */
+  public abstract int correctOffset(int currentOff);
+}

src/java/org/apache/lucene/analysis/CharTokenizer.java

@@ -90,7 +90,7 @@ public abstract class CharTokenizer extends Tokenizer {
     }
 
     termAtt.setTermLength(length);
-    offsetAtt.setOffset(start, start+length);
+    offsetAtt.setOffset(input.correctOffset(start), input.correctOffset(start+length));
     return true;
   }
 
@@ -134,8 +134,8 @@ public abstract class CharTokenizer extends Tokenizer {
     }
 
     reusableToken.setTermLength(length);
-    reusableToken.setStartOffset(start);
-    reusableToken.setEndOffset(start+length);
+    reusableToken.setStartOffset(input.correctOffset(start));
+    reusableToken.setEndOffset(input.correctOffset(start+length));
     return reusableToken;
   }
 

src/java/org/apache/lucene/analysis/KeywordTokenizer.java

@@ -59,7 +59,7 @@ public class KeywordTokenizer extends Tokenizer {
           buffer = termAtt.resizeTermBuffer(1+buffer.length);
       }
       termAtt.setTermLength(upto);
-      offsetAtt.setOffset(0, upto);
+      offsetAtt.setOffset(input.correctOffset(0), input.correctOffset(upto));
       return true;
     }
     return false;
@@ -81,8 +81,8 @@ public class KeywordTokenizer extends Tokenizer {
           buffer = reusableToken.resizeTermBuffer(1+buffer.length);
       }
       reusableToken.setTermLength(upto);
-      reusableToken.setStartOffset(0);
-      reusableToken.setEndOffset(upto);
+      reusableToken.setStartOffset(input.correctOffset(0));
+      reusableToken.setEndOffset(input.correctOffset(upto));
       
       return reusableToken;
     }

src/java/org/apache/lucene/analysis/MappingCharFilter.java

@@ -0,0 +1,140 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.lucene.analysis;
+
+import java.io.IOException;
+import java.util.LinkedList;
+
+/**
+ * {@link CharFilter} that applies the mappings contained in
+ * a {@link NormalizeCharMap} to the character stream.
+ *
+ * @version $Id$
+ */
+public class MappingCharFilter extends BaseCharFilter {
+
+  private final NormalizeCharMap normMap;
+  //private LinkedList<Character> buffer;
+  private LinkedList buffer;
+  private String replacement;
+  private int charPointer;
+  private int nextCharCounter;
+
+  public MappingCharFilter(NormalizeCharMap normMap, CharStream in) {
+    super(in);
+    this.normMap = normMap;
+  }
+
+  public int read() throws IOException {
+    while(true) {
+      if (replacement != null && charPointer < replacement.length()) {
+        return replacement.charAt(charPointer++);
+      }
+
+      int firstChar = nextChar();
+      if (firstChar == -1) return -1;
+      NormalizeCharMap nm = normMap.submap != null ?
+        (NormalizeCharMap)normMap.submap.get(Character.valueOf((char) firstChar)) : null;
+      if (nm == null) return firstChar;
+      NormalizeCharMap result = match(nm);
+      if (result == null) return firstChar;
+      replacement = result.normStr;
+      charPointer = 0;
+      if (result.diff != 0) {
+        int prevCumulativeDiff = getLastCumulativeDiff();
+        if (result.diff < 0) {
+          for(int i = 0; i < -result.diff ; i++)
+            addOffCorrectMap(nextCharCounter + i - prevCumulativeDiff, prevCumulativeDiff - 1 - i);
+        } else {
+          addOffCorrectMap(nextCharCounter - result.diff - prevCumulativeDiff, prevCumulativeDiff + result.diff);
+        }
+      }
+    }
+  }
+
+  private int nextChar() throws IOException {
+    nextCharCounter++;
+    if (buffer != null && !buffer.isEmpty()) {
+      return ((Character)buffer.removeFirst()).charValue();
+    }
+    return input.read();
+  }
+
+  private void pushChar(int c) {
+    nextCharCounter--;
+    if(buffer == null)
+      buffer = new LinkedList();
+    buffer.addFirst(new Character((char) c));
+  }
+
+  private void pushLastChar(int c) {
+    if (buffer == null) {
+      buffer = new LinkedList();
+    }
+    buffer.addLast(new Character((char) c));
+  }
+
+  private NormalizeCharMap match(NormalizeCharMap map) throws IOException {
+    NormalizeCharMap result = null;
+    if (map.submap != null) {
+      int chr = nextChar();
+      if (chr != -1) {
+        NormalizeCharMap subMap = (NormalizeCharMap) map.submap.get(Character.valueOf((char) chr));
+        if (subMap != null) {
+          result = match(subMap);
+        }
+        if (result == null) {
+          pushChar(chr);
+        }
+      }
+    }
+    if (result == null && map.normStr != null) {
+      result = map;
+    }
+    return result;
+  }
+
+  public int read(char[] cbuf, int off, int len) throws IOException {
+    char[] tmp = new char[len];
+    int l = input.read(tmp, 0, len);
+    if (l != -1) {
+      for(int i = 0; i < l; i++)
+        pushLastChar(tmp[i]);
+    }
+    l = 0;
+    for(int i = off; i < off + len; i++) {
+      int c = read();
+      if (c == -1) break;
+      cbuf[i] = (char) c;
+      l++;
+    }
+    return l == 0 ? -1 : l;
+  }
+
+  public boolean markSupported() {
+    return false;
+  }
+
+  public void mark(int readAheadLimit) throws IOException {
+    throw new IOException("mark/reset not supported");
+  }
+
+  public void reset() throws IOException {
+    throw new IOException("mark/reset not supported");
+  }
+}

src/java/org/apache/lucene/analysis/NormalizeCharMap.java

@@ -0,0 +1,55 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.lucene.analysis;
+
+import java.util.HashMap;
+import java.util.Map;
+
+/**
+ * Holds a map of String input to String output, to be used
+ * with {@link MappingCharFilter}.
+ * @version $Id$
+ */
+public class NormalizeCharMap {
+
+  //Map<Character, NormalizeMap> submap;
+  Map submap;
+  String normStr;
+  int diff;
+
+  public void add(String singleMatch, String replacement) {
+    NormalizeCharMap currMap = this;
+    for(int i = 0; i < singleMatch.length(); i++) {
+      char c = singleMatch.charAt(i);
+      if (currMap.submap == null) {
+        currMap.submap = new HashMap(1);
+      }
+      NormalizeCharMap map = (NormalizeCharMap) currMap.submap.get(Character.valueOf(c));
+      if (map == null) {
+        map = new NormalizeCharMap();
+        currMap.submap.put(new Character(c), map);
+      }
+      currMap = map;
+    }
+    if (currMap.normStr != null) {
+      throw new RuntimeException("MappingCharFilter: there is already a mapping for " + singleMatch);
+    }
+    currMap.normStr = replacement;
+    currMap.diff = singleMatch.length() - replacement.length();
+  }
+}

src/java/org/apache/lucene/analysis/Tokenizer.java

@@ -45,16 +45,20 @@ import java.io.IOException;
 
 public abstract class Tokenizer extends TokenStream {
   /** The text source for this Tokenizer. */
-  protected Reader input;
+  protected CharStream input;
 
   /** Construct a tokenizer with null input. */
   protected Tokenizer() {}
 
   /** Construct a token stream processing the given input. */
   protected Tokenizer(Reader input) {
-    this.input = input;
+    this.input = CharReader.get(input);
   }
 
+  protected Tokenizer(CharStream input) {
+    this.input = input;
+  }
+  
   /** By default, closes the input Reader. */
   public void close() throws IOException {
     input.close();
@@ -64,6 +68,10 @@ public abstract class Tokenizer extends TokenStream {
    *  analyzer (in its reusableTokenStream method) will use
    *  this to re-use a previously created tokenizer. */
   public void reset(Reader input) throws IOException {
+    this.input = CharReader.get(input);
+  }
+
+  public void reset(CharStream input) throws IOException {
     this.input = input;
   }
 }

src/java/org/apache/lucene/analysis/standard/StandardTokenizer.java

@@ -20,6 +20,7 @@ package org.apache.lucene.analysis.standard;
 import java.io.IOException;
 import java.io.Reader;
 
+import org.apache.lucene.analysis.CharReader;
 import org.apache.lucene.analysis.Token;
 import org.apache.lucene.analysis.Tokenizer;
 import org.apache.lucene.analysis.tokenattributes.OffsetAttribute;
@@ -91,7 +92,7 @@ public class StandardTokenizer extends Tokenizer {
   private boolean replaceInvalidAcronym;
     
   void setInput(Reader reader) {
-    this.input = reader;
+    input = CharReader.get(reader);
   }
 
   private int maxTokenLength = StandardAnalyzer.DEFAULT_MAX_TOKEN_LENGTH;
@@ -126,7 +127,7 @@ public class StandardTokenizer extends Tokenizer {
    */
   public StandardTokenizer(Reader input, boolean replaceInvalidAcronym) {
     this.replaceInvalidAcronym = replaceInvalidAcronym;
-    this.input = input;
+    setInput(input);
     this.scanner = new StandardTokenizerImpl(input);
     termAtt = (TermAttribute) addAttribute(TermAttribute.class);
     offsetAtt = (OffsetAttribute) addAttribute(OffsetAttribute.class);
@@ -161,7 +162,7 @@ public class StandardTokenizer extends Tokenizer {
         posIncrAtt.setPositionIncrement(posIncr);
         scanner.getText(termAtt);
         final int start = scanner.yychar();
-        offsetAtt.setOffset(start, start+termAtt.termLength());
+        offsetAtt.setOffset(input.correctOffset(start), input.correctOffset(start+termAtt.termLength()));
         // This 'if' should be removed in the next release. For now, it converts
         // invalid acronyms to HOST. When removed, only the 'else' part should
         // remain.
@@ -194,19 +195,19 @@ public class StandardTokenizer extends Tokenizer {
       int posIncr = 1;
 
       while(true) {
-	int tokenType = scanner.getNextToken();
+        int tokenType = scanner.getNextToken();
 
-	if (tokenType == StandardTokenizerImpl.YYEOF) {
-	    return null;
-	}
+        if (tokenType == StandardTokenizerImpl.YYEOF) {
+          return null;
+        }
 
         if (scanner.yylength() <= maxTokenLength) {
           reusableToken.clear();
           reusableToken.setPositionIncrement(posIncr);
           scanner.getText(reusableToken);
           final int start = scanner.yychar();
-          reusableToken.setStartOffset(start);
-          reusableToken.setEndOffset(start+reusableToken.termLength());
+          reusableToken.setStartOffset(input.correctOffset(start));
+          reusableToken.setEndOffset(input.correctOffset(start+reusableToken.termLength()));
           // This 'if' should be removed in the next release. For now, it converts
           // invalid acronyms to HOST. When removed, only the 'else' part should
           // remain.
@@ -234,13 +235,13 @@ public class StandardTokenizer extends Tokenizer {
      * @see org.apache.lucene.analysis.TokenStream#reset()
      */
     public void reset() throws IOException {
-	super.reset();
-	scanner.yyreset(input);
+      super.reset();
+      scanner.yyreset(input);
     }
 
     public void reset(Reader reader) throws IOException {
-        input = reader;
-        reset();
+      setInput(reader);
+      reset();
     }
 
   /**