jars/icu4j-52_1/main/classes/translit/src/com/ibm/icu/text/LowercaseTransliterator.java

   1 /*
   2  *******************************************************************************
   3  * Copyright (C) 1996-2011, International Business Machines Corporation and    *
   4  * others. All Rights Reserved.                                                *
   5  *******************************************************************************
   6  */
   7 package com.ibm.icu.text;
   8
   9 import com.ibm.icu.impl.UCaseProps;
  10 import com.ibm.icu.lang.UCharacter;
  11 import com.ibm.icu.util.ULocale;
  12
  13 /**
  14  * A transliterator that performs locale-sensitive toLower()
  15  * case mapping.
  16  */
  17 class LowercaseTransliterator extends Transliterator{
  18
  19     /**
  20      * Package accessible ID.
  21      */
  22     static final String _ID = "Any-Lower";
  23
  24     // TODO: Add variants for tr, az, lt, default = default locale
  25
  26     /**
  27      * System registration hook.
  28      */
  29     static void register() {
  30         Transliterator.registerFactory(_ID, new Transliterator.Factory() {
  31             public Transliterator getInstance(String ID) {
  32                 return new LowercaseTransliterator(ULocale.US);
  33             }
  34         });
  35
  36         Transliterator.registerSpecialInverse("Lower", "Upper", true);
  37     }
  38
  39     private ULocale locale;
  40
  41     private UCaseProps csp;
  42     private ReplaceableContextIterator iter;
  43     private StringBuilder result;
  44     private int[] locCache;
  45
  46     /**
  47      * Constructs a transliterator.
  48      */
  49
  50     public LowercaseTransliterator(ULocale loc) {
  51         super(_ID, null);
  52         locale = loc;
  53         csp=UCaseProps.INSTANCE;
  54         iter=new ReplaceableContextIterator();
  55         result = new StringBuilder();
  56         locCache = new int[1];
  57         locCache[0]=0;
  58     }
  59
  60     /**
  61      * Implements {@link Transliterator#handleTransliterate}.
  62      */
  63     protected synchronized void handleTransliterate(Replaceable text,
  64                                        Position offsets, boolean isIncremental) {
  65         if(csp==null) {
  66             return;
  67         }
  68
  69         if(offsets.start >= offsets.limit) {
  70             return;
  71         }
  72
  73         iter.setText(text);
  74         result.setLength(0);
  75         int c, delta;
  76
  77         // Walk through original string
  78         // If there is a case change, modify corresponding position in replaceable
  79
  80         iter.setIndex(offsets.start);
  81         iter.setLimit(offsets.limit);
  82         iter.setContextLimits(offsets.contextStart, offsets.contextLimit);
  83         while((c=iter.nextCaseMapCP())>=0) {
  84             c=csp.toFullLower(c, iter, result, locale, locCache);
  85
  86             if(iter.didReachLimit() && isIncremental) {
  87                 // the case mapping function tried to look beyond the context limit
  88                 // wait for more input
  89                 offsets.start=iter.getCaseMapCPStart();
  90                 return;
  91             }
  92
  93             /* decode the result */
  94             if(c<0) {
  95                 /* c mapped to itself, no change */
  96                 continue;
  97             } else if(c<=UCaseProps.MAX_STRING_LENGTH) {
  98                 /* replace by the mapping string */
  99                 delta=iter.replace(result.toString());
 100                 result.setLength(0);
 101             } else {
 102                 /* replace by single-code point mapping */
 103                 delta=iter.replace(UTF16.valueOf(c));
 104             }
 105
 106             if(delta!=0) {
 107                 offsets.limit += delta;
 108                 offsets.contextLimit += delta;
 109             }
 110         }
 111         offsets.start = offsets.limit;
 112     }
 113
 114     // NOTE: normally this would be static, but because the results vary by locale....
 115     SourceTargetUtility sourceTargetUtility = null;
 116
 117     /* (non-Javadoc)
 118      * @see com.ibm.icu.text.Transliterator#addSourceTargetSet(com.ibm.icu.text.UnicodeSet, com.ibm.icu.text.UnicodeSet, com.ibm.icu.text.UnicodeSet)
 119      */
 120     @Override
 121     public void addSourceTargetSet(UnicodeSet inputFilter, UnicodeSet sourceSet, UnicodeSet targetSet) {
 122         synchronized (this) {
 123             if (sourceTargetUtility == null) {
 124                 sourceTargetUtility = new SourceTargetUtility(new Transform<String,String>() {
 125                     public String transform(String source) {
 126                         return UCharacter.toLowerCase(locale, source);
 127                     }
 128                 });
 129             }
 130         }
 131         sourceTargetUtility.addSourceTargetSet(this, inputFilter, sourceSet, targetSet);
 132     }
 133 }