icuSources/i18n/unicode/sortkey.h

   1 /*
   2  *****************************************************************************
   3  * Copyright (C) 1996-2006, International Business Machines Corporation and others.
   4  * All Rights Reserved.
   5  *****************************************************************************
   6  *
   7  * File sortkey.h
   8  *
   9  * Created by: Helena Shih
  10  *
  11  * Modification History:
  12  *
  13  *  Date         Name          Description
  14  *
  15  *  6/20/97     helena      Java class name change.
  16  *  8/18/97     helena      Added internal API documentation.
  17  *  6/26/98     erm         Changed to use byte arrays and memcmp.
  18  *****************************************************************************
  19  */
  20
  21 #ifndef SORTKEY_H
  22 #define SORTKEY_H
  23
  24 #include "unicode/utypes.h"
  25
  26 /**
  27  * \file
  28  * \brief C++ API: Keys for comparing strings multiple times.
  29  */
  30
  31 #if !UCONFIG_NO_COLLATION
  32
  33 #include "unicode/uobject.h"
  34 #include "unicode/unistr.h"
  35 #include "unicode/coll.h"
  36
  37 U_NAMESPACE_BEGIN
  38
  39 /* forward declaration */
  40 class RuleBasedCollator;
  41
  42 /**
  43  *
  44  * Collation keys are generated by the Collator class.  Use the CollationKey objects
  45  * instead of Collator to compare strings multiple times.  A CollationKey
  46  * preprocesses the comparison information from the Collator object to
  47  * make the comparison faster.  If you are not going to comparing strings
  48  * multiple times, then using the Collator object is generally faster,
  49  * since it only processes as much of the string as needed to make a
  50  * comparison.
  51  * <p> For example (with strength == tertiary)
  52  * <p>When comparing "Abernathy" to "Baggins-Smythworthy", Collator
  53  * only needs to process a couple of characters, while a comparison
  54  * with CollationKeys will process all of the characters.  On the other hand,
  55  * if you are doing a sort of a number of fields, it is much faster to use
  56  * CollationKeys, since you will be comparing strings multiple times.
  57  * <p>Typical use of CollationKeys are in databases, where you store a CollationKey
  58  * in a hidden field, and use it for sorting or indexing.
  59  *
  60  * <p>Example of use:
  61  * <pre>
  62  * \code
  63  *     UErrorCode success = U_ZERO_ERROR;
  64  *     Collator* myCollator = Collator::createInstance(success);
  65  *     CollationKey* keys = new CollationKey [3];
  66  *     myCollator->getCollationKey("Tom", keys[0], success );
  67  *     myCollator->getCollationKey("Dick", keys[1], success );
  68  *     myCollator->getCollationKey("Harry", keys[2], success );
  69  *
  70  *     // Inside body of sort routine, compare keys this way:
  71  *     CollationKey tmp;
  72  *     if(keys[0].compareTo( keys[1] ) > 0 ) {
  73  *         tmp = keys[0]; keys[0] = keys[1]; keys[1] = tmp;
  74  *     }
  75  *     //...
  76  * \endcode
  77  * </pre>
  78  * <p>Because Collator::compare()'s algorithm is complex, it is faster to sort
  79  * long lists of words by retrieving collation keys with Collator::getCollationKey().
  80  * You can then cache the collation keys and compare them using CollationKey::compareTo().
  81  * <p>
  82  * <strong>Note:</strong> <code>Collator</code>s with different Locale,
  83  * CollationStrength and DecompositionMode settings will return different
  84  * CollationKeys for the same set of strings. Locales have specific
  85  * collation rules, and the way in which secondary and tertiary differences
  86  * are taken into account, for example, will result in different CollationKeys
  87  * for same strings.
  88  * <p>
  89
  90  * @see          Collator
  91  * @see          RuleBasedCollator
  92  * @version      1.3 12/18/96
  93  * @author       Helena Shih
  94  * @deprecated ICU 2.8 Use Collator::getSortKey(...) instead
  95  */
  96 class U_I18N_API CollationKey : public UObject {
  97 public:
  98     /**
  99     * This creates an empty collation key based on the null string.  An empty
 100     * collation key contains no sorting information.  When comparing two empty
 101     * collation keys, the result is Collator::EQUAL.  Comparing empty collation key
 102     * with non-empty collation key is always Collator::LESS.
 103     * @deprecated ICU 2.8 Use Collator::getSortKey(...) instead
 104     */
 105     CollationKey();
 106
 107
 108     /**
 109     * Creates a collation key based on the collation key values.
 110     * @param values the collation key values
 111     * @param count number of collation key values, including trailing nulls.
 112     * @deprecated ICU 2.8 Use Collator::getSortKey(...) instead
 113     */
 114     CollationKey(const  uint8_t*    values,
 115                 int32_t     count);
 116
 117     /**
 118     * Copy constructor.
 119     * @param other    the object to be copied.
 120     * @deprecated ICU 2.8 Use Collator::getSortKey(...) instead
 121     */
 122     CollationKey(const CollationKey& other);
 123
 124     /**
 125     * Sort key destructor.
 126     * @deprecated ICU 2.8 Use Collator::getSortKey(...) instead
 127     */
 128     virtual ~CollationKey();
 129
 130     /**
 131     * Assignment operator
 132     * @param other    the object to be copied.
 133     * @deprecated ICU 2.8 Use Collator::getSortKey(...) instead
 134     */
 135     const   CollationKey&   operator=(const CollationKey& other);
 136
 137     /**
 138     * Compare if two collation keys are the same.
 139     * @param source the collation key to compare to.
 140     * @return Returns true if two collation keys are equal, false otherwise.
 141     * @deprecated ICU 2.8 Use Collator::getSortKey(...) instead
 142     */
 143     UBool                   operator==(const CollationKey& source) const;
 144
 145     /**
 146     * Compare if two collation keys are not the same.
 147     * @param source the collation key to compare to.
 148     * @return Returns TRUE if two collation keys are different, FALSE otherwise.
 149     * @deprecated ICU 2.8 Use Collator::getSortKey(...) instead
 150     */
 151     UBool                   operator!=(const CollationKey& source) const;
 152
 153
 154     /**
 155     * Test to see if the key is in an invalid state. The key will be in an
 156     * invalid state if it couldn't allocate memory for some operation.
 157     * @return Returns TRUE if the key is in an invalid, FALSE otherwise.
 158     * @deprecated ICU 2.8 Use Collator::getSortKey(...) instead
 159     */
 160     UBool                   isBogus(void) const;
 161
 162     /**
 163     * Returns a pointer to the collation key values. The storage is owned
 164     * by the collation key and the pointer will become invalid if the key
 165     * is deleted.
 166     * @param count the output parameter of number of collation key values,
 167     * including any trailing nulls.
 168     * @return a pointer to the collation key values.
 169     * @deprecated ICU 2.8 Use Collator::getSortKey(...) instead
 170     */
 171     const    uint8_t*       getByteArray(int32_t& count) const;
 172
 173 #ifdef U_USE_COLLATION_KEY_DEPRECATES
 174     /**
 175     * Extracts the collation key values into a new array. The caller owns
 176     * this storage and should free it.
 177     * @param count the output parameter of number of collation key values,
 178     * including any trailing nulls.
 179     * @obsolete ICU 2.6. Use getByteArray instead since this API will be removed in that release.
 180     */
 181     uint8_t*                toByteArray(int32_t& count) const;
 182 #endif
 183
 184     /**
 185     * Convenience method which does a string(bit-wise) comparison of the
 186     * two collation keys.
 187     * @param target target collation key to be compared with
 188     * @return Returns Collator::LESS if sourceKey &lt; targetKey,
 189     * Collator::GREATER if sourceKey > targetKey and Collator::EQUAL
 190     * otherwise.
 191     * @deprecated ICU 2.6 use the overload with error code
 192     */
 193     Collator::EComparisonResult compareTo(const CollationKey& target) const;
 194
 195     /**
 196     * Convenience method which does a string(bit-wise) comparison of the
 197     * two collation keys.
 198     * @param target target collation key to be compared with
 199     * @param status error code
 200     * @return Returns UCOL_LESS if sourceKey &lt; targetKey,
 201     * UCOL_GREATER if sourceKey > targetKey and UCOL_EQUAL
 202     * otherwise.
 203     * @deprecated ICU 2.8 Use Collator::getSortKey(...) instead
 204     */
 205     UCollationResult compareTo(const CollationKey& target, UErrorCode &status) const;
 206
 207     /**
 208     * Creates an integer that is unique to the collation key.  NOTE: this
 209     * is not the same as String.hashCode.
 210     * <p>Example of use:
 211     * <pre>
 212     * .    UErrorCode status = U_ZERO_ERROR;
 213     * .    Collator *myCollation = Collator::createInstance(Locale::US, status);
 214     * .    if (U_FAILURE(status)) return;
 215     * .    CollationKey key1, key2;
 216     * .    UErrorCode status1 = U_ZERO_ERROR, status2 = U_ZERO_ERROR;
 217     * .    myCollation->getCollationKey("abc", key1, status1);
 218     * .    if (U_FAILURE(status1)) { delete myCollation; return; }
 219     * .    myCollation->getCollationKey("ABC", key2, status2);
 220     * .    if (U_FAILURE(status2)) { delete myCollation; return; }
 221     * .    // key1.hashCode() != key2.hashCode()
 222     * </pre>
 223     * @return the hash value based on the string's collation order.
 224     * @see UnicodeString#hashCode
 225     * @deprecated ICU 2.8 Use Collator::getSortKey(...) instead
 226     */
 227     int32_t                 hashCode(void) const;
 228
 229     /**
 230      * ICU "poor man's RTTI", returns a UClassID for the actual class.
 231      * @deprecated ICU 2.8 Use Collator::getSortKey(...) instead
 232      */
 233     virtual UClassID getDynamicClassID() const;
 234
 235     /**
 236      * ICU "poor man's RTTI", returns a UClassID for this class.
 237      * @deprecated ICU 2.8 Use Collator::getSortKey(...) instead
 238      */
 239     static UClassID U_EXPORT2 getStaticClassID();
 240
 241 private:
 242     /**
 243     * Returns an array of the collation key values as 16-bit integers.
 244     * The caller owns the storage and must delete it.
 245     * @param values Output param of the collation key values.
 246     * @param count output parameter of the number of collation key values
 247     * @return a pointer to an array of 16-bit collation key values.
 248     */
 249     void adopt(uint8_t *values, int32_t count);
 250
 251     /*
 252     * Creates a collation key with a string.
 253     */
 254
 255     /**
 256     * If this CollationKey has capacity less than newSize,
 257     * its internal capacity will be increased to newSize.
 258     * @param newSize minimum size this CollationKey has to have
 259     * @return this CollationKey
 260     */
 261     CollationKey&           ensureCapacity(int32_t newSize);
 262     /**
 263     * Set the CollationKey to a "bogus" or invalid state
 264     * @return this CollationKey
 265     */
 266     CollationKey&           setToBogus(void);
 267     /**
 268     * Resets this CollationKey to an empty state
 269     * @return this CollationKey
 270     */
 271     CollationKey&           reset(void);
 272
 273     /**
 274     * Allow private access to RuleBasedCollator
 275     */
 276     friend  class           RuleBasedCollator;
 277     /**
 278     * Bogus status
 279     */
 280     UBool                   fBogus;
 281     /**
 282     * Size of fBytes used to store the sortkey. i.e. up till the
 283     * null-termination.
 284     */
 285     int32_t                 fCount;
 286     /**
 287     * Full size of the fBytes
 288     */
 289     int32_t                 fCapacity;
 290     /**
 291     * Unique hash value of this CollationKey
 292     */
 293     int32_t                 fHashCode;
 294     /**
 295     * Array to store the sortkey
 296     */
 297     uint8_t*                fBytes;
 298
 299 };
 300
 301 inline UBool
 302 CollationKey::operator!=(const CollationKey& other) const
 303 {
 304     return !(*this == other);
 305 }
 306
 307 inline UBool
 308 CollationKey::isBogus() const
 309 {
 310     return fBogus;
 311 }
 312
 313 inline const uint8_t*
 314 CollationKey::getByteArray(int32_t &count) const
 315 {
 316     count = fCount;
 317     return fBytes;
 318 }
 319
 320 U_NAMESPACE_END
 321
 322 #endif /* #if !UCONFIG_NO_COLLATION */
 323
 324 #endif