summaryrefslogtreecommitdiffstats
path: root/intl/icu/source/i18n/sortkey.cpp
blob: f6f1873d14d764df12876a42f44e8165994238d9 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
// Copyright (C) 2016 and later: Unicode, Inc. and others.
// License & terms of use: http://www.unicode.org/copyright.html
/*
*******************************************************************************
* Copyright (C) 1996-2012, International Business Machines Corporation and
* others. All Rights Reserved.
*******************************************************************************
*/
//===============================================================================
//
// File sortkey.cpp
//
//
//
// Created by: Helena Shih
//
// Modification History:
//
//  Date         Name          Description
//
//  6/20/97      helena        Java class name change.
//  6/23/97      helena        Added comments to make code more readable.
//  6/26/98      erm           Canged to use byte arrays instead of UnicodeString
//  7/31/98      erm           hashCode: minimum inc should be 2 not 1,
//                             Cleaned up operator=
// 07/12/99      helena        HPUX 11 CC port.
// 03/06/01      synwee        Modified compareTo, to handle the result of
//                             2 string similar in contents, but one is longer
//                             than the other
//===============================================================================

#include "unicode/utypes.h"

#if !UCONFIG_NO_COLLATION

#include "unicode/sortkey.h"
#include "cmemory.h"
#include "uelement.h"
#include "ustr_imp.h"

U_NAMESPACE_BEGIN

// A hash code of kInvalidHashCode indicates that the hash code needs
// to be computed. A hash code of kEmptyHashCode is used for empty keys
// and for any key whose computed hash code is kInvalidHashCode.
static const int32_t kInvalidHashCode = 0;
static const int32_t kEmptyHashCode = 1;
// The "bogus hash code" replaces a separate fBogus flag.
static const int32_t kBogusHashCode = 2;

UOBJECT_DEFINE_RTTI_IMPLEMENTATION(CollationKey)

CollationKey::CollationKey()
    : UObject(), fFlagAndLength(0),
      fHashCode(kEmptyHashCode)
{
}

// Create a collation key from a bit array.
CollationKey::CollationKey(const uint8_t* newValues, int32_t count)
    : UObject(), fFlagAndLength(count),
      fHashCode(kInvalidHashCode)
{
    if (count < 0 || (newValues == NULL && count != 0) ||
            (count > getCapacity() && reallocate(count, 0) == NULL)) {
        setToBogus();
        return;
    }

    if (count > 0) {
        uprv_memcpy(getBytes(), newValues, count);
    }
}

CollationKey::CollationKey(const CollationKey& other)
    : UObject(other), fFlagAndLength(other.getLength()),
      fHashCode(other.fHashCode)
{
    if (other.isBogus())
    {
        setToBogus();
        return;
    }

    int32_t length = fFlagAndLength;
    if (length > getCapacity() && reallocate(length, 0) == NULL) {
        setToBogus();
        return;
    }

    if (length > 0) {
        uprv_memcpy(getBytes(), other.getBytes(), length);
    }
}

CollationKey::~CollationKey()
{
    if(fFlagAndLength < 0) { uprv_free(fUnion.fFields.fBytes); }
}

uint8_t *CollationKey::reallocate(int32_t newCapacity, int32_t length) {
    uint8_t *newBytes = static_cast<uint8_t *>(uprv_malloc(newCapacity));
    if(newBytes == NULL) { return NULL; }
    if(length > 0) {
        uprv_memcpy(newBytes, getBytes(), length);
    }
    if(fFlagAndLength < 0) { uprv_free(fUnion.fFields.fBytes); }
    fUnion.fFields.fBytes = newBytes;
    fUnion.fFields.fCapacity = newCapacity;
    fFlagAndLength |= 0x80000000;
    return newBytes;
}

void CollationKey::setLength(int32_t newLength) {
    // U_ASSERT(newLength >= 0 && newLength <= getCapacity());
    fFlagAndLength = (fFlagAndLength & 0x80000000) | newLength;
    fHashCode = kInvalidHashCode;
}

// set the key to an empty state
CollationKey&
CollationKey::reset()
{
    fFlagAndLength &= 0x80000000;
    fHashCode = kEmptyHashCode;

    return *this;
}

// set the key to a "bogus" or invalid state
CollationKey&
CollationKey::setToBogus()
{
    fFlagAndLength &= 0x80000000;
    fHashCode = kBogusHashCode;

    return *this;
}

UBool
CollationKey::operator==(const CollationKey& source) const
{
    return getLength() == source.getLength() &&
            (this == &source ||
             uprv_memcmp(getBytes(), source.getBytes(), getLength()) == 0);
}

const CollationKey&
CollationKey::operator=(const CollationKey& other)
{
    if (this != &other)
    {
        if (other.isBogus())
        {
            return setToBogus();
        }

        int32_t length = other.getLength();
        if (length > getCapacity() && reallocate(length, 0) == NULL) {
            return setToBogus();
        }
        if (length > 0) {
            uprv_memcpy(getBytes(), other.getBytes(), length);
        }
        fFlagAndLength = (fFlagAndLength & 0x80000000) | length;
        fHashCode = other.fHashCode;
    }

    return *this;
}

// Bitwise comparison for the collation keys.
Collator::EComparisonResult
CollationKey::compareTo(const CollationKey& target) const
{
    UErrorCode errorCode = U_ZERO_ERROR;
    return static_cast<Collator::EComparisonResult>(compareTo(target, errorCode));
}

// Bitwise comparison for the collation keys.
UCollationResult
CollationKey::compareTo(const CollationKey& target, UErrorCode &status) const
{
  if(U_SUCCESS(status)) {
    const uint8_t *src = getBytes();
    const uint8_t *tgt = target.getBytes();

    // are we comparing the same string
    if (src == tgt)
        return  UCOL_EQUAL;

    UCollationResult result;

    // are we comparing different lengths?
    int32_t minLength = getLength();
    int32_t targetLength = target.getLength();
    if (minLength < targetLength) {
        result = UCOL_LESS;
    } else if (minLength == targetLength) {
        result = UCOL_EQUAL;
    } else {
        minLength = targetLength;
        result = UCOL_GREATER;
    }

    if (minLength > 0) {
        int diff = uprv_memcmp(src, tgt, minLength);
        if (diff > 0) {
            return UCOL_GREATER;
        }
        else
            if (diff < 0) {
                return UCOL_LESS;
            }
    }

    return result;
  } else {
    return UCOL_EQUAL;
  }
}

#ifdef U_USE_COLLATION_KEY_DEPRECATES
// Create a copy of the byte array.
uint8_t*
CollationKey::toByteArray(int32_t& count) const
{
    uint8_t *result = (uint8_t*) uprv_malloc( sizeof(uint8_t) * fCount );

    if (result == NULL)
    {
        count = 0;
    }
    else
    {
        count = fCount;
        if (count > 0) {
            uprv_memcpy(result, fBytes, fCount);
        }
    }

    return result;
}
#endif

static int32_t
computeHashCode(const uint8_t *key, int32_t  length) {
    const char *s = reinterpret_cast<const char *>(key);
    int32_t hash;
    if (s == NULL || length == 0) {
        hash = kEmptyHashCode;
    } else {
        hash = ustr_hashCharsN(s, length);
        if (hash == kInvalidHashCode || hash == kBogusHashCode) {
            hash = kEmptyHashCode;
        }
    }
    return hash;
}

int32_t
CollationKey::hashCode() const
{
    // (Cribbed from UnicodeString)
    // We cache the hashCode; when it becomes invalid, due to any change to the
    // string, we note this by setting it to kInvalidHashCode. [LIU]

    // Note: This method is semantically const, but physically non-const.

    if (fHashCode == kInvalidHashCode)
    {
        fHashCode = computeHashCode(getBytes(), getLength());
    }

    return fHashCode;
}

U_NAMESPACE_END

U_CAPI int32_t U_EXPORT2
ucol_keyHashCode(const uint8_t *key, 
                       int32_t  length)
{
    return icu::computeHashCode(key, length);
}

#endif /* #if !UCONFIG_NO_COLLATION */