Coverage Report

Created: 2026-07-30 07:17

next uncovered line (L), next uncovered region (R), next uncovered branch (B)
/src/libzip/lib/zip_hash.c
Line
Count
Source
1
/*
2
  zip_hash.c -- hash table string -> uint64
3
  Copyright (C) 2015-2024 Dieter Baron and Thomas Klausner
4
5
  This file is part of libzip, a library to manipulate ZIP archives.
6
  The authors can be contacted at <info@libzip.org>
7
8
  Redistribution and use in source and binary forms, with or without
9
  modification, are permitted provided that the following conditions
10
  are met:
11
  1. Redistributions of source code must retain the above copyright
12
     notice, this list of conditions and the following disclaimer.
13
  2. Redistributions in binary form must reproduce the above copyright
14
     notice, this list of conditions and the following disclaimer in
15
     the documentation and/or other materials provided with the
16
     distribution.
17
  3. The names of the authors may not be used to endorse or promote
18
     products derived from this software without specific prior
19
     written permission.
20
21
  THIS SOFTWARE IS PROVIDED BY THE AUTHORS ``AS IS'' AND ANY EXPRESS
22
  OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
23
  WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24
  ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY
25
  DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
26
  DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE
27
  GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
28
  INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER
29
  IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
30
  OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN
31
  IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
32
*/
33
34
#include "zipint.h"
35
#include <stdlib.h>
36
#include <string.h>
37
38
/* parameter for the string hash function */
39
274k
#define HASH_MULTIPLIER 33
40
13.6k
#define HASH_START 5381
41
42
/* hash table's fill ratio is kept between these by doubling/halfing its size as necessary */
43
15.8k
#define HASH_MAX_FILL .75
44
0
#define HASH_MIN_FILL .01
45
46
/* but hash table size is kept between these */
47
0
#define HASH_MIN_SIZE 256
48
832
#define HASH_MAX_SIZE 0x80000000ul
49
50
struct zip_hash_entry {
51
    const zip_uint8_t *name;
52
    zip_int64_t orig_index;
53
    zip_int64_t current_index;
54
    struct zip_hash_entry *next;
55
    zip_uint32_t hash_value;
56
};
57
typedef struct zip_hash_entry zip_hash_entry_t;
58
59
struct zip_hash {
60
    zip_uint32_t table_size;
61
    zip_uint64_t nentries;
62
    zip_hash_entry_t **table;
63
};
64
65
66
/* free list of entries */
67
6.20k
static void free_list(zip_hash_entry_t *entry) {
68
13.6k
    while (entry != NULL) {
69
7.48k
        zip_hash_entry_t *next = entry->next;
70
7.48k
        free(entry);
71
7.48k
        entry = next;
72
7.48k
    }
73
6.20k
}
74
75
76
/* compute hash of string, full 32 bit value */
77
13.6k
static zip_uint32_t hash_string(const zip_uint8_t *name) {
78
13.6k
    zip_uint64_t value = HASH_START;
79
80
13.6k
    if (name == NULL) {
81
0
        return 0;
82
0
    }
83
84
287k
    while (*name != 0) {
85
274k
        value = (zip_uint64_t)(((value * HASH_MULTIPLIER) + (zip_uint8_t)*name) % 0x100000000ul);
86
274k
        name++;
87
274k
    }
88
89
13.6k
    return (zip_uint32_t)value;
90
13.6k
}
91
92
93
/* resize hash table; new_size must be a power of 2, can be larger or smaller than current size */
94
832
static bool hash_resize(zip_hash_t *hash, zip_uint32_t new_size, zip_error_t *error) {
95
832
    zip_hash_entry_t **new_table;
96
97
832
    if (new_size == hash->table_size) {
98
0
        return true;
99
0
    }
100
101
832
    if ((new_table = (zip_hash_entry_t **)calloc(new_size, sizeof(zip_hash_entry_t *))) == NULL) {
102
0
        zip_error_set(error, ZIP_ER_MEMORY, 0);
103
0
        return false;
104
0
    }
105
106
832
    if (hash->nentries > 0) {
107
0
        zip_uint32_t i;
108
109
0
        for (i = 0; i < hash->table_size; i++) {
110
0
            zip_hash_entry_t *entry = hash->table[i];
111
0
            while (entry) {
112
0
                zip_hash_entry_t *next = entry->next;
113
114
0
                zip_uint32_t new_index = entry->hash_value % new_size;
115
116
0
                entry->next = new_table[new_index];
117
0
                new_table[new_index] = entry;
118
119
0
                entry = next;
120
0
            }
121
0
        }
122
0
    }
123
124
832
    free(hash->table);
125
832
    hash->table = new_table;
126
832
    hash->table_size = new_size;
127
128
832
    return true;
129
832
}
130
131
132
832
static zip_uint32_t size_for_capacity(zip_uint64_t capacity) {
133
832
    double needed_size = capacity / HASH_MAX_FILL;
134
832
    zip_uint32_t v;
135
136
832
    if (needed_size > ZIP_UINT32_MAX) {
137
0
        v = ZIP_UINT32_MAX;
138
0
    }
139
832
    else {
140
832
        v = (zip_uint32_t)needed_size;
141
832
    }
142
143
832
    if (v > HASH_MAX_SIZE) {
144
0
        return HASH_MAX_SIZE;
145
0
    }
146
147
    /* From Bit Twiddling Hacks by Sean Eron Anderson <seander@cs.stanford.edu>
148
     (http://graphics.stanford.edu/~seander/bithacks.html#RoundUpPowerOf2). */
149
150
832
    v--;
151
832
    v |= v >> 1;
152
832
    v |= v >> 2;
153
832
    v |= v >> 4;
154
832
    v |= v >> 8;
155
832
    v |= v >> 16;
156
832
    v++;
157
158
832
    return v;
159
832
}
160
161
162
1.16k
zip_hash_t *_zip_hash_new(zip_error_t *error) {
163
1.16k
    zip_hash_t *hash;
164
165
1.16k
    if ((hash = (zip_hash_t *)malloc(sizeof(zip_hash_t))) == NULL) {
166
0
        zip_error_set(error, ZIP_ER_MEMORY, 0);
167
0
        return NULL;
168
0
    }
169
170
1.16k
    hash->table_size = 0;
171
1.16k
    hash->nentries = 0;
172
1.16k
    hash->table = NULL;
173
174
1.16k
    return hash;
175
1.16k
}
176
177
178
1.16k
void _zip_hash_free(zip_hash_t *hash) {
179
1.16k
    zip_uint32_t i;
180
181
1.16k
    if (hash == NULL) {
182
0
        return;
183
0
    }
184
185
1.16k
    if (hash->table != NULL) {
186
14.1k
        for (i = 0; i < hash->table_size; i++) {
187
13.3k
            if (hash->table[i] != NULL) {
188
6.20k
                free_list(hash->table[i]);
189
6.20k
            }
190
13.3k
        }
191
832
        free(hash->table);
192
832
    }
193
1.16k
    free(hash);
194
1.16k
}
195
196
197
/* insert into hash, return error on existence or memory issues */
198
7.48k
bool _zip_hash_add(zip_hash_t *hash, const zip_uint8_t *name, zip_uint64_t index, zip_flags_t flags, zip_error_t *error) {
199
7.48k
    zip_uint32_t hash_value, table_index;
200
7.48k
    zip_hash_entry_t *entry;
201
202
7.48k
    if (hash == NULL || name == NULL || index > ZIP_INT64_MAX) {
203
0
        zip_error_set(error, ZIP_ER_INVAL, 0);
204
0
        return false;
205
0
    }
206
207
7.48k
    if (hash->table_size == 0) {
208
0
        if (!hash_resize(hash, HASH_MIN_SIZE, error)) {
209
0
            return false;
210
0
        }
211
0
    }
212
213
7.48k
    hash_value = hash_string(name);
214
7.48k
    table_index = hash_value % hash->table_size;
215
216
8.79k
    for (entry = hash->table[table_index]; entry != NULL; entry = entry->next) {
217
1.30k
        if (entry->hash_value == hash_value && strcmp((const char *)name, (const char *)entry->name) == 0) {
218
2
            if (((flags & ZIP_FL_UNCHANGED) && entry->orig_index != -1) || entry->current_index != -1) {
219
2
                zip_error_set(error, ZIP_ER_EXISTS, 0);
220
2
                return false;
221
2
            }
222
0
            else {
223
0
                break;
224
0
            }
225
2
        }
226
1.30k
    }
227
228
7.48k
    if (entry == NULL) {
229
7.48k
        if ((entry = (zip_hash_entry_t *)malloc(sizeof(zip_hash_entry_t))) == NULL) {
230
0
            zip_error_set(error, ZIP_ER_MEMORY, 0);
231
0
            return false;
232
0
        }
233
7.48k
        entry->name = name;
234
7.48k
        entry->next = hash->table[table_index];
235
7.48k
        hash->table[table_index] = entry;
236
7.48k
        entry->hash_value = hash_value;
237
7.48k
        entry->orig_index = -1;
238
7.48k
        hash->nentries++;
239
7.48k
        if (hash->nentries > hash->table_size * HASH_MAX_FILL && hash->table_size < HASH_MAX_SIZE) {
240
0
            if (!hash_resize(hash, hash->table_size * 2, error)) {
241
0
                return false;
242
0
            }
243
0
        }
244
7.48k
    }
245
246
7.48k
    if (flags & ZIP_FL_UNCHANGED) {
247
7.48k
        entry->orig_index = (zip_int64_t)index;
248
7.48k
    }
249
7.48k
    entry->current_index = (zip_int64_t)index;
250
251
7.48k
    return true;
252
7.48k
}
253
254
255
/* remove entry from hash, error if not found */
256
0
bool _zip_hash_delete(zip_hash_t *hash, const zip_uint8_t *name, zip_error_t *error) {
257
0
    zip_uint32_t hash_value, index;
258
0
    zip_hash_entry_t *entry, *previous;
259
260
0
    if (hash == NULL || name == NULL) {
261
0
        zip_error_set(error, ZIP_ER_INVAL, 0);
262
0
        return false;
263
0
    }
264
265
0
    if (hash->nentries > 0) {
266
0
        hash_value = hash_string(name);
267
0
        index = hash_value % hash->table_size;
268
0
        previous = NULL;
269
0
        entry = hash->table[index];
270
0
        while (entry) {
271
0
            if (entry->hash_value == hash_value && strcmp((const char *)name, (const char *)entry->name) == 0) {
272
0
                if (entry->orig_index == -1) {
273
0
                    if (previous) {
274
0
                        previous->next = entry->next;
275
0
                    }
276
0
                    else {
277
0
                        hash->table[index] = entry->next;
278
0
                    }
279
0
                    free(entry);
280
0
                    hash->nentries--;
281
0
                    if (hash->nentries < hash->table_size * HASH_MIN_FILL && hash->table_size > HASH_MIN_SIZE) {
282
0
                        if (!hash_resize(hash, hash->table_size / 2, error)) {
283
0
                            return false;
284
0
                        }
285
0
                    }
286
0
                }
287
0
                else {
288
0
                    entry->current_index = -1;
289
0
                }
290
0
                return true;
291
0
            }
292
0
            previous = entry;
293
0
            entry = entry->next;
294
0
        }
295
0
    }
296
297
0
    zip_error_set(error, ZIP_ER_NOENT, 0);
298
0
    return false;
299
0
}
300
301
302
/* find value for entry in hash, -1 if not found */
303
6.15k
zip_int64_t _zip_hash_lookup(zip_hash_t *hash, const zip_uint8_t *name, zip_flags_t flags, zip_error_t *error) {
304
6.15k
    zip_uint32_t hash_value, index;
305
6.15k
    zip_hash_entry_t *entry;
306
307
6.15k
    if (hash == NULL || name == NULL) {
308
0
        zip_error_set(error, ZIP_ER_INVAL, 0);
309
0
        return -1;
310
0
    }
311
312
6.15k
    if (hash->nentries > 0) {
313
6.15k
        hash_value = hash_string(name);
314
6.15k
        index = hash_value % hash->table_size;
315
9.93k
        for (entry = hash->table[index]; entry != NULL; entry = entry->next) {
316
5.72k
            if (strcmp((const char *)name, (const char *)entry->name) == 0) {
317
1.94k
                if (flags & ZIP_FL_UNCHANGED) {
318
1.94k
                    if (entry->orig_index != -1) {
319
1.94k
                        return entry->orig_index;
320
1.94k
                    }
321
1.94k
                }
322
0
                else {
323
0
                    if (entry->current_index != -1) {
324
0
                        return entry->current_index;
325
0
                    }
326
0
                }
327
0
                break;
328
1.94k
            }
329
5.72k
        }
330
6.15k
    }
331
332
4.21k
    zip_error_set(error, ZIP_ER_NOENT, 0);
333
4.21k
    return -1;
334
6.15k
}
335
336
337
832
bool _zip_hash_reserve_capacity(zip_hash_t *hash, zip_uint64_t capacity, zip_error_t *error) {
338
832
    zip_uint32_t new_size;
339
340
832
    if (capacity == 0) {
341
0
        return true;
342
0
    }
343
344
832
    new_size = size_for_capacity(capacity);
345
346
832
    if (new_size <= hash->table_size) {
347
0
        return true;
348
0
    }
349
350
832
    if (!hash_resize(hash, new_size, error)) {
351
0
        return false;
352
0
    }
353
354
832
    return true;
355
832
}
356
357
358
0
bool _zip_hash_revert(zip_hash_t *hash, zip_error_t *error) {
359
0
    zip_uint32_t i;
360
0
    zip_hash_entry_t *entry, *previous;
361
362
0
    for (i = 0; i < hash->table_size; i++) {
363
0
        previous = NULL;
364
0
        entry = hash->table[i];
365
0
        while (entry) {
366
0
            if (entry->orig_index == -1) {
367
0
                zip_hash_entry_t *p;
368
0
                if (previous) {
369
0
                    previous->next = entry->next;
370
0
                }
371
0
                else {
372
0
                    hash->table[i] = entry->next;
373
0
                }
374
0
                p = entry;
375
0
                entry = entry->next;
376
                /* previous does not change */
377
0
                free(p);
378
0
                hash->nentries--;
379
0
            }
380
0
            else {
381
0
                entry->current_index = entry->orig_index;
382
0
                previous = entry;
383
0
                entry = entry->next;
384
0
            }
385
0
        }
386
0
    }
387
388
0
    if (hash->nentries < hash->table_size * HASH_MIN_FILL && hash->table_size > HASH_MIN_SIZE) {
389
0
        zip_uint32_t new_size = hash->table_size / 2;
390
0
        while (hash->nentries < new_size * HASH_MIN_FILL && new_size > HASH_MIN_SIZE) {
391
0
            new_size /= 2;
392
0
        }
393
0
        if (!hash_resize(hash, new_size, error)) {
394
0
            return false;
395
0
        }
396
0
    }
397
398
0
    return true;
399
0
}