Coverage Report

Created: 2025-08-26 06:17

/src/libzip/lib/zip_hash.c
Line
Count
Source (jump to first uncovered line)
1
/*
2
  zip_hash.c -- hash table string -> uint64
3
  Copyright (C) 2015-2022 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
415k
#define HASH_MULTIPLIER 33
40
26.1k
#define HASH_START 5381
41
42
/* hash table's fill ratio is kept between these by doubling/halfing its size as necessary */
43
14.5k
#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
3.19k
#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
static void
68
5.30k
free_list(zip_hash_entry_t *entry) {
69
11.6k
    while (entry != NULL) {
70
6.31k
        zip_hash_entry_t *next = entry->next;
71
6.31k
        free(entry);
72
6.31k
        entry = next;
73
6.31k
    }
74
5.30k
}
75
76
77
/* compute hash of string, full 32 bit value */
78
static zip_uint32_t
79
26.1k
hash_string(const zip_uint8_t *name) {
80
26.1k
    zip_uint64_t value = HASH_START;
81
82
26.1k
    if (name == NULL) {
83
0
        return 0;
84
0
    }
85
86
441k
    while (*name != 0) {
87
415k
        value = (zip_uint64_t)(((value * HASH_MULTIPLIER) + (zip_uint8_t)*name) % 0x100000000ul);
88
415k
        name++;
89
415k
    }
90
91
26.1k
    return (zip_uint32_t)value;
92
26.1k
}
93
94
95
/* resize hash table; new_size must be a power of 2, can be larger or smaller than current size */
96
static bool
97
3.19k
hash_resize(zip_hash_t *hash, zip_uint32_t new_size, zip_error_t *error) {
98
3.19k
    zip_hash_entry_t **new_table;
99
100
3.19k
    if (new_size == hash->table_size) {
101
0
        return true;
102
0
    }
103
104
3.19k
    if ((new_table = (zip_hash_entry_t **)calloc(new_size, sizeof(zip_hash_entry_t *))) == NULL) {
105
0
        zip_error_set(error, ZIP_ER_MEMORY, 0);
106
0
        return false;
107
0
    }
108
109
3.19k
    if (hash->nentries > 0) {
110
1.30k
        zip_uint32_t i;
111
112
2.63k
        for (i = 0; i < hash->table_size; i++) {
113
1.33k
            zip_hash_entry_t *entry = hash->table[i];
114
2.66k
            while (entry) {
115
1.33k
                zip_hash_entry_t *next = entry->next;
116
117
1.33k
                zip_uint32_t new_index = entry->hash_value % new_size;
118
119
1.33k
                entry->next = new_table[new_index];
120
1.33k
                new_table[new_index] = entry;
121
122
1.33k
                entry = next;
123
1.33k
            }
124
1.33k
        }
125
1.30k
    }
126
127
3.19k
    free(hash->table);
128
3.19k
    hash->table = new_table;
129
3.19k
    hash->table_size = new_size;
130
131
3.19k
    return true;
132
3.19k
}
133
134
135
static zip_uint32_t
136
1.88k
size_for_capacity(zip_uint64_t capacity) {
137
1.88k
    double needed_size = capacity / HASH_MAX_FILL;
138
1.88k
    zip_uint32_t v;
139
140
1.88k
    if (needed_size > ZIP_UINT32_MAX) {
141
0
        v = ZIP_UINT32_MAX;
142
0
    }
143
1.88k
    else {
144
1.88k
        v = (zip_uint32_t)needed_size;
145
1.88k
    }
146
147
1.88k
    if (v > HASH_MAX_SIZE) {
148
0
        return HASH_MAX_SIZE;
149
0
    }
150
151
    /* From Bit Twiddling Hacks by Sean Eron Anderson <seander@cs.stanford.edu>
152
     (http://graphics.stanford.edu/~seander/bithacks.html#RoundUpPowerOf2). */
153
154
1.88k
    v--;
155
1.88k
    v |= v >> 1;
156
1.88k
    v |= v >> 2;
157
1.88k
    v |= v >> 4;
158
1.88k
    v |= v >> 8;
159
1.88k
    v |= v >> 16;
160
1.88k
    v++;
161
162
1.88k
    return v;
163
1.88k
}
164
165
166
zip_hash_t *
167
4.88k
_zip_hash_new(zip_error_t *error) {
168
4.88k
    zip_hash_t *hash;
169
170
4.88k
    if ((hash = (zip_hash_t *)malloc(sizeof(zip_hash_t))) == NULL) {
171
0
        zip_error_set(error, ZIP_ER_MEMORY, 0);
172
0
        return NULL;
173
0
    }
174
175
4.88k
    hash->table_size = 0;
176
4.88k
    hash->nentries = 0;
177
4.88k
    hash->table = NULL;
178
179
4.88k
    return hash;
180
4.88k
}
181
182
183
void
184
4.88k
_zip_hash_free(zip_hash_t *hash) {
185
4.88k
    zip_uint32_t i;
186
187
4.88k
    if (hash == NULL) {
188
0
        return;
189
0
    }
190
191
4.88k
    if (hash->table != NULL) {
192
54.7k
        for (i = 0; i < hash->table_size; i++) {
193
52.8k
            if (hash->table[i] != NULL) {
194
5.30k
                free_list(hash->table[i]);
195
5.30k
            }
196
52.8k
        }
197
1.88k
        free(hash->table);
198
1.88k
    }
199
4.88k
    free(hash);
200
4.88k
}
201
202
203
/* insert into hash, return error on existence or memory issues */
204
bool
205
26.1k
_zip_hash_add(zip_hash_t *hash, const zip_uint8_t *name, zip_uint64_t index, zip_flags_t flags, zip_error_t *error) {
206
26.1k
    zip_uint32_t hash_value, table_index;
207
26.1k
    zip_hash_entry_t *entry;
208
209
26.1k
    if (hash == NULL || name == NULL || index > ZIP_INT64_MAX) {
210
0
        zip_error_set(error, ZIP_ER_INVAL, 0);
211
0
        return false;
212
0
    }
213
214
26.1k
    if (hash->table_size == 0) {
215
0
        if (!hash_resize(hash, HASH_MIN_SIZE, error)) {
216
0
            return false;
217
0
        }
218
0
    }
219
220
26.1k
    hash_value = hash_string(name);
221
26.1k
    table_index = hash_value % hash->table_size;
222
223
27.8k
    for (entry = hash->table[table_index]; entry != NULL; entry = entry->next) {
224
21.5k
        if (entry->hash_value == hash_value && strcmp((const char *)name, (const char *)entry->name) == 0) {
225
19.8k
            if (((flags & ZIP_FL_UNCHANGED) && entry->orig_index != -1) || entry->current_index != -1) {
226
19.8k
                zip_error_set(error, ZIP_ER_EXISTS, 0);
227
19.8k
                return false;
228
19.8k
            }
229
0
            else {
230
0
                break;
231
0
            }
232
19.8k
        }
233
21.5k
    }
234
235
6.31k
    if (entry == NULL) {
236
6.31k
        if ((entry = (zip_hash_entry_t *)malloc(sizeof(zip_hash_entry_t))) == NULL) {
237
0
            zip_error_set(error, ZIP_ER_MEMORY, 0);
238
0
            return false;
239
0
        }
240
6.31k
        entry->name = name;
241
6.31k
        entry->next = hash->table[table_index];
242
6.31k
        hash->table[table_index] = entry;
243
6.31k
        entry->hash_value = hash_value;
244
6.31k
        entry->orig_index = -1;
245
6.31k
        hash->nentries++;
246
6.31k
        if (hash->nentries > hash->table_size * HASH_MAX_FILL && hash->table_size < HASH_MAX_SIZE) {
247
1.30k
            if (!hash_resize(hash, hash->table_size * 2, error)) {
248
0
                return false;
249
0
            }
250
1.30k
        }
251
6.31k
    }
252
253
6.31k
    if (flags & ZIP_FL_UNCHANGED) {
254
6.31k
        entry->orig_index = (zip_int64_t)index;
255
6.31k
    }
256
6.31k
    entry->current_index = (zip_int64_t)index;
257
258
6.31k
    return true;
259
6.31k
}
260
261
262
/* remove entry from hash, error if not found */
263
bool
264
0
_zip_hash_delete(zip_hash_t *hash, const zip_uint8_t *name, zip_error_t *error) {
265
0
    zip_uint32_t hash_value, index;
266
0
    zip_hash_entry_t *entry, *previous;
267
268
0
    if (hash == NULL || name == NULL) {
269
0
        zip_error_set(error, ZIP_ER_INVAL, 0);
270
0
        return false;
271
0
    }
272
273
0
    if (hash->nentries > 0) {
274
0
        hash_value = hash_string(name);
275
0
        index = hash_value % hash->table_size;
276
0
        previous = NULL;
277
0
        entry = hash->table[index];
278
0
        while (entry) {
279
0
            if (entry->hash_value == hash_value && strcmp((const char *)name, (const char *)entry->name) == 0) {
280
0
                if (entry->orig_index == -1) {
281
0
                    if (previous) {
282
0
                        previous->next = entry->next;
283
0
                    }
284
0
                    else {
285
0
                        hash->table[index] = entry->next;
286
0
                    }
287
0
                    free(entry);
288
0
                    hash->nentries--;
289
0
                    if (hash->nentries < hash->table_size * HASH_MIN_FILL && hash->table_size > HASH_MIN_SIZE) {
290
0
                        if (!hash_resize(hash, hash->table_size / 2, error)) {
291
0
                            return false;
292
0
                        }
293
0
                    }
294
0
                }
295
0
                else {
296
0
                    entry->current_index = -1;
297
0
                }
298
0
                return true;
299
0
            }
300
0
            previous = entry;
301
0
            entry = entry->next;
302
0
        }
303
0
    }
304
305
0
    zip_error_set(error, ZIP_ER_NOENT, 0);
306
0
    return false;
307
0
}
308
309
310
/* find value for entry in hash, -1 if not found */
311
zip_int64_t
312
0
_zip_hash_lookup(zip_hash_t *hash, const zip_uint8_t *name, zip_flags_t flags, zip_error_t *error) {
313
0
    zip_uint32_t hash_value, index;
314
0
    zip_hash_entry_t *entry;
315
316
0
    if (hash == NULL || name == NULL) {
317
0
        zip_error_set(error, ZIP_ER_INVAL, 0);
318
0
        return -1;
319
0
    }
320
321
0
    if (hash->nentries > 0) {
322
0
        hash_value = hash_string(name);
323
0
        index = hash_value % hash->table_size;
324
0
        for (entry = hash->table[index]; entry != NULL; entry = entry->next) {
325
0
            if (strcmp((const char *)name, (const char *)entry->name) == 0) {
326
0
                if (flags & ZIP_FL_UNCHANGED) {
327
0
                    if (entry->orig_index != -1) {
328
0
                        return entry->orig_index;
329
0
                    }
330
0
                }
331
0
                else {
332
0
                    if (entry->current_index != -1) {
333
0
                        return entry->current_index;
334
0
                    }
335
0
                }
336
0
                break;
337
0
            }
338
0
        }
339
0
    }
340
341
0
    zip_error_set(error, ZIP_ER_NOENT, 0);
342
0
    return -1;
343
0
}
344
345
346
bool
347
2.18k
_zip_hash_reserve_capacity(zip_hash_t *hash, zip_uint64_t capacity, zip_error_t *error) {
348
2.18k
    zip_uint32_t new_size;
349
350
2.18k
    if (capacity == 0) {
351
301
        return true;
352
301
    }
353
354
1.88k
    new_size = size_for_capacity(capacity);
355
356
1.88k
    if (new_size <= hash->table_size) {
357
0
        return true;
358
0
    }
359
360
1.88k
    if (!hash_resize(hash, new_size, error)) {
361
0
        return false;
362
0
    }
363
364
1.88k
    return true;
365
1.88k
}
366
367
368
bool
369
0
_zip_hash_revert(zip_hash_t *hash, zip_error_t *error) {
370
0
    zip_uint32_t i;
371
0
    zip_hash_entry_t *entry, *previous;
372
373
0
    for (i = 0; i < hash->table_size; i++) {
374
0
        previous = NULL;
375
0
        entry = hash->table[i];
376
0
        while (entry) {
377
0
            if (entry->orig_index == -1) {
378
0
                zip_hash_entry_t *p;
379
0
                if (previous) {
380
0
                    previous->next = entry->next;
381
0
                }
382
0
                else {
383
0
                    hash->table[i] = entry->next;
384
0
                }
385
0
                p = entry;
386
0
                entry = entry->next;
387
                /* previous does not change */
388
0
                free(p);
389
0
                hash->nentries--;
390
0
            }
391
0
            else {
392
0
                entry->current_index = entry->orig_index;
393
0
                previous = entry;
394
0
                entry = entry->next;
395
0
            }
396
0
        }
397
0
    }
398
399
0
    if (hash->nentries < hash->table_size * HASH_MIN_FILL && hash->table_size > HASH_MIN_SIZE) {
400
0
        zip_uint32_t new_size = hash->table_size / 2;
401
0
        while (hash->nentries < new_size * HASH_MIN_FILL && new_size > HASH_MIN_SIZE) {
402
0
            new_size /= 2;
403
0
        }
404
0
        if (!hash_resize(hash, new_size, error)) {
405
0
            return false;
406
0
        }
407
0
    }
408
409
0
    return true;
410
0
}