jk-ferret 0.11.8.2
Sign up to get free protection for your applications and to get access to all the features.
- data/CHANGELOG +24 -0
- data/MIT-LICENSE +20 -0
- data/README +90 -0
- data/RELEASE_CHANGES +137 -0
- data/RELEASE_NOTES +60 -0
- data/Rakefile +443 -0
- data/TODO +109 -0
- data/TUTORIAL +231 -0
- data/bin/ferret-browser +79 -0
- data/ext/BZLIB_blocksort.c +1094 -0
- data/ext/BZLIB_bzlib.c +1578 -0
- data/ext/BZLIB_compress.c +672 -0
- data/ext/BZLIB_crctable.c +104 -0
- data/ext/BZLIB_decompress.c +626 -0
- data/ext/BZLIB_huffman.c +205 -0
- data/ext/BZLIB_randtable.c +84 -0
- data/ext/STEMMER_api.c +66 -0
- data/ext/STEMMER_libstemmer.c +93 -0
- data/ext/STEMMER_stem_ISO_8859_1_danish.c +337 -0
- data/ext/STEMMER_stem_ISO_8859_1_dutch.c +624 -0
- data/ext/STEMMER_stem_ISO_8859_1_english.c +1117 -0
- data/ext/STEMMER_stem_ISO_8859_1_finnish.c +762 -0
- data/ext/STEMMER_stem_ISO_8859_1_french.c +1246 -0
- data/ext/STEMMER_stem_ISO_8859_1_german.c +503 -0
- data/ext/STEMMER_stem_ISO_8859_1_hungarian.c +1230 -0
- data/ext/STEMMER_stem_ISO_8859_1_italian.c +1065 -0
- data/ext/STEMMER_stem_ISO_8859_1_norwegian.c +297 -0
- data/ext/STEMMER_stem_ISO_8859_1_porter.c +749 -0
- data/ext/STEMMER_stem_ISO_8859_1_portuguese.c +1017 -0
- data/ext/STEMMER_stem_ISO_8859_1_spanish.c +1093 -0
- data/ext/STEMMER_stem_ISO_8859_1_swedish.c +307 -0
- data/ext/STEMMER_stem_ISO_8859_2_romanian.c +998 -0
- data/ext/STEMMER_stem_KOI8_R_russian.c +700 -0
- data/ext/STEMMER_stem_UTF_8_danish.c +339 -0
- data/ext/STEMMER_stem_UTF_8_dutch.c +634 -0
- data/ext/STEMMER_stem_UTF_8_english.c +1125 -0
- data/ext/STEMMER_stem_UTF_8_finnish.c +768 -0
- data/ext/STEMMER_stem_UTF_8_french.c +1256 -0
- data/ext/STEMMER_stem_UTF_8_german.c +509 -0
- data/ext/STEMMER_stem_UTF_8_hungarian.c +1234 -0
- data/ext/STEMMER_stem_UTF_8_italian.c +1073 -0
- data/ext/STEMMER_stem_UTF_8_norwegian.c +299 -0
- data/ext/STEMMER_stem_UTF_8_porter.c +755 -0
- data/ext/STEMMER_stem_UTF_8_portuguese.c +1023 -0
- data/ext/STEMMER_stem_UTF_8_romanian.c +1004 -0
- data/ext/STEMMER_stem_UTF_8_russian.c +694 -0
- data/ext/STEMMER_stem_UTF_8_spanish.c +1097 -0
- data/ext/STEMMER_stem_UTF_8_swedish.c +309 -0
- data/ext/STEMMER_stem_UTF_8_turkish.c +2205 -0
- data/ext/STEMMER_utilities.c +478 -0
- data/ext/analysis.c +1710 -0
- data/ext/analysis.h +266 -0
- data/ext/api.h +26 -0
- data/ext/array.c +125 -0
- data/ext/array.h +62 -0
- data/ext/bitvector.c +96 -0
- data/ext/bitvector.h +594 -0
- data/ext/bzlib.h +282 -0
- data/ext/bzlib_private.h +503 -0
- data/ext/compound_io.c +384 -0
- data/ext/config.h +52 -0
- data/ext/document.c +159 -0
- data/ext/document.h +63 -0
- data/ext/except.c +102 -0
- data/ext/except.h +176 -0
- data/ext/extconf.rb +15 -0
- data/ext/ferret.c +416 -0
- data/ext/ferret.h +94 -0
- data/ext/field_index.c +262 -0
- data/ext/field_index.h +52 -0
- data/ext/filter.c +157 -0
- data/ext/fs_store.c +493 -0
- data/ext/global.c +458 -0
- data/ext/global.h +302 -0
- data/ext/hash.c +524 -0
- data/ext/hash.h +515 -0
- data/ext/hashset.c +192 -0
- data/ext/hashset.h +215 -0
- data/ext/header.h +58 -0
- data/ext/helper.c +63 -0
- data/ext/helper.h +21 -0
- data/ext/index.c +6804 -0
- data/ext/index.h +935 -0
- data/ext/internal.h +1019 -0
- data/ext/lang.c +10 -0
- data/ext/lang.h +68 -0
- data/ext/libstemmer.h +79 -0
- data/ext/mempool.c +88 -0
- data/ext/mempool.h +43 -0
- data/ext/modules.h +190 -0
- data/ext/multimapper.c +351 -0
- data/ext/multimapper.h +60 -0
- data/ext/posh.c +1006 -0
- data/ext/posh.h +973 -0
- data/ext/priorityqueue.c +149 -0
- data/ext/priorityqueue.h +155 -0
- data/ext/q_boolean.c +1621 -0
- data/ext/q_const_score.c +162 -0
- data/ext/q_filtered_query.c +212 -0
- data/ext/q_fuzzy.c +280 -0
- data/ext/q_match_all.c +149 -0
- data/ext/q_multi_term.c +673 -0
- data/ext/q_parser.c +3103 -0
- data/ext/q_phrase.c +1206 -0
- data/ext/q_prefix.c +98 -0
- data/ext/q_range.c +682 -0
- data/ext/q_span.c +2390 -0
- data/ext/q_term.c +337 -0
- data/ext/q_wildcard.c +167 -0
- data/ext/r_analysis.c +2626 -0
- data/ext/r_index.c +3468 -0
- data/ext/r_qparser.c +635 -0
- data/ext/r_search.c +4490 -0
- data/ext/r_store.c +513 -0
- data/ext/r_utils.c +1131 -0
- data/ext/ram_store.c +476 -0
- data/ext/scanner.c +895 -0
- data/ext/scanner.h +36 -0
- data/ext/scanner_mb.c +6701 -0
- data/ext/scanner_utf8.c +4415 -0
- data/ext/search.c +1864 -0
- data/ext/search.h +953 -0
- data/ext/similarity.c +151 -0
- data/ext/similarity.h +89 -0
- data/ext/sort.c +786 -0
- data/ext/stem_ISO_8859_1_danish.h +16 -0
- data/ext/stem_ISO_8859_1_dutch.h +16 -0
- data/ext/stem_ISO_8859_1_english.h +16 -0
- data/ext/stem_ISO_8859_1_finnish.h +16 -0
- data/ext/stem_ISO_8859_1_french.h +16 -0
- data/ext/stem_ISO_8859_1_german.h +16 -0
- data/ext/stem_ISO_8859_1_hungarian.h +16 -0
- data/ext/stem_ISO_8859_1_italian.h +16 -0
- data/ext/stem_ISO_8859_1_norwegian.h +16 -0
- data/ext/stem_ISO_8859_1_porter.h +16 -0
- data/ext/stem_ISO_8859_1_portuguese.h +16 -0
- data/ext/stem_ISO_8859_1_spanish.h +16 -0
- data/ext/stem_ISO_8859_1_swedish.h +16 -0
- data/ext/stem_ISO_8859_2_romanian.h +16 -0
- data/ext/stem_KOI8_R_russian.h +16 -0
- data/ext/stem_UTF_8_danish.h +16 -0
- data/ext/stem_UTF_8_dutch.h +16 -0
- data/ext/stem_UTF_8_english.h +16 -0
- data/ext/stem_UTF_8_finnish.h +16 -0
- data/ext/stem_UTF_8_french.h +16 -0
- data/ext/stem_UTF_8_german.h +16 -0
- data/ext/stem_UTF_8_hungarian.h +16 -0
- data/ext/stem_UTF_8_italian.h +16 -0
- data/ext/stem_UTF_8_norwegian.h +16 -0
- data/ext/stem_UTF_8_porter.h +16 -0
- data/ext/stem_UTF_8_portuguese.h +16 -0
- data/ext/stem_UTF_8_romanian.h +16 -0
- data/ext/stem_UTF_8_russian.h +16 -0
- data/ext/stem_UTF_8_spanish.h +16 -0
- data/ext/stem_UTF_8_swedish.h +16 -0
- data/ext/stem_UTF_8_turkish.h +16 -0
- data/ext/stopwords.c +410 -0
- data/ext/store.c +698 -0
- data/ext/store.h +799 -0
- data/ext/symbol.c +10 -0
- data/ext/symbol.h +23 -0
- data/ext/term_vectors.c +73 -0
- data/ext/threading.h +31 -0
- data/ext/win32.h +62 -0
- data/lib/ferret.rb +30 -0
- data/lib/ferret/browser.rb +246 -0
- data/lib/ferret/browser/s/global.js +192 -0
- data/lib/ferret/browser/s/style.css +148 -0
- data/lib/ferret/browser/views/document/list.rhtml +49 -0
- data/lib/ferret/browser/views/document/show.rhtml +27 -0
- data/lib/ferret/browser/views/error/index.rhtml +7 -0
- data/lib/ferret/browser/views/help/index.rhtml +8 -0
- data/lib/ferret/browser/views/home/index.rhtml +29 -0
- data/lib/ferret/browser/views/layout.rhtml +22 -0
- data/lib/ferret/browser/views/term-vector/index.rhtml +4 -0
- data/lib/ferret/browser/views/term/index.rhtml +199 -0
- data/lib/ferret/browser/views/term/termdocs.rhtml +1 -0
- data/lib/ferret/browser/webrick.rb +14 -0
- data/lib/ferret/document.rb +130 -0
- data/lib/ferret/field_infos.rb +44 -0
- data/lib/ferret/field_symbol.rb +87 -0
- data/lib/ferret/index.rb +973 -0
- data/lib/ferret/number_tools.rb +157 -0
- data/lib/ferret/version.rb +3 -0
- data/setup.rb +1555 -0
- data/test/long_running/largefile/tc_largefile.rb +46 -0
- data/test/test_all.rb +5 -0
- data/test/test_helper.rb +29 -0
- data/test/test_installed.rb +1 -0
- data/test/threading/number_to_spoken.rb +132 -0
- data/test/threading/thread_safety_index_test.rb +88 -0
- data/test/threading/thread_safety_read_write_test.rb +73 -0
- data/test/threading/thread_safety_test.rb +133 -0
- data/test/unit/analysis/tc_analyzer.rb +550 -0
- data/test/unit/analysis/tc_token_stream.rb +653 -0
- data/test/unit/index/tc_index.rb +867 -0
- data/test/unit/index/tc_index_reader.rb +699 -0
- data/test/unit/index/tc_index_writer.rb +447 -0
- data/test/unit/index/th_doc.rb +332 -0
- data/test/unit/query_parser/tc_query_parser.rb +238 -0
- data/test/unit/search/tc_filter.rb +156 -0
- data/test/unit/search/tc_fuzzy_query.rb +147 -0
- data/test/unit/search/tc_index_searcher.rb +67 -0
- data/test/unit/search/tc_multi_searcher.rb +128 -0
- data/test/unit/search/tc_multiple_search_requests.rb +58 -0
- data/test/unit/search/tc_search_and_sort.rb +179 -0
- data/test/unit/search/tc_sort.rb +49 -0
- data/test/unit/search/tc_sort_field.rb +27 -0
- data/test/unit/search/tc_spans.rb +190 -0
- data/test/unit/search/tm_searcher.rb +436 -0
- data/test/unit/store/tc_fs_store.rb +115 -0
- data/test/unit/store/tc_ram_store.rb +35 -0
- data/test/unit/store/tm_store.rb +34 -0
- data/test/unit/store/tm_store_lock.rb +68 -0
- data/test/unit/tc_document.rb +81 -0
- data/test/unit/tc_field_symbol.rb +26 -0
- data/test/unit/ts_analysis.rb +2 -0
- data/test/unit/ts_index.rb +2 -0
- data/test/unit/ts_largefile.rb +4 -0
- data/test/unit/ts_query_parser.rb +2 -0
- data/test/unit/ts_search.rb +2 -0
- data/test/unit/ts_store.rb +2 -0
- data/test/unit/ts_utils.rb +2 -0
- data/test/unit/utils/tc_bit_vector.rb +295 -0
- data/test/unit/utils/tc_number_tools.rb +117 -0
- data/test/unit/utils/tc_priority_queue.rb +106 -0
- data/test/utils/content_generator.rb +226 -0
- metadata +319 -0
data/ext/store.c
ADDED
@@ -0,0 +1,698 @@
|
|
1
|
+
#include "store.h"
|
2
|
+
#include <string.h>
|
3
|
+
#include "internal.h"
|
4
|
+
|
5
|
+
#define VINT_MAX_LEN 10
|
6
|
+
#define VINT_END BUFFER_SIZE - VINT_MAX_LEN
|
7
|
+
|
8
|
+
/*
|
9
|
+
* TODO: add try finally
|
10
|
+
*/
|
11
|
+
void with_lock(Lock *lock, void (*func)(void *arg), void *arg)
|
12
|
+
{
|
13
|
+
if (!lock->obtain(lock)) {
|
14
|
+
RAISE(IO_ERROR, "couldn't obtain lock \"%s\"", lock->name);
|
15
|
+
}
|
16
|
+
func(arg);
|
17
|
+
lock->release(lock);
|
18
|
+
}
|
19
|
+
|
20
|
+
/*
|
21
|
+
* TODO: add try finally
|
22
|
+
*/
|
23
|
+
void with_lock_name(Store *store, const char *lock_name,
|
24
|
+
void (*func)(void *arg), void *arg)
|
25
|
+
{
|
26
|
+
Lock *lock = store->open_lock_i(store, lock_name);
|
27
|
+
if (!lock->obtain(lock)) {
|
28
|
+
RAISE(LOCK_ERROR, "couldn't obtain lock \"%s\"", lock->name);
|
29
|
+
}
|
30
|
+
func(arg);
|
31
|
+
lock->release(lock);
|
32
|
+
store->close_lock_i(lock);
|
33
|
+
}
|
34
|
+
|
35
|
+
void store_deref(Store *store)
|
36
|
+
{
|
37
|
+
mutex_lock(&store->mutex_i);
|
38
|
+
if (--store->ref_cnt <= 0) {
|
39
|
+
store->close_i(store);
|
40
|
+
}
|
41
|
+
else {
|
42
|
+
mutex_unlock(&store->mutex_i);
|
43
|
+
}
|
44
|
+
}
|
45
|
+
|
46
|
+
Lock *open_lock(Store *store, const char *lockname)
|
47
|
+
{
|
48
|
+
Lock *lock = store->open_lock_i(store, lockname);
|
49
|
+
hs_add(store->locks, lock);
|
50
|
+
return lock;
|
51
|
+
}
|
52
|
+
|
53
|
+
void close_lock(Lock *lock)
|
54
|
+
{
|
55
|
+
hs_del(lock->store->locks, lock);
|
56
|
+
}
|
57
|
+
|
58
|
+
static void close_lock_i(Lock *lock)
|
59
|
+
{
|
60
|
+
lock->store->close_lock_i(lock);
|
61
|
+
}
|
62
|
+
|
63
|
+
/**
|
64
|
+
* Create a store struct initializing the mutex.
|
65
|
+
*/
|
66
|
+
Store *store_new()
|
67
|
+
{
|
68
|
+
Store *store = ALLOC(Store);
|
69
|
+
store->ref_cnt = 1;
|
70
|
+
mutex_init(&store->mutex_i, NULL);
|
71
|
+
mutex_init(&store->mutex, NULL);
|
72
|
+
store->locks = hs_new_ptr((free_ft)&close_lock_i);
|
73
|
+
return store;
|
74
|
+
}
|
75
|
+
|
76
|
+
/**
|
77
|
+
* Destroy the store freeing allocated resources
|
78
|
+
*
|
79
|
+
* @param store the store struct to free
|
80
|
+
*/
|
81
|
+
void store_destroy(Store *store)
|
82
|
+
{
|
83
|
+
mutex_destroy(&store->mutex_i);
|
84
|
+
mutex_destroy(&store->mutex);
|
85
|
+
hs_destroy(store->locks);
|
86
|
+
free(store);
|
87
|
+
}
|
88
|
+
|
89
|
+
/**
|
90
|
+
* Create a newly allocated and initialized OutStream object
|
91
|
+
*
|
92
|
+
* @return a newly allocated and initialized OutStream object
|
93
|
+
*/
|
94
|
+
OutStream *os_new()
|
95
|
+
{
|
96
|
+
OutStream *os = ALLOC(OutStream);
|
97
|
+
os->buf.start = 0;
|
98
|
+
os->buf.pos = 0;
|
99
|
+
os->buf.len = 0;
|
100
|
+
return os;
|
101
|
+
}
|
102
|
+
|
103
|
+
/**
|
104
|
+
* Flush the countents of the OutStream's buffers
|
105
|
+
*
|
106
|
+
* @param the OutStream to flush
|
107
|
+
*/
|
108
|
+
INLINE void os_flush(OutStream *os)
|
109
|
+
{
|
110
|
+
os->m->flush_i(os, os->buf.buf, os->buf.pos);
|
111
|
+
os->buf.start += os->buf.pos;
|
112
|
+
os->buf.pos = 0;
|
113
|
+
}
|
114
|
+
|
115
|
+
void os_close(OutStream *os)
|
116
|
+
{
|
117
|
+
os_flush(os);
|
118
|
+
os->m->close_i(os);
|
119
|
+
free(os);
|
120
|
+
}
|
121
|
+
|
122
|
+
off_t os_pos(OutStream *os)
|
123
|
+
{
|
124
|
+
return os->buf.start + os->buf.pos;
|
125
|
+
}
|
126
|
+
|
127
|
+
void os_seek(OutStream *os, off_t new_pos)
|
128
|
+
{
|
129
|
+
os_flush(os);
|
130
|
+
os->buf.start = new_pos;
|
131
|
+
os->m->seek_i(os, new_pos);
|
132
|
+
}
|
133
|
+
|
134
|
+
/**
|
135
|
+
* Unsafe alternative to os_write_byte. Only use this method if you know there
|
136
|
+
* is no chance of buffer overflow.
|
137
|
+
*/
|
138
|
+
#define write_byte(os, b) os->buf.buf[os->buf.pos++] = (uchar)b
|
139
|
+
|
140
|
+
/**
|
141
|
+
* Write a single byte +b+ to the OutStream +os+
|
142
|
+
*
|
143
|
+
* @param os the OutStream to write to
|
144
|
+
* @param b the byte to write
|
145
|
+
* @raise IO_ERROR if there is an IO error writing to the filesystem
|
146
|
+
*/
|
147
|
+
INLINE void os_write_byte(OutStream *os, uchar b)
|
148
|
+
{
|
149
|
+
if (os->buf.pos >= BUFFER_SIZE) {
|
150
|
+
os_flush(os);
|
151
|
+
}
|
152
|
+
write_byte(os, b);
|
153
|
+
}
|
154
|
+
|
155
|
+
void os_write_bytes(OutStream *os, const uchar *buf, int len)
|
156
|
+
{
|
157
|
+
if (os->buf.pos > 0) { /* flush buffer */
|
158
|
+
os_flush(os);
|
159
|
+
}
|
160
|
+
|
161
|
+
if (len < BUFFER_SIZE) {
|
162
|
+
os->m->flush_i(os, buf, len);
|
163
|
+
os->buf.start += len;
|
164
|
+
}
|
165
|
+
else {
|
166
|
+
int pos = 0;
|
167
|
+
int size;
|
168
|
+
while (pos < len) {
|
169
|
+
if (len - pos < BUFFER_SIZE) {
|
170
|
+
size = len - pos;
|
171
|
+
}
|
172
|
+
else {
|
173
|
+
size = BUFFER_SIZE;
|
174
|
+
}
|
175
|
+
os->m->flush_i(os, buf + pos, size);
|
176
|
+
pos += size;
|
177
|
+
os->buf.start += size;
|
178
|
+
}
|
179
|
+
}
|
180
|
+
}
|
181
|
+
|
182
|
+
/**
|
183
|
+
* Create a newly allocated and initialized InStream
|
184
|
+
*
|
185
|
+
* @return a newly allocated and initialized InStream
|
186
|
+
*/
|
187
|
+
InStream *is_new()
|
188
|
+
{
|
189
|
+
InStream *is = ALLOC(InStream);
|
190
|
+
is->buf.start = 0;
|
191
|
+
is->buf.pos = 0;
|
192
|
+
is->buf.len = 0;
|
193
|
+
is->ref_cnt_ptr = ALLOC_AND_ZERO(int);
|
194
|
+
return is;
|
195
|
+
}
|
196
|
+
|
197
|
+
/**
|
198
|
+
* Refill the InStream's buffer from the store source (filesystem or memory).
|
199
|
+
*
|
200
|
+
* @param is the InStream to refill
|
201
|
+
* @raise IO_ERROR if there is a error reading from the filesystem
|
202
|
+
* @raise EOF_ERROR if there is an attempt to read past the end of the file
|
203
|
+
*/
|
204
|
+
static void is_refill(InStream *is)
|
205
|
+
{
|
206
|
+
off_t start = is->buf.start + is->buf.pos;
|
207
|
+
off_t last = start + BUFFER_SIZE;
|
208
|
+
off_t flen = is->m->length_i(is);
|
209
|
+
|
210
|
+
if (last > flen) { /* don't read past EOF */
|
211
|
+
last = flen;
|
212
|
+
}
|
213
|
+
|
214
|
+
is->buf.len = last - start;
|
215
|
+
if (is->buf.len <= 0) {
|
216
|
+
RAISE(EOF_ERROR, "current pos = %"OFF_T_PFX"d, "
|
217
|
+
"file length = %"OFF_T_PFX"d", start, flen);
|
218
|
+
}
|
219
|
+
|
220
|
+
is->m->read_i(is, is->buf.buf, is->buf.len);
|
221
|
+
|
222
|
+
is->buf.start = start;
|
223
|
+
is->buf.pos = 0;
|
224
|
+
}
|
225
|
+
|
226
|
+
/**
|
227
|
+
* Unsafe alternative to is_read_byte. Only use this method when you know
|
228
|
+
* there is no chance that you will read past the end of the InStream's
|
229
|
+
* buffer.
|
230
|
+
*/
|
231
|
+
#define read_byte(is) is->buf.buf[is->buf.pos++]
|
232
|
+
|
233
|
+
/**
|
234
|
+
* Read a singly byte (unsigned char) from the InStream +is+.
|
235
|
+
*
|
236
|
+
* @param is the Instream to read from
|
237
|
+
* @return a single unsigned char read from the InStream +is+
|
238
|
+
* @raise IO_ERROR if there is a error reading from the filesystem
|
239
|
+
* @raise EOF_ERROR if there is an attempt to read past the end of the file
|
240
|
+
*/
|
241
|
+
INLINE uchar is_read_byte(InStream *is)
|
242
|
+
{
|
243
|
+
if (is->buf.pos >= is->buf.len) {
|
244
|
+
is_refill(is);
|
245
|
+
}
|
246
|
+
|
247
|
+
return read_byte(is);
|
248
|
+
}
|
249
|
+
|
250
|
+
off_t is_pos(InStream *is)
|
251
|
+
{
|
252
|
+
return is->buf.start + is->buf.pos;
|
253
|
+
}
|
254
|
+
|
255
|
+
uchar *is_read_bytes(InStream *is, uchar *buf, int len)
|
256
|
+
{
|
257
|
+
int i;
|
258
|
+
off_t start;
|
259
|
+
|
260
|
+
if ((is->buf.pos + len) < is->buf.len) {
|
261
|
+
for (i = 0; i < len; i++) {
|
262
|
+
buf[i] = read_byte(is);
|
263
|
+
}
|
264
|
+
}
|
265
|
+
else { /* read all-at-once */
|
266
|
+
start = is_pos(is);
|
267
|
+
is->m->seek_i(is, start);
|
268
|
+
is->m->read_i(is, buf, len);
|
269
|
+
|
270
|
+
is->buf.start = start + len; /* adjust stream variables */
|
271
|
+
is->buf.pos = 0;
|
272
|
+
is->buf.len = 0; /* trigger refill on read */
|
273
|
+
}
|
274
|
+
return buf;
|
275
|
+
}
|
276
|
+
|
277
|
+
void is_seek(InStream *is, off_t pos)
|
278
|
+
{
|
279
|
+
if (pos >= is->buf.start && pos < (is->buf.start + is->buf.len)) {
|
280
|
+
is->buf.pos = pos - is->buf.start; /* seek within buffer */
|
281
|
+
}
|
282
|
+
else {
|
283
|
+
is->buf.start = pos;
|
284
|
+
is->buf.pos = 0;
|
285
|
+
is->buf.len = 0; /* trigger refill() on read() */
|
286
|
+
is->m->seek_i(is, pos);
|
287
|
+
}
|
288
|
+
}
|
289
|
+
|
290
|
+
void is_close(InStream *is)
|
291
|
+
{
|
292
|
+
if (--(*(is->ref_cnt_ptr)) < 0) {
|
293
|
+
is->m->close_i(is);
|
294
|
+
free(is->ref_cnt_ptr);
|
295
|
+
}
|
296
|
+
free(is);
|
297
|
+
}
|
298
|
+
|
299
|
+
InStream *is_clone(InStream *is)
|
300
|
+
{
|
301
|
+
InStream *new_index_i = ALLOC(InStream);
|
302
|
+
memcpy(new_index_i, is, sizeof(InStream));
|
303
|
+
(*(new_index_i->ref_cnt_ptr))++;
|
304
|
+
return new_index_i;
|
305
|
+
}
|
306
|
+
|
307
|
+
i32 is_read_i32(InStream *is)
|
308
|
+
{
|
309
|
+
return ((i32)is_read_byte(is) << 24) |
|
310
|
+
((i32)is_read_byte(is) << 16) |
|
311
|
+
((i32)is_read_byte(is) << 8) |
|
312
|
+
((i32)is_read_byte(is));
|
313
|
+
}
|
314
|
+
|
315
|
+
i64 is_read_i64(InStream *is)
|
316
|
+
{
|
317
|
+
return ((i64)is_read_byte(is) << 56) |
|
318
|
+
((i64)is_read_byte(is) << 48) |
|
319
|
+
((i64)is_read_byte(is) << 40) |
|
320
|
+
((i64)is_read_byte(is) << 32) |
|
321
|
+
((i64)is_read_byte(is) << 24) |
|
322
|
+
((i64)is_read_byte(is) << 16) |
|
323
|
+
((i64)is_read_byte(is) << 8) |
|
324
|
+
((i64)is_read_byte(is));
|
325
|
+
}
|
326
|
+
|
327
|
+
u32 is_read_u32(InStream *is)
|
328
|
+
{
|
329
|
+
return ((u32)is_read_byte(is) << 24) |
|
330
|
+
((u32)is_read_byte(is) << 16) |
|
331
|
+
((u32)is_read_byte(is) << 8) |
|
332
|
+
((u32)is_read_byte(is));
|
333
|
+
}
|
334
|
+
|
335
|
+
u64 is_read_u64(InStream *is)
|
336
|
+
{
|
337
|
+
return ((u64)is_read_byte(is) << 56) |
|
338
|
+
((u64)is_read_byte(is) << 48) |
|
339
|
+
((u64)is_read_byte(is) << 40) |
|
340
|
+
((u64)is_read_byte(is) << 32) |
|
341
|
+
((u64)is_read_byte(is) << 24) |
|
342
|
+
((u64)is_read_byte(is) << 16) |
|
343
|
+
((u64)is_read_byte(is) << 8) |
|
344
|
+
((u64)is_read_byte(is));
|
345
|
+
}
|
346
|
+
|
347
|
+
/* optimized to use unchecked read_byte if there is definitely space */
|
348
|
+
INLINE unsigned int is_read_vint(InStream *is)
|
349
|
+
{
|
350
|
+
register unsigned int res, b;
|
351
|
+
register int shift = 7;
|
352
|
+
|
353
|
+
if (is->buf.pos > (is->buf.len - VINT_MAX_LEN)) {
|
354
|
+
b = is_read_byte(is);
|
355
|
+
res = b & 0x7F; /* 0x7F = 0b01111111 */
|
356
|
+
|
357
|
+
while ((b & 0x80) != 0) { /* 0x80 = 0b10000000 */
|
358
|
+
b = is_read_byte(is);
|
359
|
+
res |= (b & 0x7F) << shift;
|
360
|
+
shift += 7;
|
361
|
+
}
|
362
|
+
}
|
363
|
+
else { /* unchecked optimization */
|
364
|
+
b = read_byte(is);
|
365
|
+
res = b & 0x7F; /* 0x7F = 0b01111111 */
|
366
|
+
|
367
|
+
while ((b & 0x80) != 0) { /* 0x80 = 0b10000000 */
|
368
|
+
b = read_byte(is);
|
369
|
+
res |= (b & 0x7F) << shift;
|
370
|
+
shift += 7;
|
371
|
+
}
|
372
|
+
}
|
373
|
+
|
374
|
+
return res;
|
375
|
+
}
|
376
|
+
|
377
|
+
/* optimized to use unchecked read_byte if there is definitely space */
|
378
|
+
INLINE off_t is_read_voff_t(InStream *is)
|
379
|
+
{
|
380
|
+
register off_t res, b;
|
381
|
+
register int shift = 7;
|
382
|
+
|
383
|
+
if (is->buf.pos > (is->buf.len - VINT_MAX_LEN)) {
|
384
|
+
b = is_read_byte(is);
|
385
|
+
res = b & 0x7F; /* 0x7F = 0b01111111 */
|
386
|
+
|
387
|
+
while ((b & 0x80) != 0) { /* 0x80 = 0b10000000 */
|
388
|
+
b = is_read_byte(is);
|
389
|
+
res |= (b & 0x7F) << shift;
|
390
|
+
shift += 7;
|
391
|
+
}
|
392
|
+
}
|
393
|
+
else { /* unchecked optimization */
|
394
|
+
b = read_byte(is);
|
395
|
+
res = b & 0x7F; /* 0x7F = 0b01111111 */
|
396
|
+
|
397
|
+
while ((b & 0x80) != 0) { /* 0x80 = 0b10000000 */
|
398
|
+
b = read_byte(is);
|
399
|
+
res |= (b & 0x7F) << shift;
|
400
|
+
shift += 7;
|
401
|
+
}
|
402
|
+
}
|
403
|
+
|
404
|
+
return res;
|
405
|
+
}
|
406
|
+
|
407
|
+
/* optimized to use unchecked read_byte if there is definitely space */
|
408
|
+
INLINE u64 is_read_vll(InStream *is)
|
409
|
+
{
|
410
|
+
register u64 res, b;
|
411
|
+
register int shift = 7;
|
412
|
+
|
413
|
+
if (is->buf.pos > (is->buf.len - VINT_MAX_LEN)) {
|
414
|
+
b = is_read_byte(is);
|
415
|
+
res = b & 0x7F; /* 0x7F = 0b01111111 */
|
416
|
+
|
417
|
+
while ((b & 0x80) != 0) { /* 0x80 = 0b10000000 */
|
418
|
+
b = is_read_byte(is);
|
419
|
+
res |= (b & 0x7F) << shift;
|
420
|
+
shift += 7;
|
421
|
+
}
|
422
|
+
}
|
423
|
+
else { /* unchecked optimization */
|
424
|
+
b = read_byte(is);
|
425
|
+
res = b & 0x7F; /* 0x7F = 0b01111111 */
|
426
|
+
|
427
|
+
while ((b & 0x80) != 0) { /* 0x80 = 0b10000000 */
|
428
|
+
b = read_byte(is);
|
429
|
+
res |= (b & 0x7F) << shift;
|
430
|
+
shift += 7;
|
431
|
+
}
|
432
|
+
}
|
433
|
+
|
434
|
+
return res;
|
435
|
+
}
|
436
|
+
|
437
|
+
INLINE void is_skip_vints(InStream *is, register int cnt)
|
438
|
+
{
|
439
|
+
for (; cnt > 0; cnt--) {
|
440
|
+
while ((is_read_byte(is) & 0x80) != 0) {
|
441
|
+
}
|
442
|
+
}
|
443
|
+
}
|
444
|
+
|
445
|
+
/*
|
446
|
+
* FIXME: Not used. Do we need/want this?
|
447
|
+
static INLINE void is_read_chars(InStream *is, char *buffer,
|
448
|
+
int off, int len)
|
449
|
+
{
|
450
|
+
int end, i;
|
451
|
+
|
452
|
+
end = off + len;
|
453
|
+
|
454
|
+
for (i = off; i < end; i++) {
|
455
|
+
buffer[i] = is_read_byte(is);
|
456
|
+
}
|
457
|
+
}
|
458
|
+
*/
|
459
|
+
|
460
|
+
char *is_read_string(InStream *is)
|
461
|
+
{
|
462
|
+
register int length = (int) is_read_vint(is);
|
463
|
+
char *str = ALLOC_N(char, length + 1);
|
464
|
+
str[length] = '\0';
|
465
|
+
|
466
|
+
if (is->buf.pos > (is->buf.len - length)) {
|
467
|
+
register int i;
|
468
|
+
for (i = 0; i < length; i++) {
|
469
|
+
str[i] = is_read_byte(is);
|
470
|
+
}
|
471
|
+
}
|
472
|
+
else { /* unchecked optimization */
|
473
|
+
memcpy(str, is->buf.buf + is->buf.pos, length);
|
474
|
+
is->buf.pos += length;
|
475
|
+
}
|
476
|
+
|
477
|
+
return str;
|
478
|
+
}
|
479
|
+
|
480
|
+
char *is_read_string_safe(InStream *is)
|
481
|
+
{
|
482
|
+
register int length = (int) is_read_vint(is);
|
483
|
+
char *str = ALLOC_N(char, length + 1);
|
484
|
+
str[length] = '\0';
|
485
|
+
|
486
|
+
TRY
|
487
|
+
if (is->buf.pos > (is->buf.len - length)) {
|
488
|
+
register int i;
|
489
|
+
for (i = 0; i < length; i++) {
|
490
|
+
str[i] = is_read_byte(is);
|
491
|
+
}
|
492
|
+
}
|
493
|
+
else { /* unchecked optimization */
|
494
|
+
memcpy(str, is->buf.buf + is->buf.pos, length);
|
495
|
+
is->buf.pos += length;
|
496
|
+
}
|
497
|
+
XCATCHALL
|
498
|
+
free(str);
|
499
|
+
XENDTRY
|
500
|
+
|
501
|
+
return str;
|
502
|
+
}
|
503
|
+
|
504
|
+
void os_write_i32(OutStream *os, i32 num)
|
505
|
+
{
|
506
|
+
os_write_byte(os, (uchar)((num >> 24) & 0xFF));
|
507
|
+
os_write_byte(os, (uchar)((num >> 16) & 0xFF));
|
508
|
+
os_write_byte(os, (uchar)((num >> 8) & 0xFF));
|
509
|
+
os_write_byte(os, (uchar)(num & 0xFF));
|
510
|
+
}
|
511
|
+
|
512
|
+
void os_write_i64(OutStream *os, i64 num)
|
513
|
+
{
|
514
|
+
os_write_byte(os, (uchar)((num >> 56) & 0xFF));
|
515
|
+
os_write_byte(os, (uchar)((num >> 48) & 0xFF));
|
516
|
+
os_write_byte(os, (uchar)((num >> 40) & 0xFF));
|
517
|
+
os_write_byte(os, (uchar)((num >> 32) & 0xFF));
|
518
|
+
os_write_byte(os, (uchar)((num >> 24) & 0xFF));
|
519
|
+
os_write_byte(os, (uchar)((num >> 16) & 0xFF));
|
520
|
+
os_write_byte(os, (uchar)((num >> 8) & 0xFF));
|
521
|
+
os_write_byte(os, (uchar)(num & 0xFF));
|
522
|
+
}
|
523
|
+
|
524
|
+
void os_write_u32(OutStream *os, u32 num)
|
525
|
+
{
|
526
|
+
os_write_byte(os, (uchar)((num >> 24) & 0xFF));
|
527
|
+
os_write_byte(os, (uchar)((num >> 16) & 0xFF));
|
528
|
+
os_write_byte(os, (uchar)((num >> 8) & 0xFF));
|
529
|
+
os_write_byte(os, (uchar)(num & 0xFF));
|
530
|
+
}
|
531
|
+
|
532
|
+
void os_write_u64(OutStream *os, u64 num)
|
533
|
+
{
|
534
|
+
os_write_byte(os, (uchar)((num >> 56) & 0xFF));
|
535
|
+
os_write_byte(os, (uchar)((num >> 48) & 0xFF));
|
536
|
+
os_write_byte(os, (uchar)((num >> 40) & 0xFF));
|
537
|
+
os_write_byte(os, (uchar)((num >> 32) & 0xFF));
|
538
|
+
os_write_byte(os, (uchar)((num >> 24) & 0xFF));
|
539
|
+
os_write_byte(os, (uchar)((num >> 16) & 0xFF));
|
540
|
+
os_write_byte(os, (uchar)((num >> 8) & 0xFF));
|
541
|
+
os_write_byte(os, (uchar)(num & 0xFF));
|
542
|
+
}
|
543
|
+
|
544
|
+
/* optimized to use an unchecked write if there is space */
|
545
|
+
INLINE void os_write_vint(OutStream *os, register unsigned int num)
|
546
|
+
{
|
547
|
+
if (os->buf.pos > VINT_END) {
|
548
|
+
while (num > 127) {
|
549
|
+
os_write_byte(os, (uchar)((num & 0x7f) | 0x80));
|
550
|
+
num >>= 7;
|
551
|
+
}
|
552
|
+
os_write_byte(os, (uchar)(num));
|
553
|
+
}
|
554
|
+
else {
|
555
|
+
while (num > 127) {
|
556
|
+
write_byte(os, (uchar)((num & 0x7f) | 0x80));
|
557
|
+
num >>= 7;
|
558
|
+
}
|
559
|
+
write_byte(os, (uchar)(num));
|
560
|
+
}
|
561
|
+
}
|
562
|
+
|
563
|
+
/* optimized to use an unchecked write if there is space */
|
564
|
+
INLINE void os_write_voff_t(OutStream *os, register off_t num)
|
565
|
+
{
|
566
|
+
if (os->buf.pos > VINT_END) {
|
567
|
+
while (num > 127) {
|
568
|
+
os_write_byte(os, (uchar)((num & 0x7f) | 0x80));
|
569
|
+
num >>= 7;
|
570
|
+
}
|
571
|
+
os_write_byte(os, (uchar)num);
|
572
|
+
}
|
573
|
+
else {
|
574
|
+
while (num > 127) {
|
575
|
+
write_byte(os, (uchar)((num & 0x7f) | 0x80));
|
576
|
+
num >>= 7;
|
577
|
+
}
|
578
|
+
write_byte(os, (uchar)num);
|
579
|
+
}
|
580
|
+
}
|
581
|
+
|
582
|
+
/* optimized to use an unchecked write if there is space */
|
583
|
+
INLINE void os_write_vll(OutStream *os, register u64 num)
|
584
|
+
{
|
585
|
+
if (os->buf.pos > VINT_END) {
|
586
|
+
while (num > 127) {
|
587
|
+
os_write_byte(os, (uchar)((num & 0x7f) | 0x80));
|
588
|
+
num >>= 7;
|
589
|
+
}
|
590
|
+
os_write_byte(os, (uchar)num);
|
591
|
+
}
|
592
|
+
else {
|
593
|
+
while (num > 127) {
|
594
|
+
write_byte(os, (uchar)((num & 0x7f) | 0x80));
|
595
|
+
num >>= 7;
|
596
|
+
}
|
597
|
+
write_byte(os, (uchar)num);
|
598
|
+
}
|
599
|
+
}
|
600
|
+
|
601
|
+
INLINE void os_write_string_len(OutStream *os, const char *str, int len)
|
602
|
+
{
|
603
|
+
os_write_vint(os, len);
|
604
|
+
os_write_bytes(os, (uchar *)str, len);
|
605
|
+
}
|
606
|
+
void os_write_string(OutStream *os, const char *str)
|
607
|
+
{
|
608
|
+
os_write_string_len(os, str, (int)strlen(str));
|
609
|
+
}
|
610
|
+
|
611
|
+
/**
|
612
|
+
* Determine if the filename is the name of a lock file. Return 1 if it is, 0
|
613
|
+
* otherwise.
|
614
|
+
*
|
615
|
+
* @param filename the name of the file to check
|
616
|
+
* @return 1 (true) if the file is a lock file, 0 (false) otherwise
|
617
|
+
*/
|
618
|
+
int file_is_lock(const char *filename)
|
619
|
+
{
|
620
|
+
int start = (int) strlen(filename) - 4;
|
621
|
+
return ((start > 0) && (strcmp(LOCK_EXT, &filename[start]) == 0));
|
622
|
+
}
|
623
|
+
|
624
|
+
void is2os_copy_bytes(InStream *is, OutStream *os, int cnt)
|
625
|
+
{
|
626
|
+
int len;
|
627
|
+
uchar buf[BUFFER_SIZE];
|
628
|
+
|
629
|
+
for (; cnt > 0; cnt -= BUFFER_SIZE) {
|
630
|
+
len = ((cnt > BUFFER_SIZE) ? BUFFER_SIZE : cnt);
|
631
|
+
is_read_bytes(is, buf, len);
|
632
|
+
os_write_bytes(os, buf, len);
|
633
|
+
}
|
634
|
+
}
|
635
|
+
|
636
|
+
void is2os_copy_vints(InStream *is, OutStream *os, int cnt)
|
637
|
+
{
|
638
|
+
uchar b;
|
639
|
+
for (; cnt > 0; cnt--) {
|
640
|
+
while (((b = is_read_byte(is)) & 0x80) != 0) {
|
641
|
+
os_write_byte(os, b);
|
642
|
+
}
|
643
|
+
os_write_byte(os, b);
|
644
|
+
}
|
645
|
+
}
|
646
|
+
|
647
|
+
/**
|
648
|
+
* Test argument used to test the store->each function
|
649
|
+
*/
|
650
|
+
struct FileNameListArg
|
651
|
+
{
|
652
|
+
int count;
|
653
|
+
int size;
|
654
|
+
int total_len;
|
655
|
+
char **files;
|
656
|
+
};
|
657
|
+
|
658
|
+
/**
|
659
|
+
* Test function used to test store->each function
|
660
|
+
*/
|
661
|
+
static void add_file_name(const char *fname, void *arg)
|
662
|
+
{
|
663
|
+
struct FileNameListArg *fnl = (struct FileNameListArg *)arg;
|
664
|
+
if (fnl->count >= fnl->size) {
|
665
|
+
fnl->size *= 2;
|
666
|
+
REALLOC_N(fnl->files, char *, fnl->size);
|
667
|
+
}
|
668
|
+
fnl->files[fnl->count++] = estrdup(fname);
|
669
|
+
fnl->total_len += strlen(fname) + 2;
|
670
|
+
}
|
671
|
+
|
672
|
+
char *store_to_s(Store *store)
|
673
|
+
{
|
674
|
+
struct FileNameListArg fnl;
|
675
|
+
char *buf, *b;
|
676
|
+
int i;
|
677
|
+
fnl.count = 0;
|
678
|
+
fnl.size = 16;
|
679
|
+
fnl.total_len = 10;
|
680
|
+
fnl.files = ALLOC_N(char *, 16);
|
681
|
+
|
682
|
+
store->each(store, &add_file_name, &fnl);
|
683
|
+
qsort(fnl.files, fnl.count, sizeof(char *), &scmp);
|
684
|
+
b = buf = ALLOC_N(char, fnl.total_len);
|
685
|
+
|
686
|
+
for (i = 0; i < fnl.count; i++) {
|
687
|
+
char *fn = fnl.files[i];
|
688
|
+
int len = strlen(fn);
|
689
|
+
memcpy(b, fn, len);
|
690
|
+
b += len;
|
691
|
+
*b++ = '\n';
|
692
|
+
free(fn);
|
693
|
+
}
|
694
|
+
*b = '\0';
|
695
|
+
free(fnl.files);
|
696
|
+
|
697
|
+
return buf;
|
698
|
+
}
|