sdsykes-ferret 0.11.6.19
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 +102 -0
- data/Rakefile +338 -0
- data/TODO +17 -0
- data/TUTORIAL +231 -0
- data/bin/ferret-browser +79 -0
- data/ext/analysis.c +1555 -0
- data/ext/analysis.h +219 -0
- data/ext/api.c +69 -0
- data/ext/api.h +27 -0
- data/ext/array.c +123 -0
- data/ext/array.h +53 -0
- data/ext/bitvector.c +540 -0
- data/ext/bitvector.h +272 -0
- data/ext/compound_io.c +383 -0
- data/ext/config.h +42 -0
- data/ext/document.c +156 -0
- data/ext/document.h +53 -0
- data/ext/except.c +120 -0
- data/ext/except.h +168 -0
- data/ext/extconf.rb +14 -0
- data/ext/ferret.c +402 -0
- data/ext/ferret.h +91 -0
- data/ext/filter.c +156 -0
- data/ext/fs_store.c +483 -0
- data/ext/global.c +418 -0
- data/ext/global.h +117 -0
- data/ext/hash.c +567 -0
- data/ext/hash.h +473 -0
- data/ext/hashset.c +170 -0
- data/ext/hashset.h +187 -0
- data/ext/header.h +58 -0
- data/ext/helper.c +62 -0
- data/ext/helper.h +13 -0
- data/ext/inc/lang.h +48 -0
- data/ext/inc/threading.h +31 -0
- data/ext/index.c +6425 -0
- data/ext/index.h +961 -0
- data/ext/lang.h +66 -0
- data/ext/libstemmer.c +92 -0
- data/ext/libstemmer.h +79 -0
- data/ext/mempool.c +87 -0
- data/ext/mempool.h +35 -0
- data/ext/modules.h +162 -0
- data/ext/multimapper.c +310 -0
- data/ext/multimapper.h +51 -0
- data/ext/posh.c +1006 -0
- data/ext/posh.h +1007 -0
- data/ext/priorityqueue.c +151 -0
- data/ext/priorityqueue.h +143 -0
- data/ext/q_boolean.c +1608 -0
- data/ext/q_const_score.c +161 -0
- data/ext/q_filtered_query.c +209 -0
- data/ext/q_fuzzy.c +268 -0
- data/ext/q_match_all.c +148 -0
- data/ext/q_multi_term.c +677 -0
- data/ext/q_parser.c +2825 -0
- data/ext/q_phrase.c +1126 -0
- data/ext/q_prefix.c +100 -0
- data/ext/q_range.c +350 -0
- data/ext/q_span.c +2402 -0
- data/ext/q_term.c +337 -0
- data/ext/q_wildcard.c +171 -0
- data/ext/r_analysis.c +2575 -0
- data/ext/r_index.c +3472 -0
- data/ext/r_qparser.c +585 -0
- data/ext/r_search.c +4105 -0
- data/ext/r_store.c +513 -0
- data/ext/r_utils.c +963 -0
- data/ext/ram_store.c +471 -0
- data/ext/search.c +1741 -0
- data/ext/search.h +885 -0
- data/ext/similarity.c +150 -0
- data/ext/similarity.h +82 -0
- data/ext/sort.c +983 -0
- data/ext/stem_ISO_8859_1_danish.c +338 -0
- data/ext/stem_ISO_8859_1_danish.h +16 -0
- data/ext/stem_ISO_8859_1_dutch.c +635 -0
- data/ext/stem_ISO_8859_1_dutch.h +16 -0
- data/ext/stem_ISO_8859_1_english.c +1156 -0
- data/ext/stem_ISO_8859_1_english.h +16 -0
- data/ext/stem_ISO_8859_1_finnish.c +792 -0
- data/ext/stem_ISO_8859_1_finnish.h +16 -0
- data/ext/stem_ISO_8859_1_french.c +1276 -0
- data/ext/stem_ISO_8859_1_french.h +16 -0
- data/ext/stem_ISO_8859_1_german.c +512 -0
- data/ext/stem_ISO_8859_1_german.h +16 -0
- data/ext/stem_ISO_8859_1_italian.c +1091 -0
- data/ext/stem_ISO_8859_1_italian.h +16 -0
- data/ext/stem_ISO_8859_1_norwegian.c +296 -0
- data/ext/stem_ISO_8859_1_norwegian.h +16 -0
- data/ext/stem_ISO_8859_1_porter.c +776 -0
- data/ext/stem_ISO_8859_1_porter.h +16 -0
- data/ext/stem_ISO_8859_1_portuguese.c +1035 -0
- data/ext/stem_ISO_8859_1_portuguese.h +16 -0
- data/ext/stem_ISO_8859_1_spanish.c +1119 -0
- data/ext/stem_ISO_8859_1_spanish.h +16 -0
- data/ext/stem_ISO_8859_1_swedish.c +307 -0
- data/ext/stem_ISO_8859_1_swedish.h +16 -0
- data/ext/stem_KOI8_R_russian.c +701 -0
- data/ext/stem_KOI8_R_russian.h +16 -0
- data/ext/stem_UTF_8_danish.c +344 -0
- data/ext/stem_UTF_8_danish.h +16 -0
- data/ext/stem_UTF_8_dutch.c +653 -0
- data/ext/stem_UTF_8_dutch.h +16 -0
- data/ext/stem_UTF_8_english.c +1176 -0
- data/ext/stem_UTF_8_english.h +16 -0
- data/ext/stem_UTF_8_finnish.c +808 -0
- data/ext/stem_UTF_8_finnish.h +16 -0
- data/ext/stem_UTF_8_french.c +1296 -0
- data/ext/stem_UTF_8_french.h +16 -0
- data/ext/stem_UTF_8_german.c +526 -0
- data/ext/stem_UTF_8_german.h +16 -0
- data/ext/stem_UTF_8_italian.c +1113 -0
- data/ext/stem_UTF_8_italian.h +16 -0
- data/ext/stem_UTF_8_norwegian.c +302 -0
- data/ext/stem_UTF_8_norwegian.h +16 -0
- data/ext/stem_UTF_8_porter.c +794 -0
- data/ext/stem_UTF_8_porter.h +16 -0
- data/ext/stem_UTF_8_portuguese.c +1055 -0
- data/ext/stem_UTF_8_portuguese.h +16 -0
- data/ext/stem_UTF_8_russian.c +709 -0
- data/ext/stem_UTF_8_russian.h +16 -0
- data/ext/stem_UTF_8_spanish.c +1137 -0
- data/ext/stem_UTF_8_spanish.h +16 -0
- data/ext/stem_UTF_8_swedish.c +313 -0
- data/ext/stem_UTF_8_swedish.h +16 -0
- data/ext/stopwords.c +401 -0
- data/ext/store.c +692 -0
- data/ext/store.h +777 -0
- data/ext/term_vectors.c +352 -0
- data/ext/threading.h +31 -0
- data/ext/utilities.c +446 -0
- data/ext/win32.h +54 -0
- data/lib/ferret.rb +29 -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/index.rb +786 -0
- data/lib/ferret/number_tools.rb +157 -0
- data/lib/ferret_version.rb +3 -0
- data/setup.rb +1555 -0
- data/test/test_all.rb +5 -0
- data/test/test_helper.rb +24 -0
- data/test/threading/number_to_spoken.rb +132 -0
- data/test/threading/thread_safety_index_test.rb +79 -0
- data/test/threading/thread_safety_read_write_test.rb +76 -0
- data/test/threading/thread_safety_test.rb +133 -0
- data/test/unit/analysis/tc_analyzer.rb +548 -0
- data/test/unit/analysis/tc_token_stream.rb +646 -0
- data/test/unit/index/tc_index.rb +762 -0
- data/test/unit/index/tc_index_reader.rb +699 -0
- data/test/unit/index/tc_index_writer.rb +437 -0
- data/test/unit/index/th_doc.rb +315 -0
- data/test/unit/largefile/tc_largefile.rb +46 -0
- data/test/unit/query_parser/tc_query_parser.rb +238 -0
- data/test/unit/search/tc_filter.rb +135 -0
- data/test/unit/search/tc_fuzzy_query.rb +147 -0
- data/test/unit/search/tc_index_searcher.rb +61 -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 +384 -0
- data/test/unit/store/tc_fs_store.rb +77 -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/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
- metadata +285 -0
data/ext/bitvector.h
ADDED
@@ -0,0 +1,272 @@
|
|
1
|
+
#ifndef FRT_BIT_VECTOR_H
|
2
|
+
#define FRT_BIT_VECTOR_H
|
3
|
+
|
4
|
+
#include "global.h"
|
5
|
+
|
6
|
+
#define BV_INIT_CAPA 256
|
7
|
+
typedef struct BitVector
|
8
|
+
{
|
9
|
+
/** The bits are held in an array of 32-bit integers */
|
10
|
+
f_u32 *bits;
|
11
|
+
|
12
|
+
/** size is equal to 1 + the highest order bit set */
|
13
|
+
int size;
|
14
|
+
|
15
|
+
/** capa is the number of words (U32) allocated for the bits */
|
16
|
+
int capa;
|
17
|
+
|
18
|
+
/** count is the running count of bits set. This is kept up to date by
|
19
|
+
*bv_set and bv_unset. You can reset this value by calling bv_recount */
|
20
|
+
int count;
|
21
|
+
|
22
|
+
/** curr_bit is used by scan_next to record the previously scanned bit */
|
23
|
+
int curr_bit;
|
24
|
+
|
25
|
+
bool extends_as_ones : 1;
|
26
|
+
int ref_cnt;
|
27
|
+
} BitVector;
|
28
|
+
|
29
|
+
/**
|
30
|
+
* Create a new BitVector with a capacity of +BV_INIT_CAPA+. Note that the
|
31
|
+
* BitVector is growable and will adjust it's capacity when you use bv_set.
|
32
|
+
*
|
33
|
+
* @return BitVector with a capacity of +BV_INIT_CAPA+.
|
34
|
+
*/
|
35
|
+
extern BitVector *bv_new();
|
36
|
+
|
37
|
+
/**
|
38
|
+
* Create a new BitVector with a capacity of +capa+. Note that the BitVector
|
39
|
+
* is growable and will adjust it's capacity when you use bv_set.
|
40
|
+
*
|
41
|
+
* @param capa the initial capacity of the BitVector
|
42
|
+
* @return BitVector with a capacity of +capa+.
|
43
|
+
*/
|
44
|
+
extern BitVector *bv_new_capa(int capa);
|
45
|
+
|
46
|
+
/**
|
47
|
+
* Destroy a BitVector, freeing all memory allocated to that BitVector
|
48
|
+
*
|
49
|
+
* @param bv BitVector to destroy
|
50
|
+
*/
|
51
|
+
extern void bv_destroy(BitVector *bv);
|
52
|
+
|
53
|
+
/**
|
54
|
+
* Set the bit at position +index+. If +index+ is outside of the range of the
|
55
|
+
* BitVector, that is >= BitVector.size, BitVector.size will be set to +index+
|
56
|
+
* + 1. If it is greater than the capacity of the BitVector, the capacity will
|
57
|
+
* be expanded to accomodate.
|
58
|
+
*
|
59
|
+
* @param bv the BitVector to set the bit in
|
60
|
+
* @param index the index of the bit to set
|
61
|
+
*/
|
62
|
+
extern void bv_set(BitVector *bv, int index);
|
63
|
+
|
64
|
+
/**
|
65
|
+
* Unsafely set the bit at position +index+. If you choose to use this
|
66
|
+
* function you must create the BitVector with a large enough capacity to
|
67
|
+
* accomodate all of the bv_set_fast operations. You must also set bits in
|
68
|
+
* order and only one time per bit. Otherwise, use the safe bv_set function.
|
69
|
+
*
|
70
|
+
* So this is ok;
|
71
|
+
* <pre>
|
72
|
+
* BitVector *bv = bv_new_capa(1000);
|
73
|
+
* bv_set_fast(bv, 900);
|
74
|
+
* bv_set_fast(bv, 920);
|
75
|
+
* bv_set_fast(bv, 999);
|
76
|
+
* </pre>
|
77
|
+
*
|
78
|
+
* While these are not ok;
|
79
|
+
* <pre>
|
80
|
+
* BitVector *bv = bv_new_capa(90);
|
81
|
+
* bv_set_fast(bv, 80);
|
82
|
+
* bv_set_fast(bv, 79); // <= Bad: Out of Order
|
83
|
+
* bv_set_fast(bv, 80); // <= Bad: Already set
|
84
|
+
* bv_set_fast(bv, 90); // <= Bad: Out of Range. index must be < capa
|
85
|
+
* </pre>
|
86
|
+
*
|
87
|
+
* @param bv the BitVector to set the bit in
|
88
|
+
* @param index the index of the bit to set
|
89
|
+
*/
|
90
|
+
extern void bv_set_fast(BitVector *bv, int bit);
|
91
|
+
|
92
|
+
/**
|
93
|
+
* Return 1 if the bit at +index+ was set or 0 otherwise. If +index+ is out of
|
94
|
+
* range, that is greater then the BitVectors capacity, it will also return 0.
|
95
|
+
*
|
96
|
+
* @param bv the BitVector to check in
|
97
|
+
* @param index the index of the bit to check
|
98
|
+
* @return 1 if the bit was set, 0 otherwise
|
99
|
+
*/
|
100
|
+
extern int bv_get(BitVector *bv, int index);
|
101
|
+
|
102
|
+
/**
|
103
|
+
* Unset the bit at position +index+. If the +index+ was out of range, that is
|
104
|
+
* greater than the BitVectors capacity then do nothing. (bv_get will return 0
|
105
|
+
* in this case anyway).
|
106
|
+
*
|
107
|
+
* @param bv the BitVector to unset the bit in
|
108
|
+
* @param index the index of the bit to unset
|
109
|
+
*/
|
110
|
+
extern void bv_unset(BitVector *bv, int bit);
|
111
|
+
|
112
|
+
/**
|
113
|
+
* Clear all set bits. This function will set all set bits to 0.
|
114
|
+
*
|
115
|
+
* @param bv the BitVector to clear
|
116
|
+
*/
|
117
|
+
extern void bv_clear(BitVector *bv);
|
118
|
+
|
119
|
+
/**
|
120
|
+
* Resets the set bit count by running through the whole BitVector and
|
121
|
+
* counting all set bits. A running count of the bits is kept by bv_set,
|
122
|
+
*bv_get and bv_set_fast so this function is only necessary if the count could
|
123
|
+
* have been corrupted somehow or if the BitVector has been constructed in a
|
124
|
+
* different way (for example being read from the file_system).
|
125
|
+
*
|
126
|
+
* @param bv the BitVector to count the bits in
|
127
|
+
* @return the number of set bits in the BitVector. BitVector.count is also
|
128
|
+
* set
|
129
|
+
*/
|
130
|
+
extern int bv_recount(BitVector *bv);
|
131
|
+
|
132
|
+
/**
|
133
|
+
* Reset the BitVector for scanning. This function should be called before
|
134
|
+
* using bv_scan_next to scan through all set bits in the BitVector. This is
|
135
|
+
* not necessary when using bv_scan_next_from.
|
136
|
+
*
|
137
|
+
* @param bv the BitVector to reset for scanning
|
138
|
+
*/
|
139
|
+
extern void bv_scan_reset(BitVector *bv);
|
140
|
+
|
141
|
+
/**
|
142
|
+
* Scan the BitVector for the next set bit. Before using this function you
|
143
|
+
* should reset the BitVector for scanning using +bv_scan_reset+. You can the
|
144
|
+
* repeated call bv_scan_next to get each set bit until it finally returns
|
145
|
+
* -1.
|
146
|
+
*
|
147
|
+
* @param bv the BitVector to scan
|
148
|
+
* @return the next set bits index or -1 if no more bits are set
|
149
|
+
*/
|
150
|
+
extern int bv_scan_next(BitVector *bv);
|
151
|
+
|
152
|
+
/**
|
153
|
+
* Scan the BitVector for the next set bit after +from+. If no more bits are
|
154
|
+
* set then return -1, otherwise return the index of teh next set bit.
|
155
|
+
*
|
156
|
+
* @param bv the BitVector to scan
|
157
|
+
* @return the next set bit's index or -1 if no more bits are set
|
158
|
+
*/
|
159
|
+
|
160
|
+
extern int bv_scan_next_from(BitVector *bv, register const int from);
|
161
|
+
/**
|
162
|
+
* Scan the BitVector for the next unset bit. Before using this function you
|
163
|
+
* should reset the BitVector for scanning using +bv_scan_reset+. You can the
|
164
|
+
* repeated call bv_scan_next to get each unset bit until it finally returns
|
165
|
+
* -1.
|
166
|
+
*
|
167
|
+
* @param bv the BitVector to scan
|
168
|
+
* @return the next unset bits index or -1 if no more bits are unset
|
169
|
+
*/
|
170
|
+
extern int bv_scan_next_unset(BitVector *bv);
|
171
|
+
|
172
|
+
/**
|
173
|
+
* Scan the BitVector for the next unset bit after +from+. If no more bits are
|
174
|
+
* unset then return -1, otherwise return the index of teh next unset bit.
|
175
|
+
*
|
176
|
+
* @param bv the BitVector to scan
|
177
|
+
* @return the next unset bit's index or -1 if no more bits are unset
|
178
|
+
*/
|
179
|
+
extern int bv_scan_next_unset_from(BitVector *bv, register const int from);
|
180
|
+
|
181
|
+
/**
|
182
|
+
* Check whether the two BitVectors have the same bits set.
|
183
|
+
*
|
184
|
+
* @param bv1 first BitVector to compare
|
185
|
+
* @param bv2 second BitVectors to compare
|
186
|
+
* @return true if bv1 == bv2
|
187
|
+
*/
|
188
|
+
extern int bv_eq(BitVector *bv1, BitVector *bv2);
|
189
|
+
|
190
|
+
/**
|
191
|
+
* Determines a hash value for the BitVector
|
192
|
+
*
|
193
|
+
* @param bv the BitVector to hash
|
194
|
+
* @return A hash value for the BitVector
|
195
|
+
*/
|
196
|
+
extern unsigned long bv_hash(BitVector *bv);
|
197
|
+
|
198
|
+
/**
|
199
|
+
* ANDs two BitVectors (+bv1+ and +bv2+) together and return the resultant
|
200
|
+
* BitVector
|
201
|
+
*
|
202
|
+
* @param bv1 first BitVector to AND
|
203
|
+
* @param bv2 second BitVector to AND
|
204
|
+
* @return A BitVector with all bits set that are set in both bv1 and bv2
|
205
|
+
*/
|
206
|
+
extern BitVector *bv_and(BitVector *bv1, BitVector *bv2);
|
207
|
+
|
208
|
+
/**
|
209
|
+
* ORs two BitVectors (+bv1+ and +bv2+) together and return the resultant
|
210
|
+
* BitVector
|
211
|
+
*
|
212
|
+
* @param bv1 first BitVector to OR
|
213
|
+
* @param bv2 second BitVector to OR
|
214
|
+
* @return A BitVector with all bits set that are set in both bv1 and bv2
|
215
|
+
*/
|
216
|
+
extern BitVector *bv_or(BitVector *bv1, BitVector *bv2);
|
217
|
+
|
218
|
+
/**
|
219
|
+
* XORs two BitVectors (+bv1+ and +bv2+) together and return the resultant
|
220
|
+
* BitVector
|
221
|
+
*
|
222
|
+
* @param bv1 first BitVector to XOR
|
223
|
+
* @param bv2 second BitVector to XOR
|
224
|
+
* @return A BitVector with all bits set that are equal in bv1 and bv2
|
225
|
+
*/
|
226
|
+
extern BitVector *bv_xor(BitVector *bv1, BitVector *bv2);
|
227
|
+
|
228
|
+
/**
|
229
|
+
* Returns BitVector with all of +bv+'s bits flipped
|
230
|
+
*
|
231
|
+
* @param bv BitVector to flip
|
232
|
+
* @return A BitVector with all bits set that are set in both bv1 and bv2
|
233
|
+
*/
|
234
|
+
extern BitVector *bv_not(BitVector *bv);
|
235
|
+
|
236
|
+
/**
|
237
|
+
* ANDs two BitVectors together +bv1+ and +bv2+ in place of +bv1+
|
238
|
+
*
|
239
|
+
* @param bv1 first BitVector to AND
|
240
|
+
* @param bv2 second BitVector to AND
|
241
|
+
* @return A BitVector
|
242
|
+
* @return bv1 with all bits set that where set in both bv1 and bv2
|
243
|
+
*/
|
244
|
+
extern BitVector *bv_and_x(BitVector *bv1, BitVector *bv2);
|
245
|
+
|
246
|
+
/**
|
247
|
+
* ORs two BitVectors together
|
248
|
+
*
|
249
|
+
* @param bv1 first BitVector to OR
|
250
|
+
* @param bv2 second BitVector to OR
|
251
|
+
* @return bv1
|
252
|
+
*/
|
253
|
+
extern BitVector *bv_or_x(BitVector *bv1, BitVector *bv2);
|
254
|
+
|
255
|
+
/**
|
256
|
+
* XORs two BitVectors together +bv1+ and +bv2+ in place of +bv1+
|
257
|
+
*
|
258
|
+
* @param bv1 first BitVector to XOR
|
259
|
+
* @param bv2 second BitVector to XOR
|
260
|
+
* @return bv1
|
261
|
+
*/
|
262
|
+
extern BitVector *bv_xor_x(BitVector *bv1, BitVector *bv2);
|
263
|
+
|
264
|
+
/**
|
265
|
+
* Flips all bits in the BitVector +bv+
|
266
|
+
*
|
267
|
+
* @param bv BitVector to flip
|
268
|
+
* @return A +bv+ with all it's bits flipped
|
269
|
+
*/
|
270
|
+
extern BitVector *bv_not_x(BitVector *bv);
|
271
|
+
|
272
|
+
#endif
|
data/ext/compound_io.c
ADDED
@@ -0,0 +1,383 @@
|
|
1
|
+
#include "index.h"
|
2
|
+
#include "array.h"
|
3
|
+
|
4
|
+
extern void store_destroy(Store *store);
|
5
|
+
extern InStream *is_new();
|
6
|
+
extern Store *store_new();
|
7
|
+
|
8
|
+
/****************************************************************************
|
9
|
+
*
|
10
|
+
* CompoundStore
|
11
|
+
*
|
12
|
+
****************************************************************************/
|
13
|
+
|
14
|
+
typedef struct FileEntry {
|
15
|
+
off_t offset;
|
16
|
+
off_t length;
|
17
|
+
} FileEntry;
|
18
|
+
|
19
|
+
static void cmpd_touch(Store *store, char *file_name)
|
20
|
+
{
|
21
|
+
store->dir.cmpd->store->touch(store->dir.cmpd->store, file_name);
|
22
|
+
}
|
23
|
+
|
24
|
+
static int cmpd_exists(Store *store, char *file_name)
|
25
|
+
{
|
26
|
+
if (h_get(store->dir.cmpd->entries, file_name) != NULL) {
|
27
|
+
return true;
|
28
|
+
}
|
29
|
+
else {
|
30
|
+
return false;
|
31
|
+
}
|
32
|
+
}
|
33
|
+
|
34
|
+
/**
|
35
|
+
* @throws UNSUPPORTED_ERROR
|
36
|
+
*/
|
37
|
+
static int cmpd_remove(Store *store, char *file_name)
|
38
|
+
{
|
39
|
+
(void)store;
|
40
|
+
(void)file_name;
|
41
|
+
RAISE(UNSUPPORTED_ERROR, UNSUPPORTED_ERROR_MSG);
|
42
|
+
return 0;
|
43
|
+
}
|
44
|
+
|
45
|
+
/**
|
46
|
+
* @throws UNSUPPORTED_ERROR
|
47
|
+
*/
|
48
|
+
static void cmpd_rename(Store *store, char *from, char *to)
|
49
|
+
{
|
50
|
+
(void)store;
|
51
|
+
(void)from;
|
52
|
+
(void)to;
|
53
|
+
RAISE(UNSUPPORTED_ERROR, UNSUPPORTED_ERROR_MSG);
|
54
|
+
}
|
55
|
+
|
56
|
+
static int cmpd_count(Store *store)
|
57
|
+
{
|
58
|
+
return store->dir.cmpd->entries->size;
|
59
|
+
}
|
60
|
+
|
61
|
+
static void cmpd_each(Store *store,
|
62
|
+
void (*func)(char *fname, void *arg), void *arg)
|
63
|
+
{
|
64
|
+
HashTable *ht = store->dir.cmpd->entries;
|
65
|
+
int i;
|
66
|
+
for (i = 0; i <= ht->mask; i++) {
|
67
|
+
char *fn = (char *)ht->table[i].key;
|
68
|
+
if (fn) {
|
69
|
+
func(fn, arg);
|
70
|
+
}
|
71
|
+
}
|
72
|
+
}
|
73
|
+
|
74
|
+
|
75
|
+
/**
|
76
|
+
* @throws UNSUPPORTED_ERROR
|
77
|
+
*/
|
78
|
+
static void cmpd_clear(Store *store)
|
79
|
+
{
|
80
|
+
(void)store;
|
81
|
+
RAISE(UNSUPPORTED_ERROR, UNSUPPORTED_ERROR_MSG);
|
82
|
+
}
|
83
|
+
|
84
|
+
static void cmpd_close_i(Store *store)
|
85
|
+
{
|
86
|
+
CompoundStore *cmpd = store->dir.cmpd;
|
87
|
+
if (cmpd->stream == NULL) {
|
88
|
+
RAISE(IO_ERROR, "Tried to close already closed compound store");
|
89
|
+
}
|
90
|
+
|
91
|
+
h_destroy(cmpd->entries);
|
92
|
+
|
93
|
+
is_close(cmpd->stream);
|
94
|
+
cmpd->stream = NULL;
|
95
|
+
free(store->dir.cmpd);
|
96
|
+
store_destroy(store);
|
97
|
+
}
|
98
|
+
|
99
|
+
static off_t cmpd_length(Store *store, char *file_name)
|
100
|
+
{
|
101
|
+
FileEntry *fe = h_get(store->dir.cmpd->entries, file_name);
|
102
|
+
if (fe != NULL) {
|
103
|
+
return fe->length;
|
104
|
+
}
|
105
|
+
else {
|
106
|
+
return 0;
|
107
|
+
}
|
108
|
+
}
|
109
|
+
|
110
|
+
static void cmpdi_seek_i(InStream *is, off_t pos)
|
111
|
+
{
|
112
|
+
(void)is;
|
113
|
+
(void)pos;
|
114
|
+
}
|
115
|
+
|
116
|
+
static void cmpdi_close_i(InStream *is)
|
117
|
+
{
|
118
|
+
free(is->d.cis);
|
119
|
+
}
|
120
|
+
|
121
|
+
static off_t cmpdi_length_i(InStream *is)
|
122
|
+
{
|
123
|
+
return (is->d.cis->length);
|
124
|
+
}
|
125
|
+
|
126
|
+
/*
|
127
|
+
* raises: EOF_ERROR
|
128
|
+
*/
|
129
|
+
static void cmpdi_read_i(InStream *is, uchar *b, int len)
|
130
|
+
{
|
131
|
+
CompoundInStream *cis = is->d.cis;
|
132
|
+
off_t start = is_pos(is);
|
133
|
+
|
134
|
+
if ((start + len) > cis->length) {
|
135
|
+
RAISE(EOF_ERROR, "Tried to read past end of file. File length is "
|
136
|
+
"<%"F_OFF_T_PFX"d> and tried to read to <%"F_OFF_T_PFX"d>",
|
137
|
+
cis->length, start + len);
|
138
|
+
}
|
139
|
+
|
140
|
+
is_seek(cis->sub, cis->offset + start);
|
141
|
+
is_read_bytes(cis->sub, b, len);
|
142
|
+
}
|
143
|
+
|
144
|
+
static const struct InStreamMethods CMPD_IN_STREAM_METHODS = {
|
145
|
+
cmpdi_read_i,
|
146
|
+
cmpdi_seek_i,
|
147
|
+
cmpdi_length_i,
|
148
|
+
cmpdi_close_i
|
149
|
+
};
|
150
|
+
|
151
|
+
static InStream *cmpd_create_input(InStream *sub_is, off_t offset, off_t length)
|
152
|
+
{
|
153
|
+
InStream *is = is_new();
|
154
|
+
CompoundInStream *cis = ALLOC(CompoundInStream);
|
155
|
+
|
156
|
+
cis->sub = sub_is;
|
157
|
+
cis->offset = offset;
|
158
|
+
cis->length = length;
|
159
|
+
is->d.cis = cis;
|
160
|
+
is->m = &CMPD_IN_STREAM_METHODS;
|
161
|
+
|
162
|
+
return is;
|
163
|
+
}
|
164
|
+
|
165
|
+
static InStream *cmpd_open_input(Store *store, const char *file_name)
|
166
|
+
{
|
167
|
+
FileEntry *entry;
|
168
|
+
CompoundStore *cmpd = store->dir.cmpd;
|
169
|
+
InStream *is;
|
170
|
+
|
171
|
+
mutex_lock(&store->mutex);
|
172
|
+
if (cmpd->stream == NULL) {
|
173
|
+
mutex_unlock(&store->mutex);
|
174
|
+
RAISE(IO_ERROR, "Can't open compound file input stream. Parent "
|
175
|
+
"stream is closed.");
|
176
|
+
}
|
177
|
+
|
178
|
+
entry = h_get(cmpd->entries, file_name);
|
179
|
+
if (entry == NULL) {
|
180
|
+
mutex_unlock(&store->mutex);
|
181
|
+
RAISE(IO_ERROR, "File %s does not exist: ", file_name);
|
182
|
+
}
|
183
|
+
|
184
|
+
is = cmpd_create_input(cmpd->stream, entry->offset, entry->length);
|
185
|
+
mutex_unlock(&store->mutex);
|
186
|
+
|
187
|
+
return is;
|
188
|
+
}
|
189
|
+
|
190
|
+
static OutStream *cmpd_new_output(Store *store, const char *file_name)
|
191
|
+
{
|
192
|
+
(void)store;
|
193
|
+
(void)file_name;
|
194
|
+
RAISE(UNSUPPORTED_ERROR, UNSUPPORTED_ERROR_MSG);
|
195
|
+
return NULL;
|
196
|
+
}
|
197
|
+
|
198
|
+
static Lock *cmpd_open_lock_i(Store *store, char *lock_name)
|
199
|
+
{
|
200
|
+
(void)store;
|
201
|
+
(void)lock_name;
|
202
|
+
RAISE(UNSUPPORTED_ERROR, UNSUPPORTED_ERROR_MSG);
|
203
|
+
return NULL;
|
204
|
+
}
|
205
|
+
|
206
|
+
static void cmpd_close_lock_i(Lock *lock)
|
207
|
+
{
|
208
|
+
(void)lock;
|
209
|
+
RAISE(UNSUPPORTED_ERROR, UNSUPPORTED_ERROR_MSG);
|
210
|
+
}
|
211
|
+
|
212
|
+
Store *open_cmpd_store(Store *store, const char *name)
|
213
|
+
{
|
214
|
+
int count, i;
|
215
|
+
off_t offset;
|
216
|
+
char *fname;
|
217
|
+
FileEntry *entry = NULL;
|
218
|
+
Store *new_store = NULL;
|
219
|
+
CompoundStore *volatile cmpd = NULL;
|
220
|
+
InStream *volatile is = NULL;
|
221
|
+
|
222
|
+
TRY
|
223
|
+
cmpd = ALLOC_AND_ZERO(CompoundStore);
|
224
|
+
|
225
|
+
cmpd->store = store;
|
226
|
+
cmpd->name = name;
|
227
|
+
cmpd->entries = h_new_str(&free, &free);
|
228
|
+
is = cmpd->stream = store->open_input(store, cmpd->name);
|
229
|
+
|
230
|
+
/* read the directory and init files */
|
231
|
+
count = is_read_vint(is);
|
232
|
+
entry = NULL;
|
233
|
+
for (i = 0; i < count; i++) {
|
234
|
+
offset = (off_t)is_read_i64(is);
|
235
|
+
fname = is_read_string(is);
|
236
|
+
|
237
|
+
if (entry != NULL) {
|
238
|
+
/* set length of the previous entry */
|
239
|
+
entry->length = offset - entry->offset;
|
240
|
+
}
|
241
|
+
|
242
|
+
entry = ALLOC(FileEntry);
|
243
|
+
entry->offset = offset;
|
244
|
+
h_set(cmpd->entries, fname, entry);
|
245
|
+
}
|
246
|
+
XCATCHALL
|
247
|
+
if (is) is_close(is);
|
248
|
+
if (cmpd->entries) h_destroy(cmpd->entries);
|
249
|
+
free(cmpd);
|
250
|
+
XENDTRY
|
251
|
+
|
252
|
+
/* set the length of the final entry */
|
253
|
+
if (entry != NULL) {
|
254
|
+
entry->length = is_length(is) - entry->offset;
|
255
|
+
}
|
256
|
+
|
257
|
+
new_store = store_new();
|
258
|
+
new_store->dir.cmpd = cmpd;
|
259
|
+
new_store->touch = &cmpd_touch;
|
260
|
+
new_store->exists = &cmpd_exists;
|
261
|
+
new_store->remove = &cmpd_remove;
|
262
|
+
new_store->rename = &cmpd_rename;
|
263
|
+
new_store->count = &cmpd_count;
|
264
|
+
new_store->clear = &cmpd_clear;
|
265
|
+
new_store->length = &cmpd_length;
|
266
|
+
new_store->each = &cmpd_each;
|
267
|
+
new_store->close_i = &cmpd_close_i;
|
268
|
+
new_store->new_output = &cmpd_new_output;
|
269
|
+
new_store->open_input = &cmpd_open_input;
|
270
|
+
new_store->open_lock_i = &cmpd_open_lock_i;
|
271
|
+
new_store->close_lock_i = &cmpd_close_lock_i;
|
272
|
+
|
273
|
+
return new_store;
|
274
|
+
}
|
275
|
+
|
276
|
+
/****************************************************************************
|
277
|
+
*
|
278
|
+
* CompoundWriter
|
279
|
+
*
|
280
|
+
****************************************************************************/
|
281
|
+
|
282
|
+
CompoundWriter *open_cw(Store *store, char *name)
|
283
|
+
{
|
284
|
+
CompoundWriter *cw = ALLOC(CompoundWriter);
|
285
|
+
cw->store = store;
|
286
|
+
cw->name = name;
|
287
|
+
cw->ids = hs_new_str(&free);
|
288
|
+
cw->file_entries = ary_new_type_capa(CWFileEntry, CW_INIT_CAPA);
|
289
|
+
return cw;
|
290
|
+
}
|
291
|
+
|
292
|
+
void cw_add_file(CompoundWriter *cw, char *id)
|
293
|
+
{
|
294
|
+
id = estrdup(id);
|
295
|
+
if (hs_add(cw->ids, id) != HASH_KEY_DOES_NOT_EXIST) {
|
296
|
+
RAISE(IO_ERROR, "Tried to add file \"%s\" which has already been "
|
297
|
+
"added to the compound store", id);
|
298
|
+
}
|
299
|
+
|
300
|
+
ary_grow(cw->file_entries);
|
301
|
+
ary_last(cw->file_entries).name = id;
|
302
|
+
}
|
303
|
+
|
304
|
+
static void cw_copy_file(CompoundWriter *cw, CWFileEntry *src, OutStream *os)
|
305
|
+
{
|
306
|
+
off_t start_ptr = os_pos(os);
|
307
|
+
off_t end_ptr;
|
308
|
+
off_t remainder, length, len;
|
309
|
+
uchar buffer[BUFFER_SIZE];
|
310
|
+
|
311
|
+
InStream *is = cw->store->open_input(cw->store, src->name);
|
312
|
+
|
313
|
+
remainder = length = is_length(is);
|
314
|
+
|
315
|
+
while (remainder > 0) {
|
316
|
+
len = MIN(remainder, BUFFER_SIZE);
|
317
|
+
is_read_bytes(is, buffer, len);
|
318
|
+
os_write_bytes(os, buffer, len);
|
319
|
+
remainder -= len;
|
320
|
+
}
|
321
|
+
|
322
|
+
/* Verify that remainder is 0 */
|
323
|
+
if (remainder != 0) {
|
324
|
+
RAISE(IO_ERROR, "There seems to be an error in the compound file "
|
325
|
+
"should have read to the end but there are <%"F_OFF_T_PFX"d> "
|
326
|
+
"bytes left", remainder);
|
327
|
+
}
|
328
|
+
|
329
|
+
/* Verify that the output length diff is equal to original file */
|
330
|
+
end_ptr = os_pos(os);
|
331
|
+
len = end_ptr - start_ptr;
|
332
|
+
if (len != length) {
|
333
|
+
RAISE(IO_ERROR, "Difference in compound file output file offsets "
|
334
|
+
"<%"F_OFF_T_PFX"d> does not match the original file lenght "
|
335
|
+
"<%"F_OFF_T_PFX"d>", len, length);
|
336
|
+
}
|
337
|
+
|
338
|
+
is_close(is);
|
339
|
+
}
|
340
|
+
|
341
|
+
void cw_close(CompoundWriter *cw)
|
342
|
+
{
|
343
|
+
OutStream *os = NULL;
|
344
|
+
int i;
|
345
|
+
|
346
|
+
if (cw->ids->size <= 0) {
|
347
|
+
RAISE(STATE_ERROR, "Tried to merge compound file with no entries");
|
348
|
+
}
|
349
|
+
|
350
|
+
os = cw->store->new_output(cw->store, cw->name);
|
351
|
+
|
352
|
+
os_write_vint(os, ary_size(cw->file_entries));
|
353
|
+
|
354
|
+
/* Write the directory with all offsets at 0.
|
355
|
+
* Remember the positions of directory entries so that we can adjust the
|
356
|
+
* offsets later */
|
357
|
+
for (i = 0; i < ary_size(cw->file_entries); i++) {
|
358
|
+
cw->file_entries[i].dir_offset = os_pos(os);
|
359
|
+
os_write_u64(os, 0); /* for now */
|
360
|
+
os_write_string(os, cw->file_entries[i].name);
|
361
|
+
}
|
362
|
+
|
363
|
+
/* Open the files and copy their data into the stream. Remember the
|
364
|
+
* locations of each file's data section. */
|
365
|
+
for (i = 0; i < ary_size(cw->file_entries); i++) {
|
366
|
+
cw->file_entries[i].data_offset = os_pos(os);
|
367
|
+
cw_copy_file(cw, &cw->file_entries[i], os);
|
368
|
+
}
|
369
|
+
|
370
|
+
/* Write the data offsets into the directory of the compound stream */
|
371
|
+
for (i = 0; i < ary_size(cw->file_entries); i++) {
|
372
|
+
os_seek(os, cw->file_entries[i].dir_offset);
|
373
|
+
os_write_u64(os, cw->file_entries[i].data_offset);
|
374
|
+
}
|
375
|
+
|
376
|
+
if (os) {
|
377
|
+
os_close(os);
|
378
|
+
}
|
379
|
+
|
380
|
+
hs_destroy(cw->ids);
|
381
|
+
ary_free(cw->file_entries);
|
382
|
+
free(cw);
|
383
|
+
}
|