ferret 0.3.2 → 0.9.0
Sign up to get free protection for your applications and to get access to all the features.
- data/CHANGELOG +9 -0
- data/Rakefile +51 -25
- data/ext/analysis.c +553 -0
- data/ext/analysis.h +76 -0
- data/ext/array.c +83 -0
- data/ext/array.h +19 -0
- data/ext/bitvector.c +164 -0
- data/ext/bitvector.h +29 -0
- data/ext/compound_io.c +335 -0
- data/ext/document.c +336 -0
- data/ext/document.h +87 -0
- data/ext/ferret.c +88 -47
- data/ext/ferret.h +43 -109
- data/ext/field.c +395 -0
- data/ext/filter.c +103 -0
- data/ext/fs_store.c +352 -0
- data/ext/global.c +219 -0
- data/ext/global.h +73 -0
- data/ext/hash.c +446 -0
- data/ext/hash.h +80 -0
- data/ext/hashset.c +141 -0
- data/ext/hashset.h +37 -0
- data/ext/helper.c +11 -0
- data/ext/helper.h +5 -0
- data/ext/inc/lang.h +41 -0
- data/ext/ind.c +389 -0
- data/ext/index.h +884 -0
- data/ext/index_io.c +269 -415
- data/ext/index_rw.c +2543 -0
- data/ext/lang.c +31 -0
- data/ext/lang.h +41 -0
- data/ext/priorityqueue.c +228 -0
- data/ext/priorityqueue.h +44 -0
- data/ext/q_boolean.c +1331 -0
- data/ext/q_const_score.c +154 -0
- data/ext/q_fuzzy.c +287 -0
- data/ext/q_match_all.c +142 -0
- data/ext/q_multi_phrase.c +343 -0
- data/ext/q_parser.c +2180 -0
- data/ext/q_phrase.c +657 -0
- data/ext/q_prefix.c +75 -0
- data/ext/q_range.c +247 -0
- data/ext/q_span.c +1566 -0
- data/ext/q_term.c +308 -0
- data/ext/q_wildcard.c +146 -0
- data/ext/r_analysis.c +255 -0
- data/ext/r_doc.c +578 -0
- data/ext/r_index_io.c +996 -0
- data/ext/r_qparser.c +158 -0
- data/ext/r_search.c +2321 -0
- data/ext/r_store.c +263 -0
- data/ext/r_term.c +219 -0
- data/ext/ram_store.c +447 -0
- data/ext/search.c +524 -0
- data/ext/search.h +1065 -0
- data/ext/similarity.c +143 -39
- data/ext/sort.c +661 -0
- data/ext/store.c +35 -0
- data/ext/store.h +152 -0
- data/ext/term.c +704 -143
- data/ext/termdocs.c +599 -0
- data/ext/vector.c +594 -0
- data/lib/ferret.rb +9 -10
- data/lib/ferret/analysis/analyzers.rb +2 -2
- data/lib/ferret/analysis/standard_tokenizer.rb +1 -1
- data/lib/ferret/analysis/token.rb +14 -14
- data/lib/ferret/analysis/token_filters.rb +3 -3
- data/lib/ferret/document/field.rb +16 -17
- data/lib/ferret/index/document_writer.rb +4 -4
- data/lib/ferret/index/index.rb +39 -23
- data/lib/ferret/index/index_writer.rb +2 -2
- data/lib/ferret/index/multiple_term_doc_pos_enum.rb +1 -8
- data/lib/ferret/index/segment_term_vector.rb +4 -4
- data/lib/ferret/index/term.rb +5 -1
- data/lib/ferret/index/term_vector_offset_info.rb +6 -6
- data/lib/ferret/index/term_vectors_io.rb +5 -5
- data/lib/ferret/query_parser/query_parser.tab.rb +81 -77
- data/lib/ferret/search.rb +1 -1
- data/lib/ferret/search/boolean_query.rb +2 -1
- data/lib/ferret/search/field_sorted_hit_queue.rb +3 -3
- data/lib/ferret/search/fuzzy_query.rb +2 -1
- data/lib/ferret/search/index_searcher.rb +3 -0
- data/lib/ferret/search/{match_all_docs_query.rb → match_all_query.rb} +7 -7
- data/lib/ferret/search/multi_phrase_query.rb +6 -5
- data/lib/ferret/search/phrase_query.rb +3 -6
- data/lib/ferret/search/prefix_query.rb +4 -4
- data/lib/ferret/search/sort.rb +3 -1
- data/lib/ferret/search/sort_field.rb +9 -9
- data/lib/ferret/search/spans/near_spans_enum.rb +1 -1
- data/lib/ferret/search/spans/span_near_query.rb +1 -1
- data/lib/ferret/search/spans/span_weight.rb +1 -1
- data/lib/ferret/search/spans/spans_enum.rb +7 -7
- data/lib/ferret/store/fs_store.rb +10 -6
- data/lib/ferret/store/ram_store.rb +3 -3
- data/lib/rferret.rb +36 -0
- data/test/functional/thread_safety_index_test.rb +2 -2
- data/test/test_helper.rb +16 -2
- data/test/unit/analysis/c_token.rb +25 -0
- data/test/unit/analysis/tc_per_field_analyzer_wrapper.rb +1 -1
- data/test/unit/analysis/tc_standard_analyzer.rb +1 -1
- data/test/unit/document/{tc_document.rb → c_document.rb} +0 -0
- data/test/unit/document/c_field.rb +98 -0
- data/test/unit/document/tc_field.rb +0 -66
- data/test/unit/index/{tc_index.rb → c_index.rb} +62 -6
- data/test/unit/index/{tc_index_reader.rb → c_index_reader.rb} +51 -10
- data/test/unit/index/{tc_index_writer.rb → c_index_writer.rb} +0 -4
- data/test/unit/index/{tc_term.rb → c_term.rb} +1 -3
- data/test/unit/index/{tc_term_vector_offset_info.rb → c_term_voi.rb} +5 -5
- data/test/unit/index/tc_segment_term_vector.rb +2 -2
- data/test/unit/index/tc_term_vectors_io.rb +4 -4
- data/test/unit/query_parser/c_query_parser.rb +138 -0
- data/test/unit/search/{tc_filter.rb → c_filter.rb} +24 -24
- data/test/unit/search/{tc_fuzzy_query.rb → c_fuzzy_query.rb} +0 -0
- data/test/unit/search/{tc_index_searcher.rb → c_index_searcher.rb} +9 -26
- data/test/unit/search/{tc_search_and_sort.rb → c_search_and_sort.rb} +15 -15
- data/test/unit/search/{tc_sort.rb → c_sort.rb} +2 -1
- data/test/unit/search/c_sort_field.rb +27 -0
- data/test/unit/search/{tc_spans.rb → c_spans.rb} +0 -0
- data/test/unit/search/tc_sort_field.rb +7 -20
- data/test/unit/store/c_fs_store.rb +76 -0
- data/test/unit/store/c_ram_store.rb +35 -0
- data/test/unit/store/m_store.rb +34 -0
- data/test/unit/store/m_store_lock.rb +68 -0
- data/test/unit/store/tc_fs_store.rb +0 -53
- data/test/unit/store/tc_ram_store.rb +0 -20
- data/test/unit/store/tm_store.rb +0 -30
- data/test/unit/store/tm_store_lock.rb +0 -66
- metadata +84 -31
- data/ext/Makefile +0 -140
- data/ext/ferret_ext.so +0 -0
- data/ext/priority_queue.c +0 -232
- data/ext/ram_directory.c +0 -321
- data/ext/segment_merge_queue.c +0 -37
- data/ext/segment_term_enum.c +0 -326
- data/ext/string_helper.c +0 -42
- data/ext/tags +0 -344
- data/ext/term_buffer.c +0 -230
- data/ext/term_infos_reader.c +0 -54
- data/ext/terminfo.c +0 -160
- data/ext/token.c +0 -93
- data/ext/util.c +0 -12
data/ext/analysis.h
ADDED
@@ -0,0 +1,76 @@
|
|
1
|
+
#ifndef FRT_ANALYSIS_H
|
2
|
+
#define FRT_ANALYSIS_H
|
3
|
+
|
4
|
+
#include <global.h>
|
5
|
+
#include <hash.h>
|
6
|
+
|
7
|
+
/****************************************************************************
|
8
|
+
*
|
9
|
+
* Token
|
10
|
+
*
|
11
|
+
****************************************************************************/
|
12
|
+
|
13
|
+
typedef struct Token {
|
14
|
+
char text[MAX_WORD_SIZE];
|
15
|
+
int start;
|
16
|
+
int end;
|
17
|
+
int pos_inc;
|
18
|
+
} Token;
|
19
|
+
|
20
|
+
Token *tk_create();
|
21
|
+
void tk_destroy(void *p);
|
22
|
+
Token *tk_set(Token *tk, char *text, int tlen, int start, int end, int pos_inc);
|
23
|
+
Token *tk_set_no_len(Token *tk, char *text, int start, int end, int pos_inc);
|
24
|
+
int tk_eq(Token *tk1, Token *tk2);
|
25
|
+
int tk_cmp(Token *tk1, Token *tk2);
|
26
|
+
|
27
|
+
/****************************************************************************
|
28
|
+
*
|
29
|
+
* TokenStream
|
30
|
+
*
|
31
|
+
****************************************************************************/
|
32
|
+
|
33
|
+
typedef struct TokenStream TokenStream;
|
34
|
+
struct TokenStream {
|
35
|
+
void *data;
|
36
|
+
char *text;
|
37
|
+
int pos;
|
38
|
+
Token *token;
|
39
|
+
Token *(*next)(TokenStream *ts);
|
40
|
+
void (*reset)(TokenStream *ts, char *text);
|
41
|
+
void (*destroy)(void *p);
|
42
|
+
TokenStream *sub_ts; // used by filters
|
43
|
+
};
|
44
|
+
|
45
|
+
#define ts_next(mts) mts->next(mts)
|
46
|
+
#define ts_destroy(mts) mts->destroy(mts)
|
47
|
+
|
48
|
+
TokenStream *whitespace_tokenizer_create();
|
49
|
+
TokenStream *letter_tokenizer_create();
|
50
|
+
TokenStream *standard_tokenizer_create();
|
51
|
+
TokenStream *lowercase_filter_create(TokenStream *ts);
|
52
|
+
TokenStream *stop_filter_create_with_words(TokenStream *ts, char **words, int len);
|
53
|
+
TokenStream *stop_filter_create(TokenStream *ts);
|
54
|
+
|
55
|
+
/****************************************************************************
|
56
|
+
*
|
57
|
+
* Analyzer
|
58
|
+
*
|
59
|
+
****************************************************************************/
|
60
|
+
|
61
|
+
typedef struct Analyzer {
|
62
|
+
void *data;
|
63
|
+
TokenStream *current_ts;
|
64
|
+
TokenStream *(*get_ts)(struct Analyzer *a, char *field, char *text);
|
65
|
+
void (*destroy)(void *p);
|
66
|
+
} Analyzer;
|
67
|
+
|
68
|
+
#define a_destroy(ma) ma->destroy(ma)
|
69
|
+
#define a_get_ts(ma, field, text) ma->get_ts(ma, field, text)
|
70
|
+
|
71
|
+
Analyzer *whitespace_analyzer_create();
|
72
|
+
Analyzer *letter_analyzer_create();
|
73
|
+
Analyzer *standard_analyzer_create();
|
74
|
+
Analyzer *standard_analyzer_create_with_words(char **words, int len);
|
75
|
+
|
76
|
+
#endif
|
data/ext/array.c
ADDED
@@ -0,0 +1,83 @@
|
|
1
|
+
#include <global.h>
|
2
|
+
#include <array.h>
|
3
|
+
#include <string.h>
|
4
|
+
|
5
|
+
Array *ary_create(int allocate, void (*free_elem)(void *p))
|
6
|
+
{
|
7
|
+
Array *ary = ALLOC(Array);
|
8
|
+
if (allocate == 0) {
|
9
|
+
ary->elems = NULL;
|
10
|
+
} else {
|
11
|
+
ary->elems = ALLOC_N(void *, allocate);
|
12
|
+
memset(ary->elems, 0, sizeof(void *) * allocate);
|
13
|
+
}
|
14
|
+
ary->size = 0;
|
15
|
+
ary->allocated = allocate;
|
16
|
+
ary->free_elem = free_elem;
|
17
|
+
|
18
|
+
return ary;
|
19
|
+
}
|
20
|
+
|
21
|
+
void ary_destroy(void *p)
|
22
|
+
{
|
23
|
+
Array *ary = (Array *)p;
|
24
|
+
int i;
|
25
|
+
for (i = 0; i < ary->size; i++) {
|
26
|
+
if (ary->free_elem != NULL && ary->elems[i] != NULL)
|
27
|
+
ary->free_elem(ary->elems[i]);
|
28
|
+
}
|
29
|
+
free(ary->elems);
|
30
|
+
free(ary);
|
31
|
+
}
|
32
|
+
|
33
|
+
void ary_set(Array *ary, int index, void *value)
|
34
|
+
{
|
35
|
+
if (index >= ary->allocated) {
|
36
|
+
ary->allocated = (index + 1)*2;
|
37
|
+
REALLOC_N(ary->elems, void *, (ary->allocated));
|
38
|
+
memset((&ary->elems[ary->size]), 0, sizeof(void *) * (ary->allocated - ary->size));
|
39
|
+
}
|
40
|
+
|
41
|
+
if (index >= ary->size)
|
42
|
+
ary->size = index + 1;
|
43
|
+
|
44
|
+
if (ary->free_elem != NULL && ary->elems[index] != NULL)
|
45
|
+
ary->free_elem(ary->elems[index]);
|
46
|
+
|
47
|
+
ary->elems[index] = value;
|
48
|
+
}
|
49
|
+
|
50
|
+
void ary_append(Array *ary, void *value)
|
51
|
+
{
|
52
|
+
ary_set(ary, ary->size, value);
|
53
|
+
}
|
54
|
+
|
55
|
+
void *ary_get(Array *ary, int index)
|
56
|
+
{
|
57
|
+
if (index >= ary->size)
|
58
|
+
return NULL;
|
59
|
+
return ary->elems[index];
|
60
|
+
}
|
61
|
+
|
62
|
+
void ary_delete(Array *ary, int index)
|
63
|
+
{
|
64
|
+
if (index >= ary->size)
|
65
|
+
return;
|
66
|
+
ary->free_elem(ary->elems[index]);
|
67
|
+
ary->elems[index] = NULL;
|
68
|
+
if (index == ary->size - 1)
|
69
|
+
ary->size--;
|
70
|
+
}
|
71
|
+
|
72
|
+
void *ary_remove(Array *ary, int index)
|
73
|
+
{
|
74
|
+
void *p;
|
75
|
+
if (index >= ary->size)
|
76
|
+
return NULL;
|
77
|
+
p = ary->elems[index];
|
78
|
+
ary->elems[index] = NULL;
|
79
|
+
ary->size--;
|
80
|
+
memmove(&ary->elems[index], &ary->elems[index + 1],
|
81
|
+
sizeof(void *) *(ary->size - index));
|
82
|
+
return p;
|
83
|
+
}
|
data/ext/array.h
ADDED
@@ -0,0 +1,19 @@
|
|
1
|
+
#ifndef FRT_ARRAY_H
|
2
|
+
#define FRT_ARRAY_H
|
3
|
+
|
4
|
+
typedef struct Array {
|
5
|
+
void **elems;
|
6
|
+
int size;
|
7
|
+
int allocated;
|
8
|
+
void (*free_elem)(void *p);
|
9
|
+
} Array;
|
10
|
+
|
11
|
+
Array *ary_create(int size, void (*free_elem)(void *p));
|
12
|
+
void ary_destroy(void *p);
|
13
|
+
void ary_set(Array *ary, int index, void *value);
|
14
|
+
void ary_append(Array *ary, void *value);
|
15
|
+
void *ary_get(Array *ary, int index);
|
16
|
+
void ary_delete(Array *ary, int index);
|
17
|
+
void *ary_remove(Array *ary, int index);
|
18
|
+
|
19
|
+
#endif
|
data/ext/bitvector.c
ADDED
@@ -0,0 +1,164 @@
|
|
1
|
+
#include <bitvector.h>
|
2
|
+
#include <string.h>
|
3
|
+
|
4
|
+
BitVector *bv_create_size(int size)
|
5
|
+
{
|
6
|
+
BitVector *bv = ALLOC(BitVector);
|
7
|
+
|
8
|
+
bv->capa = (size >> 3) + 1;
|
9
|
+
bv->bits = ALLOC_N(uchar, bv->capa);
|
10
|
+
memset(bv->bits, 0, bv->capa);
|
11
|
+
|
12
|
+
bv->size = 0;
|
13
|
+
bv->count = 0;
|
14
|
+
bv->curr_bit = -1;
|
15
|
+
return bv;
|
16
|
+
}
|
17
|
+
|
18
|
+
BitVector *bv_create()
|
19
|
+
{
|
20
|
+
return bv_create_size(BV_INIT_CAPA);
|
21
|
+
}
|
22
|
+
|
23
|
+
void bv_destroy(void *p)
|
24
|
+
{
|
25
|
+
BitVector *bv = (BitVector *)p;
|
26
|
+
free(bv->bits);
|
27
|
+
free(bv);
|
28
|
+
}
|
29
|
+
|
30
|
+
void bv_set(BitVector *bv, int bit)
|
31
|
+
{
|
32
|
+
int byte = bit>>3;
|
33
|
+
uchar bitmask = 1<<(bit&7);
|
34
|
+
if (bv->size <= byte) {
|
35
|
+
bv->size = byte + 1;
|
36
|
+
if (bv->size >= bv->capa) {
|
37
|
+
int capa = bv->capa * 2;
|
38
|
+
while (capa < bv->size) capa *= 2;
|
39
|
+
REALLOC_N(bv->bits, uchar, capa);
|
40
|
+
memset(bv->bits + bv->capa, 0, capa - bv->capa);
|
41
|
+
bv->capa = capa;
|
42
|
+
}
|
43
|
+
}
|
44
|
+
uchar *byte_p = &(bv->bits[byte]);
|
45
|
+
if ((bitmask & *byte_p) == 0) {
|
46
|
+
bv->count++;
|
47
|
+
*byte_p |= bitmask;
|
48
|
+
}
|
49
|
+
}
|
50
|
+
|
51
|
+
int bv_get(BitVector *bv, int bit)
|
52
|
+
{
|
53
|
+
int byte = bit>>3;
|
54
|
+
if (byte >= bv->size) return 0;
|
55
|
+
return (bv->bits[byte]>>(bit&7))&1;
|
56
|
+
}
|
57
|
+
|
58
|
+
void bv_clear(BitVector *bv)
|
59
|
+
{
|
60
|
+
memset(bv->bits, 0, bv->size);
|
61
|
+
bv->count = 0;
|
62
|
+
}
|
63
|
+
|
64
|
+
void bv_unset(BitVector *bv, int bit)
|
65
|
+
{
|
66
|
+
int byte = bit>>3;
|
67
|
+
if (byte >= bv->size) return;
|
68
|
+
uchar *byte_p = &(bv->bits[byte]);
|
69
|
+
uchar bitmask = 1<<(bit&7);
|
70
|
+
if ((bitmask & *byte_p) > 0) {
|
71
|
+
bv->count--;
|
72
|
+
*byte_p &= ~bitmask;
|
73
|
+
}
|
74
|
+
}
|
75
|
+
|
76
|
+
void bv_write(BitVector *bv, Store *store, char *name)
|
77
|
+
{
|
78
|
+
OutStream *os = store->create_output(store, name);
|
79
|
+
os_write_vint(os, bv->size);
|
80
|
+
os_write_bytes(os, bv->bits, bv->size);
|
81
|
+
os_close(os);
|
82
|
+
}
|
83
|
+
|
84
|
+
const uchar BYTE_COUNTS[] = { // table of bits/char
|
85
|
+
0, 1, 1, 2, 1, 2, 2, 3, 1, 2, 2, 3, 2, 3, 3, 4,
|
86
|
+
1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
|
87
|
+
1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
|
88
|
+
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
|
89
|
+
1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
|
90
|
+
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
|
91
|
+
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
|
92
|
+
3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
|
93
|
+
1, 2, 2, 3, 2, 3, 3, 4, 2, 3, 3, 4, 3, 4, 4, 5,
|
94
|
+
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
|
95
|
+
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
|
96
|
+
3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
|
97
|
+
2, 3, 3, 4, 3, 4, 4, 5, 3, 4, 4, 5, 4, 5, 5, 6,
|
98
|
+
3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
|
99
|
+
3, 4, 4, 5, 4, 5, 5, 6, 4, 5, 5, 6, 5, 6, 6, 7,
|
100
|
+
4, 5, 5, 6, 5, 6, 6, 7, 5, 6, 6, 7, 6, 7, 7, 8
|
101
|
+
};
|
102
|
+
|
103
|
+
int bv_count(BitVector *bv)
|
104
|
+
{
|
105
|
+
// if the vector has been modified
|
106
|
+
int i, c = 0;
|
107
|
+
uchar *bytes = bv->bits;
|
108
|
+
for (i = 0; i < bv->size; i++)
|
109
|
+
c += BYTE_COUNTS[bytes[i]]; // sum bits per char
|
110
|
+
bv->count = c;
|
111
|
+
return c;
|
112
|
+
}
|
113
|
+
|
114
|
+
BitVector *bv_read(Store *store, char *name)
|
115
|
+
{
|
116
|
+
BitVector *bv = ALLOC(BitVector);
|
117
|
+
InStream *is = store->open_input(store, name);
|
118
|
+
bv->capa = bv->size = is_read_vint(is);
|
119
|
+
bv->bits = ALLOC_N(uchar, bv->capa);
|
120
|
+
is_read_bytes(is, bv->bits, 0, bv->size);
|
121
|
+
is_close(is);
|
122
|
+
bv_count(bv);
|
123
|
+
return bv;
|
124
|
+
}
|
125
|
+
|
126
|
+
void bv_scan_reset(BitVector *bv)
|
127
|
+
{
|
128
|
+
bv->curr_bit = -1;
|
129
|
+
}
|
130
|
+
|
131
|
+
inline int bv_scan_next_from(BitVector *bv, register const int from)
|
132
|
+
{
|
133
|
+
register const uchar *const bits = bv->bits;
|
134
|
+
register const int size = bv->size;
|
135
|
+
register int byte_pos = (from) >> 3;
|
136
|
+
register int inc = ((from) & 7);
|
137
|
+
register int bit = 1 << inc;
|
138
|
+
register int mask = 0xff << inc;
|
139
|
+
register int byte;
|
140
|
+
|
141
|
+
if (byte_pos >= size) return -1;
|
142
|
+
if ((bits[byte_pos]&mask) == 0) {
|
143
|
+
inc = 0;
|
144
|
+
bit = 1;
|
145
|
+
do {
|
146
|
+
byte_pos++;
|
147
|
+
if (byte_pos >= size) return -1;
|
148
|
+
} while (bits[byte_pos] == 0);
|
149
|
+
}
|
150
|
+
|
151
|
+
byte = bits[byte_pos];
|
152
|
+
while ((byte & bit) == 0) {
|
153
|
+
bit <<= 1;
|
154
|
+
inc++;
|
155
|
+
}
|
156
|
+
|
157
|
+
return bv->curr_bit = ((byte_pos << 3) + inc);
|
158
|
+
}
|
159
|
+
|
160
|
+
inline int bv_scan_next(BitVector *bv)
|
161
|
+
{
|
162
|
+
return bv_scan_next_from(bv, bv->curr_bit+1);
|
163
|
+
}
|
164
|
+
|
data/ext/bitvector.h
ADDED
@@ -0,0 +1,29 @@
|
|
1
|
+
#ifndef FRT_BIT_VECTOR_H
|
2
|
+
#define FRT_BIT_VECTOR_H
|
3
|
+
|
4
|
+
#include <global.h>
|
5
|
+
#include <store.h>
|
6
|
+
|
7
|
+
#define BV_INIT_CAPA 256
|
8
|
+
typedef struct BitVector {
|
9
|
+
uchar *bits;
|
10
|
+
int size;
|
11
|
+
int capa;
|
12
|
+
int count;
|
13
|
+
int curr_bit;
|
14
|
+
} BitVector;
|
15
|
+
|
16
|
+
BitVector *bv_create();;
|
17
|
+
BitVector *bv_create_size(int size);
|
18
|
+
void bv_destroy(void *bv);
|
19
|
+
void bv_set(BitVector *bv, int bit);
|
20
|
+
int bv_get(BitVector *bv, int bit);
|
21
|
+
void bv_clear(BitVector *bv);
|
22
|
+
void bv_unset(BitVector *bv, int bit);
|
23
|
+
void bv_write(BitVector *bv, Store *store, char *name);
|
24
|
+
BitVector *bv_read(Store *store, char *name);
|
25
|
+
void bv_scan_reset(BitVector *bv);
|
26
|
+
int bv_scan_next(BitVector *bv);
|
27
|
+
int bv_scan_next_from(BitVector *bv, int from);
|
28
|
+
|
29
|
+
#endif
|
data/ext/compound_io.c
ADDED
@@ -0,0 +1,335 @@
|
|
1
|
+
#include "index.h"
|
2
|
+
|
3
|
+
/****************************************************************************
|
4
|
+
*
|
5
|
+
* CompoundStore
|
6
|
+
*
|
7
|
+
****************************************************************************/
|
8
|
+
|
9
|
+
typedef struct FileEntry {
|
10
|
+
int offset;
|
11
|
+
int length;
|
12
|
+
} FileEntry;
|
13
|
+
|
14
|
+
void cmpd_touch(Store *store, char *filename)
|
15
|
+
{
|
16
|
+
store->dir.cmpd->store->touch(store->dir.cmpd->store, filename);
|
17
|
+
}
|
18
|
+
|
19
|
+
int cmpd_exists(Store *store, char *filename)
|
20
|
+
{
|
21
|
+
if (h_get(store->dir.cmpd->entries, filename) != NULL)
|
22
|
+
return true;
|
23
|
+
else
|
24
|
+
return false;
|
25
|
+
}
|
26
|
+
|
27
|
+
int cmpd_remove(Store *store, char *filename)
|
28
|
+
{
|
29
|
+
eprintf(UNSUPPORTED_ERROR, "Unsupported operation");
|
30
|
+
return 0;
|
31
|
+
}
|
32
|
+
|
33
|
+
int cmpd_rename(Store *store, char *from, char *to)
|
34
|
+
{
|
35
|
+
eprintf(UNSUPPORTED_ERROR, "Unsupported operation");
|
36
|
+
return 0;
|
37
|
+
}
|
38
|
+
|
39
|
+
int cmpd_count(Store *store)
|
40
|
+
{
|
41
|
+
return store->dir.cmpd->entries->used;
|
42
|
+
}
|
43
|
+
|
44
|
+
void cmpd_clear(Store *store)
|
45
|
+
{
|
46
|
+
eprintf(UNSUPPORTED_ERROR, "Unsupported operation");
|
47
|
+
}
|
48
|
+
|
49
|
+
void cmpd_close(Store *store)
|
50
|
+
{
|
51
|
+
mutex_lock(&store->mutex);
|
52
|
+
CompoundStore *cmpd = store->dir.cmpd;
|
53
|
+
if (cmpd->stream == NULL)
|
54
|
+
eprintf(IO_ERROR, "Already closed");
|
55
|
+
|
56
|
+
h_destroy(cmpd->entries);
|
57
|
+
|
58
|
+
is_close(cmpd->stream);
|
59
|
+
cmpd->stream = NULL;
|
60
|
+
free(store->dir.cmpd);
|
61
|
+
store_destroy(store);
|
62
|
+
}
|
63
|
+
|
64
|
+
int cmpd_length(Store *store, char *filename)
|
65
|
+
{
|
66
|
+
FileEntry *fe = (FileEntry *)h_get(store->dir.cmpd->entries, filename);
|
67
|
+
if (fe != NULL)
|
68
|
+
return fe->length;
|
69
|
+
else
|
70
|
+
return 0;
|
71
|
+
}
|
72
|
+
|
73
|
+
void cmpdi_seek_internal(InStream *is, int pos) {}
|
74
|
+
void cmpdi_close_internal(InStream *is)
|
75
|
+
{
|
76
|
+
//is_close(is->d.cis->sub);
|
77
|
+
free(is->d.cis);
|
78
|
+
}
|
79
|
+
|
80
|
+
void cmpdi_clone_internal(InStream *is, InStream *new_is)
|
81
|
+
{
|
82
|
+
CompoundInStream *cis = ALLOC(CompoundInStream);
|
83
|
+
//cis->sub = is_clone(is->d.cis->sub);
|
84
|
+
cis->sub = is->d.cis->sub;
|
85
|
+
cis->offset = is->d.cis->offset;
|
86
|
+
cis->length = is->d.cis->length;
|
87
|
+
new_is->d.cis = cis;
|
88
|
+
}
|
89
|
+
|
90
|
+
int cmpdi_length_internal(InStream *is)
|
91
|
+
{
|
92
|
+
return (is->d.cis->length);
|
93
|
+
}
|
94
|
+
|
95
|
+
void cmpdi_read_internal(InStream *is, uchar *b, int offset, int len)
|
96
|
+
{
|
97
|
+
CompoundInStream *cis = is->d.cis;
|
98
|
+
int start = is_pos(is);
|
99
|
+
if ((start + len) > cis->length)
|
100
|
+
eprintf(EOF_ERROR, "read past EOF");
|
101
|
+
is_seek(cis->sub, cis->offset + start);
|
102
|
+
is_read_bytes(cis->sub, b, offset, len);
|
103
|
+
}
|
104
|
+
|
105
|
+
InStream *cmpd_create_input(InStream *sub_is, int offset, int length)
|
106
|
+
{
|
107
|
+
InStream *is = is_create();
|
108
|
+
CompoundInStream *cis = ALLOC(CompoundInStream);
|
109
|
+
//cis->sub = is_clone(sub_is);
|
110
|
+
cis->sub = sub_is;
|
111
|
+
cis->offset = offset;
|
112
|
+
cis->length = length;
|
113
|
+
is->d.cis = cis;
|
114
|
+
is->file = NULL;
|
115
|
+
|
116
|
+
is->read_internal = &cmpdi_read_internal;
|
117
|
+
is->seek_internal = &cmpdi_seek_internal;
|
118
|
+
is->close_internal = &cmpdi_close_internal;
|
119
|
+
is->clone_internal = &cmpdi_clone_internal;
|
120
|
+
is->length_internal = &cmpdi_length_internal;
|
121
|
+
return is;
|
122
|
+
}
|
123
|
+
|
124
|
+
InStream *cmpd_open_input(Store *store, const char *filename)
|
125
|
+
{
|
126
|
+
CompoundStore *cmpd = store->dir.cmpd;
|
127
|
+
InStream *is;
|
128
|
+
|
129
|
+
mutex_lock(&store->mutex);
|
130
|
+
if (cmpd->stream == NULL) {
|
131
|
+
mutex_unlock(&store->mutex);
|
132
|
+
eprintf(IO_ERROR, "Stream closed");
|
133
|
+
}
|
134
|
+
|
135
|
+
FileEntry *entry = (FileEntry *)h_get(cmpd->entries, filename);
|
136
|
+
if (entry == NULL) {
|
137
|
+
mutex_unlock(&store->mutex);
|
138
|
+
eprintf(IO_ERROR, "No sub-file with id <%s> found", filename);
|
139
|
+
}
|
140
|
+
|
141
|
+
is = cmpd_create_input(cmpd->stream, entry->offset, entry->length);
|
142
|
+
mutex_unlock(&store->mutex);
|
143
|
+
|
144
|
+
return is;
|
145
|
+
}
|
146
|
+
|
147
|
+
OutStream *cmpd_create_output(Store *store, const char *filename)
|
148
|
+
{
|
149
|
+
eprintf(UNSUPPORTED_ERROR, "Unsupported operation");
|
150
|
+
return NULL;
|
151
|
+
}
|
152
|
+
|
153
|
+
Lock *cmpd_open_lock(Store *store, char *lockname)
|
154
|
+
{
|
155
|
+
eprintf(UNSUPPORTED_ERROR, "Unsupported operation");
|
156
|
+
return NULL;
|
157
|
+
}
|
158
|
+
|
159
|
+
void cmpd_close_lock(Lock *lock)
|
160
|
+
{
|
161
|
+
eprintf(UNSUPPORTED_ERROR, "Unsupported operation");
|
162
|
+
}
|
163
|
+
|
164
|
+
Store *open_cmpd_store(Store *store, const char *name)
|
165
|
+
{
|
166
|
+
CompoundStore *cmpd = ALLOC(CompoundStore);
|
167
|
+
Store *new_store = store_create();
|
168
|
+
|
169
|
+
cmpd->store = store;
|
170
|
+
cmpd->name = name;
|
171
|
+
cmpd->entries = h_new_str(&efree, &efree);
|
172
|
+
InStream *is = cmpd->stream = store->open_input(store, cmpd->name);
|
173
|
+
|
174
|
+
// read the directory and init files
|
175
|
+
int count = is_read_vint(is);
|
176
|
+
FileEntry *entry = NULL;
|
177
|
+
int i, offset;
|
178
|
+
char *fname;
|
179
|
+
for (i = 0; i < count; i++) {
|
180
|
+
offset = is_read_long(is);
|
181
|
+
fname = is_read_string(is);
|
182
|
+
|
183
|
+
if (entry != NULL) {
|
184
|
+
// set length of the previous entry
|
185
|
+
entry->length = offset - entry->offset;
|
186
|
+
}
|
187
|
+
|
188
|
+
entry = ALLOC(FileEntry);
|
189
|
+
entry->offset = offset;
|
190
|
+
h_set(cmpd->entries, fname, entry);
|
191
|
+
}
|
192
|
+
|
193
|
+
// set the length of the final entry
|
194
|
+
if (entry != NULL)
|
195
|
+
entry->length = is_length(is) - entry->offset;
|
196
|
+
|
197
|
+
new_store->dir.cmpd = cmpd;
|
198
|
+
new_store->touch = &cmpd_touch;
|
199
|
+
new_store->exists = &cmpd_exists;
|
200
|
+
new_store->remove = &cmpd_remove;
|
201
|
+
new_store->rename = &cmpd_rename;
|
202
|
+
new_store->count = &cmpd_count;
|
203
|
+
new_store->close = &cmpd_close;
|
204
|
+
new_store->clear = &cmpd_clear;
|
205
|
+
new_store->length = &cmpd_length;
|
206
|
+
new_store->create_output = &cmpd_create_output;
|
207
|
+
new_store->open_input = &cmpd_open_input;
|
208
|
+
new_store->open_lock = &cmpd_open_lock;
|
209
|
+
new_store->close_lock = &cmpd_close_lock;
|
210
|
+
return new_store;
|
211
|
+
}
|
212
|
+
|
213
|
+
/****************************************************************************
|
214
|
+
*
|
215
|
+
* CompoundWriter
|
216
|
+
*
|
217
|
+
****************************************************************************/
|
218
|
+
|
219
|
+
typedef struct WFileEntry {
|
220
|
+
char *name;
|
221
|
+
int dir_offset;
|
222
|
+
int data_offset;
|
223
|
+
} WFileEntry;
|
224
|
+
|
225
|
+
WFileEntry *wfe_create(char *name)
|
226
|
+
{
|
227
|
+
WFileEntry *wfe = ALLOC(WFileEntry);
|
228
|
+
wfe->name = name;
|
229
|
+
return wfe;
|
230
|
+
}
|
231
|
+
|
232
|
+
void wfe_destroy(void *p)
|
233
|
+
{
|
234
|
+
WFileEntry *wfe = (WFileEntry *)p;
|
235
|
+
efree(wfe);
|
236
|
+
}
|
237
|
+
|
238
|
+
CompoundWriter *open_cw(Store *store, char *name)
|
239
|
+
{
|
240
|
+
CompoundWriter *cw = ALLOC(CompoundWriter);
|
241
|
+
cw->store = store;
|
242
|
+
cw->name = name;
|
243
|
+
cw->ids = hs_str_create(NULL);
|
244
|
+
cw->file_entries = ary_create(1, &wfe_destroy);
|
245
|
+
cw->merged = false;
|
246
|
+
return cw;
|
247
|
+
}
|
248
|
+
|
249
|
+
void cw_add_file(CompoundWriter *cw, char *id)
|
250
|
+
{
|
251
|
+
if (cw->merged) eprintf(STATE_ERROR, "Already merged");
|
252
|
+
if (hs_add(cw->ids, id) != HASH_KEY_DOES_NOT_EXIST)
|
253
|
+
eprintf(STATE_ERROR, "Already merged");
|
254
|
+
|
255
|
+
hs_add(cw->ids, id);
|
256
|
+
ary_append(cw->file_entries, wfe_create(id));
|
257
|
+
}
|
258
|
+
|
259
|
+
void cw_copy_file(CompoundWriter *cw, WFileEntry *src, OutStream *os)
|
260
|
+
{
|
261
|
+
|
262
|
+
int start_ptr = os_pos(os);
|
263
|
+
|
264
|
+
InStream *is = cw->store->open_input(cw->store, src->name);
|
265
|
+
int remainder, length, len;
|
266
|
+
remainder = length = is_length(is);
|
267
|
+
|
268
|
+
uchar buffer[BUFFER_SIZE];
|
269
|
+
while (remainder > 0) {
|
270
|
+
len = MIN(remainder, BUFFER_SIZE);
|
271
|
+
is_read_bytes(is, buffer, 0, len);
|
272
|
+
os_write_bytes(os, buffer, len);
|
273
|
+
remainder -= len;
|
274
|
+
}
|
275
|
+
|
276
|
+
// Verify that remainder is 0
|
277
|
+
if (remainder != 0)
|
278
|
+
eprintf(IO_ERROR, "Non-zero remainder length after copying: %ld "
|
279
|
+
"(id:%s, length: %ld, buffer size: %ld\n", remainder,
|
280
|
+
src->name, length, BUFFER_SIZE);
|
281
|
+
|
282
|
+
// Verify that the output length diff is equal to original file
|
283
|
+
int end_ptr = os_pos(os);
|
284
|
+
int diff = end_ptr - start_ptr;
|
285
|
+
if (diff != length)
|
286
|
+
eprintf(IO_ERROR, "Difference in the output file offsets %ld "
|
287
|
+
" does not match the original file length ", diff, length);
|
288
|
+
|
289
|
+
is_close(is);
|
290
|
+
}
|
291
|
+
|
292
|
+
void cw_close(CompoundWriter *cw)
|
293
|
+
{
|
294
|
+
if (cw->merged) eprintf(STATE_ERROR, "Already merged");
|
295
|
+
if (cw->ids->size <= 0)
|
296
|
+
eprintf(STATE_ERROR, "No Files to merge into the compound file");
|
297
|
+
|
298
|
+
cw->merged = true;
|
299
|
+
|
300
|
+
OutStream *os = cw->store->create_output(cw->store, cw->name);
|
301
|
+
os_write_vint(os, cw->file_entries->size);
|
302
|
+
|
303
|
+
/* Write the directory with all offsets at 0.
|
304
|
+
* Remember the positions of directory entries so that we can adjust the
|
305
|
+
* offsets later */
|
306
|
+
int i;
|
307
|
+
WFileEntry *wfe;
|
308
|
+
for (i = 0; i < cw->file_entries->size; i++) {
|
309
|
+
wfe = (WFileEntry *)cw->file_entries->elems[i];
|
310
|
+
wfe->dir_offset = os_pos(os);
|
311
|
+
os_write_long(os, 0); // for now
|
312
|
+
os_write_string(os, wfe->name);
|
313
|
+
}
|
314
|
+
|
315
|
+
/* Open the files and copy their data into the stream. Remember the
|
316
|
+
* locations of each file's data section. */
|
317
|
+
for (i = 0; i < cw->file_entries->size; i++) {
|
318
|
+
wfe = (WFileEntry *)cw->file_entries->elems[i];
|
319
|
+
wfe->data_offset = os_pos(os);
|
320
|
+
cw_copy_file(cw, wfe, os);
|
321
|
+
}
|
322
|
+
|
323
|
+
/* Write the data offsets into the directory of the compound stream */
|
324
|
+
for (i = 0; i < cw->file_entries->size; i++) {
|
325
|
+
wfe = (WFileEntry *)cw->file_entries->elems[i];
|
326
|
+
os_seek(os, wfe->dir_offset);
|
327
|
+
os_write_long(os, wfe->data_offset);
|
328
|
+
}
|
329
|
+
|
330
|
+
os_close(os);
|
331
|
+
hs_destroy(cw->ids);
|
332
|
+
ary_destroy(cw->file_entries);
|
333
|
+
free(cw);
|
334
|
+
}
|
335
|
+
|