1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
|
#include <stdio.h>
#include <assert.h>
#include <glib.h>
#include "memory_chunk.h"
#include "novel_types.h"
#include "phrase_index.h"
#include "ngram.h"
/* export interpolation model as textual format */
void gen_unigram(FILE * output, FacadePhraseIndex * phrase_index);
void gen_bigram(FILE * output, FacadePhraseIndex * phrase_index, Bigram * bigram);
/* consider moving the following function to utils/storage/utility.h */
char * token_to_string(FacadePhraseIndex * phrase_index, phrase_token_t token);
void begin_data(FILE * file){
fprintf(file, "\\data\n");
}
void end_data(FILE * file){
fprintf(file, "\\end\n");
}
int main(int argc, char * argv[]){
FILE * file = stdout;
const char * bigram_filename = "../../data/bigram.db";
FacadePhraseIndex phrase_index;
//gb_char binary file
MemoryChunk * chunk = new MemoryChunk;
chunk->load("../../data/gb_char.bin");
phrase_index.load(1, chunk);
//gbk_char binary file
chunk = new MemoryChunk;
chunk->load("../../data/gbk_char.bin");
phrase_index.load(2, chunk);
Bigram bigram;
bigram.attach(bigram_filename, NULL);
begin_data(file);
gen_unigram(stdout, &phrase_index);
gen_bigram(stdout, &phrase_index, &bigram);
end_data(stdout);
return 0;
}
void gen_unigram(FILE * output, FacadePhraseIndex * phrase_index) {
fprintf(output, "\\1-gram\n");
for ( size_t i = 0; i < PHRASE_INDEX_LIBRARY_COUNT; i++) {
PhraseIndexRange range;
int result = phrase_index->get_range(i, range);
if ( result )
continue;
PhraseItem item;
for ( size_t j = range.m_range_begin; j < range.m_range_end; j++) {
int result = phrase_index->get_phrase_item(j, item);
if ( result == ERROR_NO_ITEM )
continue;
assert( result == ERROR_OK);
size_t freq = item.get_unigram_frequency();
char * phrase = token_to_string(phrase_index, j);
if ( phrase )
fprintf(output, "\\item %s count %d\n", phrase, freq);
g_free(phrase);
}
}
}
void gen_bigram(FILE * output, FacadePhraseIndex * phrase_index, Bigram * bigram){
fprintf(output, "\\2-gram\n");
/* Retrieve all user items. */
GArray * system_items = g_array_new(FALSE, FALSE, sizeof(phrase_token_t));
GArray * user_items = g_array_new(FALSE, FALSE, sizeof(phrase_token_t));
bigram->get_all_items(system_items, user_items);
assert(0 == user_items->len);
g_array_free(user_items, TRUE);
PhraseItem item;
utf16_t buffer[MAX_PHRASE_LENGTH];
for(int i = 0; i < system_items->len; i++){
phrase_token_t token = g_array_index(system_items, phrase_token_t, i);
SingleGram * system = NULL, * user = NULL;
bigram->load(token, system, user);
assert(NULL == user);
BigramPhraseWithCountArray array = g_array_new(FALSE, FALSE, sizeof(BigramPhraseItemWithCount));
system->retrieve_all(array);
for(int j = 0; j < array->len; j++) {
BigramPhraseItemWithCount * item = &g_array_index(array, BigramPhraseItemWithCount, j);
char * word1 = token_to_string(phrase_index, token);
char * word2 = token_to_string(phrase_index, item->m_token);
guint32 freq = item->m_count;
if ( word1 && word2)
fprintf(output, "\\item %s %s count %d\n", word1, word2, freq);
g_free(word1); g_free(word2);
}
g_array_free(array, TRUE);
}
g_array_free(system_items, TRUE);
}
static const char * special_token_to_string(phrase_token_t token){
struct token_pair{
phrase_token_t token;
const char * string;
};
static const token_pair tokens [] = {
{sentence_start, "<start>"},
{0, NULL}
};
const token_pair * pair = tokens;
while (pair->token) {
if ( token == pair->token )
return pair->string;
}
return NULL;
}
char * token_to_string(FacadePhraseIndex * phrase_index, phrase_token_t token) {
PhraseItem item;
utf16_t buffer[MAX_PHRASE_LENGTH];
gchar * phrase;
/* deal with the special phrase index, for "<start>..." */
if ( PHRASE_INDEX_LIBRARY_INDEX(token) == 0 ) {
return g_strdup(special_token_to_string(token));
}
int result = phrase_index->get_phrase_item(token, item);
if (result != ERROR_OK)
return NULL;
item.get_phrase_string(buffer);
guint8 length = item.get_phrase_length();
phrase = g_utf16_to_utf8(buffer, length, NULL, NULL, NULL);
return phrase;
}
|
