summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
Diffstat
-rw-r--r--src/storage/flexible_ngram.h17
-rw-r--r--tests/storage/test_flexible_ngram.cpp69
-rw-r--r--tests/storage/test_ngram.cpp72
3 files changed, 102 insertions, 56 deletions
diff --git a/src/storage/flexible_ngram.h b/src/storage/flexible_ngram.h
index ff366da..10bad8f 100644
--- a/src/storage/flexible_ngram.h
+++ b/src/storage/flexible_ngram.h
@@ -162,7 +162,7 @@ public:
return true;
}
if ( cur_item->m_token == token ){
- cur_item.m_item = item;
+ cur_item->m_item = item;
return true;
}
}
@@ -251,7 +251,7 @@ public:
}
bool store(phrase_token_t index,
- FlexibleSingleGram<ArrayHeader, ArrayItem> * & single_gram){
+ FlexibleSingleGram<ArrayHeader, ArrayItem> * single_gram){
if ( !m_db )
return false;
@@ -287,7 +287,10 @@ public:
/* Iterate over the database, retrieving each record in turn. */
while ((ret = cursorp->c_get(cursorp, &key, &data, DB_NEXT)) == 0 ){
- assert(key.size == sizeof(phrase_token_t));
+ if (key.size > sizeof(phrase_token_t)){
+ /* skip magic header. */
+ continue;
+ }
phrase_token_t * token = (phrase_token_t *) key.data;
g_array_append_val(items, *token);
}
@@ -305,9 +308,6 @@ public:
/* get/set magic header. */
bool get_magic_header(MagicHeader & header){
- /* Note: remove the below statement later? */
- assert(sizeof(m_magic_header_index) == 2 * sizeof(phrase_token_t));
-
if ( !m_db )
return false;
@@ -327,9 +327,6 @@ public:
}
bool set_magic_header(const MagicHeader & header){
- /* Note: remove the below statement later? */
- assert(sizeof(m_magic_header_index) == 2 * sizeof(phrase_token_t));
-
if ( !m_db )
return false;
@@ -339,7 +336,7 @@ public:
db_key.size = sizeof(m_magic_header_index);
DBT db_data;
memset(&db_data, 0, sizeof(DBT));
- db_data.data = &header;
+ db_data.data = (void *) &header;
db_data.size = sizeof(MagicHeader);
int ret = m_db->put(m_db, NULL, &db_key, &db_data, 0);
diff --git a/tests/storage/test_flexible_ngram.cpp b/tests/storage/test_flexible_ngram.cpp
index 744187f..f915a1f 100644
--- a/tests/storage/test_flexible_ngram.cpp
+++ b/tests/storage/test_flexible_ngram.cpp
@@ -1,5 +1,72 @@
#include "pinyin.h"
int main(int argc, char * argv[]) {
- FlexibleBigram<guint64, guint32, guint32> test_bigram;
+ FlexibleSingleGram<guint32, guint32> single_gram;
+ typedef FlexibleSingleGram<guint32, guint32>::ArrayItemWithToken array_item_t;
+
+ const guint32 total_freq = 16;
+ assert(single_gram.set_array_header(total_freq));
+
+ phrase_token_t tokens[6] = { 2, 6, 4, 3, 1, 3 };
+ guint32 freqs[6] = { 1, 2, 4, 8, 16, 32};
+
+ for ( size_t i = 0; i < 6; ++i ){
+ single_gram.set_array_item(tokens[i], freqs[i]);
+ }
+
+ guint32 freq;
+ single_gram.get_array_item(3, freq);
+ assert(freq == 32);
+
+ printf("--------------------------------------------------------\n");
+ PhraseIndexRange range;
+ FlexibleBigramPhraseArray array = g_array_new(FALSE, FALSE, sizeof(array_item_t));
+ range.m_range_begin = 0; range.m_range_end = 8;
+ single_gram.search(&range, array);
+ for ( size_t i = 0; i < array->len; ++i ){
+ array_item_t * item = &g_array_index(array, array_item_t, i);
+ printf("item:%d:%d\n", item->m_token, item->m_item);
+ }
+
+ assert(single_gram.get_array_header(freq));
+ assert(freq == total_freq);
+
+ FlexibleBigram<guint32, guint32, guint32> bigram;
+ assert(bigram.attach("/tmp/training.db"));
+ bigram.store(1, &single_gram);
+ single_gram.set_array_item(5, 8);
+ single_gram.set_array_header(32);
+ single_gram.get_array_header(freq);
+ printf("new array header:%d\n", freq);
+
+ bigram.store(2, &single_gram);
+
+ for (int m = 1; m <= 2; ++m ){
+ printf("--------------------------------------------------------\n");
+ FlexibleSingleGram<guint32, guint32> * train_gram;
+ bigram.load(m, train_gram);
+ g_array_set_size(array, 0);
+ range.m_range_begin = 0; range.m_range_end = 8;
+ train_gram->search(&range, array);
+ for ( size_t i = 0; i < array->len; ++i ){
+ array_item_t * item = &g_array_index(array, array_item_t, i);
+ printf("item:%d:%d\n", item->m_token, item->m_item);
+ }
+ delete train_gram;
+ }
+ g_array_free(array, TRUE);
+
+ GArray * items = g_array_new(FALSE, FALSE, sizeof(phrase_token_t));
+ bigram.get_all_items(items);
+ printf("-----------------------train----------------------------\n");
+ for ( size_t i = 0; i < items->len; ++i ){
+ phrase_token_t * token = &g_array_index(items, phrase_token_t, i);
+ printf("item:%d\n", *token);
+ }
+
+ printf("-----------------------magic header---------------------\n");
+ bigram.set_magic_header(total_freq);
+ bigram.get_magic_header(freq);
+ assert(total_freq == freq);
+ printf("magic header:%d\n", freq);
}
diff --git a/tests/storage/test_ngram.cpp b/tests/storage/test_ngram.cpp
index 0b8f199..1dd65a8 100644
--- a/tests/storage/test_ngram.cpp
+++ b/tests/storage/test_ngram.cpp
@@ -43,55 +43,37 @@ int main(int argc, char * argv[]){
bigram.store(2, &single_gram);
- printf("--------------------------------------------------------\n");
- SingleGram * system, * user;
- bigram.load(1, system, user);
- assert(NULL == system);
- g_array_set_size(array, 0);
- range.m_range_begin = 0; range.m_range_end = 8;
- user->search(&range,array);
- for ( size_t i = 0; i < array->len; ++i){
- BigramPhraseItem * item = &g_array_index(array, BigramPhraseItem, i);
- printf("item:%d:%f\n", item->m_token, item->m_freq);
- }
- delete user;
- printf("--------------------------------------------------------\n");
- bigram.load(2, system, user);
- assert(NULL == system);
- g_array_set_size(array, 0);
- range.m_range_begin = 0; range.m_range_end = 8;
- user->search(&range,array);
- for ( size_t i = 0; i < array->len; ++i){
- BigramPhraseItem * item = &g_array_index(array, BigramPhraseItem, i);
- printf("item:%d:%f\n", item->m_token, item->m_freq);
- }
- delete user;
+ SingleGram * system, * user;
+ for ( int m = 1; m <= 2; ++m ){
+ printf("--------------------------------------------------------\n");
+ bigram.load(m, system, user);
+ assert(NULL == system);
+ g_array_set_size(array, 0);
+ range.m_range_begin = 0; range.m_range_end = 8;
+ user->search(&range,array);
+ for ( size_t i = 0; i < array->len; ++i){
+ BigramPhraseItem * item = &g_array_index(array, BigramPhraseItem, i);
+ printf("item:%d:%f\n", item->m_token, item->m_freq);
+ }
+ delete user;
+ }
bigram.attach("/tmp/system.db", NULL);
- printf("--------------------------------------------------------\n");
- bigram.load(1, system, user);
- assert(NULL == user);
- g_array_set_size(array, 0);
- range.m_range_begin = 0; range.m_range_end = 8;
- system->search(&range,array);
- for ( size_t i = 0; i < array->len; ++i){
- BigramPhraseItem * item = &g_array_index(array, BigramPhraseItem, i);
- printf("item:%d:%f\n", item->m_token, item->m_freq);
- }
- delete system;
-
- printf("--------------------------------------------------------\n");
- bigram.load(2, system, user);
- assert(NULL == user);
- g_array_set_size(array, 0);
- range.m_range_begin = 0; range.m_range_end = 8;
- system->search(&range,array);
- for ( size_t i = 0; i < array->len; ++i){
- BigramPhraseItem * item = &g_array_index(array, BigramPhraseItem, i);
- printf("item:%d:%f\n", item->m_token, item->m_freq);
+
+ for ( int m = 1; m <=2; ++m ){
+ printf("--------------------------------------------------------\n");
+ bigram.load(m, system, user);
+ assert(NULL == user);
+ g_array_set_size(array, 0);
+ range.m_range_begin = 0; range.m_range_end = 8;
+ system->search(&range,array);
+ for ( size_t i = 0; i < array->len; ++i){
+ BigramPhraseItem * item = &g_array_index(array, BigramPhraseItem, i);
+ printf("item:%d:%f\n", item->m_token, item->m_freq);
+ }
+ delete system;
}
- delete system;
printf("--------------------------------------------------------\n");
single_gram.prune();
generated by cgit (git 2.34.1) at 2024-05-04 08:58:37 +0000