write validate k mixture model tool

3 files changed, 142 insertions, 3 deletions

diff --git a/utils/training/Makefile.am b/utils/training/Makefile.am
index 0915479..a8b5478 100644
--- a/utils/training/Makefile.am
+++ b/utils/training/Makefile.am
@@ -35,7 +35,8 @@ noinst_PROGRAMS		= gen_ngram \
 			  prune_k_mixture_model \
 			  import_k_mixture_model \
 			  export_k_mixture_model \
-			  k_mixture_model_to_interpolation
+			  k_mixture_model_to_interpolation \
+			  validate_k_mixture_model
 
 gen_ngram_SOURCES	= gen_ngram.cpp
 
@@ -79,4 +80,8 @@ export_k_mixture_model_LDADD = ../../src/libpinyin.la @GLIB2_LDFLAGS@
 
 k_mixture_model_to_interpolation_SOURCES = k_mixture_model_to_interpolation.cpp
 
-k_mixture_model_to_interpolation_LDADD = ../../src/libpinyin.la @GLIB2_LDFLAGS@
\ No newline at end of file
+k_mixture_model_to_interpolation_LDADD = ../../src/libpinyin.la @GLIB2_LDFLAGS@
+
+validate_k_mixture_model_SOURCES = validate_k_mixture_model.cpp
+
+validate_k_mixture_model_LDADD = ../../src/libpinyin.la @GLIB2_LDFLAGS@
diff --git a/utils/training/export_k_mixture_model.cpp b/utils/training/export_k_mixture_model.cpp
index f42df96..b10ee43 100644
--- a/utils/training/export_k_mixture_model.cpp
+++ b/utils/training/export_k_mixture_model.cpp
@@ -48,7 +48,7 @@ bool print_k_mixture_model_array_headers(FILE * output,
     for (size_t i = 0; i < items->len; ++i) {
         phrase_token_t * token = &g_array_index(items, phrase_token_t, i);
         KMixtureModelArrayHeader array_header;
-        bigram->get_array_header(*token, array_header);
+        assert(bigram->get_array_header(*token, array_header));
         char * phrase = taglib_token_to_string(phrase_index, *token);
         if ( phrase )
             fprintf(output, "\\item %s count %d\n", phrase, array_header.m_WC);
diff --git a/utils/training/validate_k_mixture_model.cpp b/utils/training/validate_k_mixture_model.cpp
new file mode 100644
index 0000000..3e5458d
--- /dev/null
+++ b/utils/training/validate_k_mixture_model.cpp
@@ -0,0 +1,134 @@
+/* 
+ *  libpinyin
+ *  Library to deal with pinyin.
+ *  
+ *  Copyright (C) 2011 Peng Wu <alexepico@gmail.com>
+ *  
+ *  This program is free software; you can redistribute it and/or modify
+ *  it under the terms of the GNU General Public License as published by
+ *  the Free Software Foundation; either version 2 of the License, or
+ *  (at your option) any later version.
+ * 
+ *  This program is distributed in the hope that it will be useful,
+ *  but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ *  GNU General Public License for more details.
+ *  
+ *  You should have received a copy of the GNU General Public License
+ *  along with this program; if not, write to the Free Software
+ *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
+ */
+
+#include "pinyin.h"
+#include "k_mixture_model.h"
+
+void print_help(){
+    printf("Usage: validate_k_mixture_model <FILENAME>\n");
+}
+
+bool validate_unigram(KMixtureModelBigram * bigram){
+    KMixtureModelMagicHeader magic_header;
+    if( !bigram->get_magic_header(magic_header) ){
+        fprintf(stderr, "no magic header in k mixture model.\n");
+        return false;
+    }
+
+    guint32 expected_sum = magic_header.m_WC;
+    if ( 0 == expected_sum ){
+        fprintf(stderr, "word count in magic header is unexpected zero.\n");
+        return false;
+    }
+    
+    GArray * items = g_array_new(FALSE, FALSE, sizeof(phrase_token_t));
+    bigram->get_all_items(items);
+
+    guint32 sum = 0;
+    for (size_t i = 0; i < items->len; ++i) {
+        phrase_token_t * token = &g_array_index(items, phrase_token_t, i);
+        KMixtureModelArrayHeader array_header;
+        assert(bigram->get_array_header(*token, array_header));
+        sum += array_header.m_WC;
+    }
+
+    if ( sum != expected_sum ){
+        fprintf(stderr, "word count in magic header:%d\n", expected_sum);
+        fprintf(stderr, "sum of word count in array headers:%d\n", sum);
+        fprintf(stderr, "the sum differs from word count.\n");
+        return false;
+    }
+
+    g_array_free(items, TRUE);
+    return true;
+}
+
+bool validate_bigram(KMixtureModelBigram * bigram){
+    bool result = true;
+
+    GArray * items = g_array_new(FALSE, FALSE, sizeof(phrase_token_t));
+    bigram->get_all_items(items);
+
+    for (size_t i = 0; i < items->len; ++i) {
+        phrase_token_t * token = &g_array_index(items, phrase_token_t, i);
+        KMixtureModelSingleGram * single_gram = NULL;
+        assert(bigram->load(*token, single_gram));
+        FlexibleBigramPhraseArray array = g_array_new
+            (FALSE, FALSE, sizeof(KMixtureModelArrayItemWithToken));
+        single_gram->retrieve_all(array);
+
+        KMixtureModelArrayHeader array_header;
+        assert(single_gram->get_array_header(array_header));
+
+        guint32 expected_sum = array_header.m_WC;
+        if ( 0 == expected_sum ){
+            fprintf(stderr, "in the array header of token %d:\n", *token);
+            fprintf(stderr, "word count is unexpected zero.\n");
+            result = false;
+        }
+
+        guint32 sum = 0;
+        for (size_t m = 0; m< array->len; ++m){
+            KMixtureModelArrayItemWithToken * item = &g_array_index(array, KMixtureModelArrayItemWithToken, m);
+
+            sum += item->m_item.m_WC;
+        }
+
+        if ( sum != expected_sum ){
+            fprintf(stderr, "word count in array header:%d\n", expected_sum);
+            fprintf(stderr, "sum of word count in array items:%d\n", sum);
+            fprintf(stderr, "the sum differs from word count.\n");
+            result = false;
+        }
+    }
+
+    g_array_free(items, TRUE);
+    return result;
+}
+
+int main(int argc, char * argv[]){
+    int i = 1;
+    const char * k_mixture_model_filename = NULL;
+
+    while ( i < argc ){
+        if ( strcmp ("--help", argv[i]) == 0 ){
+            print_help();
+            exit(0);
+        } else {
+            k_mixture_model_filename = argv[i];
+        }
+    }
+
+    KMixtureModelBigram bigram(K_MIXTURE_MODEL_MAGIC_NUMBER);
+    bigram.attach(k_mixture_model_filename, ATTACH_READONLY);
+
+    if (!validate_unigram(&bigram)) {
+        fprintf(stderr, "k mixture model validation failed.\n");
+        exit(ENODATA);
+    }
+
+    if (!validate_bigram(&bigram)) {
+        fprintf(stderr, "k mixture model validation failed.\n");
+        exit(ENODATA);
+    }
+
+    return 0;
+}