Пример #1
0
int main(int argc, char * argv[]) {
    SystemTableInfo2 system_table_info;

    bool retval = system_table_info.load("../../data/table.conf");
    if (!retval) {
        fprintf(stderr, "load table.conf failed.\n");
        exit(ENOENT);
    }

    pinyin_option_t options = PINYIN_CORRECT_ALL | PINYIN_AMB_ALL | PINYIN_INCOMPLETE;

    FacadeChewingTable2 largetable;
    largetable.load("../../data/pinyin_index.bin", NULL);

    const pinyin_table_info_t * phrase_files =
        system_table_info.get_default_tables();

    FacadePhraseIndex phrase_index;
    if (!load_phrase_index(phrase_files, &phrase_index))
        exit(ENOENT);

    PhoneticParser2 * parser = new FullPinyinParser2();
    ChewingKeyVector keys = g_array_new(FALSE, FALSE, sizeof(ChewingKey));
    ChewingKeyRestVector key_rests =
        g_array_new(FALSE, FALSE, sizeof(ChewingKeyRest));

    PhoneticKeyMatrix matrix;

    char* linebuf = NULL; size_t size = 0; ssize_t read;
    while( (read = getline(&linebuf, &size, stdin)) != -1 ){
        if ( '\n' == linebuf[strlen(linebuf) - 1] ) {
            linebuf[strlen(linebuf) - 1] = '\0';
        }

        if ( strcmp ( linebuf, "quit" ) == 0)
            break;

        int len = 0;
        guint32 start_time = record_time();
        for (size_t i = 0; i < bench_times; ++i) {
            matrix.clear_all();

            len = parser->parse(options, keys, key_rests,
                                linebuf, strlen(linebuf));

            fill_matrix(&matrix, keys, key_rests, len);

            resplit_step(options, &matrix);

            inner_split_step(options, &matrix);

            fuzzy_syllable_step(options, &matrix);
        }
        print_time(start_time, bench_times);

        printf("parsed %d chars, %d keys.\n", len, keys->len);

        dump_matrix(&matrix);

        PhraseIndexRanges ranges;
        memset(ranges, 0, sizeof(PhraseIndexRanges));

        phrase_index.prepare_ranges(ranges);

        for (size_t i = 0; i < matrix.size(); ++i) {
            for (size_t j = i + 1; j < matrix.size(); ++j) {
                phrase_index.clear_ranges(ranges);

                printf("search index: start %ld\t end %ld\n", i, j);
                int retval = search_matrix(&largetable, &matrix, i, j, ranges);

#if 0
                if (retval & SEARCH_OK) {
                    dump_ranges(&phrase_index, ranges);
                    test_pronunciation_possibility
                        (&matrix, i, j, &phrase_index, ranges);
                }
#endif

                if (!(retval & SEARCH_CONTINUED))
                    break;
            }
        }

        phrase_index.destroy_ranges(ranges);
    }

    if (linebuf)
        free(linebuf);

    delete parser;

    g_array_free(key_rests, TRUE);
    g_array_free(keys, TRUE);

    return 0;
}
Пример #2
0
int main(int argc, char * argv[]) {
    GError * error = NULL;
    GOptionContext * context;

    context = g_option_context_new("- test pinyin parser");
    g_option_context_add_main_entries(context, entries, NULL);
    if (!g_option_context_parse(context, &argc, &argv, &error)) {
        g_print("option parsing failed:%s\n", error->message);
        exit(EINVAL);
    }

    pinyin_option_t options = PINYIN_CORRECT_ALL | USE_TONE | USE_RESPLIT_TABLE;
    if (incomplete)
        options |= PINYIN_INCOMPLETE | ZHUYIN_INCOMPLETE;

    PhoneticParser2 * parser = NULL;
    ChewingKeyVector keys = g_array_new(FALSE, FALSE, sizeof(ChewingKey));
    ChewingKeyRestVector key_rests =
        g_array_new(FALSE, FALSE, sizeof(ChewingKeyRest));

    /* create the parser */
    if (strcmp("fullpinyin", parsername) == 0) {
        parser = new FullPinyinParser2();
    } else if (strcmp("doublepinyin", parsername) == 0) {
        parser = new DoublePinyinParser2();
    } else if (strcmp("zhuyin", parsername) == 0) {
        if (strcmp("standard", schemename) == 0) {
            parser = new ZhuyinSimpleParser2();
        } else if (strcmp("hsu", schemename) == 0) {
            parser = new ZhuyinDiscreteParser2();
        } else if (strcmp("dachen26", schemename) == 0) {
            parser = new ZhuyinDaChenCP26Parser2();
        }
    } else if (strcmp("pinyindirect", parsername) == 0) {
        parser = new PinyinDirectParser2();
    } else if (strcmp("zhuyindirect", parsername) == 0) {
        parser = new ZhuyinDirectParser2();
    }


    if (!parser)
        parser = new FullPinyinParser2();

    char* linebuf = NULL; size_t size = 0; ssize_t read;
    while( (read = getline(&linebuf, &size, stdin)) != -1 ){
        if ( '\n' == linebuf[strlen(linebuf) - 1] ) {
            linebuf[strlen(linebuf) - 1] = '\0';
        }

        if ( strcmp ( linebuf, "quit" ) == 0)
            break;

#if 0
        ChewingKey key;
        bool success = parser->parse_one_key(options, key,
                                             linebuf, strlen(linebuf));
        if (success) {
            gchar * pinyins = key.get_pinyin_string();
            printf("pinyin:%s\n", pinyins);
            g_free(pinyins);
        }
#endif

#if 1
        int len = 0;
        guint32 start_time = record_time();
        for ( size_t i = 0; i < bench_times; ++i)
            len = parser->parse(options, keys, key_rests,
                                linebuf, strlen(linebuf));

        print_time(start_time, bench_times);

        printf("parsed %d chars, %d keys.\n", len, keys->len);

        assert(keys->len == key_rests->len);

        for (size_t i = 0; i < keys->len; ++i) {
            ChewingKey * key =
                &g_array_index(keys, ChewingKey, i);
            ChewingKeyRest * key_rest =
                &g_array_index(key_rests, ChewingKeyRest, i);

            gchar * pinyins = key->get_pinyin_string();
            printf("%s %d %d\t", pinyins,
                   key_rest->m_raw_begin, key_rest->m_raw_end);
            g_free(pinyins);
        }
        printf("\n");
#endif

    }

    if (linebuf)
        free(linebuf);

    delete parser;

    g_array_free(key_rests, TRUE);
    g_array_free(keys, TRUE);

    return 0;
}