/src/im/gpinyin/include/spellingtable.h

  1/*
  2 * Copyright (C) 2009 The Android Open Source Project
  3 *
  4 * Licensed under the Apache License, Version 2.0 (the "License");
  5 * you may not use this file except in compliance with the License.
  6 * You may obtain a copy of the License at
  7 *
  8 *      http://www.apache.org/licenses/LICENSE-2.0
  9 *
 10 * Unless required by applicable law or agreed to in writing, software
 11 * distributed under the License is distributed on an "AS IS" BASIS,
 12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 13 * See the License for the specific language governing permissions and
 14 * limitations under the License.
 15 */
 16
 17#ifndef PINYINIME_INCLUDE_SPELLINGTABLE_H__
 18#define PINYINIME_INCLUDE_SPELLINGTABLE_H__
 19
 20#include <stdlib.h>
 21#include "./dictdef.h"
 22
 23namespace ime_pinyin {
 24
 25#ifdef ___BUILD_MODEL___
 26
 27const unsigned kMaxSpellingSize = kMaxPinyinSize;
 28
 29typedef struct {
 30  char str[kMaxSpellingSize + 1];
 31  double freq;
 32} RawSpelling, *PRawSpelling;
 33
 34// This class is used to store the spelling strings
 35// The length of the input spelling string should be less or equal to the
 36// spelling_size_ (set by init_table). If the input string is too long,
 37// we only keep its first spelling_size_ chars.
 38class SpellingTable {
 39 private:
 40  static const unsigned kNotSupportNum = 3;
 41  static const char kNotSupportList[kNotSupportNum][kMaxSpellingSize + 1];
 42
 43  bool need_score_;
 44
 45  unsigned spelling_max_num_;
 46
 47  RawSpelling *raw_spellings_;
 48
 49  // Used to store spelling strings. If the spelling table needs to calculate
 50  // score, an extra char after each spelling string is the score.
 51  // An item with a lower score has a higher probability.
 52  char *spelling_buf_;
 53  unsigned spelling_size_;
 54
 55  double total_freq_;
 56
 57  unsigned spelling_num_;
 58
 59  double score_amplifier_;
 60
 61  unsigned char average_score_;
 62
 63  // If frozen is true, put_spelling() and contain() are not allowed to call.
 64  bool frozen_;
 65
 66  unsigned get_hash_pos(const char* spelling_str);
 67  unsigned hash_pos_next(unsigned hash_pos);
 68  void free_resource();
 69 public:
 70  SpellingTable();
 71  ~SpellingTable();
 72
 73  // pure_spl_size is the pure maximum spelling string size. For example,
 74  // "zhuang" is the longgest item in Pinyin, so pure_spl_size should be 6.
 75  // spl_max_num is the maximum number of spelling strings to store.
 76  // need_score is used to indicate whether the caller needs to calculate a
 77  // score for each spelling.
 78  bool init_table(unsigned pure_spl_size, unsigned spl_max_num, bool need_score);
 79
 80  // Put a spelling string to the table.
 81  // It always returns false if called after arrange() withtout a new
 82  // init_table() operation.
 83  // freq is the spelling's occuring count.
 84  // If the spelling has been in the table, occuring count will accumulated.
 85  bool put_spelling(const char* spelling_str, double spl_count);
 86
 87  // Test whether a spelling string is in the table.
 88  // It always returns false, when being called after arrange() withtout a new
 89  // init_table() operation.
 90  bool contain(const char* spelling_str);
 91
 92  // Sort the spelling strings and put them from the begin of the buffer.
 93  // Return the pointer of the sorted spelling strings.
 94  // item_size and spl_num return the item size and number of spelling.
 95  // Because each spelling uses a '\0' as terminator, the returned item_size is
 96  // at least one char longer than the spl_size parameter specified by
 97  // init_table(). If the table is initialized to calculate score, item_size
 98  // will be increased by 1, and current_spl_str[item_size - 1] stores an
 99  // unsinged char score.
100  // An item with a lower score has a higher probability.
101  // Do not call put_spelling() and contains() after arrange().
102  const char* arrange(unsigned *item_size, unsigned *spl_num);
103
104  float get_score_amplifier();
105
106  unsigned char get_average_score();
107};
108#endif  // ___BUILD_MODEL___
109}
110
111#endif  // PINYINIME_INCLUDE_SPELLINGTABLE_H__