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_SPLPARSER_H__
18#define PINYINIME_INCLUDE_SPLPARSER_H__
19
20#include "./dictdef.h"
21#include "./spellingtrie.h"
22
23namespace ime_pinyin {
24
25class SpellingParser {
26 protected:
27  const SpellingTrie *spl_trie_;
28
29 public:
30  SpellingParser();
31
32  // Given a string, parse it into a spelling id stream.
33  // If the whole string are sucessfully parsed, last_is_pre will be true;
34  // if the whole string is not fullly parsed, last_is_pre will return whether
35  // the last part of the string is a prefix of a full spelling string. For
36  // example, given string "zhengzhon", "zhon" is not a valid speling, but it is
37  // the prefix of "zhong".
38  //
39  // If splstr starts with a character not in ['a'-z'] (it is a split char),
40  // return 0.
41  // Split char can only appear in the middle of the string or at the end.
42  uint16 splstr_to_idxs(const char *splstr, uint16 str_len, uint16 splidx[],
43                        uint16 start_pos[], uint16 max_size, bool &last_is_pre);
44
45  // Similar to splstr_to_idxs(), the only difference is that splstr_to_idxs()
46  // convert single-character Yunmus into half ids, while this function converts
47  // them into full ids.
48  uint16 splstr_to_idxs_f(const char *splstr, uint16 str_len, uint16 splidx[],
49          uint16 start_pos[], uint16 max_size, bool &last_is_pre);
50
51  // Similar to splstr_to_idxs(), the only difference is that this function
52  // uses char16 instead of char8.
53  uint16 splstr16_to_idxs(const char16 *splstr, uint16 str_len, uint16 splidx[],
54                        uint16 start_pos[], uint16 max_size, bool &last_is_pre);
55
56  // Similar to splstr_to_idxs_f(), the only difference is that this function
57  // uses char16 instead of char8.
58  uint16 splstr16_to_idxs_f(const char16 *splstr16, uint16 str_len,
59                            uint16 splidx[], uint16 start_pos[],
60                            uint16 max_size, bool &last_is_pre);
61
62  // If the given string is a spelling, return the id, others, return 0.
63  // If the give string is a single char Yunmus like "A", and the char is
64  // enabled in ShouZiMu mode, the returned spelling id will be a half id.
65  // When the returned spelling id is a half id, *is_pre returns whether it
66  // is a prefix of a full spelling string.
67  uint16 get_splid_by_str(const char *splstr, uint16 str_len, bool *is_pre);
68
69  // If the given string is a spelling, return the id, others, return 0.
70  // If the give string is a single char Yunmus like "a", no matter the char
71  // is enabled in ShouZiMu mode or not, the returned spelling id will be
72  // a full id.
73  // When the returned spelling id is a half id, *p_is_pre returns whether it
74  // is a prefix of a full spelling string.
75  uint16 get_splid_by_str_f(const char *splstr, uint16 str_len, bool *is_pre);
76
77  // Splitter chars are not included.
78  bool is_valid_to_parse(char ch);
79
80  // When auto-correction is not enabled, get_splid_by_str() will be called to
81  // return the single result. When auto-correction is enabled, this function
82  // will be called to get the results. Auto-correction is not ready.
83  // full_id_num returns number of full spelling ids.
84  // is_pre returns whether the given string is the prefix of a full spelling
85  // string.
86  // If splstr starts with a character not in [a-zA-Z] (it is a split char),
87  // return 0.
88  // Split char can only appear in the middle of the string or at the end.
89  // The caller should guarantee NULL != splstr && str_len > 0 && NULL != splidx
90  uint16 get_splids_parallel(const char *splstr, uint16 str_len,
91                             uint16 splidx[], uint16 max_size,
92                             uint16 &full_id_num, bool &is_pre);
93};
94}
95
96#endif  // PINYINIME_INCLUDE_SPLPARSER_H__
97