lexicon.h
1.4 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
// sherpa-onnx/csrc/lexicon.h
//
// Copyright (c) 2022-2023 Xiaomi Corporation
#ifndef SHERPA_ONNX_CSRC_LEXICON_H_
#define SHERPA_ONNX_CSRC_LEXICON_H_
#include <cstdint>
#include <string>
#include <unordered_map>
#include <unordered_set>
#include <vector>
namespace sherpa_onnx {
// TODO(fangjun): Refactor it to an abstract class
class Lexicon {
public:
Lexicon(const std::string &lexicon, const std::string &tokens,
const std::string &punctuations, const std::string &language);
std::vector<int64_t> ConvertTextToTokenIds(const std::string &text) const;
private:
std::vector<int64_t> ConvertTextToTokenIdsEnglish(
const std::string &text) const;
std::vector<int64_t> ConvertTextToTokenIdsChinese(
const std::string &text) const;
void InitLanguage(const std::string &lang);
void InitTokens(const std::string &tokens);
void InitLexicon(const std::string &lexicon);
void InitPunctuations(const std::string &punctuations);
private:
enum class Language {
kEnglish,
kChinese,
kUnknown,
};
private:
std::unordered_map<std::string, std::vector<int32_t>> word2ids_;
std::unordered_set<std::string> punctuations_;
std::unordered_map<std::string, int32_t> token2id_;
int32_t blank_ = -1; // ID for the blank token
Language language_;
//
};
} // namespace sherpa_onnx
#endif // SHERPA_ONNX_CSRC_LEXICON_H_