mirror of
https://github.com/ggml-org/llama.cpp.git
synced 2025-10-28 08:31:25 +00:00
b43272afa2
* Replace CODEPOINT_TYPE_* with codepoint_flags * Update and bugfix brute force random test * Deterministic brute force random test * Unicode normalization NFD * Get rid of BOM
64 lines
2.0 KiB
C++
64 lines
2.0 KiB
C++
#pragma once
|
|
|
|
#include <cstdint>
|
|
#include <string>
|
|
#include <vector>
|
|
|
|
struct codepoint_flags {
|
|
enum {
|
|
UNDEFINED = 0x0001,
|
|
NUMBER = 0x0002, // regex: \p{N}
|
|
LETTER = 0x0004, // regex: \p{L}
|
|
SEPARATOR = 0x0008, // regex: \p{Z}
|
|
ACCENT_MARK = 0x0010, // regex: \p{M}
|
|
PUNCTUATION = 0x0020, // regex: \p{P}
|
|
SYMBOL = 0x0040, // regex: \p{S}
|
|
CONTROL = 0x0080, // regex: \p{C}
|
|
MASK_CATEGORIES = 0x00FF,
|
|
};
|
|
|
|
// codepoint type
|
|
uint16_t is_undefined : 1;
|
|
uint16_t is_number : 1; // regex: \p{N}
|
|
uint16_t is_letter : 1; // regex: \p{L}
|
|
uint16_t is_separator : 1; // regex: \p{Z}
|
|
uint16_t is_accent_mark : 1; // regex: \p{M}
|
|
uint16_t is_punctuation : 1; // regex: \p{P}
|
|
uint16_t is_symbol : 1; // regex: \p{S}
|
|
uint16_t is_control : 1; // regex: \p{C}
|
|
// helper flags
|
|
uint16_t is_whitespace : 1; // regex: \s
|
|
uint16_t is_lowercase : 1;
|
|
uint16_t is_uppercase : 1;
|
|
uint16_t is_nfd : 1;
|
|
|
|
// decode from uint16
|
|
inline codepoint_flags(const uint16_t flags=0) {
|
|
*reinterpret_cast<uint16_t*>(this) = flags;
|
|
}
|
|
|
|
inline uint16_t as_uint() const {
|
|
return *reinterpret_cast<const uint16_t*>(this);
|
|
}
|
|
|
|
inline uint16_t category_flag() const {
|
|
return this->as_uint() & MASK_CATEGORIES;
|
|
}
|
|
};
|
|
|
|
|
|
std::string unicode_cpt_to_utf8(uint32_t cp);
|
|
std::vector<uint32_t> unicode_cpts_from_utf8(const std::string & utf8);
|
|
|
|
std::vector<uint32_t> unicode_cpts_normalize_nfd(const std::vector<uint32_t> & cpts);
|
|
|
|
codepoint_flags unicode_cpt_flags(const uint32_t cp);
|
|
codepoint_flags unicode_cpt_flags(const std::string & utf8);
|
|
|
|
std::string unicode_byte_to_utf8(uint8_t byte);
|
|
uint8_t unicode_utf8_to_byte(const std::string & utf8);
|
|
|
|
char32_t unicode_tolower(char32_t cp);
|
|
|
|
std::vector<std::string> unicode_regex_split(const std::string & text, const std::vector<std::string> & regex_exprs);
|