src/basic/utf8.h

   1 /* SPDX-License-Identifier: LGPL-2.1+ */
   2 #pragma once
   3
   4 /***
   5   This file is part of systemd.
   6
   7   Copyright 2012 Lennart Poettering
   8 ***/
   9
  10 #include <stdbool.h>
  11 #include <stddef.h>
  12 #include <stdint.h>
  13 #include <uchar.h>
  14
  15 #include "macro.h"
  16 #include "missing.h"
  17
  18 #define UTF8_REPLACEMENT_CHARACTER "\xef\xbf\xbd"
  19 #define UTF8_BYTE_ORDER_MARK "\xef\xbb\xbf"
  20
  21 bool unichar_is_valid(char32_t c);
  22
  23 const char *utf8_is_valid(const char *s) _pure_;
  24 char *ascii_is_valid(const char *s) _pure_;
  25
  26 bool utf8_is_printable_newline(const char* str, size_t length, bool newline) _pure_;
  27 #define utf8_is_printable(str, length) utf8_is_printable_newline(str, length, true)
  28
  29 char *utf8_escape_invalid(const char *s);
  30 char *utf8_escape_non_printable(const char *str);
  31
  32 size_t utf8_encode_unichar(char *out_utf8, char32_t g);
  33 char *utf16_to_utf8(const void *s, size_t length);
  34
  35 int utf8_encoded_valid_unichar(const char *str);
  36 int utf8_encoded_to_unichar(const char *str, char32_t *ret_unichar);
  37
  38 static inline bool utf16_is_surrogate(char16_t c) {
  39         return (0xd800 <= c && c <= 0xdfff);
  40 }
  41
  42 static inline bool utf16_is_trailing_surrogate(char16_t c) {
  43         return (0xdc00 <= c && c <= 0xdfff);
  44 }
  45
  46 static inline char32_t utf16_surrogate_pair_to_unichar(char16_t lead, char16_t trail) {
  47                 return ((lead - 0xd800) << 10) + (trail - 0xdc00) + 0x10000;
  48 }
  49
  50 size_t utf8_n_codepoints(const char *str);