X-Git-Url: https://www.chiark.greenend.org.uk/ucgi/~ianmdlvl/git?p=elogind.git;a=blobdiff_plain;f=src%2Fshared%2Futf8.h;h=e745649f06f5b06fcbe7b0869a798f068820d403;hp=f805ea6b590608a988aa85fe4307577f53b6176e;hb=71573c4556aef59cb31c9cf48e8e8f74d01df537;hpb=31f7bf1994523f5b8fd014c69b97e09d7043d9ff diff --git a/src/shared/utf8.h b/src/shared/utf8.h index f805ea6b5..e745649f0 100644 --- a/src/shared/utf8.h +++ b/src/shared/utf8.h @@ -25,12 +25,33 @@ #include "macro.h" -char *utf8_is_valid(const char *s) _pure_; +#define UTF8_REPLACEMENT_CHARACTER "\xef\xbf\xbd" + +bool unichar_is_valid(uint32_t c); + +const char *utf8_is_valid(const char *s) _pure_; char *ascii_is_valid(const char *s) _pure_; -bool utf8_is_printable(const char* str, size_t length) _pure_; +bool utf8_is_printable_newline(const char* str, size_t length, bool newline) _pure_; +#define utf8_is_printable(str, length) utf8_is_printable_newline(str, length, true) -char *utf8_filter(const char *s); -char *ascii_filter(const char *s); +char *utf8_escape_invalid(const char *s); +char *utf8_escape_non_printable(const char *str); +size_t utf8_encode_unichar(char *out_utf8, uint32_t g); char *utf16_to_utf8(const void *s, size_t length); + +int utf8_encoded_valid_unichar(const char *str); +int utf8_encoded_to_unichar(const char *str); + +static inline bool utf16_is_surrogate(uint16_t c) { + return (0xd800 <= c && c <= 0xdfff); +} + +static inline bool utf16_is_trailing_surrogate(uint16_t c) { + return (0xdc00 <= c && c <= 0xdfff); +} + +static inline uint32_t utf16_surrogate_pair_to_unichar(uint16_t lead, uint16_t trail) { + return ((lead - 0xd800) << 10) + (trail - 0xdc00) + 0x10000; +}