diff options
author | Rich Felker <dalias@aerifal.cx> | 2014-10-13 20:59:42 -0400 |
---|---|---|
committer | Rich Felker <dalias@aerifal.cx> | 2014-10-13 20:59:42 -0400 |
commit | ab9672ae73248f51e30f4553c4b8878525e46383 (patch) | |
tree | fd3cfee2bcc10a64b83009af92f0de09703dfa75 /src/multibyte | |
parent | 00548408398ced546c540dab773ea66cea4fe1c2 (diff) | |
download | musl-ab9672ae73248f51e30f4553c4b8878525e46383.tar.gz musl-ab9672ae73248f51e30f4553c4b8878525e46383.tar.xz musl-ab9672ae73248f51e30f4553c4b8878525e46383.zip |
implement uchar.h (C11 UTF-16/32 conversion) interfaces
Diffstat (limited to 'src/multibyte')
-rw-r--r-- | src/multibyte/c16rtomb.c | 33 | ||||
-rw-r--r-- | src/multibyte/c32rtomb.c | 7 | ||||
-rw-r--r-- | src/multibyte/mbrtoc16.c | 28 | ||||
-rw-r--r-- | src/multibyte/mbrtoc32.c | 11 |
4 files changed, 79 insertions, 0 deletions
diff --git a/src/multibyte/c16rtomb.c b/src/multibyte/c16rtomb.c new file mode 100644 index 00000000..2e8ec970 --- /dev/null +++ b/src/multibyte/c16rtomb.c @@ -0,0 +1,33 @@ +#include <uchar.h> +#include <errno.h> +#include <wchar.h> + +size_t c16rtomb(char *restrict s, char16_t c16, mbstate_t *restrict ps) +{ + unsigned *x = (unsigned *)ps; + wchar_t wc; + + if (!s) { + if (*x) goto ilseq; + return 1; + } + + if (!*x && c16 - 0xd800u < 0x400) { + *x = c16 - 0xd7c0 << 10; + return 0; + } + + if (*x) { + if (c16 - 0xdc00u >= 0x400) goto ilseq; + else wc = *x + c16 - 0xdc00; + *x = 0; + } else { + wc = c16; + } + return wcrtomb(s, wc, 0); + +ilseq: + *x = 0; + errno = EILSEQ; + return -1; +} diff --git a/src/multibyte/c32rtomb.c b/src/multibyte/c32rtomb.c new file mode 100644 index 00000000..67851328 --- /dev/null +++ b/src/multibyte/c32rtomb.c @@ -0,0 +1,7 @@ +#include <uchar.h> +#include <wchar.h> + +size_t c32rtomb(char *restrict s, char32_t c32, mbstate_t *restrict ps) +{ + return wcrtomb(s, c32, ps); +} diff --git a/src/multibyte/mbrtoc16.c b/src/multibyte/mbrtoc16.c new file mode 100644 index 00000000..74b7d77e --- /dev/null +++ b/src/multibyte/mbrtoc16.c @@ -0,0 +1,28 @@ +#include <uchar.h> +#include <wchar.h> + +size_t mbrtoc16(char16_t *restrict pc16, const char *restrict s, size_t n, mbstate_t *restrict ps) +{ + unsigned *pending = (unsigned *)ps; + + if (!s) return mbrtoc16(0, "", 1, ps); + + /* mbrtowc states for partial UTF-8 characters have the high bit set; + * we use nonzero states without high bit for pending surrogates. */ + if ((int)*pending > 0) { + if (pc16) *pc16 = *pending; + *pending = 0; + return -3; + } + + wchar_t wc; + size_t ret = mbrtowc(&wc, s, n, ps); + if (ret <= 4) { + if (wc >= 0x10000) { + *pending = (wc & 0x3ff) + 0xdc00; + wc = 0xd7c0 + (wc >> 10); + } + if (pc16) *pc16 = wc; + } + return ret; +} diff --git a/src/multibyte/mbrtoc32.c b/src/multibyte/mbrtoc32.c new file mode 100644 index 00000000..c6d20824 --- /dev/null +++ b/src/multibyte/mbrtoc32.c @@ -0,0 +1,11 @@ +#include <uchar.h> +#include <wchar.h> + +size_t mbrtoc32(char32_t *restrict pc32, const char *restrict s, size_t n, mbstate_t *restrict ps) +{ + if (!s) return mbrtoc32(0, "", 1, ps); + wchar_t wc; + size_t ret = mbrtowc(&wc, s, n, ps); + if (ret <= 4 && pc32) *pc32 = wc; + return ret; +} |