X-Git-Url: http://www.chiark.greenend.org.uk/ucgi/~ianmdlvl/git?a=blobdiff_plain;f=src%2Fshared%2Futf8.c;h=013c110f07c072426615bd3b9f226038b2605d49;hb=d01efa07444e49aafd78e0aa822168da688d53f5;hp=4469a73751c3b5624d10c01a0eac59e74eb4e453;hpb=6ed62be0b15ac92d9687cff93c63d7b06d11b9f0;p=elogind.git diff --git a/src/shared/utf8.c b/src/shared/utf8.c index 4469a7375..013c110f0 100644 --- a/src/shared/utf8.c +++ b/src/shared/utf8.c @@ -263,39 +263,91 @@ char *ascii_is_valid(const char *str) { return (char*) str; } +/** + * utf8_encode_unichar() - Encode single UCS-4 character as UTF-8 + * @out_utf8: output buffer of at least 4 bytes or NULL + * @g: UCS-4 character to encode + * + * This encodes a single UCS-4 character as UTF-8 and writes it into @out_utf8. + * The length of the character is returned. It is not zero-terminated! If the + * output buffer is NULL, only the length is returned. + * + * Returns: The length in bytes that the UTF-8 representation does or would + * occupy. + */ +size_t utf8_encode_unichar(char *out_utf8, uint32_t g) { + if (g < (1 << 7)) { + if (out_utf8) + out_utf8[0] = g & 0x7f; + return 1; + } else if (g < (1 << 11)) { + if (out_utf8) { + out_utf8[0] = 0xc0 | ((g >> 6) & 0x1f); + out_utf8[1] = 0x80 | (g & 0x3f); + } + return 2; + } else if (g < (1 << 16)) { + if (out_utf8) { + out_utf8[0] = 0xe0 | ((g >> 12) & 0x0f); + out_utf8[1] = 0x80 | ((g >> 6) & 0x3f); + out_utf8[2] = 0x80 | (g & 0x3f); + } + return 3; + } else if (g < (1 << 21)) { + if (out_utf8) { + out_utf8[0] = 0xf0 | ((g >> 18) & 0x07); + out_utf8[1] = 0x80 | ((g >> 12) & 0x3f); + out_utf8[2] = 0x80 | ((g >> 6) & 0x3f); + out_utf8[3] = 0x80 | (g & 0x3f); + } + return 4; + } else { + return 0; + } +} + char *utf16_to_utf8(const void *s, size_t length) { - char *r; const uint8_t *f; - uint8_t *t; + char *r, *t; - r = new(char, (length*3+1)/2 + 1); + r = new(char, (length * 4 + 1) / 2 + 1); if (!r) return NULL; - t = (uint8_t*) r; + f = s; + t = r; - for (f = s; f < (const uint8_t*) s + length; f += 2) { - uint16_t c; + while (f < (const uint8_t*) s + length) { + uint16_t w1, w2; - c = (f[1] << 8) | f[0]; + /* see RFC 2781 section 2.2 */ - if (c == 0) { - *t = 0; - return r; - } else if (c < 0x80) { - *(t++) = (uint8_t) c; - } else if (c < 0x800) { - *(t++) = (uint8_t) (0xc0 | (c >> 6)); - *(t++) = (uint8_t) (0x80 | (c & 0x3f)); - } else { - *(t++) = (uint8_t) (0xe0 | (c >> 12)); - *(t++) = (uint8_t) (0x80 | ((c >> 6) & 0x3f)); - *(t++) = (uint8_t) (0x80 | (c & 0x3f)); + w1 = f[1] << 8 | f[0]; + f += 2; + + if (!utf16_is_surrogate(w1)) { + t += utf8_encode_unichar(t, w1); + + continue; } + + if (utf16_is_trailing_surrogate(w1)) + continue; + else if (f >= (const uint8_t*) s + length) + break; + + w2 = f[1] << 8 | f[0]; + f += 2; + + if (!utf16_is_trailing_surrogate(w2)) { + f -= 2; + continue; + } + + t += utf8_encode_unichar(t, utf16_surrogate_pair_to_unichar(w1, w2)); } *t = 0; - return r; }