diff options
Diffstat (limited to 'lib/charset.c')
| -rw-r--r-- | lib/charset.c | 57 |
1 files changed, 56 insertions, 1 deletions
diff --git a/lib/charset.c b/lib/charset.c index ff76e88c77..8cd17ea1cb 100644 --- a/lib/charset.c +++ b/lib/charset.c @@ -6,7 +6,6 @@ * SPDX-License-Identifier: GPL-2.0+ */ -#include <common.h> #include <charset.h> #include <malloc.h> @@ -99,3 +98,59 @@ uint8_t *utf16_to_utf8(uint8_t *dest, const uint16_t *src, size_t size) return dest; } + +uint16_t *utf8_to_utf16(uint16_t *dest, const uint8_t *src, size_t size) +{ + while (size--) { + int extension_bytes; + uint32_t code; + + extension_bytes = 0; + if (*src <= 0x7f) { + code = *src++; + /* Exit on zero byte */ + if (!code) + size = 0; + } else if (*src <= 0xbf) { + /* Illegal code */ + code = '?'; + } else if (*src <= 0xdf) { + code = *src++ & 0x1f; + extension_bytes = 1; + } else if (*src <= 0xef) { + code = *src++ & 0x0f; + extension_bytes = 2; + } else if (*src <= 0xf7) { + code = *src++ & 0x07; + extension_bytes = 3; + } else { + /* Illegal code */ + code = '?'; + } + + for (; extension_bytes && size; --size, --extension_bytes) { + if ((*src & 0xc0) == 0x80) { + code <<= 6; + code |= *src++ & 0x3f; + } else { + /* Illegal code */ + code = '?'; + ++src; + --size; + break; + } + } + + if (code < 0x10000) { + *dest++ = code; + } else { + /* + * Simplified expression for + * (((code - 0x10000) >> 10) & 0x3ff) | 0xd800 + */ + *dest++ = (code >> 10) + 0xd7c0; + *dest++ = (code & 0x3ff) | 0xdc00; + } + } + return dest; +} |