diff options
Diffstat (limited to 'src/utf8.c')
| -rw-r--r-- | src/utf8.c | 74 |
1 files changed, 74 insertions, 0 deletions
diff --git a/src/utf8.c b/src/utf8.c new file mode 100644 index 0000000..07b751e --- /dev/null +++ b/src/utf8.c @@ -0,0 +1,74 @@ +#include "utf8.h" +#include <unicode/ustring.h> +#include <unicode/uchar.h> +#include <unicode/ubrk.h> +#include <string.h> +#include <stdlib.h> + +size_t utf8_display_width(const char *str) { + if (str == NULL) { + return 0; + } + + UErrorCode status = U_ZERO_ERROR; + int32_t utf8_len = (int32_t)strlen(str); + + // Convert UTF-8 to UTF-16 for ICU processing + UChar *utf16 = NULL; + int32_t utf16_len = 0; + + // First call to get required length + u_strFromUTF8(NULL, 0, &utf16_len, str, utf8_len, &status); + if (status != U_BUFFER_OVERFLOW_ERROR && U_FAILURE(status)) { + return 0; + } + + status = U_ZERO_ERROR; + utf16 = (UChar *)malloc((utf16_len + 1) * sizeof(UChar)); + if (utf16 == NULL) { + return 0; + } + + u_strFromUTF8(utf16, utf16_len + 1, NULL, str, utf8_len, &status); + if (U_FAILURE(status)) { + free(utf16); + return 0; + } + + // Use break iterator to count grapheme clusters + UBreakIterator *bi = ubrk_open(UBRK_CHARACTER, NULL, utf16, utf16_len, &status); + if (U_FAILURE(status)) { + free(utf16); + return 0; + } + + size_t width = 0; + int32_t start = ubrk_first(bi); + int32_t end = ubrk_next(bi); + + while (end != UBRK_DONE) { + // Get the first code point of this grapheme cluster + UChar32 ch; + int32_t idx = start; + U16_NEXT(utf16, idx, utf16_len, ch); + + // Determine display width based on East Asian Width property + int32_t ea_width = u_getIntPropertyValue(ch, UCHAR_EAST_ASIAN_WIDTH); + + // Full-width and Wide characters take 2 columns + if (ea_width == U_EA_FULLWIDTH || ea_width == U_EA_WIDE) { + width += 2; + } else { + // All others (Narrow, Halfwidth, Neutral, Ambiguous) take 1 column + width += 1; + } + + start = end; + end = ubrk_next(bi); + } + + ubrk_close(bi); + free(utf16); + + return width; +} |