4-byte utf-8 sequences (emoji) are broken inside litehtml about litehtml HOT 4 CLOSED

You are on Windows and you are using litehtml_to_wchar/utf8_to_wchar in your container, right? Those functions incorrectly work with Unicode characters above 0xFFFF on Windows.

from litehtml.

Just checked this issue on Windows. Actually litehtml converts from/to UTF-32. On Linux sizeof(wchar_t) == 4, on Windows sizeof(wchar_t) == 2. That's why there are issues with unicode characters above 0xFFFF on Windows.

Replaced wchar_t with char32_t and renamed utf8_to_wchar to utf8_to_utf32 and whar_to_utf8 to utf32_to_utf8

from litehtml.

Nice. Finally all vestiges of wchar_t are eradicated.

from litehtml.

@stasoid @tordex thanks for the fix, but there're also 5-6 bytes long utf-8 sequences, so anyway you'd have to deal with UTF32-LE instead of UTF16-LE in the previous case

from litehtml.