Closed veksha closed 3 years ago
Thanks for letting me know. This is clearly a regression. I'll look into it.
Issue should be fixed by the latest commit (ae67ad9). It was a regression during a code re-organization between 0.4 and 0.5, a check in the function which changes glyphmaps was omitted, which previously was forcing a flush of any buffered text drawing if the new glyphmap is different. The omission resulted in any leftover buffered text being rendered using the incorrect new glyphmap.
Hello. When I compile unicode example (commented out cjk stuff) I get the following result: Russian characters are drawing corrupted except for the last word.
But when make only one call to dtx_prepare_range like so:
dtx_prepare_range(fntmain, FONT_SZ, 0, 0x500); /* cyrilic */
(from zero to 0x500) then russian characters are drawing as expected: