Fix Font.getGlyph(s) with emoji characters

[tobil4sk]

The original code didn't work for emojis, but the stack overflow discussion since then includes a fix for this issue. See the comment:

https://stackoverflow.com/questions/2948308/how-do-i-read-utf-8-characters-via-a-pointer#comment72311507_2953960

Now readNextChar can extract the unicode value correctly, however, it still does not result in a correct glyph index being found.

[tobil4sk]

I was able to get it working properly by adding in FT_Select_Charmap((FT_Face)face, FT_ENCODING_UNICODE);. We already have some code to set the char map, so that probably needs to be updated or replaced:

lime/project/src/text/Font.cpp

Lines 402 to 422 in a9f72d6

	/* Set charmap
	*
	* See http://www.microsoft.com/typography/otspec/name.htm for a list of
	* some possible platform-encoding pairs. We're interested in 0-3 aka 3-1
	* - UCS-2. Otherwise, fail. If a font has some unicode map, but lacks
	* UCS-2 - it is a broken or irrelevant font. What exactly Freetype will
	* select on face load (it promises most wide unicode, and if that will be
	* slower that UCS-2 - left as an excercise to check.
	*/
	for (int i = 0; i < ((FT_Face)face)->num_charmaps; i++) {
	FT_UShort pid = ((FT_Face)face)->charmaps[i]->platform_id;
	FT_UShort eid = ((FT_Face)face)->charmaps[i]->encoding_id;
	if (((pid == 0) && (eid == 3)) || ((pid == 3) && (eid == 1))) {
	FT_Set_Charmap ((FT_Face)face, ((FT_Face)face)->charmaps[i]);
	}
	}