Added FreeTypeFont has_characters()

[radarhere]

Resolves #9158. The issue requested a method to check if glyphs are present in a font. This question has arisen previously in #7380 (comment), and I think this would also help users from #4808, in a less elegant way.

This adds FreeTypeFont.has_characters(text). I expect that the issue was only thinking about checking a single character at a time, but if user wanted to check multiple characters at once, I don't see any reason not to let them. If any character in the string is missing, then False is returned.

The C code in here is based on

Pillow/src/_imagingft.c

Lines 432 to 463 in 9d39fe6

	if (PyUnicode_Check(string)) {
	count = PyUnicode_GET_LENGTH(string);
	} else if (PyBytes_Check(string)) {
	PyBytes_AsStringAndSize(string, &buffer, &count);
	} else {
	PyErr_SetString(PyExc_TypeError, "expected string or bytes");
	return 0;
	}
	if (count == 0) {
	return 0;
	}
	(*glyph_info) = PyMem_New(GlyphInfo, count);
	if ((*glyph_info) == NULL) {
	PyErr_SetString(PyExc_MemoryError, "PyMem_New() failed");
	return 0;
	}
	load_flags = FT_LOAD_DEFAULT;
	if (mask) {
	load_flags |= FT_LOAD_TARGET_MONO;
	}
	if (color) {
	load_flags |= FT_LOAD_COLOR;
	}
	for (i = 0; i < count; i++) {
	if (buffer) {
	ch = buffer[i];
	} else {
	ch = PyUnicode_READ_CHAR(string, i);
	}
	(*glyph_info)[i].index = FT_Get_Char_Index(self->face, ch);

[hugovk]

I'm on the fence with this. It feels outside the core focus of Pillow, and it's not too tricky to do this with Pillow in Python:

#9158 (comment)

Plus it's a even easier with a dedicated FreeType library:

#9158 (comment)

[glebm]

A font is roughly a collection of images and checking if an image exists in the collection is pretty basic functionality?

Of course it's doable in other ways but:

1. The pillow way suggested is verbose and inefficient.
2. The freetype way requires loading the font twice (once in Pillow, once in Freetype).

[radarhere]

If there is no consensus on this, should the issue be closed? We can always return to the idea again if someone else raises it in the future.

[aclark4life]

Any downside assuming we determine it's in scope?

[radarhere]

To adding the code? Only the usual theoretical maintenance burden. But it's a new method, so it doesn't affect anything else, and it's relatively simple code, so I can't foresee any problems.

To returning to it later? No. I don't see us refactoring the underlying font handling that this method relies on anytime soon.