Make `GraphemeCat` and `grapheme_category` public

[Hedgehogo]

This is necessary so that you can analyze the grapheme you have already received, determine its structure, and categorize it. For example, to determine whether it is alphabetic, even when there are diacritics. This analysis itself is outside the area of responsibility of your library, I only ask you to make the existing structure public so that others do not have to duplicate, maintain and synchronize all the relevant tables.

[Hedgehogo]

And a couple of functions for the incb tables.

[Manishearth]

I'm not convinced this is actually useful for other processes: the categories are designed specifically for segmentation and do not necessarily have meaning outside of that.

For things like "determine the class of a grapheme" I would recommend using the ccc property to ignore diacritics, and the General_Category property for categorization.

icu_properties from ICU4X has these properties (it also exposes GraphemeBreak though again I do not think that property is generally useful)

[Hedgehogo]

However, icu_properties does not provide the Indic_Conjunct_Break property, as well as any other crates in general. But it seems to me that this is important enough for analysis

[Manishearth]

ICU4X might be convinced to expose that.

But this issue would not solve that problem for you either, GraphemeCat doesn't cover InCB.

(and I'm not convinced InCB has general uses outside of the grapheme algorithm, anyway)