#37 Rewrite the buffer gap to use `uint32_t`

Right now the buffer gap uses char_t because Anthony's Editor was written that way. I handle unicode values by splitting them up into chars and then termbox handles the displaying. Since termbox handles the displaying and uses uint32_t I'm pretty sure we can simplify the code and the utf8 detection stuff that Hugh Barney added in by doing this. Probably will take a long time to do.

In order to really support unicode or multi-byte UTF-8, this needs to happen.

I'm not sure the last comment is accurate because UTF-8 and unicode are working currently but there are just some issues. I'm not sure the squeeze is worth the juice to actually do this.

This won't be necessary.