What data structures for indexes should be used for various data characteristics?

Question

Database

For example, for data with uniformly distributed unique (or near-unique) keys, btree is a good option, if sequential selections are often made on such data - can be improved to b+tree.
And what to use for an index, if one key is not unique, and one key has more records than btree dimension?
What is the best data structure to implement a full-text index?
I would be happy with a good answer, or a link to a good answer, a link to a good book on the topic, I will also be glad