In this chapter, we started by examining the possible techniques used to represent heterogeneous data in a relational database. All the approaches we discussed have some pros and some cons, so the best choice depends on our needs.

Then, we discussed dynamic columns. They solve the problem in an easy way, by treating a table column as a dynamic object. In fact, different rows can have different dynamic columns. Attributes can even be added or dropped at any time. We examined the functions used to do this. We also demonstrated that dynamic columns can be nested.

In the last section, we built a PERSISTENT virtual column on a dynamic column, and then indexed the PERSISTENT column. This is a very important technique, because it is the only way to index a dynamic column.

In the next chapter, you will learn how to make full-text search queries with MariaDB.