Support Array of citext `citext[]` column

Nobody has claimed this yet.

Avram currently supports `Array(String)`, but if you need to search an array with case insensitive, then things get a little hairy

(assuming you have a GIN index on this column...)

```
=# select username from users where tags @> array['Lucky'];
username
-------------
jeremy
(1 row)

Time: 0.438 ms
=# select username from users where tags @> array['LUCKY'];
username
-------------
(0 rows)

Time: 0.815 ms
=# select username from users where lower(tags::text)::text[] @> array['lucky'];
username
-------------
jeremy
(1 row)

Time: 36.088 ms
```

Now if you use a `citext[]` column instead..

```
=# select username from users where tags @> array['LUCKY']::citext[];
username
-------------
jeremy
(1 row)

Time: 0.395 ms
=# select username from users where tags @> array['lUcKy']::citext[];
username
-------------
jeremy
(1 row)

Time: 0.408 ms
```

Start by tracing how Avram currently supports Array(String) columns and how PostgreSQL citext types are represented. Use the query examples in this issue to define the expected case-insensitive behavior for citext[]; done when such columns can be queried without the lower(tags::text) workaround.

Written by the indexing model from the issue text.