Add new `having` methods

I think using HAVING isn't super common, but when you need it, it's pretty nice.

I had a thought about how this might look, but since I rarely use this function, I'm not sure if my idea here covers a wide enough case or not.. I think to start we could have 2 methods. The first being having which might work similar to where(raw_sql : String). This would let you just write what you need as an escape hatch.

The second (and subsequent) methods would work similar to a column method, but also piggy back off the aggregate functions.

# or maybe just having_count ?
def having_select_count(exact_match : Int)
  having_select_count.eq(exact_match)
end

def having_select_count
end

Then it could look like this:

# give me all users that have exactly 4 draft posts
UserQuery.new.where_posts(PostQuery.new.published(false).having_select_count(4)).group(&.id)

# give me all users that have at least 2 published posts
UserQuery.new.where_posts(PostQuery.new.published(true).having_count.gte(2)).group(&.id)

This would give us

SELECT users.*
FROM users
INNER JOIN posts on posts.user_id = users.id
WHERE posts.published = 'f'
GROUP BY users.id
HAVING COUNT(posts.*) = 4

SELECT users.*
FROM users
INNER JOIN posts on posts.user_id = users.id
WHERE posts.published = 't'
GROUP BY users.id
HAVING COUNT(posts.*) >= 2

Or of course you could do UserQuery.new.where_posts(...).group(&.id).having("COUNT(posts.*) > ?", 4)

The others could be like having_sum having_min, having_max, etc...

No source files, tests, or entry points are named. Start by locating the existing query, grouping, column, and aggregate APIs, then compare them with the proposed having methods and raw SQL escape hatch. Done should include an agreed scope for HAVING support and tests showing the generated queries for the examples.