Selecting a DB system for a Scala/Slick fan

Questions & Help General Questions
1

I have a colleague who is excited to use Clojure, but is not a fan of the Yesql approach. At the same time, none of us want the Hibernate approach.

Is there anything similar to Slick for Clojure? Slick is inspired by LINQ.

2

To your question there is https://github.com/tatut/specql which is pretty cool

I would personally love something that looks like this:

(deft posts {:id “serial primary key”
             :title “text”
             :body “text”
             :published “boolean not null default false“})

(defn published? [post]
  (true? (:published post)))

(sql/filter published? posts) ; => [“select * from posts where published = ?” true]

I have no idea how this would work or if it’s even possible. So the deft would be a “smart” migration and there would be analogous functions for map/filter/reduce that emit sql and maybe you could traverse through a relational database like a huge nested map, automatically creating join queries along the way. :sweat_smile:

Or I guess just use datomic haha

3

If you want a SQL DSL, you can use HoneySql or Sqlingvo or Korma.

None of them will be like Linq, because Linq is also a Dsl. So you can use the same DSL for sql and datastructure querying. While map/reduce is just plain old functions in Clojure.

4

Building non trivial queries with yesql approach can become tedious quickly. Walkable is a new (and efficient) approach to SQL in Clojure.


Basically you define the schema and write queries in plain Clojure, so I guess it matches Slick’s Scala-ness and composability.

5

So this is more about selecting a SQL wrapper than a DB, correct? Otherwise, you should check out Datomic :slight_smile:

6

That might be too much of a leap at this point, but it’s a great idea.

7

I would also like to mention that you can use datahike as a FOSS Datomic replacement for simple needs: https://github.com/replikativ/datahike

2 Likes
8

Datahike is super interesting! I’m going to give it a shot

9