Beast mode 🦍

You can now put edn stuff and sql stuff side by side in where clauses 🎉

Like this:

(q [:select :post/published-at
    :joins :post/member
    :where ["post.published_at > ?" start-date]
           [:post/member '?member/id]
    :order :post/published-at :desc]
   {:member/id member-id})

It's always a hassle to make a sql file and call defq just because of a > or a between ? and ? NOW YOU DON'T HAVE TO

Bugfixes & Improvements

This was a nasty bug that got me right in the kisser, tried to update two boolean columns at once, I foolishly put a distinct in the where args section. Terrible.

Better late than never ❄️

Let's take a look at the last time I took the time to make a formal release

Goals:

Add defq
Add support for regular sql queries and coast schema inference from the database schema
Add migration rollbacks
Now that I know the major differences between sqlite and postgres, add a layer between the final sql output for migrations and queries to work with sqlite and postgres, and possibly mysql, but probably not

As you can plainly see from the strikethroughs, I've made some serious progress porting over missing features from coast including postgres support 🙌

I've made a few improvements too around date/time parsing and adding * support to queries and full on sql strings in where clauses!

Might as well take this space up to set a new "roadmap"

• Add a watered-down transact (upsert) #1
• Add better support for strings in where clauses
• Add support for count, avg, sum and possibly other aggregate-y things
• Make it easier to run regular sql (as strings), i.e. don't require a whole other sql file to run a query (maybe)
• Add mysql support (maybe)

First release 🎉

This is really a re-hash of coast's code to work for sqlite, but I wanted to split it up into it's own library for marketing purposes and to learn git submodules, it will be nice to have each major part of coast (db, background jobs, caching) be it's own git repo but act just like folders in coast

Goals:

• Add defq
• Add support for regular sql queries and coast schema inference from the database schema
• Add migration rollbacks
• Now that I know the major differences between sqlite and postgres, add a layer between the final sql output for migrations and queries to work with sqlite and postgres, and possibly mysql, but probably not