Helper functions for working with PostgresSQL hstore data type in Clojure via JDBC
Clojure
Switch branches/tags
Nothing to show
Latest commit bd9a5a1 Nov 8, 2012 @blakesmith Merge pull request #2 from citizenparker/master
Updated documentation on Korma integration

README.md

pghstore-clj

What is it?

pghstore-clj is a small helper library to help you work with PostgresSQL's hstore data type with the postgresql JDBC adapter (I use version 9.1-901.jdbc4) in Clojure. It provides two simple helpers to help you get your data into and out of a hash-map.

Installation

Add pghstore-clj to your project.clj file in leiningen:

    [pghstore-clj "0.1.0"]

Usage

Once installed, you can use it like so:

user> (use '[pghstore-clj.core])
nil
user> (def h (to-hstore {:color "blue" :size "small"}))
#'user/h
user> h
#<PGobject "color"=>"blue", "size"=>"small">
user> (from-hstore h)
{:color "blue", :size "small"}
user>

Korma example

If you have a table named "products" with an hstore column called "attributes", a korma insertion statement might look like this:

    (insert products
      (values {:name "computer"
               :attributes (to-hstore {:color "black" :manufacturer "samsung"})}))

When you pull the row back out again, just call "from-hstore" on the value at attributes.

user> row
{:name "computer", :attributes #<PGobject "color"=>"black", "manufacturer"=>"samsung">}
user> (from-hstore (:attributes row))
{:color "black", :manufacturer "samsung"}
user>

If you are using Korma's defentity function, you can apply the hstore translation at the entity rather than at query-time. For the above products example, your entity might look something like so:

    (defentity products
      (prepare #(update-in % [:attributes] to-hstore))
      (transform #(update-in % [:attributes] from-hstore)))

With this in place, the translation is transparent and you can insert plain data structures:

    (insert products
      (values {:name "computer"
               :attributes {:color "black" :manufacturer "samsung"}}))

Wishlist

Ideally, these transformation functions would be applied at the JDBC level, so that you only have to work with hashes - This is going to take a bit more work.

License

Copyright © 2012 Blake Smith

Distributed under the Eclipse Public License, the same as Clojure.