IO abstractions to enable rapid research and prototyping.



Clojars Project

Simple IO library to enable rapid development on at least files, http/https, s3 with optional caching.


The philosophy is to split the difference between do-as-I-want programming which is what I do at the repl and do-as-I-say programming which matches system programming much closer. "" is DWIW, but it is built on an set of protocols to enable very precise DWIS programming where surprises are minimal.

Do What I Want:

(io/get-nippy "s3://blah/blah.nippy")

(io/copy "s3://blah/blah.nippy" "file://blah.nippy") 

(io/put-image! "s3://blah/blah.jpg" buffered-image)

This layer has a global variable you can use to override the way the system maps from url->io-provider. Using this variable you can setup global caching so that anything downloaded or read will get cached, regardless of where it came from (http, https, file, s3). You can also setup redirection where a directory looks like s3 for unit tests that read/write to s3 urls.

Do What I Say

The intention is to provide a simple and minimal base io abstraction and then add capabilities via chaining and layering abstractions.

The base abstraction is the IOProvider. There are basic implementations of this abstraction in base.clj. An aws s3 layer is also provided.

Given this abstraction we can implement a caching layer which will cache anything to a defined point in the filesystem. We can also implement a redirection layer which will redirect requests to a defined point on the filesystem.


Please see the tests.


Copyright © 2018 TechAscent, LLC

Distributed under the Eclipse Public License either version 1.0 or (at your option) any later version.