sbt-s3 is a simple sbt plugin that can manipulate objects on Amazon S3.
This is a fork of https://github.com/sbt/sbt-s3. This version allows specifying the S3 endpoint (read region) and bucket to be specified explicitly instead of co-opting '''host''' - which didn't work for non-US standard regions.
Also tries to avoid downloading stuff when a local version already exists. Currently aborted the use of ETags to detect similarity in favour of content length since some of my files are multi-part uploaded and the ETags seem to have different semantics for such files. Will consider adding that in later.
- add to your project/plugin.sbt the line:
addSbtPlugin("lakshmanan.ca.sbt" % "sbt-s3" % "0.8") - then add to your build.sbt the line:
s3Settings
You will then be able to use the tasks s3-upload, s3-download, and s3-delete, defined
in the nested object com.typesafe.sbt.S3Plugin.S3 as upload, download, and delete, respectively.
All these operations will use HTTPS as a transport protocol.
Please check the Scaladoc API of the S3Plugin object, and of its nested S3 object,
to get additional documentation of the available sbt tasks.
Here is a complete example:
project/plugin.sbt:
addSbtPlugin("lakshmanan.ca.sbt" % "sbt-s3" % "0.8")
build.sbt:
import S3._
s3Settings
mappings in upload := Seq((new java.io.File("a"),"zipa.txt"),(new java.io.File("b"),"pongo/zipb.jar"))
endpoint in upload := "s3-us-west-2.amazonaws.com"
bucket in upload := "blah-blah"
credentials += Credentials(Path.userHome / ".s3credentials")
~/.s3credentials:
realm=Amazon S3
host=s3-us-west-2.amazonaws.com # You need this regardless, because sbt Credentials needs it :\
user=<Access Key ID>
password=<Secret Access Key>
Just create two sample files called "a" and "b" in the same directory that contains build.sbt, then try:
$ sbt s3-upload
You can also see progress while uploading:
$ sbt
> set S3.progress in S3.upload := true
> s3-upload
[==================================================] 100% zipa.txt
[=====================================> ] 74% zipb.jar