Getting your library into Clojars is fairly straightforward as is documented near the end of the Leiningen tutorial. However, deploying is not always as straightforward as the process described there.
Sometimes you'll need to publish libraries that you don't directly maintain, either because the original maintainer hasn't published it or because you need some bugfixes that haven't been applied upstream yet. Don't do this if it's at all possible to get the upstream project to release a snapshot somewhere with the changes you need, but in some cases it's unavoidable.
In this case you don't want to publish it under its original group-id, since either you won't have permission to do so (if it's already on Clojars) or it will conflict with the same artifact on other repositories. You should use "org.clojars.$USERNAME" as the group-id instead.
If it's a Clojure project that already has a project.clj file, it's
easy enough to just follow the regular lein jar, pom; scp [...]
path. If you don't have a readily-available pom, you can create a
dummy project with lein new. Edit project.clj to include your
org.clojars.$USERNAME group-id, the project's original artifact
name, and the version. Then you can use the output from lein pom to
upload to clojars.
There may be times when you want to make a library available to your team without making it public. This is best done by setting up a private Maven repository.
The simplest way to do this is by hosting a repository on Amazon's S3 service. You can use the s3-wagon-private plugin to deploy and consume from S3. Once you've signed up for an Amazon account, create a "bucket" to contain your repositories.
It's best to keep snapshots and released versions separate, so add
both to the :repositories in project.clj; in this example the
"s3p" URL scheme is used for S3 private repositories:
:repositories {"releases" "s3p://mybucket/releases/"
"snapshots" "s3p://mybucket/snapshots/"}If you don't mind running your own server, both
Archiva and
Nexus will allow you to set up private,
password-protected repositories. These also provide proxying to other
repositories, so you can set :omit-default-repositories in
project.clj, and dependency downloads will speed up by quite a bit
with only one server to check.
If you are are deploying to a repository that is only used for deployment
and never for dependency resolution, then it should be specified in a
:deploy-repositories slot instead of included in the more general-purpose
:repositories map; the former is checked by lein deploy before the latter.
Deployment-only repositories useful across a number of locally developed
projects may also be specified in the settings map in ~/.lein/init.clj:
(def settings {:deploy-repositories { ... }})Private repositories will need authentication credentials. You'll need
to provide a :username and :password or :passphrase depending on
the repository. In order to avoid putting sensitive information into
your project.clj file, you should store authentication information
in ~/.lein/init.clj as a leiningen-auth map keyed off the
repository's URL:
(def leiningen-auth {"http://localhost:8080/archiva/repository/internal/"
{:username "milgrim" :password "locative.2"}})This also allows different users using the same checkout to upload using different credentials.
Once you've set up a private repository and configured project.clj appropriately, you can deploy to it:
$ lein deploy
If the project's current version is a SNAPSHOT, it will deploy to the
snapshots repository; otherwise it will go to releases. The
deploy task also takes a repository name as an argument that will be
looked up in the :deploy-repositories and :repositories maps if
you want to override this.