Gradle S3 Plugin

This is a fork of skhatri/gradle-s3-plugin - as that project maintainer was unresponsive to our pull requests, we forked it and released our own version of it. Changes are kept to bare minimum (for now).

The plugin can be used to download from and upload to S3 buckets. The link configuration attribute can be used to create amz-website-redirect-location based redirects. Technically this is not a symbolic link, but it can behave like one when trying to reference same file from multiple locations.

It creates two tasks:

s3Upload

As the name suggests, the s3Upload task can be used to upload a local file to a bucket with a given key name.

s3Download

Similarly, the s3Download task can be used to download a given file from a chosen bucket to a local directory.

The plugin uses a ProfileCredentialsProvider strategy to figure out AWS access and secret key. The name of the credentials profile to use when connecting to AWS can be specified by the awsProfile property.

See com.github.skhatri.s3aws.client.S3Client for implementation.

Pull requests are welcome if you want to modify to your needs.

Plugin configuration in your build.gradle

buildscript {
    dependencies {
        classpath 'net.sympower.gradle:gradle-s3-plugin:1.0.6'
    }
    repositories {
        mavenCentral()
    }
}
apply plugin: 's3'

s3 {
    bucket = 'sample-bucket'
    awsProfile = 'default'
    upload {
        //"key" is the name of the target file in S3 (in this example we create a versioned file)
        key =  new org.joda.time.LocalDate().toString('yyyy/MM/dd')+'/gradle-s3-plugin-1.0.2-SNAPSHOT.jar'
        
        // "file" is the name of the local source file
        file = '../build/libs/gradle-s3-plugin-1.0.2-SNAPSHOT.jar'
        
        //"link" will create http redirect to the latest uploaded version of the file
        // This will redirect is only available via the "s3 static webpage" option. 
        // walkthorugh link from AWS on how to set it up
        // https://docs.aws.amazon.com/AmazonS3/latest/dev/website-hosting-custom-domain-walkthrough.html
        link = 'latest/gradle-plugin.jar'
    }
    download {
        key = 'latest/gradle-plugin.jar'
        saveTo = 'gradle-pugin.jar'
    }
}

Using Upload Task to upload more artifacts

task uploadAppJar(type: com.github.skhatri.s3aws.plugin.S3UploadTask) {
    key = 'gradle-s3-plugin-1.0.0-something.jar'
    file = '../build/libs/gradle-s3-plugin-master-1.0.1-SNAPSHOT.jar'
    link = 'latest/gradle-plugin.jar'
}

Then I can call gradle uploadAppJar to upload yet another artifact to S3.

This can be useful, if uploading hash checksums of the artifacts. For instance, I can upload sha1 value of the artifact so my automated artifact deployment task could download checksum file first to decide whether downloading the big artifact is worth it.

task writeHash() {
    String hashValue = computeHash('../build/libs/gradle-s3-plugin-master-1.0.1-SNAPSHOT.jar')
    file('../build/libs/gradle-s3-plugin-master-1.0.1-SNAPSHOT.sha1').write(hashValue)
}

task uploadHash(type: com.github.skhatri.s3aws.plugin.S3UploadTask) {
    key = 'gradle-s3-plugin-1.0.0-something.sha1'
    file = '../build/libs/gradle-s3-plugin-master-1.0.1-SNAPSHOT.sha1'
    link = 'latest/gradle-plugin.sha1'
}

task awsUpload(dependsOn:['uploadAppJar', 'writeHash', 'uploadHash']) {
}

Now, calling awsUpload will upload jar and the hash file.

Downloading more than one artifact

Similarly, I can download more artifacts by simply creating ad-hoc S3DownloadTask. I am downloading the previously uploaded sha1 file of the artifact using the task below.

task awsDownload(type: com.github.skhatri.s3aws.plugin.S3DownloadTask) {
    key = 'latest/gradle-plugin.sha1'
    saveTo = 'gradle-plugin.txt'
}