goccp (go cloud copy)
An attempt to learn about golang by creating a tool to copy file from and to AWS S3 and Google GCS
The reason this exists is to have no-dependency binary that can easily copy-pasted inside containers or your own server.
An UPX-ed golang binary should be small enough for the purpose (I add upx and gox to the build process)
UPX files compared to non-UPX files (19 megs in size)
19M Mar 21 01:38 goccp
5.5M Mar 21 01:40 goccp_darwin_amd64
5.5M Mar 21 01:40 goccp_linux_amd64
5.0M Mar 21 01:40 goccp_linux_arm
AWS Credentials
The credential will use Amazon SDK style of storing credential (use aws on your home, on your ENV var, or on your EC2 binded role - if you run it on EC2) - More info here https://github.com/aws/aws-sdk-go/wiki/Getting-Started-Credentials
GCS Credentials
This also will use Google SDK style of credentials - more info here
https://cloud.google.com/docs/authentication/production
Usage
# upload
goccp ./file.txt s3:us-east-1:bucketname:/file.txt
goccp -vv -perm public-read ./test ap-southeast-1:bucketname:/targetpath/
# download
goccp gs://bucket/path/file.zip ./
goccp s3:us-east-1:bucket:/file.zip ./
TODO
-
Refactor the provider to an interface :P (so we can support digical ocean's for example, or Webdav or FTP or other :|)
-
For now copy from remote source cannot walk through nested-folders since that will incur additional API calls for listing the files