FTP-like client for interacting with cloud storage
The cftp package provides an ftp-like interface to cloud-based storage services. It is basically a wrapper. It abstracts the user from the particulars of specific cloud storage services by exposing a common interface. Accessing the interface is simple. Instantiate a class corresponding to a desired cloud storage service (eg, Amazon S3). The class has methods corresponding to common ftp client commands.
The current version includes the four files noted below.
Over time, this package may be extended to include an ftp-like client interface to the DropBox storage services. That functionality would be implemented as a sub-class of base.py, just as s3.py is a sub-class. The same is true for other cloud-based storage services that offer an API.
Install using pip or easy_install. The cftp package does require Amazon’s boto3 package, which exposes an API for controlling AWS objects using python. The boto3 package will be automatically installed (via pip or easy_install), since it is listed as a dependency in this package.
This package also includes a tool called s3ftp that provides command line, ftp-like access to the Amazon S3 service.
Before using this package, you will need to setup AWS authentication credentials as required by boto3. See the following URL:
This distribution also includes a file called, s3ftp.json. It contains default parameters used when reading or writing objects in an S3 bucket. For example, it can be used to specify if/how objects are encrypted or to add metadata to objects. Edit this file to reflect desired default parameters. Then copy it to your home directory and rename to .s3ftp.json. The parameters include those allowed in the extraArgs argument to the S3Transfer functions download_file and upload_file.
The S3FtpClient constructor looks for a .s3ftp.json file first in the current working directory and then in the user’s home directory. If it finds this file, then it loads default parameters for S3 objects from there. The constructor may also be called with an argument that specifies default object parameters. If so, these are merged with defaults read from the .s3ftp.json file. If a particular default is specified in both the .s3ftp.json file and the constructor argument, then the latter takes precedence.
Note that when running s3ftp interactively, it is not yet possible to change these parameters without quitting, editing .s3ftp.json and restarting. However, the S3FtpClient class does have methods to get and set the default parameters.
Handling of incorrectly specified defaults is rudimentary right now.
The following illustrates using the interactive s3ftp command line utility. It assumes the existence of an Amazon S3 bucket, in this this case a notional bucket named com.s3ftp.test.
The following connects to a bucket, lists its contents, changes the remote working directory (in the bucket), changes the local working directory, creates a directory in the bucket, uploads a file to it, multiple gets several files, deletes a file from the bucket, closes the bucket, opens a new bucket, lists its contents, and quits. Below does not show s3ftp output.
The Amazon S3 client above does not support creation or deletion of S3 buckets. It assumes the bucket already exists. This is consistent with behavior of a traditional ftp client in that it is accessing existing storage (on an ftp server).
This software has been tested on Linux but not Windows or Mac OS platforms.
The S3FtpClient object (and associated s3ftp utility) do not provide the ability to examine S3 object parameters, such as metadata and content-type. This could be added in the future as an auxiliary (non-ftp) command.
Rearchitected the internal design of the BaseFtpClient and the S3FtpClient. Moved more implementation-independent functionality into the BaseFTPClient. Created abstract methods in the base client that can be overridden to implement functionality specific to Amazon S3 or other cloud providers.
Made some internal bug fixes to handling of the extraArgs parameters in the S3FtpClient.