tuxput 0.0.1

The Serverless File Uploader

tuxput

The Serverless File Uploader

tuxput is a python/flask application which provides a file uploading interface to S3, and is designed to be run serverlessly with Zappa.

Configuration

Configuration is handled through environment variables. The following configuration variables are used.

• S3_BUCKET
  • required: True
  • description: S3 bucket name containing the files to serve. Example: storage.staging.lkft.org
• S3_REGION
  • required: True
  • description: Region containing the S3 bucket. Example:us-east-1
• S3_ROOT
  • required: False
  • description: An optional path that can be used to prefix the folder in S3 to where the files will be uploaded.
• SITE_TITLE
  • required: False
  • description: Defaults to Tuxput. Set to anything you like for a global site title.
• ALLOW_UPLOAD_OVERWRITE
  • required: False
  • description: boolean, defaults to False. Allows uploaders to overwrite existing files.
• AUTH_BUCKET
  • required: False
  • description: String that is the name of a separate S3 bucket that contains the AUTH_FILE.
• AUTH_FILE
  • required: False
  • description: String that is the key (filename) of a json file containing tokens and paths to which they are authorized to upload. Defaults to 'tuxput.passwd'

Run Locally

To run locally, install tuxput, ensure AWS access is available environmentally, and run:

S3_BUCKET=testing-tuxput.ctt.linaro.org S3_REGION=us-east-1 FLASK_APP=tuxput flask run

Run with Zappa

This application is intended to be ran and deployed with Zappa and hosted by AWS API Gateway and Lambda.

To use with Zappa, create an app shim named zappa_init.py.

For an example, please see:

https://gitlab.com/kspoon-linaro/tuxput-deploy-ctt/

Features

Auth File

Authorization is handled very simply. Tokens are created and put into a CSV file specified by the AUTH_FILE option. Each entry in the file lists a token and a unix-style file pattern to which the token is allowed to upload. When a client attempts to upload a file, the path will be checked against the authorized paths associated with the token and access will be granted to the first match that is found.

If the AUTH_FILE is located in an S3 bucket different from S3_ROOT, you may specify it with the AUTH_BUCKET option.

token,path
qza1234,somecorp/someproduct/*

Only the 'token' and 'path' fields are currently required.

The 'path' setting in this example will allow the user to create any filepath under the somecorp/someproduct folder in S3. Note that the "*" is required or else the auth rule would restrict the user to just creating the object somecorp/someproduct