datasafe 0.1.0

Setup file for datasafe.

DataSafe - Utilities for data encryption / decryption

Table of contents

• Installation
  • Docker
• Command-line
  • Encrypt
  • Decrypt
• Python
  • Pandas
  • Files
• License
• Contact

Installation

pip install datasafe

Docker

git clone --depth 1 https://github.com/StephenRicher/datasafe.git
cd datasafe/
docker build -t datasafe .
docker run datasafe --help

Command-line

Encrypt

datasafe encrypt data.csv > data.encrypted

Decrypt

datasafe decrypt data.encrypted > data.decrypted.csv

Python

datasafe can be imported as a python module to encrypt and decrypt files.

Pandas

If a Pandas DataFrame is provided to datasafe.encrypt then it will be encrypted in .parquet format. Following decryption, an in-memory buffer is returned which should be passed to pd.read_parquet to recover the dataframe and datatypes.

import pandas as pd
from datasafe import encrypt, decrypt

df = pd.DataFrame({
    'A': [1, 2, 3],
    'B': ['dog', 'cat', 'bat']
})

encrypt(df, 'df-encrypted.pq')

df = pd.read_parquet(decrypt('df-encrypted.pq'))

Files

The command line functionality can also be achieved within Python. In addition the datasafe.decrypt function returns an in-memory buffer which can be read directly.

Encrypt and write encrypted data to file

import pandas as pd
from datasafe import encrypt, decrypt

with open('data.encrypted', 'wb') as fh:
    fh.write(encrypt('data.csv'))

Decrypt and write decrypted data to file

with open('data.decrypted.csv', 'w') as fh:
    fh.write(decrypt('data.encrypted').getvalue())

Decrypt and read in-memory buffer to Pandas

df = pd.read_csv(decrypt('data.encrypted'))

License

Distributed under the MIT License. See LICENSE for more information.

Contact

If you have any other questions please contact the author Stephen Richer at [e-mail]