Credbl: database settings and credentials simplified

There are many moving parts when setting up programmatic access to a database. Credbl divides and conquers those bits and pieces the following way:

• [where] database server specific settings are read from a YAML configuration file provided by the database admin

• [who] user's credentials are requested and stored in keyring or winreg (on Windows).

  The first time a user connects to the database with connect_* functions, the credentials will be requested and stored. Subsequent times, they will be silently retrieved and used to authenticate. If authentication fails due to wrong credentials, the user will be asked to enter credentials again.

• [how] database drivers are chosen based on user's Operation system.

Currently credbl focuses on MS SQL Server and MongoDB connections. You are welcome to submit issues and pull requests for other database types.

Installation

Python package

from terminal shell (standard):

• latest release: pip install credbl

• bleeding edge: pip install -U git+git://github.com/BCHSI/credbl.git#egg=credbl (you might also need to add --user)

• development mode

    git clone https://github.com/BCHSI/credbl
    cd credbl
    pip install -e .
  

on Mac, you might need to install libsodium using homebrew: brew install libsodium

R package

from R:

install.packages('devtools')
library(devtools)
install_github('https://github.com/BCHSI/credbl')

or terminal + R:

git clone https://github.com/BCHSI/credbl
cd credbl
R
> install.packages('devtools')
> library(devtools)
> install('.')

get help:

?credbl

Usage examples:

connecting to a MS SQL Server

from credbl import connect_mssql
conn = connect_mssql("tp-mssql-settings.yaml")

Contents of database configuration file

In a previous example, "tp-mssql-settings.yaml" must contain:

server:    12.34.56.78 (OR) mydatabase.mybusiness.com
port:      1234
database:  tp-inventory
driver:    FreeTDS (optional)

Low-level interface:

You might need to understand it if you would like to use SQL drivers or driver wrappers other than pyodbc, such as SQLAlchemy.

with pyodbc

import pyodbc
from credbl import get_mssql_connection_string 

# if called for the first time, will request credentials
# second time may ask for your _system_ credentials; mark "always allow"

connection_str = get_mssql_connection_string("tp-mssql-settings.yaml")
# you'll be requested to enter your credentials when running it for the first time
conn = pyodbc.connect(connection_str)

if you believe you've entered wrong credentials first time, call with reset=True

connection_str = get_mssql_connection_string("tp-mssql-settings.yaml", reset=True)

conn = pyodbc.connect(connection_str)

with SqlAlchemy

import sqlalchemy
from credbl import get_mssql_connection_string 

connection_str_encoded = get_mssql_connection_string('covid19_omop.yaml',
                                                 urlencode=True)
connection_uri = 'mssql+pyodbc:///?odbc_connect={}'.format(connection_str_encoded)
conn = sqlalchemy.create_engine(connection_uri)

connecting to mongodb

from credbl import connect_mongodb

# if called for the first time, will request database credentials
mdb = connect_mongodb("mongo-settings.yaml")

mdb.list_collection_names()

The "mongo-settings.yaml" file must contain following:

url: mongodb://10.20.30.40:27017
db: 'databasename'

Alternatively / optionally to URL, server or IP address and port can be provided:

server: xyz.company.org
ip: 10.20.30.40
port: 27017

storing credentials in keyring (Mac, Unix) or Windows key storage:

In [1]: import credbl

In [2]: credbl.get_credentials('something')
enter user name for 'something': []: myname
enter password for 'myname':
Out[2]: ('myname', 'xyz')

In [3]: credbl.get_credentials('something')
Out[3]: ('myname', 'xyz')

Files

• dbfunctions.R authentication (keyring and Windows ODBC) and dbplyr-based shortcuts
• credentials.py authentication
• mongodb_utils.py wrapper for db connection and authentication for MongoDB