splicer
makes the entire world look like a SQL database.
It is a python module for working with data from disparate sources
using commands to those familiar with SQL. It aims to make quick
one off queries and ETL scripts more declarative rather than procedural.
Inspired by projects like BigQuery, Postgres Foreign Data Wrappers and Multicorn, except no database is required.
splicer
enables:
-
Analysts to create Datasets linking various foreign tables together along with User Defined Functions written in python. Once defined, the datasets can be queried via SQL Select statements to create new Views of the Data.
-
Extension Developers to create extensions that make various data sources REST endpoints, log files, NoSQL Servers, traditional Databases, CSV Files to behave like tables.
splicer
will take advantage of these various sources' capabilities where appropriate and will compensate for sources that lack basic features.For example if a database supports joins and you want to query two tables within that database,
splicer
will have that system perform the join for you. If however you're working with a less sophisticated source, like plain files,splicer
will perform the operations for you locally.
Enough reading! Try it out