Welcome to tiny-blocks’ documentation!
Tiny Blocks to build large and complex ETL pipelines!
Tiny-Blocks is a library for data engineering operations. Each pipeline is made out of tiny-blocks glued with the >> operator. This library relies on a fundamental streaming abstraction consisting of three parts: extract, transform, and load. You can view a pipeline as an extraction, followed by zero or more transformations, followed by a sink. Visually, this looks like:
extract -> transform1 -> transform2 -> ... -> transformN -> load
You can also fan-in, fan-out for more complex operations:
extract1 -> transform1 -> |-> transform2 -> ... -> | -> transformN -> load1
extract2 ---------------> | | -> load2
Tiny-Blocks use generators to stream data. Each chunk is a Pandas DataFrame. The chunksize or buffer size is adjustable per pipeline.
Basic usage
Make sure you had install the package by doing pip install tiny-blocks and then:
from tiny_blocks.extract import FromCSV
from tiny_blocks.transform import Fillna
from tiny_blocks.load import ToSQL
# ETL Blocks
from_csv = FromCSV(path='/path/to/source.csv')
fill_na = Fillna(value="Hola Mundo")
to_sql = ToSQL(dsn_conn='psycopg2+postgres://...', table_name="sink")
# Pipeline
from_csv >> fill_na >> to_sql