semolina 0.6.0

A Pythonic ORM for querying data warehouse semantic views (Snowflake, Databricks, DuckDB)

Semolina

[ Docs ]

The ORM for your Semantic Layer.

Typed models in Python, supporting IDE autocomplete, and a Django-like fluent query interface for the semantic layer of your data warehouse backend.

pip install semolina
pip install semolina[snowflake]
pip install semolina[databricks]

A model maps to a semantic view in your warehouse.

from semolina import (
    SemanticView,
    Metric,
    Dimension,
)


class Sales(SemanticView, view="sales"):
    revenue = Metric()
    cost = Metric()
    country = Dimension()
    region = Dimension()

Semolina needs an engine to talk to your warehouse. An engine owns one connection pool and the dialect for a backend. Build one with create_engine and register it before running any queries:

from semolina import create_engine, register

engine = create_engine(
    "default"
)  # reads [connections.default] from .semolina.toml
register("default", engine)

Use Model.query() to start building. Chain .metrics() and .dimensions() to select the fields you want, then call .execute():

query = (
    Sales.query()
    .metrics(Sales.revenue)
    .dimensions(Sales.country)
)

.execute() returns a SemolinaCursor. Use it as a context manager so the pooled connection is released when you are done, and call .fetchall_rows() to get Row objects that support both attribute and dict-style access:

with query.execute() as cursor:
    for row in cursor.fetchall_rows():
        print(row.country, row.revenue)  # attribute access
        print(row["country"])  # dict-style access

You should see output like:

US 1000
US
CA 2000
CA
US 500
US