saio 0.2

SQLAlchemyIO (saio): Module hack for autoloading table definitions

SQLAlchemyIO (saio)

Module hack for autoloading table definitions.

Also provides a helper function as_pandas to read an sqlalchemy.orm.query.Query into a (Geo)Pandas dataframe.

Usage

After

import saio
saio.register_schema("model_draft", engine)

one can import table declarations easily using

from saio.model_draft import lis_charging_poi as LisChargingPoi

Note that ipython and Jupyter Notebook, allow using <TAB> to auto-complete table names.

The helper function as_pandas reads a query into a GeoDataFrame:

saio.register_schema("boundaries", engine)

from saio.boundaries import bkg_vg250_2_lan as BkgLan
df = saio.as_pandas(session.query(BkgLan))
df.plot()

Installation

Until we decide to register this package with pypi and/or conda-forge, please install it directly from git using pip:

pip install git+https://github.com/coroa/saio.git#egg=saio

or add it to the environment.yaml file you are working with as

name: ...
channels:
  - [ ... ]
dependencies:
  -  [ ... ]
  - pip:
    - [ ... ]
    - git+https://github.com/coroa/saio.git#egg=saio

Implementation details

saio.register_schema instantiates a declarative base using

from sqlalchemy.ext.declarative import declarative_base
Base = declarative_base(bind=engine)
# The Base can be imported using from saio.model_draft import Base

and then whenever one imports any table from saio.model_draft, ie. by calling from saio.model_draft import lis_charging_poi as LisChargingPoi, saio does approximately

class LisChargingPoi(Base):
    __tablename__ =  'lis_charging_poi'
    __table_args__ = {'schema': 'model_draft', 'autoload': True}