shipyard-bigquery 0.1.3

A local client for connecting to and working with Google BigQuery

shipyard-bigquery

The shipyard-bigquery Python package is a local client for connecting to BigQuery to perform common operations such as loading and unloading data.

Installation

python3 -m pip install shipyard-bigquery

Usage

To initialize the client, provide the service account (in JSON format) as a string:

import os
from shipyard_bigquery import BigQueryClient 
service_account = os.getenv('GOOGLE_APPLICATION_CREDENTIALS') # in this case, service account is stored as an environment variable
client = BigQueryClient(service_account)
client.connect()

Upload a file to a table in BigQuery

The upload method allows for the loading of a local CSV file to BigQuery and supports two different types of upload: overwrite and append.

client.upload(file = "my_file.csv", dataset = "my_dataset", table = "my_table", upload_type = 'overwrite')

Additionally, a the schema of the file to be loaded can be decalred ahead of time, otherwise they will be infered

schema = '[{"name": "string_col", "type": "string"}, {"name": "char_col", "type": "string"}, {"name": "int_col", "type": "Int64"}, {"name": "float_col", "type": "Float64"}, {"name": "bool_col", "type": "Bool"}, {"name": "date_col", "type": "Date"},{"name": "datetime_col", "type": "Datetime"}]'
client.upload(file = "my_file.csv", dataset = "my_dataset", table = "my_table", upload_type = 'overwrite',schema = schema)

Fetch Query Results

The fetch method allows for the downloading of a SQL query to a pandas dataframe

df = client.fetch('select * from my_table limit 10')

Execute Remote Queries

The execute method allows for you to execute queries without return values in BigQuery. Common operations would be dropping tables, deleting tables, or altering tables

client.execute_query('drop table if exists my_table')