PyHive

PyHive is a collection of Python DB-API and SQLAlchemy interfaces for Presto and Hive.

Usage

DB-API

from pyhive import presto  # or import hive or import trino
cursor = presto.connect('localhost').cursor()
cursor.execute('SELECT * FROM my_awesome_data LIMIT 10')
print cursor.fetchone()
print cursor.fetchall()

DB-API (asynchronous)

from pyhive import hive
from TCLIService.ttypes import TOperationState
cursor = hive.connect('localhost').cursor()
cursor.execute('SELECT * FROM my_awesome_data LIMIT 10', async=True)

status = cursor.poll().operationState
while status in (TOperationState.INITIALIZED_STATE, TOperationState.RUNNING_STATE):
    logs = cursor.fetch_logs()
    for message in logs:
        print message

    # If needed, an asynchronous query can be cancelled at any time with:
    # cursor.cancel()

    status = cursor.poll().operationState

print cursor.fetchall()

In Python 3.7 async became a keyword; you can use async_ instead:

cursor.execute('SELECT * FROM my_awesome_data LIMIT 10', async_=True)

SQLAlchemy

First install this package to register it with SQLAlchemy (see setup.py).

from sqlalchemy import *
from sqlalchemy.engine import create_engine
from sqlalchemy.schema import *
# Presto
engine = create_engine('presto://localhost:8080/hive/default')
# Trino
engine = create_engine('trino://localhost:8080/hive/default')
# Hive
engine = create_engine('hive://localhost:10000/default')
logs = Table('my_awesome_data', MetaData(bind=engine), autoload=True)
print select([func.count('*')], from_obj=logs).scalar()

# Hive + HTTPS + LDAP or basic Auth
engine = create_engine('hive+https://username:password@localhost:10000/')
logs = Table('my_awesome_data', MetaData(bind=engine), autoload=True)
print select([func.count('*')], from_obj=logs).scalar()

Note: query generation functionality is not exhaustive or fully tested, but there should be no problem with raw SQL.

Passing session configuration

# DB-API
hive.connect('localhost', configuration={'hive.exec.reducers.max': '123'})
presto.connect('localhost', session_props={'query_max_run_time': '1234m'})
trino.connect('localhost',  session_props={'query_max_run_time': '1234m'})
# SQLAlchemy
create_engine(
    'presto://user@host:443/hive',
    connect_args={'protocol': 'https',
                  'session_props': {'query_max_run_time': '1234m'}}
)
create_engine(
    'trino://user@host:443/hive',
    connect_args={'protocol': 'https',
                  'session_props': {'query_max_run_time': '1234m'}}
)
create_engine(
    'hive://user@host:10000/database',
    connect_args={'configuration': {'hive.exec.reducers.max': '123'}},
)
# SQLAlchemy with LDAP
create_engine(
    'hive://user:password@host:10000/database',
    connect_args={'auth': 'LDAP'},
)

Requirements

Install using

• pip install 'pyhive[hive]' for the Hive interface and
• pip install 'pyhive[presto]' for the Presto interface.
• pip install 'pyhive[trino]' for the Trino interface

PyHive works with

• Python 2.7 / Python 3
• For Presto: Presto install
• For Trino: Trino install
• For Hive: HiveServer2 daemon

Changelog

See https://github.com/dropbox/PyHive/releases.

Contributing

• Please fill out the Dropbox Contributor License Agreement at https://opensource.dropbox.com/cla/ and note this in your pull request.
• Changes must come with tests, with the exception of trivial things like fixing comments. See .travis.yml for the test environment setup.
• Notes on project scope:
  • This project is intended to be a minimal Hive/Presto client that does that one thing and nothing else. Features that can be implemented on top of PyHive, such integration with your favorite data analysis library, are likely out of scope.
  • We prefer having a small number of generic features over a large number of specialized, inflexible features. For example, the Presto code takes an arbitrary requests_session argument for customizing HTTP calls, as opposed to having a separate parameter/branch for each requests option.

Testing

Run the following in an environment with Hive/Presto:

./scripts/make_test_tables.sh
virtualenv --no-site-packages env
source env/bin/activate
pip install -e .
pip install -r dev_requirements.txt
py.test

WARNING: This drops/creates tables named one_row, one_row_complex, and many_rows, plus a database called pyhive_test_database.

Updating TCLIService

The TCLIService module is autogenerated using a TCLIService.thrift file. To update it, the generate.py file can be used: python generate.py <TCLIServiceURL>. When left blank, the version for Hive 2.3 will be downloaded.