cTDS

.. include-documentation-begin-marker

.. image:: https://github.com/zillow/ctds/workflows/CI/CD/badge.svg?branch=master :target: https://github.com/zillow/ctds/actions

.. image:: https://ci.appveyor.com/api/projects/status/voa33r7qdnxh6wwp/branch/master?svg=true :target: https://ci.appveyor.com/project/joshuahlang/ctds/branch/master

.. image:: http://img.shields.io/pypi/v/ctds.svg :target: https://pypi.python.org/pypi/ctds/

.. image:: https://codecov.io/gh/zillow/ctds/branch/master/graph/badge.svg :target: https://codecov.io/gh/zillow/ctds

cTDS is a full Python DB API-2.0_-compliant SQL Server database library for Linux, Windows, and Mac OS X supporting both Python 2 and Python 3.

The full documentation for cTDS can be found here <https://zillow.github.io/ctds/>_.

Features

• Supports Microsoft SQL Server <http://www.microsoft.com/sqlserver/>_ 2008 and up.
• Complete DB API-2.0_ support.
• Python 2.6, Python 2.7, Python 3.3, Python 3.4, Python 3.5, Python 3.6, Python 3.7, Python 3.8, and Python 3.9 support.
• Bulk insert (bcp) support.
• Written entirely in C.

Dependencies

• FreeTDS_
• Python_

.. FreeTDS: https://www.freetds.org/ .. Python: https://www.python.org/ .. _DB API-2.0: https://www.python.org/dev/peps/pep-0249

.. include-documentation-end-marker

See installation instructions <https://zillow.github.io/ctds/install.html> for more information on installing FreeTDS.

Releasing

Publishing new versions of the egg and documentation is automated using a Github Actions <https://docs.github.com/en/actions/> workflow. Official releases are marked using git tags <https://git-scm.com/book/en/v2/Git-Basics-Tagging>. Pushing the tag to the git remote will trigger the automated deployment. E.g.

.. code-block:: console

git tag -a v1.2.3 -m 'v1.2.3'
git push --tags

Documentation

Generate documentation using the following:

.. code-block:: console

tox -e docs
# Generated to build/docs/

Documentation is hosted on GitHub Pages <https://pages.github.com/>. As such, the source code for the documentation pages must be committed to the gh-pages <https://github.com/zillow/ctds/tree/gh-pages> branch in order to update the live documentation.

Development

Local development and testing is supported on Linux-based systems running tox and Docker. Docker containers are used for running a local instance of SQL Server on Linux. Only Docker and tox are required for running tests locally on Linux or OS X systems. pyenv is recommended for managing multiple local versions of Python. By default all tests are run against the system version of FreeTDS. GNU Make targets are provided to make compiling specific FreeTDS_ versions locally for testing purposes. For example:

.. code-block:: console

# Run tests against FreeTDS version 1.1.24
make test-1.1.24

Development and testing will require an instance of SQL Server on Linux running for validation. A script, ./scripts/ensure-sqlserver.sh, is provided to start a Docker container running the database and create the login used by the tests.

.. code-block:: console

# Start a docker-based SQL Server instance.
# The default tox targets will do this automatically for you.
make start-sqlserver

# Run tests as needed ...

# Stop the docker-base SQL Server instance.
make stop-sqlserver

Testing

Testing is designed to be relatively seamless using Docker containers and SQL Server on Linux. The pytest_ framework is used for running the automated tests.

To run the tests against the system version of FreeTDS and Python, use:

.. code-block:: console

tox

GNU make targets are provided for convenience and to provide a standard method for building and installing the various versions of FreeTDS used in testing. Most targets are wrappers around tox_ or replicate some behavior in the CI/CD automation.

To run the tests against an arbitrary version of FreeTDS_:

.. code-block:: console

# Python X.Y & FreeTDS Z.ZZ.ZZ
make test_X.Y_Z.ZZ.ZZ

To run tests against all supported versions of FreeTDS and Python and additional linting and metadata checks:

.. code-block:: console

make check

Valgrind

valgrind is utilized to ensure memory is managed properly and to detect defects such as memory leaks, buffer overruns, etc. Because valgrind requires Python is compiled with specific flags, a Docker file is provided to compile Python as necessary to run the test suite under valgrind_.

To run test test suite under valgrind_:

.. code-block:: console

make valgrind

.. Docker: https://www.docker.com/ .. compile Python: https://pythonextensionpatterns.readthedocs.io/en/latest/debugging/valgrind.html .. SQL Server on Linux: https://hub.docker.com/r/microsoft/mssql-server-linux/ .. GNU make: https://www.gnu.org/software/make/ .. pyenv: https://github.com/pyenv/pyenv .. pytest: https://docs.pytest.org/en/stable/ .. tox: https://tox.readthedocs.io/en/latest/ .. valgrind: https://valgrind.org/