Installing the Extract API 2.0

This topic describes the requirements for using the Extract API and where you can get an installation file.


To work with the Extract API, you need the following:

  • A computer running a 64-bit operating system (Linux, Mac OS X, Windows). The Extract API 2.0 requires a 64-bit OS.

  • For Python, you must have Python 2.x or 3.x installed.

  • On Windows, you must have the Visual C++ runtime library installed on your computer. You might already have this library installed; for example, it is installed as part of Visual Studio. If you don't already have the runtime library installed, you can get it from the following page on the Microsoft site:

    Visual C++ Redistributable Packages for Visual Studio 2015

    Note: Even if you intend to use Java and Eclipse (that is, you are not going to use C or C++, and you are not intending to use Visual Studio), you must still make sure you have the Visual C++ runtime library installed. The Extract API requires this runtime library.

  • If you are installing on Windows, you must be running Windows Server 2008 R2 or later, or Windows 7 or later.

  • You need to use Visual Studio 2015 if you are building a C or C++ application on Windows that creates extracts.

  • If you are installing on Linux: the package is a required dependency. If you do not have the package installed, which is common on newer versions of Ubuntu, you need to install the latest package and create a symbolic link for the missing file. Follow the instructions for Missing package on the Troubleshooting with the Extract API 2.0 page.

  • If you are installing on Linux, you must allow localhost TCP connections. The Extract API will start a process locally and needs to communicate through TCP.

  • If you are installing on CentOS, you might need to install a more recent version of GCC and GLIBC. Follow the instructions for Unable to build on CentOS on the Troubleshooting with the Extract API 2.0 page.

Downloading the package

To get the Extract API, use the appropriate link from the following list.

Note: For the best performance, particularly if you are creating extracts from very large data sources, use the C or C++ libraries.


Mac OS X

  • Mac OS X (64-bit) (.tar.gz file). Use this file to install the Python version of the Extract API.

  • Mac OS X (image file, 64-bit) (.dmg file).  Use this file to install the C, C++, and Java versions of the Extract API.

    To use the .dmg file, do the following:

    1. Open the .dmg file you downloaded.

    2. Drag the three files in the Frameworks directory to /Library/Frameworks. You might be asked to enter your password.

    3. Drag the License.txt file, the Java folder, and the Samples folder to a location of your choice.


Setting environment variables

Optionally, you can set the following environment variables to specify working directories for the Extract API. If you don't set these variables, the Extract API uses the current working directory as the default location.

Note: The user identity under which code is running must have write permissions to the locations you specify in these environment variables.

  • TAB_SDK_LOGDIR The folder where the Extract API writes the DataExtract.log file.

  • TAB_SDK_TMPDIR The temporary directory where the Extract API keeps intermediate (temporary) files, such as when it's creating an extract.

If an invalid TAB_SDK_LOGDIR is provided, the Extract API will not start correctly. If you are setting shell or environment variables in Linux, and want to use a tilde (~) in the path, such as ~/user_name, you will need to export the variable first so that the tilde expansion takes place. For example, you might enter something like the following: 

export TAB_SDK_LOGDIR=~/mydir