Installation
=============

.. warning::
    As of version 2.9.0, the required dependency ``gdal``, ``matplotlib`` and ``seaborn`` was converted into an optional dependency. Fresh installations of this library will require that you then install ``gdal``, ``matplotlib`` and ``seaborn`` yourself with ``pip install gdal matplotlib seaborn``.


Conda installation
^^^^^^^^^^^^^^^^^^

The ``geoserver-rest`` can be installed from either ``conda-forge`` channel as below:

.. code-block:: shell

    $ conda install -c conda-forge geoserver-rest

Pip installation
^^^^^^^^^^^^^^^^

The ``geoserver-rest`` library can be installed using ``pip`` as below:

.. code-block:: shell

    $ pip install geoserver-rest

But best way to get all the functationality is to install the optional dependencies as well:

.. code-block:: shell

    $ pip install geoserver-rest[all]

If you want to install the geoserver-rest library with the optional dependencies (this will be useful if you are planning to create dynamic style files based on your dataset. Explore ``create_coveragestyle``, ``upload_style`` etc functions), you need to install the following dependencies first:

* `GDAL <https://gdal.org/>`_
* `matplotlib <https://matplotlib.org/>`_
* `seaborn <https://seaborn.pydata.org/>`_


Dependencies installation in Windows
------------------------------------

.. warning::
    As of March 2022, ``pipwin`` has been deprecated and is no longer maintained. Do not use this method.

For Windows, the ``gdal`` dependency can be complex to install. There are a handful of ways to install ``gdal`` in Windows.

One way is install the wheel directly from the `Geospatial library wheels for Python Windows <https://github.com/cgohlke/geospatial-wheels>`_ releases page. Be sure to select the wheel for your system from the latest release and install it using pip install command:

.. code-block:: shell

    # For Python3.10 on Windows 64-bit systems
    $ pip.exe install https://github.com/cgohlke/geospatial-wheels/releases/download/<release_version>/GDAL-3.7.1-cp310-cp310-win_amd64.whl
    $ pip.exe install seaborn matplotlib

Another way is to use the GDAL network installer binary package available at: `OSGeo4W <https://trac.osgeo.org/osgeo4w/>`_.


macOS installation
------------------

For macOS, we suggest using the `homebrew` package manager to install ``gdal``. Once ``homebrew`` is installed, ``gdal`` can be installed using following method:

.. code-block:: shell

    $ brew update
    $ brew install gdal
    $ pip3 install pygdal=="$(gdalinfo --version | awk '{print $2}' | sed s'/.$//')"

Linux installation
------------------

For Ubuntu specifically, we suggest installing ``gdal`` from the ``ubuntugis`` PPA:

.. code-block:: shell

    $ sudo add-apt-repository ppa:ubuntugis/ppa
    $ sudo apt update -y
    $ sudo apt upgrade -y
    $ sudo apt install gdal-bin libgdal-dev

For other versions of Linux, simply use your package manager to install ``gdal``.

.. code-block:: shell

    # Debian, Mint, etc.
    $ sudo apt install gdal-bin libgdal-dev
    # Fedora, RHEL, etc.
    $ sudo yum install gdal gdal-devel
    # Arch, Manjaro, etc.
    $ sudo pacman -S gdal
    # Void Linux
    $ sudo xbps-install -S libgdal libgdal-devel

Now the ``pygdal`` and ``geoserver-rest`` libraries can be installed using ``pip``:

.. code-block:: shell

    $ pip install pygdal=="$(gdal-config --version).*"
    $ pip install geoserver-rest