Installing LabGuruAPI
=====================

Configuring AWS
---------------

Making a profile
^^^^^^^^^^^^^^^^
You will need an AWS profile to be able to install the LabGuruAPI package from our CodeArtifact repository. If you do
not have one already, ask `Mark <mark.cerutti@grobio.com>`_ or `Rob <rob.wr@grobio.com>`_ to create one for you. We will
provide you with a ``credentials.csv`` file once your profile is active.

Installing the AWS Command Line Interface (CLI)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
.. tabs::
    .. group-tab:: Windows
        System Requirements:
            * 64-bit Windows (all GRO computers should be 64-bit)
            * Admin rights to install software (all GRO users should be able to install software on their computers)

        Process:
            1. Download and run the `CLI Installer <https://awscli.amazonaws.com/AWSCLIV2.msi>`_ from Amazon.
            2. To confirm the installation, open the *Start* menu, search for ``cmd`` to open a command prompt window,
               and at the command prompt use the ``aws --version`` command. Your output should look like:

               .. code-block:: doscon

                    C:\> aws --version
                    aws-cli/2.11.14 Python/3.11.3 Windows/10 exe/AMD64 prompt/off

    .. group-tab:: macOS
        System Requirements:
            * macOS version 10.15 or later
            * ``sudo`` permissions

        Process:
            1. Download the installer package.

               .. code-block:: console

                    $ curl "https://awscli.amazonaws.com/AWSCLIV2.pkg" -o "AWSCLIV2.pkg"
            2. Install the package. The files will be installed to ``/usr/local/aws-cli`` with a symlink created to
               ``/usr/local/bin``. You must use ``sudo`` to write into those locations.

               .. code-block:: console

                    $ sudo installer -pkg ./AWSCLIV2.pkg -target /
            3. Verify the location and version of your AWS CLI. The output should look something like this:

               .. code-block:: console

                    $ which aws
                    /usr/local/bin/aws
                    $ aws --version
                    aws-cli/2.19.1 Python/3.11.6 Darwin/23.3.0 botocore/2.4.5

If you encounter problems, please reference the `AWS Troubleshooting Document <https://docs.aws.amazon.com/cli/latest/userguide/cli-chap-troubleshooting.html>`_

Importing your credentials
^^^^^^^^^^^^^^^^^^^^^^^^^^

Open a command prompt window and use the ``aws configure import`` command with the ``--csv`` option to import the
``credentials.csv`` file `provided by a system admin <Making a profile_>`_

.. tabs::
    .. group-tab:: Windows
        Run the following command, replacing ``file://credentials.csv`` with the path to your ``credentials.csv`` file.

            .. code-block:: doscon

                C:\> aws configure import --csv file://credentials.csv

    .. group-tab:: macOS
        Run the following command, replacing ``path/to/credentials.csv`` with the path to your ``credentials.csv`` file.

            .. code-block:: console

                $ aws configure import --csv path/to/credentials.csv

Getting the package from CodeArtifact
-------------------------------------

Once you have configured your AWS CLI, you can install the ``LabGuruAPI`` package **almost** as easily as everything else.
The only difference is that you will need to log in to CodeArtifact and add the code artifact repository to ``pip``.
Luckily, this can be done in a single command via the AWS CLI. You will remain logged in for 12 hours. The shell command
is the same for each operating system.

.. code-block::

    aws codeartifact login --tool pip --repository foundry-pypi --domain grobio-foundry --domain-owner 409124030547 --region us-east-2

You can then install ``foundrybackend`` package, which includes the ``LabGuruAPI`` module.

.. code-block::

    pip install foundrybackend

.. note::
    Running this command will change the default python repository from pypi to CodeArtifact *for your entire system*.
    To switch back, run the command

    .. code-block:: none

        pip config unset global.index-url