TeselaGen Python API Client

The TeselaGen Python API Client runs on Python 3.9+

NOTE : All the following commands are supposed to be run on the base directory, unless specified.

Library Installation

This library contains the TeselaGen Python API Client.

To install it locally with pip:

pip3 install teselagen

Use and login

Import the teselagen library:

from teselagen.api import TeselaGenClient

Create an instance of the client:

client = TeselaGenClient(host_url="https://<INSTANCE NAME>.teselagen.com/")

Then, login by using your user email and One Time Password (OTP). You can get one from Settings-> API Password within the application. Alternatively ,you can use your application password.

client.login(username="my@email.com", password="<OTP OR PASSWORD>")

Once you have logged in, you don't need to do it everytime. By default login token lasts for 1 week, but you can change the duration on creation by specifying an expiration time:

client.login(username="my@email.com", password="<OTP OR PASSWORD>", expiration_time="1d")

Examples

Check out the provided examples. To be able to run them:

1. Clone or download teselagen/examples

2. Open any notebook in the examples folder with Jupyter Notebook

Development

Poetry virtual environment

For development you can just use the virtual environment created by poetry. To install the project within a virtual environment, use at the project's root folder:

poetry install

Dependencies

Install dependencies

If lock file exists (the lock file contains fixed versions of dependencies), the poetry install command will install all dependencies according to the lock file (lock file must be added to the repo). If the file does not exist, it will generate the lock file again.

Update dependencies

Use this command if you made changes on the dependencies at the pyproject.toml file:

poetry update

It is the equivalent to make an install after deleting the lock file.

Tests

1. Add your credentials

   To run the tests, you must create a .credentials file containing the test username and password, in the root folder.

   The content of .credentials file should look similar to the following:

       {
           "username" : "ReplaceWithYourUsername",
           "password" : "ReplaceWithYourPassword"
       }
   

   • NOTE 1: It should be a valid JSON file. JSON with Comments (JSONC ) format is not supported.
   • NOTE 2: You may also add a host field. It's value will be used as default host name.

   - DO NOT COMMIT THIS FILE : .credentials
   

2. Modify configuration

   You may modify some test configuration parameters by creating a .test_configuration file. This is a JSON formatted file, where you can edit the server name used for tests. This file must be stored next to .credentials file. Here is an example

   {
     "host_url" : "https://platform.teselagen.com"
   }
   

   • NOTE: It should be a valid JSON file. JSON with Comments (JSONC ) format is not supported.

3. Run the tests

   cd /home && python3 setup.py test
   

   You may use the docker environment for testing. For that, first build the environment with bash build.sh. Then just run the container with bash run_dev.sh. Once inside (docker exec -ti tgclient bash), go to home/ and you are ready to run the test command shown above.

Publishing

Publishing is limited to administrators. PyPi publishing is made by using poetry.

To publish:

1. Run poetry build from the project's root folder (same directory as pyproject.toml)

2. Be sure you have set the credentials with the api token:

poetry config pypi-token.pypi <TOKEN>

Ask for a token to administrators if needed

1. Publish (check you have set a new version tag in pyproject.toml):

poetry publish

Notes

Default shell in Ubuntu is dash, to which /bin/sh is symlinked. But dash doesn't have the pipefail option. That's why some of our shell scripts have the following line:

# pipefail is necessary to propagate exit codes (but it may not be supported by your shell)
bash | set -o pipefail > /dev/null 2>&1

For example, the following commands will list all options supported by set in each of the respective shells:

sh -c 'set -o'
dash -c 'set -o'
bash -c 'set -o'
zsh -c 'set -o'

So, in Ubuntu it may be recommended to use bash instead of sh.

bash some_shell_script.sh

Or, as follows:

. some_shell_script.sh