- Language: Python
- Stars: 0
- Repository: https://github.com/Buildly-Marketplace/sales_forecast_service
#4# Issue The current README is sufficient to explain the functionality of the service but may be cleaned up to better align with the formatting used in other Buildly marketplace services e.g. https://github.com/Buildly-Marketplace/crm_service. # Solution Adjust the Markdown formatting in the README and rewrite to lay out clear instructions - this may be modified depending on the addition of new features such as containerization and continuous integration.
#3# Issue As the microservice grows more complex and for better integration with Buildly Core, we would like there to be an autogenerated Swagger documentation for the service. # Solution Use an existing library (such as `flask-swagger`) to generate a `swagger.json` file and add an endpoint that allows that documentation to be viewed - or, refactor the API to use Flask RestPlus.
#2# Issue The individual components of the service (the machine learning model, the server, the dataset) are currently stored in the same root directory, which can lead to issues when debugging an individual component. # Solution Create separate directories for each component and make sure the references to specific files are modified so that the service still runs. # References Organizing your Flask project https://exploreflask.com/en/latest/organizing.html
#1# Issue Currently the microservice is not containerized and those who wish to run it must use their existing Python installation or a virtual environment to install the requirements, which may conflict with their existing libraries. # Solution Write a Dockerfile (preferably with Python 3.7 Alpine) and an accompanying docker-compose.yml that will run the Flask app. # References Docker Compose with Flask Apps https://runnable.com/docker/python/docker-compose-with-flask-apps