rlegorreta/udf-run

Microservice in Python to store the UDFs code

UDF-RUN

Microservice that runs un Python all User Defined Functions (UDFs)acts. The Python code is set by the microservice udf-ui, for each UDF. The Python code is defined is to defined in the udf-ui using the function in a Blockly manner and translates as Python code (the end user does not require Python knowledge).

The udf-run microservice is the one that executes the udf and is called from any UI.

UI for the definition of User Define Functions(UDF)

The UDFs functionality is composed by three microservices: udf-ui, udf-service and udf-run (in Python).

The UDF functionality consists in the extension of a data model with extra derived attributes that can be programmed without the need to ad code, reprogram or re-install eny microservices. This is done by what we define as power users but they do not need to have any programming experience.

They could be compared as macros for a spreadsheets but inside the AI Legorreta Market Pl.

Introduction

User defined functions (UDFs) is a way to extend the financial data model without the need to modify or add code to the system.

The steps to create a new UDf are the following:

• Create a new UDF in this microservice udf-ui where the end user defines the input parameters in order to do some calculations and produce a result.
• In this same fron udf-ui the end user defines the fields mapping from the input parameters of the UDF and the fields defined in a datasource.
• A datasource are a group of attributes that a microservice exports. This microservice support UDFs.
• In this microservice udf-ui also the end user can define the calculations needed in Blockly format that it is translated to a Python code and store in a file to be read and executed by the udf-run.
• In the microservices where the datasource is defined and exported can be added in any Vaadin component like a grid or any textFiled in a Vaadin form or even a report.
• The UDF calls the microservice udf-run is the piece of software that calls the Python function and returns the result to the microservice.

note: Since Python is an interpreted language no recompile, or re-initialization of udf is needed, it is done on-line and no coding is necessary.

Python project as a microservice:

For more information to create a Python project as a microservice see::

https://github.com/tiangolo/uwsgi-nginx-flask-docker

Example for a UDF

Several UDFs are defined inside the cartera demonstration microservice:

• In the portfolio grid component we add an extra column that it gives a grade for the instrument. The UDF is called 'Calific_Actinver'.
• Another UDF fo create a new textField inside the form that list a detail of the cash movement. The UDf is called 'Mov-efectivo'
• Two microservices are define cartera and ingestor.

note: The database udfdb is initialized with these data by the udf-service microservice.

Port availability

This is to detect port availability (for any micro.service is usefull)

sudo nc localhost 7076 < /dev/null; echo $?

Compilation

If we want to crate a new docker

sudo bash start.sh

Or build the image in this two steps:

(step 1) Build the image:

docker build -t udf-run .    

(step 2) Build the container

docker run -d -p 7076:80 --name=udf-run -v $PWD:/app udfrun

Run the container

And to run an already created docker :

docker start udf-run

To stop the docker

docker stop udf-run

Update the code without shutting down the container

When we want to update the Python code:

sudo touch uwsgi.ini

More information

See links:

https://www.digitalocean.com/community/tutorials/how-to-build-and-deploy-a-flask-application-using-docker-on-ubuntu-18-04

Or for Python Flask see:

https://programminghistorian.org/en/lessons/creating-apis-with-python-and-flask

Contact AI Legorreta

Feel free to reach out to AI Legorreta on web page.

Version: 2.0.0 ©LegoSoft Soluciones, S.C., 2023