Home Applications iris-embedded-python-template


Supported by InterSystems Corporation
2 reviews
IPM installs
Pull requests
The simplest template to run embedded python

What's new in this version

Thanks to Guillaume Rongier and Dmitry Maslennikov added following features to IRIS Embedded Python template:

  1. Clear Python setup (no ObjectScript needed).

  2. Merge IRIS initialization example

  3. Embedded Python IRIS initialization example with IPM support.

  4. Flask server setup example


This is a template to work with Embedded Python in InterSystems IRIS
It demonstrates how to call python libs from ObjectScript in dc.python.test class.
And it demonstrates how to deal with IRIS from python scripts - python/irisapp.py

What is Embedded Python ?

Embedded Python is a feature of InterSystems IRIS that allows you to run python code in the same process as the IRIS database engine.

The benefits of Embedded Python are:

  • Performance
    • no need to serialize data between IRIS and Python
    • speed of data processing in Python is comparable to ObjectScript
  • Simplicity
    • no need to install and configure Python separately
    • easy to deploy IRIS and Python together
    • easy access to ObjectScript code and functionality from Python
  • Security
    • no need to open any additional ports for communication between IRIS and Python

And the main benefit is that you can use all the power of Python libraries and frameworks in your InterSystems IRIS solutions.


Make sure you have git and Docker desktop installed.

Installation ZPM

Open IRIS terminal in your IRIS with installed IPM. Run the command:

USER>zpm "install iris-python-template"

Installation docker

Clone/git pull the repo into any local directory

$ git clone https://github.com/intersystems-community/iris-embedded-python-template.git

Open the terminal in this directory and run:

$ docker-compose build
  1. Run the IRIS container with your project:
$ docker-compose up -d

IRIS Initialization

In this template two approaches are provided to initialize iris: merge and python.
merge.cpf is a convenient way to setup different IRIS configuration settings. Learn more about merge.cpf.

  1. Using merge to initialize IRIS and create IRIS Database and Namespace
    Notice merge.cpf file that is being implemented during docker image build in Dockerfile
iris merge IRIS merge.cpf && \

that contains:

CreateResource:Name=%DB_IRISAPP_DATA,Description="IRISAPP_DATA database"
CreateResource:Name=%DB_IRISAPP_CODE,Description="IRISAPP_CODE database"

As you can see it creates dabasases IRISAPP_DATA and IRISAPP_CODE for data and code, the related IRISAPP namespace to access it and the related resources %IRISAPP_DATA and %IRISAPP_CODE” to manage the access.

Also it enables Callin service to make Embedded python work via ModifyService clause.
and it updates the password for the built-in user SuperUser to “SYS”. The hash for this password is obtained via the following command:

docker run --rm -it containers.intersystems.com/intersystems/passwordhash:1.1 -algorithm SHA512 -workfactor 10000
  1. Using python to initialize IRIS.
    Often we used a special iris.script file to run ObjectScript commands during the initialization - it is here just for the information.
    This template shows you how to use python for the same purpose with iris_script.pyfile.
    It is being executed via the line in Dockerfile:
irispython iris_script.py && \

the iris_script.py file contains examples how developer can initialize different services of iris via Python code.

How to test it

Working with IRIS from Embedded Python

As mentioned Embedded Python works in the same process as IRIS.

So you have 2 options to work with Embedded Python in IRIS:

  1. Run the code in IRIS container with a shared folder.
  2. Bind VsCode to the running IRIS container.

Run the python script in iris container:

# attach to the running IRIS container
docker-compose exec iris bash
# run the script
$ irispython ./python/irisapp.py

The script contains different samples of working with IRIS from python and goes through it.
it should return something like this:

Hello World
Method call:
It works!
Iris Version:
IRIS for UNIX (Ubuntu Server LTS for ARM64 Containers) 2023.2 (Build 227U) Mon Jul 31 2023 17:43:25 EDT
Creating new record in dc.python.PersistentClass
Printing one IRIS Object Dump:
+----------------- general information ---------------
|      oref value: 1
|      class name: dc.python.PersistentClass
|           %%OID: $lb("1","dc.python.PersistentClass")
| reference count: 1
+----------------- attribute values ------------------
|       %Concurrency = 1  <Set>
|               Test = "2023-09-03 10:56:45.227577"
Running SQL query Select * from dc_python.PersistentClass
[0]: ['1', '2023-09-03 10:56:45.227577']
Printing the whole global of the persistence storage for the class dc.python.PersistentClass:^dc.Package4C8F.PersistentC1A93D
key=['1']: 2023-09-03 10:56:45.227577

Bind VSCode to the running IRIS container

Open VSCode in the project directory.

Go to the docker-compose.yml file, right-click on it and select Compose Up.

Once the container is up and running you can open the docker extension and right-click on the container name and select Attach Visual Studio Code.

Working with IRIS from Embedded Python

Open VSCode in Devcontainer - this is the bell(notifications) button in the left bottom corner, where you will see the suggestion to open VSCOde in DevContainer mode.
Follow it - it will let to execute Embedded Python scripts vs IRIS and develop it at the same time.

Once devcontainer is opened go to /python/irisapp.py and run it, either with Run button in the top right corner, or in terminal via:

$ irispython /python/irisapp.py

Working with Python libs from ObjectScript

Open IRIS terminal:

$ docker-compose exec iris iris session iris

The first test demonstrates the call to a standard python library working with dates datetime

IRISAPP>d ##class(dc.python.test).Today()

Another example shows the work of a custom lib sample.py which is installed with repo or ZPM. It has function hello which returns string “world”:

IRISAPP>d ##class(dc.python.test).Hello()

Another example shows how to work with files and use pandas and numpy libs.
It calculates the mean age of Titanic passengers:

IRISAPP>d ##class(dc.python.test).TitanicMeanAge()
mean age=29.69911764705882

Develop python scripts locally

By default, the template is configured to use the shared folder ./python for python scripts to /home/irisowner/dev/python in IRIS container.

You can change the folder according to your preferences. Usually python developers name project at the root folder.

It’s recommended to work with a virtual environment.

Create a virtual environment in the project directory. Click New Terminal in the VS Code menu:

$ python3 -m venv .venv

Activate the virtual environment:

$ source .venv/bin/activate

Install the requirements:

$ pip install -r requirements.txt

Working with flask

The template also contains samples of working with flask.

Connect to the running with a bash terminal:

$ docker-compose exec iris bash

Run the following commands to start the flask server:

irispython /python/flask/app.py

That will start the flask server and you will see the following output:

 * Serving Flask app 'app'
 * Debug mode: off
WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead.
 * Running on

5000 is mapped to 55030 in docker-compose.yml

Test it

Hello world :

GET http://localhost:55030/
Accept: application/json


    "message": "Hello world"

Post a new persistent class

POST http://localhost:55030/persistentclass
Content-Type: application/json
Accept: application/json
    "test": "toto"


    "id": 1,
    "test": "toto"

Get the persistent class

GET http://localhost:55030/persistentclass/1
Accept: application/json


    "id": 1,
    "test": "toto"

Feel free to use the template for your own development just by adding new py files.

Made with
zpm install iris-python-template download archive
3.0.003 Sep, 2023
Works with
InterSystems IRIS
First published
09 Feb, 2021
Last checked by moderator
27 Jun, 2023Works