Skip to main content

​
Overview

The Containerize class is a Python decorator that automatically builds, packages, and runs your application in a container. When applied to a function, it handles all the containerization steps including Dockerfile generation, image building, and container execution.

​
Constructor

Containerize(runtime={}, package={})

​
Parameters

​
runtime
dict
default:"{}"
Runtime configuration options. See RuntimeOptions for available fields.
​
package
dict
default:"{}"
Package configuration options. See PackageOptions for available fields.

​
Usage

​
Basic Example

from metaparticle_pkg.containerize import Containerize

@Containerize(
    package={
        'repository': 'myrepo',
        'name': 'myapp'
    },
    runtime={
        'executor': 'docker',
        'ports': [8080]
    }
)
def main():
    print("Hello from container!")

if __name__ == '__main__':
    main()

​
With Publishing

@Containerize(
    package={
        'repository': 'myrepo',
        'name': 'myapp',
        'publish': True,
        'builder': 'docker'
    },
    runtime={
        'executor': 'kubernetes',
        'replicas': 3,
        'ports': [8080],
        'public': True
    }
)
def serve():
    # Your application code
    pass

​
Behavior

When the decorated function is called:
  1. Container Detection: Checks if already running inside a container
  2. Direct Execution: If in container, executes the wrapped function directly
  3. Containerization: If not in container:
    • Generates a Dockerfile (or uses custom dockerfile if provided)
    • Builds the container image
    • Publishes the image if publish=True
    • Runs the container using the configured executor
    • Returns container logs

​
Methods

​
__call__(func)

Makes the class instance callable as a decorator.
​
func
function
required
The function to containerize
Returns: A wrapped function that handles containerization

​
Properties

​
image

The full image name constructed from package repository and name. Format: {repository}/{name}:latest

​
builder

The selected builder instance based on package.builder configuration.

​
runner

The selected runner instance based on runtime.executor configuration.

​
Signal Handling

The decorator automatically handles SIGINT (Ctrl+C) by:
  • Canceling the running container
  • Cleaning up resources
  • Exiting gracefully

​
Custom Dockerfiles

You can provide a custom Dockerfile by setting the dockerfile parameter in PackageOptions: 
@Containerize(
    package={
        'repository': 'myrepo',
        'name': 'myapp',
        'dockerfile': './custom.Dockerfile'
    }
)
def main():
    pass
If not provided, a default Dockerfile is generated using the Python version specified in py_version.

Build docs developers (and LLMs) love

Get started for freeTalk to us