Lambda image directory structure / imports

Question

I'm deploying a lambda function as an image. Locally I have the following structure:

```

├── project_folder/
    ├── lambda_handler.py
    ├── utils/
        ├── function_1.py
        ├── function_2.py
    └── modules/
    ├── module_1.py
    ├── module_2.py
```

My dockerfile looks like this

```
FROM public.ecr.aws/lambda/python:3.11
COPY project_folder/ requirements.txt ./

RUN pip3 install -r requirements.txt

CMD ["lambda_handler.lambda_handler"]
```

In lambda_handler.py I'm doing:

```
from utils import function_1.py
module = importlib.import_module(module_name)
```

In the modules I'm doing:

```
from ..utils import function_2
```

In the function_1.py I'm doing:

```
import function_2
```

I can't for the life of me figure it out how the structure and imports should be. I tried multiple combinations folder structure and imports but I keep getting errors like:

[ERROR] Runtime.ImportModuleError: Unable to import module 'lambda_handler': No module named 'function_1'
[ERROR] Runtime.ImportModuleError: Unable to import module 'lambda_handler': attempted relative import beyond top-level package

With the imports I'm doing, how should the directory structure be and how to do the relative imports?

Answer

Hi.

I'm using `handler` for the entry point function name to avoid confusion in the example below.

This is my folder structure:
![Folders tree screenshot](/media/postImages/original/IMDuhTxhkYSD6t4kqejyQqqQ)

`Dockerfile`:
```docker
FROM public.ecr.aws/lambda/python:3.11
COPY project_folder/ requirements.txt ./

RUN pip install -r requirements.txt

CMD ["lambda_handler.handler"]
```

`project_folder/modules/__init__.py` is empty. See [Python packages documentation](https://docs.python.org/3/reference/import.html#regular-packages) for more details why you will need it.

`project_folder/modules/module1.py`:
```python
from utils.function_1 import get_python_version

def module_say_hello():
    return "I'm using " + get_python_version() + "!"
```

`project_folder/utils/function_1.py`:
```python
import sys
def get_python_version():
    return 'Python' + sys.version + '!'
```

`project_folder/lambda_handler.py`:
```python
from modules.module1 import module_say_hello

def handler(event, context):
    return module_say_hello()
```

`template.yaml` - SAM template for test deployment to an AWS account:
```yaml
AWSTemplateFormatVersion: 2010-09-09
Transform: AWS::Serverless-2016-10-31

Resources:
  PythonDockerImageFunction:
    Type: AWS::Serverless::Function
    Properties:
      PackageType: Image
      MemorySize: 2048
      Timeout: 900
    Metadata:
      DockerContext: ./
      Dockerfile: Dockerfile
```

Test invoke with AWS CLI:
![Output screenshot](/media/postImages/original/IMC_OOcAp2RP2NqwBjC4Opfg)

**Additional references**:
- [Deploy Python Lambda functions with container images](https://docs.aws.amazon.com/lambda/latest/dg/python-image.html)
- [Using container image support for AWS Lambda with AWS SAM](https://aws.amazon.com/blogs/compute/using-container-image-support-for-aws-lambda-with-aws-sam/)

Best regards,

Lambda image directory structure / imports

Relevant content