Reduce warm lambda processing time and add SAM framework

[andrba]

Hi, we would like to propose the following changes to your repo:

Performance optimisation

clamscan loads virus databases each time it is executed. The actual scan operation takes an order of magnitude less time to complete.

By starting ClamAV as a daemon and loading the virus databases only once, the execution time on warm lambdas improves drammatically:

Deployment optimisation

Managing the deployments and AWS resources and policies manually is hard to track and prone to human error. Infrastructure as code greatly reduces the risk of resource management and traceability.

The deployment process described in the README.md is written in a template file that can be deployed via SAM.

This PR can be split into two if necessary.

Thank you for building this tool 🙏

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[sean-redmond]

seems very sensible and useful PR

[fbuchmeier]

Hi folks,

I am currently testing this and the PR requires aws-sam-cli to be installed on the local machine (referenced in the Makefile).

So to make all work again with this PR, a

 pip install aws-sam-cli

Is required.

I tried to use a virtualenv/pipenv to install this, but this is not picked up by the Makefile.

Would it be a better option to include aws-sam-cli in the Docker Image and run it from there instead of relying on the existence on the build machine?

Regards,
Florian.