ActiveStorage::ClamAV::Analyzer

This gem packages an analyzer to perform ClamAV virus scans on uploaded ActiveStorage::Blob objects, adding the results of the scan to the blob metadata.

The actual analyzer is very simple, and can be found in lib/active_storage/clamav/analyzer if you would prefer to just drop this in app/analyzers in your codebase and prepend it to the analyzers list yourself.

Installing ClamAV

Ensure you have ClamAV installed. This gem uses these commands, but does not set them up if they are missing. On your path you should have:

• clamav
• clamscan

On most platforms, you can install ClamAV with the package name:

• Mac OS: brew install clamav (Further setup steps are necessary with Homebrewed ClamAV, see https://gist.github.com/mendozao/3ea393b91f23a813650baab9964425b9)
• Debian/Ubuntu: apt install clamav

There are plenty of other installation methods and platforms available. More information about these is available on ClamAV's website

You can also run ClamAV scans in a Docker container. The [ClamAV documentation] has an installation page dedicated to this. While you will have to tune ActiveStorage::ClamAV::Analyzer.command for this, it should work with this gem.

Installation

Add this line to your application's Gemfile:

gem 'activestorage-clamav-analyzer', require: "active_storage/clamav/analyzer"

And then execute:

$ bundle

Or install it yourself as:

$ gem install activestorage-clamav-analyzer

This gem will automatically add itself to the analyzer pipeline and run across any supported image files. If you wish to control the precise analyzer order, you can manipulate the ActiveStorage.analyzers array.

Usage

This gem automatically adds itself to the analysis pipeline, simply ensure that analysis is run on your uploaded files.

To manually analyze a particular blob, simply grab an attachment and pass the blob directly to the analyzer:

ActiveStorage::ClamAV::Analyzer.new(ActiveStorage::Attachment.first.blob).metadata
 =>  {
    "analyzed"=>true,
    "clamav": {
        "detection": true,
        "output": "test.txt: Eicar-Signature FOUND\n\n-------" #...
    }}

Recipes

Scan with ClamD

clamscan is the default command, but starts up ClamAV from scratch each time it is run, which takes several seconds. Using clamd is much faster, but requires you to have started clamd ahead of time.

An example of the speedups that are possible:

• clamscan README.md 9.68s user 0.36s system 96% cpu 10.400 total
• clamdscan README.md 0.01s user 0.00s system 36% cpu 0.026 total

If your infrastructure set up allows you to run clamd, you can adjust the command to use clamdscan, which will scan files in a fraction of the time:

# config/initializers/active_storage.rb
ActiveStorage::ClamAV::Analyzer.command = "clamdscan"

Scan with a Docker container

ClamAV has comprehensive documentation on how to scan files in a Docker container. If you'd like to do this yourself using the ClamAV analyzer, that's no problem! You'll need to build a custom command to mount the blob's tempfile into your container to get the result. ActiveStorage::ClamAV::Analyzer.command accepts anything that responds to #call, so you can customise your command:

# config/initializers/active_storage.rb

clamav_command = (tempfile) -> { "docker run --rm -v #{tempfile.path}:#{tempfile.path} clamav/clamav clamscan" }
ActiveStorage::ClamAV::Analyzer.command = clamav_command

Report detections as an exception

The on_detection setting on ActiveSupport::ClamAV::Analyzer can be used to take some action when a detection occurs. Your application may not have a defined code path for virus detections, but you still want to know when it happens. You can use on_detection for this to report detections to your exception monitoring tool of choice.

ActiveStorage::ClamAV::Analyzer.on_detection = lambda do |blob|
  err = StandardError.new("Virus detected in ActiveStorage::Blob ##{blob.id}")
  ExceptionMonitoringService.capture_exception(err)
end

Remove blobs that have a detection

This analyzer records detections, but by default takes no action. Destroying blobs from a library could surprise some users, stops further analyzers and processing from running, and also would prevent any investigation of the blob, who uploaded it, and what the exact detection is.

The analyzer does however call a callable (Proc, lambda, etc) when a detection occurs, passing the blob - so it's very simple to remove the blob automatically when a detection occurs.

ActiveStorage::ClamAV::Analyzer.on_detection = (blob) -> { blob.destroy }

You can use the same technique to take some other action - perhaps quarantine the blob in some way (make it inactive), or add it to a moderation or alert queue.

Development

After checking out the repo, run bin/setup to install dependencies. Then, run rake to run the lint checks and tests. You can also run bin/console for an interactive prompt that will allow you to experiment.

To install this gem onto your local machine, run bundle exec rake install. To release a new version, update the version number in version.rb, and then run bundle exec rake release, which will create a git tag for the version, push git commits and tags, and push the .gem file to rubygems.org.

Contributing

Bug reports and pull requests are welcome on GitHub at https://github.com/ackama/activestorage-clamav-analyzer.