We should support having an input as an image

[pombredanne]

VMDK, VDI and QCOW2 are common VM image formats for rootfs (so is squashfs but we already support this).
See aboutcode-org/extractcode#16 as this is the primary dependency.

[pombredanne]

See aboutcode-org/extractcode#20 for some progress

[tdruez]

The image extraction works but the extracted structure fails with the rootfs pipeline:

codebase/image.qcow2
codebase/image.qcow2-extract

'NoneType' object has no attribute 'merge'

Traceback:
  File "scancode.io/scanpipe/pipelines/__init__.py", line 96, in execute
    step(self)
  File "scancode.io/scanpipe/pipelines/root_filesystems.py", line 54, in find_root_filesystems
    self.root_filesystems = list(rootfs.RootFs.from_project_codebase(self.project))
  File "scancode.io/scanpipe/pipes/rootfs.py", line 101, in from_project_codebase
    yield RootFs(location=rootfs_location)
  File "<attrs generated init scanpipe.pipes.rootfs.RootFs>", line 4, in __init__
    self.__attrs_post_init__()
  File "scancode.io/scanpipe/pipes/rootfs.py", line 88, in __attrs_post_init__
    self.distro = Distro.from_rootfs(self.location)
  File "scancode.io/lib/python3.9/site-packages/container_inspector/distro.py", line 355, in from_rootfs
    return base_distro.merge(found)

[tdruez]

Support added and available in https://github.com/nexB/scancode.io/releases/tag/v21.6.10