Regression: errors out if no indy

[dbluhm]

We've inadvertently made indy a requirement again with recent changes. Using the following dockerfile:

FROM bcgovimages/von-image:py36-1.16-0
RUN pip install git+https://github.com/hyperledger/aries-cloudagent-python@main
RUN yes | pip uninstall python3-indy

ENTRYPOINT ["/bin/bash", "-c", "aca-py \"$@\"", "--"]

CMD ["start", "-it", "http", "0.0.0.0", "3000", "-ot", "http", "-e", "http://localhost:3000", "--admin", "0.0.0.0", "3001", "--admin-insecure-mode", "--no-ledger"]

The following error occurs:

Traceback (most recent call last):
  File "/home/indy/.pyenv/versions/3.6.13/lib/python3.6/runpy.py", line 193, in _run_module_as_main
    "__main__", mod_spec)
  File "/home/indy/.pyenv/versions/3.6.13/lib/python3.6/runpy.py", line 85, in _run_code
    exec(code, run_globals)
  File "/home/indy/.pyenv/versions/3.6.13/lib/python3.6/site-packages/aries_cloudagent/__main__.py", line 74, in <module>
    main(sys.argv)
  File "/home/indy/.pyenv/versions/3.6.13/lib/python3.6/site-packages/aries_cloudagent/__main__.py", line 71, in main
    run(args)
  File "/home/indy/.pyenv/versions/3.6.13/lib/python3.6/site-packages/aries_cloudagent/__main__.py", line 64, in run
    run_command(command, args)
  File "/home/indy/.pyenv/versions/3.6.13/lib/python3.6/site-packages/aries_cloudagent/commands/__init__.py", line 36, in run_command
    module = load_command(command) or load_command("help")
  File "/home/indy/.pyenv/versions/3.6.13/lib/python3.6/site-packages/aries_cloudagent/commands/__init__.py", line 31, in load_command
    return import_module(module_path)
  File "/home/indy/.pyenv/versions/3.6.13/lib/python3.6/importlib/__init__.py", line 126, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "<frozen importlib._bootstrap>", line 994, in _gcd_import
  File "<frozen importlib._bootstrap>", line 971, in _find_and_load
  File "<frozen importlib._bootstrap>", line 955, in _find_and_load_unlocked
  File "<frozen importlib._bootstrap>", line 665, in _load_unlocked
  File "<frozen importlib._bootstrap_external>", line 678, in exec_module
  File "<frozen importlib._bootstrap>", line 219, in _call_with_frames_removed
  File "/home/indy/.pyenv/versions/3.6.13/lib/python3.6/site-packages/aries_cloudagent/commands/start.py", line 16, in <module>
    from ..core.conductor import Conductor
  File "/home/indy/.pyenv/versions/3.6.13/lib/python3.6/site-packages/aries_cloudagent/core/conductor.py", line 35, in <module>
    from ..protocols.out_of_band.v1_0.manager import OutOfBandManager
  File "/home/indy/.pyenv/versions/3.6.13/lib/python3.6/site-packages/aries_cloudagent/protocols/out_of_band/v1_0/manager.py", line 16, in <module>
    from ....indy.models.xform import indy_proof_req_preview2indy_requested_creds
  File "/home/indy/.pyenv/versions/3.6.13/lib/python3.6/site-packages/aries_cloudagent/indy/models/xform.py", line 5, in <module>
    from .pres_preview import IndyPresPreview
  File "/home/indy/.pyenv/versions/3.6.13/lib/python3.6/site-packages/aries_cloudagent/indy/models/pres_preview.py", line 9, in <module>
    from ...ledger.indy import IndySdkLedger
  File "/home/indy/.pyenv/versions/3.6.13/lib/python3.6/site-packages/aries_cloudagent/ledger/indy.py", line 14, in <module>
    import indy.ledger
ModuleNotFoundError: No module named 'indy'

I believe this might have been introduced with #1125 @shaangill025

[dbluhm]

@Luis-GA related to the issue we were seeing from acapy-resolver-universal

[TimoGlastra]

Maybe good to add a Github Action script that builds and start ACA-Py using the above docker script?

[dbluhm]

@TimoGlastra agreed; we should be able to catch this on PRs without too much difficulty. Or else, making the unit test runner for non-indy tests not actually have python3-indy available in the python environment.

[shaangill025]

@dbluhm Any ideas when was the last time this was working? I used the above docker script to test with edfd5b52abd3f2fb4dbf30351cde35560bbb1242 [from May] and b926cac23ed783adbd862ca49d760446acb005fb [from January] commits and get the same error. I believe this might have been an issue since the OOB protocol was implemented earlier this year. I am working on a fix.

[dbluhm]

@TimoGlastra fixed this once before with #1129 so it was working as expected semi-recently