S3 bucket for country data is not accessible

[syltruong]

When following the instructions on the README.md and after installing python dependencies and running streamlit run corona-calculator.py locally, the app fails with

KeyError: 'AWSAccessKeyId'
Traceback:
  File "/Users/struong/project/corona-calculator/.venv/lib/python3.7/site-packages/streamlit/ScriptRunner.py", line 314, in _run_script
    exec(code, module.__dict__)
  File "/Users/struong/project/corona-calculator/corona-calculator.py", line 6, in <module>
    import models
  File "/Users/struong/project/corona-calculator/models.py", line 5, in <module>
    import data.constants as constants
  File "/Users/struong/project/corona-calculator/data/constants.py", line 14, in <module>
    from s3_utils import download_file, S3_DISEASE_DATA_OBJ_NAME
  File "/Users/struong/project/corona-calculator/s3_utils.py", line 10, in <module>
    _S3_ACCESS_KEY = os.environ["AWSAccessKeyId"].replace("\r", "")
  File "/usr/local/bin/../Cellar/python/3.7.5/bin/../Frameworks/Python.framework/Versions/3.7/lib/python3.7/os.py", line 679, in __getitem__
    raise KeyError(key) from None

• Is there a way we can make it run locally? with and/or without access to the S3 bucket coronavirus-calculator-data.

• Is it possible to have that bucket public?

[archydeberker]

Hey! Very good question. @psteeves can you chip in on this?

I don't think we can make S3 public, but we should add functionality to mock it locally, I think.

[psteeves]

I see two options:

• Add a csv version of the data to the repo we can fallback on if the credentials are not available (kinda hacky)
• Add a read-only S3 access key to the repo (better option IMO). Users would not be able to update the data (requires write access) but would be able to pull the latest data.

[archydeberker]

I'd prefer to simulate S3 locally for fear of leaking S3 credentials which, even if RO, might lead to spamming and DOS.

[psteeves]

@syltruong Fixed by https://github.com/archydeberker/corona-calculator/pull/70 Will deploy to prod in the next half hour.