Add safeguard against S3 overwrites

[zstumgoren]

Add safeguard against accidental overwriting of a previously created project on S3.

Implementation

Some initial thoughts on implementation

Create S3 integration code

Create lib-level code and related data:s3:integrate command that:

• Generates a unique hash
• Checks if bucket@s3_path/.project_id exists
• If exists, errors out
• If not exists:
  • Stores hash as project_id setting in config/datakit-data.json
  • Creates bucket@s3_path/.project_id with the unique hash

Initialization

• Triggers data:s3:integrate
• If bucket@s3_path/.project_id exists:
  • Store config/datakit-data.json without the project_id setting
  • Warn user that the s3_path already exists, that they should update the setting and run data:s3:integrate
• If the hidden file does not exist, proceed with the actions of data:s3:integrate:
  • Stores hash as project_id setting in config/datakit-data.json
  • Creates bucket@s3_path/.project_id with the unique hash

Push/Pull

Update data push and pull to always run a preliminary check of project_id from local config against bucket@s3_path/.project_id

• If hashes match, proceed with operation
• If hashes don't match, exit with a warning

[zstumgoren]

Agreed we should do this for v0.2.0. Removing Question label