Allow archive to be run as its own module

[sgsmob]

[edit:krivard] Fixes #280.

Archiving can now be run from the command line as a module, e.g.,

$ python -m delphi_utils.archive --help

usage: archive.py [-h] --archive_type {git,s3} [--indicator_prefix INDICATOR_PREFIX] [--branch_name BRANCH_NAME]
                  [--override_dirty] [--commit_partial_success] [--commit_message COMMIT_MESSAGE]

optional arguments:
  -h, --help            show this help message and exit
  --archive_type {git,s3}
                        Type of archive differ to use.
  --indicator_prefix INDICATOR_PREFIX
                        The prefix for S3 keys related to this indicator. Required for `archive_type = 's3'
  --branch_name BRANCH_NAME
                        Branch to use for `archive_type` = 'git'.
  --override_dirty      Whether to allow overwriting of untracked & uncommitted changes for `archive_type` = 'git'
  --commit_partial_success
                        Whether to still commit for `archive_type` = 'git' even if some files were not archived and staged due to
                        `override_dirty` = False.
  --commit_message COMMIT_MESSAGE
                        Commit message for `archive_type` = 'git'

[sgsmob]

See #280

[huisaddison]

tl;dr : LGTM

• Read through source - code looks good - added ArchDiffer.run() and run_module() (a few diffs are line breaks to enforce character limits, which I appreciate)
• Compared the implementation of ArchiveDiffer.run() [link] against the ArchDiffer runs in delphi_jhu [link] and delphi_google_health [link]. Looks like a straightforward abstraction of the functionality to me.
• Unit tests look good to me (also looks like @sgsmob deduped some local constants into global constants :) ). Checked out the PR locally, installed module, ran unit tests, everything passes.

@krivard is there anything else you want me to look into for this code review? For @sgsmob 's context - I am not a professional software engineer, and I haven't been heavily involved in engineering since April-ish. I wrote a bunch of initial indicator code, then switched to modeling for the summer, and then I came back to eng because the structure allows me to get Phd student work done in parallel :P So my main "strength" is probably explaining why there are certain hacks in legacy code (and spending time fixing them to pay down our tech debt!)

[krivard]

Nope! Just getting some additional eyeballs on the code. Thank you!

…

On Fri, Oct 9, 2020 at 12:46 AM Addison Hu ***@***.***> wrote: tl;dr : LGTM - Read through source - code looks good - added ArchDiffer.run() and run_module() (a few diffs are line breaks to enforce character limits, which I appreciate) - Compared the implementation of ArchiveDiffer.run() [link] <https://github.com/cmu-delphi/covidcast-indicators/pull/295/files#diff-4e2435d91326a204681786df4dd349fdR270-R292> against the ArchDiffer runs in delphi_jhu [link] <https://github.com/cmu-delphi/covidcast-indicators/blob/main/jhu/delphi_jhu/run.py#L120-L135> and delphi_google_health [link] <https://github.com/cmu-delphi/covidcast-indicators/blob/main/google_health/delphi_google_health/run.py#L93-L107>. Looks like a straightforward abstraction of the functionality to me. - Unit tests look good to me (also looks like @sgsmob <https://github.com/sgsmob> deduped some local constants into global constants :) ). Checked out the PR locally, installed module, ran unit tests, everything passes. @krivard <https://github.com/krivard> is there anything else you want me to look into for this code review? For @sgsmob <https://github.com/sgsmob> 's context - I am not a professional software engineer, and I haven't been heavily involved in engineering since April-ish. I wrote a bunch of initial indicator code, then switched to modeling for the summer, and then I came back to eng because the structure allows me to get Phd student work done in parallel :P So my main strength is probably explaining why there are certain hacks in legacy code (and spending time fixing them to pay down our tech debt!) — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#295 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAI24CU7H3O2HSP2TFIO5CTSJ2IQ3ANCNFSM4SGZANLA> .