1721 restrict editing to certain submission statuses

[samobermiller]

this PR closes issue #1721. NOTE: to be merged in alongside PR #[tbd] that will enable drop down menu for status changes

Currently, a submission can continue to be edited after hitting the Submit button and there is no functionality to the UpdatesRequired submission status.

Updates:

• Edits are now only allowed if a submission's status is InProgress or UpdatesRequired
  • This means you can no longer edit a submission after hitting the Submit button (when the status is SubmittedPendingReview)
  • This is communicated by a banner at the top of each page (superseding all other banners) saying edits aren't permitted due to status
• Submissions are view-only if status is anything other than InProgress or UpdatesRequired
  • Resume button now says View with an eye symbol
• If a submission's status is UpdatesRequired and edits are made, the status becomes InProgress again
  • Montana will manually assign a status of UpdatesRequired after review via a yet to be implemented drop down
• Github issues are now only created if there is no other issue in the repo with the same submission id. If there is an existing issue, a comment is added saying it was resubmitted on x date and the issue is reopened if closed (test created in test_submission.py)
• Restrict edits via API based on submission status (test created in test_submission.py)

Note: status does not change from 'InProgress' to 'UpdatesRequired' if admin makes edits, this is intentional

[naglepuff]

Looks good! I was not able to do any functional testing for the github endpoint, but tests of the submission portal components looked good. I had a few minor code comments is all.

[pkalita-lbl]

Changes look good overall.

As I was testing I noticed that I wasn't seeing the status automatically change from UpdatesRequired to InProgress. I could see that the InProgress status was part of the PATCH /api/metadata_submission/{id} request, but the response still had UpdatesRequired in it. I think the source of that bug is in the update_submission route handler function gates the status change on checking if the logged-in user has role of owner on the submission. In my case I was editing someone else's submission as an admin. It's an edge case, but it might be good to clean that up now as we start making more extensive use of statuses.

[mslarae13]

lgtm from a totally non technical perspective.
The description's expected behaviors are accurate.