Rule "studyid_not_in_content" should gracefully pass when unable to read incoming file

[jujaga]

OCWA Issue

User Story

As a researcher,

I want the "studyid_not_in_content" rule to not fail on files which do obviously do not have study ids

so that I can submit files such as images and other binary content.

Test Case

ENV

• DEV
• TEST
• PROD

TESTCASE

• Add any image file

EXPECTED

• The studyid_not_in_content rule should pass

ACTUAL

• The studyid_not_in_content rule fails

Notes

The rule needs to be rewritten to gracefully fail upon failure to decode instead of hard failing.

[jujaga]

Rule was updated to print(not bool(re.compile(b'[\\w]{1}[\\d]{9}').search(${file.content}))) which handles binary content better and satisfies this issue.

[pripley123]

tried creating a new request in the test environment and uploaded a jpg. Saved and closed. Viewed the request. Still flagged for study ids

[pripley123]

I've tried a couple of more jpg - these are passing ... so I need to narrow this down a bit

[jujaga]

Certain image files have an explicit string inside the binary which will trip the [\w]{1}[\d]{9} regex rule as it only looks for either

1. a letter followed by exactly 9 digits or
2. exactly 10 digits

The implemented rule is behaving as expected; however until we are provided better requirements that do not match on so many potential values, some images are inevitably going to fail.

[pripley123]

Thanks for documenting the reason for images failing. Agreed it is expected behaviour for now. Since the regex used to search for study ids is configurable, this seems fine for MVP