Admin tests

[actlikewill]

• this adds tests for the judgment admin

[sentry-io]

🔍 Existing Issues For Review

Your pull request is modifying functions with the following pre-existing issues:

📄 File: peachjam/models/core_document_model.py

Function	Unhandled Issue
update_or_create_for_document	CalledProcessError: Command '['/app/peachjam/../bin/pdfjs-to-text', '/tmp/tmpklra61y4', '/tmp/tmpj6l0ot__.txt']' retu... ... Event Count: 1

---

_{Did you find this useful? React with a 👍 or 👎}

[actlikewill]

@longhotsummer these tests are already revealing some issues. There is some inconsistency with the way we set mimetypes for source documents. When adding new documents, to set the mimetype, we read from file.content_type but when we are modifying, we use the python magic library. These two dont always give the same results. It could also be because of the way Webtest is mocking the files but its making the tests that I'm writing fail.

[longhotsummer]

@longhotsummer these tests are already revealing some issues. There is some inconsistency with the way we set mimetypes for source documents. When adding new documents, to set the mimetype, we read from file.content_type but when we are modifying, we use the python magic library. These two dont always give the same results. It could also be because of the way Webtest is mocking the files but its making the tests that I'm writing fail.

This is good to learn. We file.content_type trusts the browser, whereas the python magic library actually looks at the file contents. So b"pdf judgment content" won't be considered a PDF by the magic library.

An option is to add some small PDF and DOCX files as fixture objects and post real files.

[actlikewill]

An option is to add some small PDF and DOCX files as fixture objects and post real files.

@longhotsummer I've tried this but it doesn't seem to work properly. The webtes.Upload doesn't seem to work with actual files. Still checking though

[actlikewill]

@longhotsummer this is the assertion that revealed the issue. I'm checking that the content_html has been changed, but since the mimetype is incorrect, it is not extracting the content. The mimetype is reported as text/plain

[actlikewill]

@longhotsummer this pr will merge to this one #1771 with the bug fix, the we can merge it to master