8344 file upload: default mime type

[landreev]

see the issue.

What this PR does / why we need it:

Which issue(s) this PR closes:

Closes #8344

Special notes for your reviewer:

Suggestions on how to test this:

To reproduce the issue: try to upload a valid jpeg, exactly as shown below:

cat test.jpg | curl -H X-Dataverse-key:XXX -X POST -F "file=@-" -F 'jsonData={"label":"test_stream.jpg"}' http://localhost:8080/api/datasets/NNN/add

the file will upload, but will be identified as "text/plain". The PR should fix this.

Does this PR introduce a user interface change? If mockups are available, please link/include them here:

Is there a release notes update needed for this change?:

Additional documentation:

[landreev]

testing slack activity feed. (@pdurbin)

[coveralls]

Coverage decreased (-0.001%) to 18.852% when pulling 8698f78 on 8344-file-upload-default-mime into ee7cc4b on develop.

[kcondon]

@landreev I can deploy due to a flyway conflict, does this pr need to be refreshed from develop?

[landreev]

Having looked into the reason this could not be easily reproduced on datavese-internal - my very first guess (shockingly!) appears to have been correct. The version of curl we have on the linux boxes is smarter than the one I have on my MacBook. It does actually send the default mime type header when called as described in my test scenario:

002c: Content-Disposition: form-data; name="file"; filename="-"
0067: Content-Type: application/octet-stream

... and that triggers the application to use the jhove-detected type, so the file ends up being properly typed as "image/jpeg".

To reiterate, the problem this PR was supposed to solve was not with curl. The problem was a file upload without a mime type header in the request. The curl line was just an easy way to produce such an upload that I knew of.

You should still be able to replicate it with curl, by going remotely from your own Macbook, like

cat some_valid.jpg | curl -k -H X-Dataverse-key:xxx -X POST -F "file=@-" -F 'jsonData={"label":"test_stream.jpg"}' https://dataverse-internal.iq.harvard.edu/api/datasets/NNNNN/add

(assuming your curl version is the same as mine; I would not count on it 100%)

Yet another way to test it would be using pyDataverse (that's how the issue was discovered; and that's the main use case I'm trying to fix). But then you'll have to mess with Python code, installing the module, etc.