Import job run outside webapp

[sverhoeven]

Is your feature request related to a problem? Please describe.
As a researcher I have run haddock3 on the command line and would like to look at the results using the webapp.

Describe the solution you'd like
For big jobs I would like to point to a location on webapp server where my output is located.
For medium jobs I would like to upload an archive to the webapp.

Describe alternatives you've considered

Additional context

[sverhoeven]

The webapp currently expects a job to have

1. a worklow.cfg in the root
2. a output/ directory in the root with the haddock3 output.

The bartender web service to list the jobs, so any imported job should go via bartender.

Possible implementation for archive

We could add an application to bartender webservice that unpacks the archive and does not run anything else.
Then the webapp could have a choice on the upload page to upload either a workflow.cfg with its linked files or an archive with a completed job. Someone can upload a competed job if they have the guru expertise level.

Possbile implementation for outside job

We could add a bartender cli subcommand that would symlink the dir into bartenders job root dir. This way we don't make a copy. The symlinked files will not be owned by web service which breaks running rescore and friends.

[sverhoeven]

In i-VRESSE/bartender#86 a bartender link <somedir> command was added.

[sverhoeven]

In #81 upload run radio was added.