Set up staging area + travis checks for TSVs specifying new content to be loaded into KB

[dosumis]

TBD: new repo, or perhaps re-use data repo to keep private?

Repo should have:

• SOPs for adding content including TSV specs and folder locations
• processing code calling KB_tools + some standard tsv parser (have been using my own tsv2pdm, but should probably be using Python Pandas). This will likely be superceded by new code when pipeline version 2 is in place.
• A travis job to test new content and warn of issues.

Folder locations should include:

• working folders and a staging area

TBD: How do we indicate what pattern to follow when loading content. Do we use filenames or have dedicated folders for data types, e.g. one for new images, one for new datasets, one for new sites etc.

[dosumis]

Curation platform under development here: https://github.com/VirtualFlyBrain/curation