figshare-uploader is a bulk upload program for the research archiving & storage provider, Figshare.
This is an unofficial tool and is not developed by Figshare-the-company.
It will accept spreadsheets in .xlsx format as input, upload the files listed
therein, with their associated metadata.
It is a small GUI program that works on Windows, Mac, and Linux. It supports uploading files of unlimited size.
A personal token is required to access your Figshare account. Please read the instructions for getting one.
The program works simply. You prepare your upload metadata in Excel format, following a certain schema. The settings allow flexible mapping of Excel fields.
You need three files:
Start the program, press the 'Pick' button to pick the file. Enter your token. Then go to the Settings dialogue in the menu (named Preferences on Mac systems). You'll see the field mapping menu. Choose "Load". Then load in the JSON file. Press OK.
Now press "Start upload process". The tool will create an article in your
account named Test article, with eye.jpg attached to it.
You can copy the sheet and modify it for your own purposes. All rows below the header in the spreadsheet will be treated as items to upload. (Other sheets in the workbook will be ignored at present. Only the first one will be handled.)
- Mac: There is currently a .dmg file. This is built on an Intel Mac. There's no native build for Apple Silicon yet, I do not have access to this hardware.
- There's a portable .zip file for Windows machines.
- Linux users must build it yourself.
Linux is the primary testing and development platform.
Currently the tool does not operate as a 'synchronization': that is, if you already have the articles present and uploaded, it will attempt to upload them again, potentially creating duplicates.
It does not have support for pause and resuming, of the kind that the API would allow, and dealing with intermittent connections. That kind of support would be useful when dealing with upload of very large files.
It only supports .xlsx format sheets as input.
This work was supported by the Sussex Humanities Lab.
Apache 2.0, see the LICENSE file.