CodeMeta

CodeMeta is likely the best source of metadata, as the files typically have emails and orcid ids. Since CodeMeta already provide several high quality generation tools we don’t provide an init function here, however the update function can serve to update your .tributors metadata file and the codemeta file if an email, orcid, or name is missing. If you haven’t already, make sure that you install tributors. Here we show basic commands for interacting with the allcontrib generator, and a table of optional arguments.

Optional Arguments

name description required default
--codemeta-file the codemeta file, if changed from default false codemeta.json
--log-level Log level to use, one of INFO, DEBUG, CRITICAL, ERROR, WARNING, FATAL (default INFO) false INFO
--thresh the minimum number of contributions required to add a user false 1
--force if files exist, force overwrit false code

Init codemeta

As mentioned above, there is no init defined because tools already exist! If you run an init for codemeta you’ll see this message:

$ tributors init codemeta
Codemeta provides several tools to generate this for you: https://codemeta.github.io/tools/

The .tributors file

After you run this command, you’ll also notice you have a .tributors file in your repository. You can choose to add this to version control or not - it contains shared metadata between the services. Read more about the tributors file here.

Update

If you already have a codemeta.json or codemeta.jsonld, you can update it as follows (and importantly, extract emails or orcids from it for user’s that you’ve already defined in your .tributors file):

$ tributors update codemeta
INFO:codemeta:Updating codemeta.json

You can also target a different filename;

$ tributors update codemeta --codemeta-file codemeta.jsonld

Update .tributors

Let’s say that we have a local codemeta.json, and we just want to use it to update our .tributors file. We could do:

$ tributors update-lookup codemeta

And if you want it auto-discovered (with other known files) you can just do:

$ tributors update-lookup

For both of the above, since the cache is kept based on GitHub user id, we aren’t able to store codemeta entries unless they have an email or orcid that we can match to an existing entry.