EDIT: with #80 being merged it's possible to run semantic-rs but it does not push any changes and also does not create a release on GitHub.

I gave that issue some thought and my proposal here is to have a configuration file in place which denotes the remote to be used. The other option would be to ask the user but since the user is probably going to run semantic-rs multiple times they probably don't want to name the remote every time again.

Also asking the user doens't make a lot of sense when semantic-rs is run on a CI server.

So we could search for a .semantic-rs.toml in the project directory and extract the remote's name from there.

remote="production"

TODO for this one:

• Look for a .semantic-rs.toml config file
• Extract the remote value if present and use it
• Document it in README.md

UPDATE:

Updated to toml as suggested by @badboy.

Why json? It's hard to write for a human. TOML is pretty much the go-to format in the Rust world and far easier to write.