Documentation on running a translator is unclear about dream2nix HOT 6 CLOSED

Ah, I see. Then I think an easier way might be to implement a discoverer for Scala, and then just use dream2nix on a test repo (the python discoverer is simple, so you could just modify that.). You could also do the second way I mentioned, but instead of .translate, you could build the .translateBin derivation, and use the resulting script, passing a file path containing a JSON object with source and outputFile attributes, provided your translator uses those.

from dream2nix.

okay after some initial investigation I found this commit:
489f1ab

Which seems to have removed the cli for whatever reason. But there is no explanation in the commit. If the process has changed it would be very nice to update the docs regardless.

from dream2nix.

If you are trying to test a translator, you can add sources to translate like this to your translator.

Otherwise you could also start a nix repl, :lf <path-to-dream2nix> and then you can access your translator using translator = lib.${builtins.currentSystem}.translators.translators.scala.all.<translator-name> then using the translator.translate function. You can pass an attribute set with a source specified (and a project).

This is all assuming it's a pure translator though (which seems like what you are doing).

from dream2nix.

Thank you, but unfortunately it is not a pure translator, I need to call SBT so it can fetch the Scala dependencies once, in order to be able to build the lock file.

from dream2nix.

In the mean time the contributors docs have been revamped and there is no mention of the cli app anymore.

Anyhow, the docs on writing an impure translator can still be improved, I guess.

from dream2nix.

closing in favor of: #181

from dream2nix.