@pepeiborra - Here's a first pass. Functionality seems to be the same. I don't see any major differences that would break anything.

Is the intended operation for getParsedModule to fire multiple times? It looks like when a file contains CPP it will continuously run getParsedModule, until for whatever reason it will stop and finally create the command.

Here's a sample output for running unfold in the current file on qualify in Retrie.hs: https://gist.github.com/drsooch/f63e9a45e6c17a4229c7e85bd2b79746

The PARSED and other lines, are just me trying to trace the logic. In the gist, you can see my memory usage explodes to ~10GB.

I compared with current master branch, and it appears this continuous parsing is a by-product of the file itself. So it doesn't look like this is a new issues.

Is the intended operation for getParsedModule to fire multiple times? It looks like when a file contains CPP it will continuously run getParsedModule, until for whatever reason it will stop and finally create the command.

Here's a sample output for running unfold in the current file on qualify in Retrie.hs: https://gist.github.com/drsooch/f63e9a45e6c17a4229c7e85bd2b79746

The PARSED and other lines, are just me trying to trace the logic. In the gist, you can see my memory usage explodes to ~10GB.

I compared with current master branch, and it appears this continuous parsing is a by-product of the file itself. So it doesn't look like this is a new issues.

Yes, this is how retrie deals with CPP files. The memory usage is unfortunate, maybe there's room for improvement there

Looks like there's some files that snuck through to master without formatting. Those are cleaned up here.