i sense that preserving formats is going to be an never ending game of whack-a-mole 😅

i do want to point out that this is only the case if the user doesn't have prettier installed, which is probably an uncommon occurrence. in any case, once #463 is merged, we can reuse that utility function to fix this as well

I think in the past I made sure that this worked as expected. We might need to try to remove this method for a second to see if this might already fix this.

i sense that preserving formats is going to be an never ending game of whack-a-mole 😅

I haven't looked at the code that closely, but was hoping if we got the functionality to live in the core functions that it would handle it for most cases

I think in the past I made sure that this worked as expected.

Yeah, I think that's the case with most of the stuff found in the past day or so. Understandable since there's been a lot of refactoring. I think with some tests hopefully we can avoid having to deal with too many regressions (#461)

I think in the past I made sure that this worked as expected. We might need to try to remove this method for a second to see if this might already fix this.

You're correct, it is the removal of the locations causing this! Rather than removing stripAst altogether, we can just strip a copy of the AST during comparisons, preserving the original tree