intuition check: what would you expect this Rust snippet to print?
```
use itertools::Itertools;
fn main() {
let p = std::path::Path::new("/a/b/c");
let s = p.components().map(|c| c.as_os_str().to_string_lossy()).join("/");
println!("s={s:?}");
}
```
@yossarian I don't think a "component" of a path is a precisely defined thing, but my initial intuition would be "a/b/c"
@migratory oops, that was supposed to be one of the options...fixed.
Jun 30, 2026, 8:51 PM@yossarian ugh, that's pretty surprising. Fun also that the documentation tries to cover some attempted normalization... but doesn't talk about this aspect at all.
@djc yeah, it's IMO an unfortunate wart in the API. path APIs are hard!
(the other one I always get stung on is `Path::new("foo").parent()` being `Some("")` instead of `None`)
@yossarian @djc hard blorg for both of these disturbing factoids
@yossarian I knew this one! I worked with that specific part of the standard library before, it makes more sense when looking at the windows implementation of this. The first / is considered the equivalent of C:, so it's yielded as it's own entity.
@kpcyrd yeah, I think's actually a great example of how *good* the stdlib's modeling is!
(arguably the user code here is just 100% incorrect and the stdlib doesn't even make it possible, it's just that itertools adds that `join` API)