That's an awesome explanation, thanks!

Thanks, good to know!

Who does this keep double whitespaces?? You're just joining the reversed words by a single whitespace.

it doesn't matter if he use sizeof(char), as this would only consume compile time and it would become a constant.

Fixed

This mutates the original array, which was not forbidden by the description, but is probably why this rankes more on the "clever" than on the "best practice" side.
It is really pretty clever though, I have to admit.

Nice solution. Some nitpicks:

1. I see this a lot here, but sizeof(char) is by definition always 1, so it's kinda redundant to write that instead of just 1.
2. Also for best practice, you should always check the return value of malloc, calloc and friends and null terminate strings.
3. And you could realloc(final, temp - final) in the end so you don't return a buffer that is longer than it needs to be.

In that case it should be changed in the test cases. Either way, the function name in the solution stub should not have to be changed for the tests to pass. Currently you get an error (error[E0425]: unresolved name 'DNA_strand').

https://doc.rust-lang.org/1.0.0/style/style/naming/README.html

In CamelCase, acronyms count as one word: use Uuid rather than UUID. In snake_case, acronyms are lower-cased: is_xid_start.

The Rust code misspells the function name (should be DNA_strand instead of dna_strand).