Add str literal mutation

[samgoldman]

If you're open to new mutations, I created a mutator for string literals. It replaces empty strings with a single character string and for non-empty strings it does a few things: append and prepend a single character as well as replacing the string with an empty string. I'd welcome any feedback. Thanks!

[samuelpilz]

This seems like a really good implementation of a new mutation, thank you. Documentation and tests are well done.
The only thing that concerns me is the owned String as return value from the run function. Generally, changing the type of an expression is tricky, it messes with type-inference and could result in different trait resolutions.

One potential issue in particular: Does this implementation break code like this (because the owned String is only a temp variable)?

fn a() -> usize {let x = ""; x.len()}

I suggest using a return value of &'static str. This requires baking the possible mutation literals into the source code.

[samgoldman]

Thanks for the feedback, that makes sense. I made the update, although I had to remove the appending and prepending mutations. I also added a test case for the example you provided and it doesn't break it (and it didn't break before changing to &'static str either).

[samuelpilz]

Indeed, for some reason {let x = &String::new(); x.len()} is valid Rust now (for some reason, the lifetime of the tempvar gets extended). I thought this was a minimal example of lifetime extension. Probably a better example is:

fn a() -> &'static str {""}

This code should not break as well (which is now fixed).

I really liked the string-dependent mutations. My idea is changing the signature of run roughly to:

fn run(
    mutator_id: usize,
    original_lit: &'static str,
    mutations: &[&'statc str],
    runtime: &impl Deref<Target = MutagenRuntimeConfig>,
)  -> &'static str {...}

This allows you to create the mutated strings at compile-time and have them available at run-time as &'static str

[samgoldman]

Yeah, the new example would have broken the original version. I updated to the signature you suggested and that example works (and I've added it as a test case).

[samuelpilz]

Great. One last comment: could you revert the documentation to your original proposal?
Otherwise, this is ready for merge (although I don't have the rights)

[samgoldman]

I forgot to leave a comment but that should be done.

[llogiq]

Thank you! I didn't have the time to look at this until now, but Samuel has already given all the feedback needed.

Thanks to both of you!