Is there any efficient way to implement WeakMaps in Rust? Well, technically it's not WeakMap because Rust doesn't have GC and thus "weak" doesn't make much sense, but what I'm interested in is a way to store extra information on objects with O(1) lookup, where HashMap is not really the best solution.
Another way of looking into this would be "traits with data members".
To elaborate a little bit more - I'm going to store information associated with nodes in a tree retrieved from an external crate, so just adding data members is not really an option (without forking that crate just for this purpose); recreating entire tree but with wrappers around each node also seems too expensive (and requires lots of code to handle all kinds of nodes).
The only approach that seems to allow this so far is defining custom allocator which would reserve extra space on objects, but I'm not sure I want to go down that way and hope for a simpler solution.
let conn = Connection::open_in_memory().unwrap();
the doc tells me I can open like this: open<P: AsRef<Path>>(path: P) -> Result<Connection>
e.g.
FYI this comes from the rust compiler (
rustc) which is called by cargo. You can use #[allow(dead_code)] in the code. This is called an attribute https://doc.rust-lang.org/book/attributes.html.
warning: unused import, #[warn(unused_imports)] on by default
with this.. doesn't work
"#![warn(unused_imports)]"
"#![warn(unused_imports)]"
wow
"The compilation model centers on artifacts called crates. "
not yet
the parts I have read so far
any other standard docs? I've listened to some talks recently
ok, thx, will go through that
I think it does compile to LLVM which is then compiled to binary but I'm not the best person to ask
^ this is a nice place to get news
@benjyz if you are interested in the compilation process: https://blog.rust-lang.org/2016/04/19/MIR.html
MIR.. so much to learn ;)
finally a systems language which isn't C/++
I gotta say when I have to go back to C++ it hurts
