Rust interviews evaluate understanding of the language's unique ownership system, memory management without a garbage collector, and the ability to write safe concurrent code. This guide covers essential questions, from ownership fundamentals to advanced async and concurrency patterns. Interviewers value explanations that demonstrate understanding of Rust's memory safety guarantees. Explaining how the compiler prevents bugs at compile time makes the difference. ## Ownership and Borrowing ### Question 1: Explain Rust's ownership system Ownership is Rust's central concept that enables memory management without a garbage collector while guaranteeing memory safety at compile time. ```rust // ownership_basics.rs // The three fundamental rules of ownership fn main() { // Rule 1: Each value has a single owner let s1 = String::from("hello"); // s1 is the owner // Rule 2: Only one variable can own a value at a time let s2 = s1; // s1 is MOVED to s2 // println!("{}", s1); // ERROR: s1 is no longer valid println!("{}", s2); // OK: s2 is now the owner // Rule 3: When the owner goes out of scope, the value is dropped { let s3 = String::from("world"); // s3 is valid here } // s3 goes out of scope, memory is automatically freed // Copy types: simple types are copied, not moved let x = 5; let y = x; // x is COPIED, not moved println!("x = {}, y = {}", x, y); // Both are valid } // Move in action with functions fn take_ownership(s: String) { // s takes ownership of the String println!("{}", s); } // s is dropped here, memory freed fn makes_copy(i: i32) { // i is a copy of the argument println!("{}", i); } // i goes out of scope, nothing special (Copy type) fn ownership_with_functions() { let s = String::from("hello"); take_ownership(s); // s is moved into the function // println!("{}", s); // ERROR: s is no longer valid let x = 5; makes_copy(x); // x is copied println!("{}", x); // OK: x is still valid } ``` Ownership eliminates common memory bugs: use-after-free, double-free, and memory leaks. The compiler guarantees these properties at compile time. ### Question 2: What's the difference between immutable and mutable borrowing? Borrowing allows using a value without taking ownership, with strict rules to prevent data races. ```rust // borrowing_rules.rs // Immutable and mutable references fn main() { let mut s = String::from("hello"); // IMMUTABLE REFERENCES (&T) // Can coexist in unlimited numbers let r1 = &s; // immutable reference let r2 = &s; // another immutable reference println!("{} and {}", r1, r2); // OK // MUTABLE REFERENCE (&mut T) // Only one at a time, and no simultaneous immutable references let r3 = &mut s; // mutable reference // let r4 = &s; // ERROR: cannot have both immutable and mutable // let r5 = &mut s; // ERROR: only one mutable reference allowed r3.push_str(" world"); println!("{}", r3); // Reference scopes are limited to their last use let r6 = &s; // OK because r3 is no longer used println!("{}", r6); } // Practical example: modifying a struct struct User { name: String, age: u32, } impl User { // &self: read-only access fn get_name(&self) -> &str { &self.name } // &mut self: modification access fn set_name(&mut self, name: String) { self.name = name; } // self: takes ownership (consumes the instance) fn into_name(self) -> String { self.name // The User instance no longer exists after this } } fn borrowing_with_structs() { let mut user = User { name: String::from("Alice"), age: 30, }; // Reading println!("Name: {}", user.get_name()); // Modifying user.set_name(String::from("Bob")); // Consuming let name = user.into_name(); // user.age; // ERROR: user has been consumed } ``` These rules guarantee the absence of data races at compile time. No other language offers this guarantee without performance sacrifice. Since Rust 2018, the compiler uses NLL (Non-Lexical Lifetimes) to more precisely determine when a reference is no longer used, allowing more flexibility. ### Question 3: What are lifetimes and when should they be annotated? Lifetimes are annotations that tell the compiler how long references are valid, preventing dangling references. ```rust // lifetimes.rs // Understanding and annotating lifetimes // ERROR: dangling reference // fn dangling() -> &String { // let s = String::from("hello"); // &s // s is dropped at function end, reference invalid // } // The compiler often infers lifetimes automatically fn first_word(s: &str) -> &str { // Elided lifetime: compiler understands the return // has the same lifetime as the input match s.find(' ') { Some(i) => &s[..i], None => s, } } // Explicit annotation needed with multiple references fn longest<'a>(x: &'a str, y: &'a str) -> &'a str { // 'a means: the return lives at least as long // as the shorter of the two inputs if x.len() > y.len() { x } else { y } } fn lifetime_example() { let string1 = String::from("long string"); let result; { let string2 = String::from("xyz"); result = longest(&string1, &string2); println!("Longest: {}", result); // OK here } // println!("{}", result); // ERROR: string2 is dropped } // Lifetimes in structs struct ImportantExcerpt<'a> { part: &'a str, // Struct cannot outlive part } impl<'a> ImportantExcerpt<'a> { // Method returning a reference with the same lifetime fn level(&self) -> i32 { 3 } // Elided lifetime for &self returning a new reference fn announce_and_return_part(&self, announcement: &str) -> &str { println!("Attention: {}", announcement); self.part // Returns with lifetime 'a } } // Static lifetime: lives for the entire program duration fn static_lifetime() { let s: &'static str = "hello"; // Stored in the binary // Constants have implicit 'static lifetime const MAX_POINTS: u32 = 100_000; } // Combining lifetimes and generics fn longest_with_announcement<'a, T>( x: &'a str, y: &'a str, ann: T, ) -> &'a str where T: std::fmt::Display, { println!("Announcement: {}", ann); if x.len() > y.len() { x } else { y } } ``` Lifetimes are verified at compile time. If the code compiles, references are guaranteed valid. ## Traits and Generics ### Question 4: How do traits work in Rust? Traits define shared behavior between different types, similar to interfaces but with additional features. ```rust // traits_basics.rs // Defining and implementing traits // Trait definition trait Summary { // Required method (no body) fn summarize(&self) -> String; // Method with default implementation fn summarize_author(&self) -> String { String::from("(Anonymous)") } // Default method that calls a required method fn full_summary(&self) -> String { format!("By {} - {}", self.summarize_author(), self.summarize()) } } // Implementation for different types struct NewsArticle { headline: String, location: String, author: String, content: String, } impl Summary for NewsArticle { fn summarize(&self) -> String { format!("{}, by {} ({})", self.headline, self.author, self.location) } fn summarize_author(&self) -> String { format!("@{}", self.author) } } struct Tweet { username: String, content: String, reply: bool, retweet: bool, } impl Summary for Tweet { fn summarize(&self) -> String { format!("{}: {}", self.username, self.content) } } // Trait bounds: constraining generics fn notify(item: &T) { println!("Breaking news! {}", item.summarize()); } // Alternative syntax with where fn notify_verbose(item: &T) where T: Summary, { println!("Breaking news! {}", item.summarize()); } // Multiple trait bounds fn notify_complex(item: &T) { println!("{}", item); } // Return a type that implements a trait fn create_summarizable() -> impl Summary { Tweet { username: String::from("rust_lang"), content: String::from("Rust 2026 is amazing!"), reply: false, retweet: false, } } ``` Traits enable polymorphism without class inheritance, favoring composition over inheritance. ### Question 5: Explain the difference between static and dynamic genericity Rust offers two approaches for polymorphism: monomorphization (static) and trait objects (dynamic). ```rust // static_vs_dynamic_dispatch.rs // Static vs dynamic dispatch trait Animal { fn speak(&self) -> String; fn name(&self) -> &str; } struct Dog { name: String } struct Cat { name: String } impl Animal for Dog { fn speak(&self) -> String { String::from("Woof!") } fn name(&self) -> &str { &self.name } } impl Animal for Cat { fn speak(&self) -> String { String::from("Meow!") } fn name(&self) -> &str { &self.name } } // STATIC DISPATCH (monomorphization) // Compiler generates a version for each concrete type fn make_speak_static(animal: &T) { // At compile time, becomes make_speak_Dog and make_speak_Cat println!("{} says {}", animal.name(), animal.speak()); } // Advantages: inlining possible, no runtime overhead // Disadvantages: larger binary, type must be known at compile time // DYNAMIC DISPATCH (trait objects) // Uses a vtable to resolve methods at runtime fn make_speak_dynamic(animal: &dyn Animal) { // Resolved via a pointer table (vtable) at runtime println!("{} says {}", animal.name(), animal.speak()); } // Advantages: can store different types, smaller binary // Disadvantages: indirection overhead, no inlining fn main() { let dog = Dog { name: String::from("Rex") }; let cat = Cat { name: String::from("Whiskers") }; // Static: type is known at compile time make_speak_static(&dog); make_speak_static(&cat); // Dynamic: type is resolved at runtime make_speak_dynamic(&dog); make_speak_dynamic(&cat); // Heterogeneous collection (requires dynamic dispatch) let animals: Vec> = vec![ Box::new(Dog { name: String::from("Buddy") }), Box::new(Cat { name: String::from("Luna") }), ]; for animal in animals.iter() { println!("{} says {}", animal.name(), animal.speak()); } } // Object safety: not all traits can become trait objects trait ObjectSafe { fn method(&self); // No Self in return type // No generic parameters } // NOT object safe (cannot be dyn NotObjectSafe) trait NotObjectSafe { fn create() -> Self; // Self in return fn generic(&self, t: T); // Generic } ``` Static dispatch is preferable for performance. Dynamic dispatch is useful for heterogeneous collections and flexibility. ## Error Handling ### Question 6: How to handle errors with Result and Option? Rust has no exceptions. Error handling is done via `Result` and `Option` types with pattern matching. ```rust // error_handling.rs // Idiomatic error handling in Rust use std::fs::File; use std::io::{self, Read}; // Option: presence or absence of a value fn find_user(id: u32) -> Option { match id { 1 => Some(String::from("Alice")), 2 => Some(String::from("Bob")), _ => None, // No user found } } // Result: success or error fn divide(a: f64, b: f64) -> Result { if b == 0.0 { Err(String::from("Division by zero")) } else { Ok(a / b) } } fn option_combinators() { let user = find_user(1); // Pattern matching match user { Some(name) => println!("Found: {}", name), None => println!("Not found"), } // unwrap_or: default value let name = find_user(99).unwrap_or(String::from("Unknown")); // map: transform the value if present let upper = find_user(1).map(|n| n.to_uppercase()); // and_then (flatMap): chain Options let first_char = find_user(1).and_then(|n| n.chars().next()); // if let: simplified pattern matching if let Some(name) = find_user(2) { println!("User 2 is {}", name); } } fn result_handling() -> Result<(), Box> { // The ? operator propagates errors automatically let result = divide(10.0, 2.0)?; println!("Result: {}", result); // Equivalent to: // let result = match divide(10.0, 2.0) { // Ok(v) => v, // Err(e) => return Err(e.into()), // }; Ok(()) } // File reading with error propagation fn read_file_contents(path: &str) -> Result { let mut file = File::open(path)?; // Propagates error if failure let mut contents = String::new(); file.read_to_string(&mut contents)?; Ok(contents) } // Custom errors #[derive(Debug)] enum AppError { IoError(io::Error), ParseError(String), NotFound(String), } impl std::fmt::Display for AppError { fn fmt(&self, f: &mut std::fmt::Formatter) -> std::fmt::Result { match self { AppError::IoError(e) => write!(f, "IO error: {}", e), AppError::ParseError(s) => write!(f, "Parse error: {}", s), AppError::NotFound(s) => write!(f, "Not found: {}", s), } } } impl std::error::Error for AppError {} // Automatic conversion with From impl From for AppError { fn from(error: io::Error) -> Self { AppError::IoError(error) } } fn complex_operation() -> Result { let contents = std::fs::read_to_string("config.txt")?; // Auto-convert if contents.is_empty() { return Err(AppError::NotFound(String::from("Config is empty"))); } Ok(contents) } ``` The `?` operator makes code concise while forcing explicit error handling. No surprises at runtime. `unwrap()` and `expect()` panic if the value is None or Err. Reserve them for prototyping or cases where failure is impossible. In production, prefer propagation with `?` or combinators. ### Question 7: How to create custom errors with thiserror? The `thiserror` crate simplifies creating ergonomic custom errors. ```rust // custom_errors.rs // Custom errors with thiserror use thiserror::Error; // Error definition with derive macro #[derive(Error, Debug)] pub enum DataStoreError { #[error("connection failed: {0}")] ConnectionFailed(String), #[error("query failed: {query}")] QueryFailed { query: String, source: std::io::Error }, #[error("record not found: id={id}")] NotFound { id: u64 }, #[error("invalid data: {0}")] InvalidData(#[from] serde_json::Error), #[error(transparent)] // Delegates Display to source Other(#[from] anyhow::Error), } // Implementation with rich context pub struct DataStore { connection_string: String, } impl DataStore { pub fn connect(conn_str: &str) -> Result { if conn_str.is_empty() { return Err(DataStoreError::ConnectionFailed( "Empty connection string".into() )); } Ok(Self { connection_string: conn_str.to_string() }) } pub fn get_record(&self, id: u64) -> Result { // Query simulation if id == 0 { return Err(DataStoreError::NotFound { id }); } Ok(Record { id, data: format!("Record {}", id) }) } } pub struct Record { pub id: u64, pub data: String, } // Usage with anyhow for applications use anyhow::{Context, Result}; fn application_code() -> Result<()> { let store = DataStore::connect("postgres://localhost/db") .context("Failed to connect to database")?; let record = store.get_record(42) .context("Failed to fetch user record")?; println!("Got: {}", record.data); Ok(()) } // Pattern: converting errors with context fn read_config() -> Result { let contents = std::fs::read_to_string("config.toml") .context("Failed to read config file")?; let config: Config = toml::from_str(&contents) .context("Failed to parse config file")?; Ok(config) } #[derive(Debug)] struct Config { // ... } ``` `thiserror` is ideal for libraries (typed errors), while `anyhow` suits applications (maximum flexibility). ## Smart Pointers ### Question 8: Explain Box, Rc, Arc and RefCell Smart pointers manage heap memory and enable patterns that simple ownership doesn't allow. ```rust // smart_pointers.rs // Main smart pointers in Rust use std::rc::Rc; use std::sync::Arc; use std::cell::RefCell; // BOX: heap allocation // Used for: recursive types, large types, trait objects fn box_example() { // Simple heap allocation let b = Box::new(5); println!("b = {}", b); // Recursive type (impossible without Box) #[derive(Debug)] enum List { Cons(i32, Box), Nil, } let list = List::Cons(1, Box::new(List::Cons(2, Box::new(List::Cons(3, Box::new(List::Nil)))))); println!("{:?}", list); } // RC: Reference Counting (single-threaded) // Multiple owners for the same data fn rc_example() { let data = Rc::new(vec![1, 2, 3]); // Clone increments the reference counter let data_clone1 = Rc::clone(&data); // count = 2 let data_clone2 = Rc::clone(&data); // count = 3 println!("Reference count: {}", Rc::strong_count(&data)); // 3 // Each clone can read the data println!("data_clone1: {:?}", data_clone1); // Data is freed when the last Rc is dropped } // ARC: Atomic Reference Counting (thread-safe) // Like Rc but usable across threads fn arc_example() { use std::thread; let data = Arc::new(vec![1, 2, 3, 4, 5]); let mut handles = vec![]; for i in 0..3 { let data_clone = Arc::clone(&data); let handle = thread::spawn(move || { // Each thread has its own Arc println!("Thread {}: {:?}", i, data_clone); }); handles.push(handle); } for handle in handles { handle.join().unwrap(); } } // REFCELL: Interior Mutability // Allows mutation even with an immutable reference fn refcell_example() { let data = RefCell::new(5); // borrow() returns an immutable reference println!("Value: {}", *data.borrow()); // borrow_mut() returns a mutable reference *data.borrow_mut() += 1; println!("After mutation: {}", *data.borrow()); // Borrowing rules are checked at RUNTIME // Panics if rules are violated // let r1 = data.borrow(); // let r2 = data.borrow_mut(); // PANIC: already borrowed } // Common combination: Rc> // Multiple owners with possible mutation fn rc_refcell_example() { #[derive(Debug)] struct Node { value: i32, children: Vec>>, } let node1 = Rc::new(RefCell::new(Node { value: 1, children: vec![], })); let node2 = Rc::new(RefCell::new(Node { value: 2, children: vec![Rc::clone(&node1)], // node1 is child of node2 })); // Modify node1 from anywhere node1.borrow_mut().value = 10; println!("node2 child value: {}", node2.borrow().children[0].borrow().value); // 10 } // For threads: Arc> or Arc> fn arc_mutex_example() { use std::sync::Mutex; use std::thread; let counter = Arc::new(Mutex::new(0)); let mut handles = vec![]; for _ in 0..10 { let counter = Arc::clone(&counter); let handle = thread::spawn(move || { let mut num = counter.lock().unwrap(); *num += 1; }); handles.push(handle); } for handle in handles { handle.join().unwrap(); } println!("Final count: {}", *counter.lock().unwrap()); // 10 } ``` Choose the right smart pointer for the context: `Box` for simple heap, `Rc`/`Arc` for sharing, `RefCell`/`Mutex` for interior mutability. ## Concurrency ### Question 9: How does Rust guarantee thread safety? Rust's type system prevents data races at compile time via the `Send` and `Sync` traits. ```rust // thread_safety.rs // Concurrent safety guarantees use std::thread; use std::sync::{Arc, Mutex, mpsc}; // SEND: a type can be transferred to another thread // SYNC: a type can be shared between threads via references // Most types are Send and Sync automatically // Exceptions: Rc (not Send/Sync), RefCell (not Sync), raw pointers fn send_example() { let data = vec![1, 2, 3]; // Vec is Send, so it can be moved to another thread let handle = thread::spawn(move || { println!("Data in thread: {:?}", data); }); handle.join().unwrap(); } // The compiler prevents concurrency errors fn compile_time_safety() { // This would NOT compile: // let data = std::rc::Rc::new(5); // thread::spawn(move || { // println!("{}", data); // ERROR: Rc is not Send // }); // Solution: use Arc let data = Arc::new(5); let data_clone = Arc::clone(&data); thread::spawn(move || { println!("{}", data_clone); // OK: Arc is Send }); } // Mutex for thread-safe shared mutation fn mutex_pattern() { let counter = Arc::new(Mutex::new(0)); let mut handles = vec![]; for _ in 0..10 { let counter = Arc::clone(&counter); let handle = thread::spawn(move || { // lock() blocks until exclusive access is obtained let mut num = counter.lock().unwrap(); *num += 1; // MutexGuard is dropped here, releasing the lock }); handles.push(handle); } for handle in handles { handle.join().unwrap(); } println!("Result: {}", *counter.lock().unwrap()); } // RwLock for multiple reads / exclusive write fn rwlock_example() { use std::sync::RwLock; let data = Arc::new(RwLock::new(vec![1, 2, 3])); let mut handles = vec![]; // Multiple simultaneous readers for i in 0..3 { let data = Arc::clone(&data); handles.push(thread::spawn(move || { let read = data.read().unwrap(); println!("Reader {}: {:?}", i, *read); })); } // Only one writer at a time { let data = Arc::clone(&data); handles.push(thread::spawn(move || { let mut write = data.write().unwrap(); write.push(4); println!("Writer added 4"); })); } for handle in handles { handle.join().unwrap(); } } // Channels for inter-thread communication fn channel_example() { let (tx, rx) = mpsc::channel(); // Multi-producer, single-consumer // Clone the sender for multiple producers let tx1 = tx.clone(); thread::spawn(move || { tx1.send("from thread 1").unwrap(); }); thread::spawn(move || { tx.send("from thread 2").unwrap(); }); // Receive messages for received in rx { println!("Got: {}", received); } } ``` "Fearless concurrency": if the code compiles, there are no data races. The compiler is the first line of defense. If a thread panics while holding a Mutex, the Mutex is "poisoned". Subsequent calls to `lock()` return an error that can be recovered with `into_inner()`. ### Question 10: How does async/await work in Rust? Async in Rust is based on zero-cost Futures, without a runtime built into the language. ```rust // async_await.rs // Asynchronous programming in Rust use tokio::time::{sleep, Duration}; // async fn returns a Future that must be executed async fn fetch_data(url: &str) -> Result { // await suspends execution without blocking the thread let response = reqwest::get(url).await?; let body = response.text().await?; Ok(body) } // Futures are lazy: nothing executes without await or poll async fn lazy_example() { let future = async { println!("This won't print yet"); }; // Nothing happened future.await; // Now it executes } // Parallel execution of futures async fn parallel_execution() { // join! executes multiple futures in parallel let (result1, result2) = tokio::join!( fetch_data("https://api.example.com/1"), fetch_data("https://api.example.com/2"), ); println!("Results: {:?}, {:?}", result1, result2); } // select! for the first completed future async fn race_example() { tokio::select! { result = fetch_data("https://api1.example.com") => { println!("API 1 responded first: {:?}", result); } result = fetch_data("https://api2.example.com") => { println!("API 2 responded first: {:?}", result); } _ = sleep(Duration::from_secs(5)) => { println!("Timeout!"); } } } // Streams: asynchronous iterators use tokio_stream::StreamExt; async fn stream_example() { let mut stream = tokio_stream::iter(vec![1, 2, 3, 4, 5]); while let Some(value) = stream.next().await { println!("Got: {}", value); } } // Spawn for background tasks async fn spawn_tasks() { let handle = tokio::spawn(async { sleep(Duration::from_secs(1)).await; "Task completed" }); println!("Task spawned, doing other work..."); let result = handle.await.unwrap(); println!("Result: {}", result); } // Entry point with tokio #[tokio::main] async fn main() { // The tokio runtime executes futures parallel_execution().await; } // Alternative: multi-threaded or single-threaded runtime #[tokio::main(flavor = "current_thread")] async fn main_single_thread() { // Everything runs on a single thread } #[tokio::main(flavor = "multi_thread", worker_threads = 4)] async fn main_multi_thread() { // Pool of 4 worker threads } ``` Rust async is "bring your own runtime": tokio, async-std, or smol. This flexibility allows use-case-specific optimizations. ## Advanced Patterns ### Question 11: Explain the Builder pattern in Rust The Builder pattern is idiomatic in Rust for constructing complex structures with many optional fields. ```rust // builder_pattern.rs // Idiomatic Builder pattern in Rust #[derive(Debug, Clone)] pub struct Server { host: String, port: u16, max_connections: usize, timeout_seconds: u64, tls_enabled: bool, tls_cert_path: Option, } // Builder with consuming approach (ownership) #[derive(Default)] pub struct ServerBuilder { host: String, port: u16, max_connections: usize, timeout_seconds: u64, tls_enabled: bool, tls_cert_path: Option, } impl ServerBuilder { pub fn new() -> Self { Self { host: String::from("localhost"), port: 8080, max_connections: 100, timeout_seconds: 30, tls_enabled: false, tls_cert_path: None, } } // Each method takes self and returns Self for chaining pub fn host(mut self, host: impl Into) -> Self { self.host = host.into(); self } pub fn port(mut self, port: u16) -> Self { self.port = port; self } pub fn max_connections(mut self, max: usize) -> Self { self.max_connections = max; self } pub fn timeout(mut self, seconds: u64) -> Self { self.timeout_seconds = seconds; self } pub fn enable_tls(mut self, cert_path: impl Into) -> Self { self.tls_enabled = true; self.tls_cert_path = Some(cert_path.into()); self } // build() consumes the builder and creates the final structure pub fn build(self) -> Result { if self.tls_enabled && self.tls_cert_path.is_none() { return Err("TLS enabled but no certificate path provided".into()); } Ok(Server { host: self.host, port: self.port, max_connections: self.max_connections, timeout_seconds: self.timeout_seconds, tls_enabled: self.tls_enabled, tls_cert_path: self.tls_cert_path, }) } } // Fluent usage fn create_server() -> Result { ServerBuilder::new() .host("0.0.0.0") .port(443) .max_connections(1000) .timeout(60) .enable_tls("/etc/ssl/cert.pem") .build() } // Alternative with derive macro (typed-builder crate) // #[derive(TypedBuilder)] // pub struct Config { // #[builder(default = "localhost".to_string())] // host: String, // #[builder(default = 8080)] // port: u16, // } // Pattern with type-level validation (typestate pattern) pub struct Unvalidated; pub struct Validated; pub struct Request { url: String, method: String, headers: Vec<(String, String)>, _state: std::marker::PhantomData, } impl Request { pub fn new(url: &str) -> Self { Self { url: url.to_string(), method: "GET".to_string(), headers: vec![], _state: std::marker::PhantomData, } } pub fn method(mut self, method: &str) -> Self { self.method = method.to_string(); self } // validate() changes the state type pub fn validate(self) -> Result, String> { if self.url.is_empty() { return Err("URL cannot be empty".into()); } Ok(Request { url: self.url, method: self.method, headers: self.headers, _state: std::marker::PhantomData, }) } } impl Request { // send() is only available on validated requests pub async fn send(self) -> Result { // Implementation... todo!() } } struct Response; ``` The typestate pattern guarantees at compile time that certain operations can only be called in the correct state. ### Question 12: How to implement a trait for external types? The "orphan rule" prevents implementing an external trait for an external type, but solutions exist. ```rust // newtype_pattern.rs // The Newtype pattern to work around the orphan rule use std::fmt; // ORPHAN RULE: cannot implement Display (std) for Vec (std) // impl fmt::Display for Vec { ... } // ERROR // SOLUTION 1: Newtype wrapper struct Wrapper(Vec); impl fmt::Display for Wrapper { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "[{}]", self.0.join(", ")) } } fn newtype_example() { let w = Wrapper(vec![ String::from("hello"), String::from("world"), ]); println!("{}", w); // [hello, world] } // Transparent access with Deref use std::ops::Deref; impl Deref for Wrapper { type Target = Vec; fn deref(&self) -> &Self::Target { &self.0 } } fn deref_example() { let w = Wrapper(vec![String::from("test")]); println!("Length: {}", w.len()); // Calls Vec::len via Deref } // SOLUTION 2: Extension trait (to add methods) trait VecExt { fn first_or_default(&self) -> Option<&T>; } impl VecExt for Vec { fn first_or_default(&self) -> Option<&T> { self.first() } } fn extension_trait_example() { let v = vec![1, 2, 3]; println!("First: {:?}", v.first_or_default()); } // Newtype with domain semantics #[derive(Debug, Clone, PartialEq, Eq, Hash)] struct Email(String); impl Email { pub fn new(email: &str) -> Result { if email.contains('@') && email.contains('.') { Ok(Self(email.to_string())) } else { Err("Invalid email format") } } pub fn as_str(&self) -> &str { &self.0 } } impl fmt::Display for Email { fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { write!(f, "{}", self.0) } } #[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord)] struct UserId(u64); impl UserId { pub fn new(id: u64) -> Self { Self(id) } } // Newtypes add type safety without runtime overhead fn process_user(id: UserId, email: Email) { println!("Processing user {} with email {}", id.0, email); } fn type_safety_example() { let id = UserId::new(42); let email = Email::new("user@example.com").unwrap(); process_user(id, email); // This would not compile: // process_user(email, id); // Types reversed // process_user(UserId::new(42), "string"); // String instead of Email } ``` Newtypes have zero runtime cost thanks to the identical memory representation guarantee. ### Question 13: How to use procedural macros? Procedural macros allow generating code at compile time, like custom derives. ```rust // procedural_macros.rs // Understanding procedural macros // Proc macros are defined in a separate crate with proc-macro = true // Crate: my_derive (Cargo.toml: proc-macro = true) use proc_macro::TokenStream; use quote::quote; use syn::{parse_macro_input, DeriveInput}; // DERIVE MACRO: #[derive(MyTrait)] #[proc_macro_derive(MyDebug)] pub fn my_debug_derive(input: TokenStream) -> TokenStream { // Parse input as a type definition let input = parse_macro_input!(input as DeriveInput); let name = input.ident; // Generate implementation code let expanded = quote! { impl std::fmt::Debug for #name { fn fmt(&self, f: &mut std::fmt::Formatter) -> std::fmt::Result { write!(f, stringify!(#name)) } } }; TokenStream::from(expanded) } // ATTRIBUTE MACRO: #[my_attribute] #[proc_macro_attribute] pub fn route(attr: TokenStream, item: TokenStream) -> TokenStream { // attr contains the attribute arguments // item contains the annotated element (function, struct, etc.) let method_path = attr.to_string(); // "GET, /users" let input = parse_macro_input!(item as syn::ItemFn); let fn_name = &input.sig.ident; let expanded = quote! { #input // Additionally generated code fn register_#fn_name() { println!("Registered route: {}", #method_path); } }; TokenStream::from(expanded) } // FUNCTION-LIKE MACRO: my_macro!(...) #[proc_macro] pub fn make_answer(_input: TokenStream) -> TokenStream { "fn answer() -> u32 { 42 }".parse().unwrap() } // --- Usage in client code --- // Derive macro #[derive(MyDebug)] struct Point { x: i32, y: i32, } // Attribute macro #[route("GET", "/users")] fn get_users() -> Vec { vec![] } // Function-like macro make_answer!(); // Generates fn answer() -> u32 { 42 } fn main() { let p = Point { x: 1, y: 2 }; println!("{:?}", p); // Uses our MyDebug println!("Answer: {}", answer()); // 42 } struct User; ``` Procedural macros are powerful for repetitive code: serialization, web routing, validation, etc. ## Memory Safety and Unsafe ### Question 14: When and how to use unsafe? The `unsafe` block allows bypassing certain compiler checks for low-level code. ```rust // unsafe_rust.rs // Understanding unsafe and its guarantees // The 5 superpowers of unsafe: // 1. Dereference raw pointers // 2. Call unsafe functions // 3. Access/modify mutable static variables // 4. Implement unsafe traits // 5. Access union fields // RAW POINTERS fn raw_pointers() { let mut num = 5; // Creating raw pointers is safe let r1 = &num as *const i32; let r2 = &mut num as *mut i32; // Dereferencing requires unsafe unsafe { println!("r1 is: {}", *r1); *r2 = 10; println!("r2 is: {}", *r2); } } // UNSAFE FUNCTION // The function guarantees safety IF preconditions are met unsafe fn dangerous() { // Code that assumes the caller verified invariants } fn call_dangerous() { // Must be in an unsafe block unsafe { dangerous(); } } // SAFE ABSTRACTION over unsafe code fn split_at_mut(values: &mut [i32], mid: usize) -> (&mut [i32], &mut [i32]) { let len = values.len(); let ptr = values.as_mut_ptr(); assert!(mid <= len); // Runtime check unsafe { // We know the two slices don't overlap ( std::slice::from_raw_parts_mut(ptr, mid), std::slice::from_raw_parts_mut(ptr.add(mid), len - mid), ) } } // FFI: calling C code extern "C" { fn abs(input: i32) -> i32; } fn call_c_function() { unsafe { println!("Absolute value: {}", abs(-3)); } } // Export a function for C #[no_mangle] pub extern "C" fn call_from_c() { println!("Called from C!"); } // MUTABLE STATIC static mut COUNTER: u32 = 0; fn increment_counter() { unsafe { COUNTER += 1; println!("COUNTER: {}", COUNTER); } } // UNSAFE TRAIT unsafe trait Dangerous { // Implementers guarantee invariants } unsafe impl Dangerous for i32 { // Implementer asserts respecting the trait's invariants } // Practical example: structure with internal pointer pub struct MyVec { ptr: *mut T, len: usize, capacity: usize, } impl MyVec { pub fn new() -> Self { Self { ptr: std::ptr::null_mut(), len: 0, capacity: 0, } } pub fn push(&mut self, value: T) { if self.len == self.capacity { self.grow(); } unsafe { std::ptr::write(self.ptr.add(self.len), value); } self.len += 1; } fn grow(&mut self) { // Unsafe allocation/reallocation... } } impl Drop for MyVec { fn drop(&mut self) { unsafe { // Properly free memory for i in 0..self.len { std::ptr::drop_in_place(self.ptr.add(i)); } if self.capacity > 0 { let layout = std::alloc::Layout::array::(self.capacity).unwrap(); std::alloc::dealloc(self.ptr as *mut u8, layout); } } } } ``` Minimize the unsafe code surface. Encapsulate unsafe code in safe abstractions that guarantee invariants. Unsafe code must never corrupt surrounding safe memory. ### Question 15: How does the borrow checker work? The borrow checker is the heart of the Rust compiler that verifies ownership and borrowing rules. ```rust // borrow_checker.rs // Understanding how the borrow checker works fn borrow_checker_basics() { let mut v = vec![1, 2, 3]; // RULE 1: Either multiple immutable references or one mutable let r1 = &v; let r2 = &v; println!("{:?} {:?}", r1, r2); // OK: multiple immutable references // From here, r1 and r2 are no longer used (NLL) let r3 = &mut v; // OK thanks to Non-Lexical Lifetimes r3.push(4); } // The borrow checker tracks lifetimes fn lifetime_tracking() { let mut data = String::from("hello"); let slice = &data[..]; // Immutable borrow starts // data.push_str(" world"); // ERROR: cannot mutate during borrow println!("{}", slice); // Last use of slice data.push_str(" world"); // OK: borrow ended } // Common problems and solutions mod common_patterns { // Problem: borrowing two mutable fields struct Data { field1: Vec, field2: Vec, } fn problem(data: &mut Data) { // This sometimes doesn't compile directly: // let f1 = &mut data.field1; // let f2 = &mut data.field2; // Solution: destructuring let Data { field1, field2 } = data; field1.push(1); field2.push(2); } // Problem: iterate and modify fn iterate_and_modify() { let mut v = vec![1, 2, 3, 4, 5]; // Does not compile: // for &x in &v { // if x % 2 == 0 { // v.push(x * 2); // ERROR: borrowed by iterator // } // } // Solution 1: collect indices first let to_add: Vec = v.iter() .filter(|&&x| x % 2 == 0) .map(|&x| x * 2) .collect(); v.extend(to_add); // Solution 2: use explicit indices let len = v.len(); for i in 0..len { if v[i] % 2 == 0 { let new_val = v[i] * 2; v.push(new_val); } } } // Problem: self-referential struct // struct SelfRef { // data: String, // slice: &str, // Reference to data - IMPOSSIBLE // } // Solution: use indices or crates like ouroboros struct SafeSelfRef { data: String, slice_start: usize, slice_end: usize, } impl SafeSelfRef { fn get_slice(&self) -> &str { &self.data[self.slice_start..self.slice_end] } } } // Patterns to work around limitations mod workarounds { use std::cell::RefCell; // Interior mutability when borrow checker is too restrictive struct Graph { nodes: RefCell>, } struct Node { value: i32, } impl Graph { fn add_node(&self, value: i32) { // Mutation possible despite &self self.nodes.borrow_mut().push(Node { value }); } fn get_node(&self, index: usize) -> Option { self.nodes.borrow().get(index).map(|n| n.value) } } } ``` The borrow checker may seem restrictive at first, but these constraints eliminate entire categories of bugs present in other languages. ## Conclusion Rust interviews evaluate deep understanding of the ownership system, memory safety guarantees, and the ability to write concurrent code without data races. Mastering these concepts distinguishes developers who can leverage Rust's unique advantages. ### Preparation Checklist - ✅ Understand ownership, borrowing and the three fundamental rules - ✅ Know when and how to annotate lifetimes - ✅ Master traits and the difference between static and dynamic dispatch - ✅ Handle errors idiomatically with Result and Option - ✅ Choose the right smart pointer for the context - ✅ Write concurrent code with Arc, Mutex and channels - ✅ Understand async/await and runtimes like tokio - ✅ Know when and how to use unsafe safely Preparing for Rust interviews requires practice with the language's unique ownership concepts. Exercises on Exercism, personal projects, and contributing to the Rust ecosystem consolidate this knowledge for the most demanding technical interviews.