Implementing a custom logging framework in Rust with `log` and `fern` results in panics
I'm trying to create a custom logging framework in Rust using the `log` and `fern` crates, but I'm working with panics during runtime when I try to set the logging level or when logging messages... I have the following code: ```rust use fern::Dispatch; use log::{info, LevelFilter}; fn setup_logging() -> Result<(), fern::InitError> { Dispatch::new() .format(|out, message, _| { out.finish(format_args!("{}: {}", chrono::Local::now().format("%Y-%m-%d %H:%M:%S"), message)); }) .level(LevelFilter::Debug) .chain(std::io::stdout()) .apply()? Ok(()) } fn main() { setup_logging().expect("Failed to set up logging"); info!("This is an info message"); } ``` When I run this, I get the following panic: ``` panicked at 'called `Option::unwrap()` on a `None` value', src/main.rs:7:9 ``` I've ensured that the `log` and `fern` versions in my `Cargo.toml` are set correctly: ```toml [dependencies] log = "0.4" fern = "0.6" chrono = "0.4" ``` The panic seems to indicate that something isn't being initialized properly, but I need to pinpoint the scenario. I've tried rearranging the order of operations and checking for `None` cases, but without success. What could be causing this panic in my logging setup, and what are the best practices for ensuring robust logging in Rust applications? My team is using Rust for this REST API. Any ideas what could be causing this? My development environment is CentOS. Hoping someone can shed some light on this.