# Errors

If you've ever programmed on a blockchain, you've probably been frustrated by either non existent or opaque error codes. Anchor attempts to address this by providing the #[error_code] attribute, which can be used to create typed Errors with descriptive messages that automatically propagate to the client.

# Defining a Program

For example,

use anchor_lang::prelude::*;
mod errors {
    use super::*;
    pub fn hello(_ctx: Context<Hello>) -> Result<()> {
pub struct Hello {}
pub enum ErrorCode {
    #[msg("This is an error message clients will automatically display")]

Observe the #[error_code] (opens new window) attribute on the ErrorCode enum. This macro generates internal anchor code that helps anchor turn the error code into an error and display it properly.

To create an error, use the error! (opens new window) macro together with an error code. This macro creates an AnchorError (opens new window) that includes helpful information like the file and line the error was created in.

To make writing errors even easier, anchor also provides the err! (opens new window) and the require! (opens new window) macros.

# Using the Client

When using the client, we get the error message.

try {
  const tx = await program.rpc.hello();
} catch (err) {
  const errMsg = "This is an error message clients will automatically display";
  assert.equal(err.toString(), errMsg);

It's that easy. 😃

To run the full example, go here (opens new window).