Skip to main content

Error Handling

This details how to handle errors in ambiorix.

When an error occurs server-side it should send the client a response with a status starting in 5 to indicates that was the case.


If you created your project using the ambiorix.generator or the ambiorix-cli, ambiorix defaults to using the following handler on error:

render_500 <- \(req, res, error) {
res$status <- 500L
"Internal server error"

Otherwise, no error handler is put in place and you have to create one.


One can specify the handler to use when an error occurs anywhere in the application.

The error handler should be a function that takes 3 arguments: req, res and the error object.

A basic example:

app$error <- \(req, res, error){
res$send("There was a server error :(")

You have absolute full control in what you choose to do with the error object. In most cases, you will need to log errors.

Here is a more involved and realistic example:

error_handler <- \(req, res, error = NULL) {
if (!is.null(error)) {
error_msg <- conditionMessage(error)
cli::cli_alert_danger("Error: {error_msg}")
response <- list(
code = 500L,
msg = "A server error occurred!"

There are 2 ways you can use the handler:

  1. Via the set_error() method:
    The advantage of this is that you can chain other operations.
  2. The old way, via the error field:
    app <- Ambiorix$new()
    app$error <- error_handler

Route Specific

Alternatively one can specify errors specific to certain routes, if these are not specified the global handler (above) is used.

app$get("/error", \(req, res){
}, \(req, res, error){
res$send("This is an error on /error", status = 500L)