One way to handle exceptions in Perl is with Syntax::Keyword::Try which adds try and catch keywords to Perl. As always there are trade offs and it depends on your goals and preferences.

It has no dependencies and plays nicely with perltidy and perlcritic. However, it is 2.6 times slower than eval() and it’s not included in the Perl core so you will need to install it.

Alternate
Solutions
There is more than one way to handle exceptions in Perl and several modules on MetaCPAN which add try/catch keywords to Perl. To find out which solution is best for you, see Whats the best way to handle exceptions?

Throwing exceptions

To throw an exception call die(). If the exception is not caught, an error message is displayed to STDERR and the process exits with a non-zero value. Syntax::Keyword::Try doesn’t change how exceptions are thrown.

die "Something bad happened";
# prints "Something bad happened at line 123." 
# and then the process exits;

Catching exceptions

use Syntax::Keyword::Try

try {
    die "its only a flesh wound";
}
catch {
    warn "something bad happened: $@";
}

The catch block can inspect the exception by looking at $@.

Exceptions as objects

Exceptions are usually strings, but you can throw objects too.

try {
    die My::Exception->new(
        error    => 'Something bad happened',
        request  => $request,
        response => $response,
    );
}
catch {
    warn "something bad happened: " . $@->error();
}

Pitfalls for Perl programmers

These are not really pitfalls. Syntax::Keyword::Try works the same as try/catch works in other languages. However if you are used to handling exceptions with die/eval or Try::Tiny here are a few things that you may have forgotten about try/catch:

1. A try block must be followed by catch or finally or both.

try { do_something_risk() } # syntax error

2. A try block does not catch exceptions. The catch block catches exceptions.

try     { die "its just a flesh wound" }
finally { say "hello" }
# no exception is caught
# prints "its just a flesh wound" to STDERR
# process exits

3. A return statement will exit the containing function – not the try block.

4. Loop control statements like redo, next, and last act on any containing loops.