12  Errors and Error Handling

12 Errors and Error Handling

Errors can roughly be divided into four different types:

When the compiler fails to compile the program, for example a syntax error.
When a program does not behave as intended, but does not crash. An example is that nothing happens when a button in a graphical user interface is clicked.
When a crash occurs. An example is when an operator is applied to arguments of the wrong type. The Erlang programming language has built-in features for handling of run-time errors. A run-time error can also be emulated by calling error(Reason). Run-time errors are exceptions of class error.
When the code itself calls exit/1 or throw/1. Generated errors are exceptions of class exit or throw.

When an exception occurs in Erlang, execution of the process that evaluated the erroneous expression is stopped. This is referred to as a failure, that execution or evaluation fails, or that the process fails, terminates, or exits. Notice that a process can terminate/exit for other reasons than a failure.

A process that terminates emits an exit signal with an exit reason that describes why the process terminated. Normally, some information about any erroneous termination is printed to the terminal. See Process Termination in the Processes chapter for more details on termination.

Exceptions are run-time errors or generated errors and are of three different classes, with different origins. The try expression can distinguish between the different classes, whereas the catch expression cannot. try and catch are described in Expressions.

Class Origin
error Run-time error, for example, 1+a, or the process called error/1,2
exit The process called exit/1
throw The process called throw/1

Table 12.1:   Exception Classes.

All of the above exceptions can also be generated by calling erlang:raise/3.

An exception consists of its class, an exit reason (see Exit Reason), and a stack trace (which aids in finding the code location of the exception).

The stack trace can be bound to a variable from within a try expression for any exception class, or as part of the exit reason when a run-time error is caught by a catch. Example:

> {'EXIT',{test,Stacktrace}} = (catch error(test)), Stacktrace.
[{shell,apply_fun,3,[]},
 {erl_eval,do_apply,6,[]},
 ...]
> try throw(test) catch Class:Reason:Stacktrace -> Stacktrace end.
[{shell,apply_fun,3,[]},
 {erl_eval,do_apply,6,[]},
 ...]

The stack back-trace (stacktrace) is a list contains {Module, Function, Arity, ExtraInfo} and/or {Fun, Arity, ExtraInfo} tuples. The field Arity in the tuple can be the argument list of that function call instead of an arity integer, depending on the exception.

ExtraInfo is a (possibly empty) list of two-element tuples in any order that provides additional information about the exception. The first element is an atom describing the type of information in the second element. The following items can occur:

The second element of the tuple is a map providing additional information about what caused the exception. This information can be created by calling error/3 and is used by erl_error:format_exception/4.
The second element of the tuple is a string (list of characters) representing the filename of the source file of the function.
The second element of the tuple is the line number (an integer > 0) in the source file where the exception occurred or the function was called.
Warning

Developers should rely on stacktrace entries only for debugging purposes.

The VM performs tail call optimization, which does not add new entries to the stacktrace, and also limits stacktraces to a certain depth. Furthermore, compiler options, optimizations and future changes may add or remove stacktrace entries, causing any code that expects the stacktrace to be in a certain order or contain specific items to fail.

The only exception to this rule is the class error with the reason undef which is guaranteed to include the Module, Function and Arity of the attempted function as the first stacktrace entry.

It is possible to prevent run-time errors and other exceptions from causing the process to terminate by using catch or try, see Expressions about catch and try.

Processes can monitor other processes and detect process terminations, see Processes.

When a run-time error occurs, that is an exception of class error. The exit reason is a tuple {Reason,Stack}, where Reason is a term indicating the type of error:

Reason Type of Error
badarg Bad argument. The argument is of wrong data type, or is otherwise badly formed.
badarith Bad argument in an arithmetic expression.
{badmatch,V} Evaluation of a match expression failed. The value V did not match.
function_clause No matching function clause is found when evaluating a function call.
{case_clause,V} No matching branch is found when evaluating a case expression. The value V did not match.
if_clause No true branch is found when evaluating an if expression.
{try_clause,V} No matching branch is found when evaluating the of-section of a try expression. The value V did not match.
undef The function cannot be found when evaluating a function call.
{badfun,F} Something is wrong with a fun F.
{badarity,F} A fun is applied to the wrong number of arguments. F describes the fun and the arguments.
timeout_value The timeout value in a receive..after expression is evaluated to something else than an integer or infinity.
noproc Trying to link or monitor to a non-existing process or port.
noconnection A link or monitor to a remote process was broken because a connection between the nodes could not be established or was severed.
{nocatch,V} Trying to evaluate a throw outside a catch. V is the thrown term.
system_limit A system limit has been reached. See Efficiency Guide for information about system limits.

Table 12.2:   Exit Reasons

Stack is the stack of function calls being evaluated when the error occurred, given as a list of tuples {Module,Name,Arity,ExtraInfo} with the most recent function call first. The most recent function call tuple can in some cases be {Module,Name,[Arg],ExtraInfo}.