Undefined values: Difference between revisions

=={{header|ActionScript}}==

ActionScript has a special '''undefined''' value which applies to untyped variables and properties of dynamic classes which have not been initialized.

var bar:*; // explicitly untyped

if (foo == undefined)

ActionScript also has a '''null''' value: see [[Null object#ActionScript]].

This pragma is required to be applied to the whole partition, which would require recompilation of the run-time library. For this reason, the presented example uses another pragma Initialize_Scalars.

This one has the effect similar to Normalize_Scalars, but is [[GNAT]]-specific:

pragma Initialize_Scalars;

with Ada.Text_IO; use Ada.Text_IO;

end if;

end Invalid_Value;

Sample output:

<pre>

Note: Some implementations (eg [[ALGOL 68C]]) also have a procedure named ''undefined'' that is called to indicated that the behaviour at a particular point in a program is unexpected, undefined, or non-standard.

R r := NIL;

(VOID):print(("u is EMPTY", new line))

OUT print(("u isnt EMPTY", new line))

Output:

<pre>

=={{header|Arturo}}==

{{out}}

=={{header|BBC BASIC}}==

A scalar variable (numeric or string) cannot have an 'undefined' value; if an attempt is made to read a variable which has never been defined a 'No such variable' error results. By trapping errors this condition can be detected:

ON ERROR LOCAL IF ERR<>26 REPORT : END ELSE ok% = FALSE

IF ok% THEN

PRINT "Not defined"

ENDIF

{{works with|BBC BASIC for Windows}}

Arrays and structures however '''can''' have an undefined state; for example after having been declared as LOCAL or PRIVATE but before being defined using DIM. This condition can be detected and manipulated:

END

!^array() = 0 : REM Set array to undefined state

=={{header|C}}==

C programs can also read garbage values from uninitialized memory. There is no way to test for these garbage values, because they might equal anything.

#include <stdlib.h>

printf("*junkp: %d\n", *junkp);

return 0;

=={{header|C sharp|C#}}==

In C# it's important to see the difference between reference and value types. For reference types (class instances) there is <code>null</code> as a general undefined reference.

Dereferencing a null reference will throw a <code>NullReferenceException</code>.

This can't be used normally for value types (<code>int</code>, <code>double</code>, <code>DateTime</code>, etc.) since they are no ''references,'' so the following is a compiler error:

With .NET 2.0 there is an additional <code>Nullable<T></code> structure which enables those semantics for value types as well:

if (answer == null) {

answer = 42;

There is a bit syntactic sugar involved here. The <code>?</code> after the type name signals the compiler that it's a ''nullable'' type. This only works for value types since reference types are nullable due to their very nature.

But since value types still can't actually ''have'' a <code>null</code> value this gets converted into the following code by the compiler:

if (!answer.HasValue) {

answer = new Nullable<int>(42);

So it's a little compiler magic but in the end works just as one would expect.

the behavior is undefined, the compiler can do whatever it wants. Most compilers will give a warning.

int main()

std::cout << "not 42";

}

{{out}}

<pre>

In Lisp, there are three possibilities with regard to a dynamic variable. It may exist, and have a value. It may exist, but not have a value. Or it may not exist at all. These situations can be probed with certain functions that take a symbol as an argument:

;; assumption: none of these variables initially exist

(boundp '*y*) -> T

Furthermore, a variable which is bound can be made unbound:

(makunbound '*y*) ;; *y* no longer has a value; it is erroneous to evaluate *y*

By contrast, lexical variables never lack a binding. Without an initializer, they are initialized to nil.

The same goes for local re-binding of special variables:

(defvar *dyn*) ;; special, no binding

(list (boundp '*dyn*) *dyn* (boundp 'lex) lex)) -> (T NIL NIL NIL)

Here we can see that inside the scope of the let, the special variable has a binding (to the value <code>NIL</code>) and so <code>(boundp '*dyn*)</code> yields <code>T</code>. But <code>boundp</code> does not "see" the lexical variable <code>lex</code>; it reports that <code>lex</code> is unbound. Local binding constructs never leave a variable without a value, be it dynamic or lexical: both <code>*dyn*</code> and <code>lex</code> evaluate to NIL, but using very different mechanisms.

=={{header|D}}==

In D variables are initialized either with an explicit Initializer or are set to the default value for the type of the variable. If the Initializer is void, however, the variable is not initialized. If its value is used before it is set, undefined program behavior will result. "void" initializers can be used to avoid the overhead of default initialization in performance critical code.

// Initialized:

int a = 5;

double[] bbb = void;

int[3] eee = void;

=={{header|Delphi}}==

For Referenced data types like pointers, classes and interfaces a reference can be explicitely set to undefined by assigning the NIL value for it. No memory management like garbage collection (except for interfaces) is done.

P: PInteger;

begin

Dispose(P); //Release memory allocated by New

end;

If P was a Class only the Assigned function would be available; in addition Dispose would have to be replaced by FreeAndNil or calling the .Free method of the instance. For Interfaces no such last call would be necessary as simple removal of the reference would be sufficient to trigger the Garbage Collector.

=={{header|Déjà Vu}}==

There is no undefined value in Déjà Vu. Instead, trying to access an undefined variable raises an exception.

bogus

catch name-error:

!print "There is *no* :bogus in the current context"

return

If you need to declare a local variable, but don't have a value for it yet, there is the standard function <code>undef</code>, which raises an exception when called but is an actual value that can be passed around and assigned to names.

Each variable's behavior is defined by a <em>slot</em> object with <code>get</code> and <code>put</code> methods. If the slot throws when invoked, then a program may contain a reference to that variable, but not actually access it. A notable way for the slot to throw is for the slot to itself be a broken reference. This may occur when slots are being passed around as part of metaprogramming or export/import; it is also used in certain control structures. For example:

...

The slot for <var>baz</var> is broken if the left side of the <code>||</code> was true and the right side was therefore not evaluated.

=={{header|Erlang}}==

In Erlang a variable is created by assigning to it, so all variables have a value. To get undefined values you have to use a record. The default value of a record member is undefined.

-module( undefined_values ).

io:fwrite( "Record member_1 ~p, member_2 ~p~n", [Record#a_record.member_1, Record#a_record.member_2] ),

io:fwrite( "Member_2 is undefined ~p~n", [Record#a_record.member_2 =:= undefined] ).

{{out}}

<pre>

=={{header|Factor}}==

Factor does not have undefined values. Most values used in Factor live on the data stack. They are not named, so it is not possible for them to be undefined:

Tuple slots are always initialized with <code>f</code> or other values like <code>0</code> when the slot has been class-restricted:

foo new bar>> . ! f

TUPLE: my-tuple { n integer } ;

Dynamic variables (and indeed all words) are initialized to <code>f</code>:

n get . ! f

Finally, Factor does not allow lexical variables to be declared without initialization:

2 :> n ! There is no other way!

0 1 :> ( a b )

=={{header|Fortran}}==

More modern Fortrans recognise the NaN state of floating-point variables on computers whose floating-point arithmetic follows the IEEE standard, via the logical function

This is the only safe way to detect them as tests to detect the special behaviour of NaN states such as ''if x = x then...else aha!;'' might be optimised away by the compiler, and other tests may behave oddly. For instance x ¬= 0 might be compiled as ¬(x = 0) and the special NaN behaviour will not be as expected. Such NaN values can come via READ statements, because "NaN" is a recognised numerical input option, and could be used instead of "999" in a F3.0 data field, etc. Or, your system's input processing might recognise "?" or other special indicators and so on.

=={{header|FreeBASIC}}==

In FreeBASIC all variables are given a default value (zero for numbers, false for boolean and empty for strings) when they are declared unless they are assigned a different value at that time or are specifically left uninitialized (using the 'Any' keyword). If the latter are used before they have been initialized, then they will contain a 'garbage' value i.e. whatever value happens to be in the associated memory:

Dim i As Integer '' initialized to 0 by default

Print i, j, k

{{out}}

=={{header|GAP}}==

# true

IsBound(a);

=={{header|Go}}==

# Successful (non panicking) use of initialized objects.

# One more quirky little feature involving a type switch on a nil interface.

import "fmt"

close(c)

fmt.Println("channel closed")

Output:

<pre>

In Haskell, there is a semantic concept called [http://www.haskell.org/haskellwiki/Bottom "bottom"], which is a computation that never terminates or runs into an error. So <code>undefined</code> is not a proper value at all; it is a bottom that causes an exception when evaluated. For example,

-- When run in GHC:

This isn't quite as dangerous as it sounds because of Haskell's laziness. For example, this program:

prints <code>3</code>, since <code>length</code> doesn't need to evaluate any of the elements of its input.

In practice, one uses <code>undefined</code> less often than <code>error</code>, which behaves exactly the same except that it lets you choose the error message. So if you say

then if you make the mistake of writing your program such that it at some point requires the value of <code>resurrect 0</code>, you'll get the error message "I'm out of orange smoke!". <code>undefined</code> may be defined in the same way:

Since <code>undefined</code> causes an exception, the usual exception handling mechanism can be used to catch it:

import System.IO.Unsafe (unsafePerformIO)

import Prelude hiding (catch)

main = do

print $ safeHead ([] :: String)

== Icon and Unicon ==

==={{header|Unicon}}===

procedure main(arglist)

write((localnames|paramnames|staticnames|globalnames)(&current,0)) # ... visible in the current co-expression at this calling level (0)

return

The output looks like:

J does not have a concept of an "undefined value" as such, but J does allow treatment of undefined names. The verb <code>nc</code> finds the (syntactic) class of a name. This result is negative one for names which have not been defined.

foo=: 3

nc;:'foo bar'

From this we can infer that <code>foo</code> has a definition (and its definition is a noun, since 0 is the syntactic [http://www.jsoftware.com/help/dictionary/dx004.htm name class] for nouns), and we can also infer that <code>bar</code> does not have a definition.

This task also asked that we ''identify and exercise .. mechanisms for ... manipulating a variable's value's status as being undefined''. So: a name can be made to be undefined using the verb <code>erase</code>. The undefined status can be removed by assigning a value to the name.

erase;:'foo bar'

1 1

bar=:99

nc;:'foo bar'

=={{header|Java}}==

Java has a special null type, the type of the expression <code>null</code>, that can be cast to any reference type; in practice the null type can be ignored and <code>null</code> can be treated as a special literal that can be of any reference type. When a reference variable has the special value <code>null</code> it refers to no object, meaning that it is undefined.

System.out.println(string); //prints "null" to std out

Variables of primitive types cannot be assigned the special value <code>null</code>, but there are wrapper classes corresponding to the primitive types (eg. <code>Boolean</code>, <code>Integer</code>, <code>Long</code>, <code>Double</code>, &c), that can be used instead of the corresponding primitive; since they can have the special value <code>null</code> it can be used to identify the variable as undefined.

if (i == null) { // compilation error: incomparable types: int and <nulltype>

i = 1;

But this piece of code can be made valid by replacing <code>int</code> with <code>Integer</code>, and thanks to the automatic conversion between primitive types and their wrapper classes (called autoboxing) the only change required is in the declaration of the variable:

if (i == null) {

i = 1;

=={{header|JavaScript}}==

In Javascript undefined is a property of the global object, i.e. it is a variable in global scope. The initial value of undefined is the primitive value undefined. The problem with using undefined is that undefined is mutable. Instead we can use typeof to check if a value is undefined.

typeof(a) === "undefined";

obj.c === 42;

delete obj.c;

We can also use the prefix keyword void, it always returns undefined. But this will throw a error if the variable has not been defined.

a === void 0; // true

=={{header|jq}}==

Given a JSON object, o, and a key, k, that is not present in that object, then o[k] evaluates to null, e.g.

In an important sense, therefore, null in jq represents an undefined value. However, it should be noted that in jq, 1/0 does not yield null:

It should also be noted that in jq, null can combine with other values to form non-null values. Specifically, for any JSON entity, e,

For example, suppose it is agreed that the "sum" of the elements of an empty array should be null. Then one can simply write:

=={{header|Julia}}==

</code>

2. For variables that are defined for the Julia program but have undefined values, there are two types of undefined value in Julia (version > 0.7): <code> nothing </code> and <code> missing </code>. "nothing" and "missing" are typed constants used by convention to refer to either (with <code> nothing </code>) an absent result, such as a search with nothing found, or in the case of <code> missing, </code> a data location containing a missing value, such as a data table with missing values. <code>nothing</code> generally produces an error if any calculations incorporate it, but <code>missing</code> can be propagated along a calculation:

julia> arr = [1, 2, nothing, 3]

4-element Array{Union{Nothing, Int64},1}:

missing

8

=={{header|Kotlin}}==

Here are some simple examples illustrating these points:

class SomeClass

println(e)

}

{{out}}

=={{header|Lingo}}==

In Lingo an undefined variable has the value <Void>. If a variable is <Void> (i.e. undefined) can be checked by comparing it with the constant VOID, or by using the function voidP():

-- <Void>

put var=VOID

var = 23

put voidP(var)

=={{header|Logo}}==

UCB Logo has separate namespaces for procedures and variables ("names"). There is no distinction between a proc/name with no value and an undefined proc/name.

to square :x

output :x * :x

ern "n

show name? "n ; false

=={{header|LOLCODE}}==

LOLCODE's nil value is called <tt>NOOB</tt>, to which all uninitialized variables evaluate, and which is distinct from <tt>FAIL</tt>, the false value.

I HAS A foo BTW, INISHULIZD TO NOOB

OIC

=={{header|Lua}}==

local b

b = 5

end

Output:

<pre>nil

=={{header|Mathematica}}/{{header|Wolfram Language}}==

Mathematica is a symbolic mathematical software. Variables without given values are treated as symbols.

-> a

a + a

-> 5

ValueQ[a]

Mathematica also has a build-in symbol "Undefined", representing a quantity with no defined value.

Mathematical expressions containing Undefined evaluate to Undefined:

Of course you can assign Undefined to be the value of a variable. Here "Undefined" is itself a value.

-> Undefined

a

-> Undefined

ValueQ[a]

=={{header|MATLAB}} / {{header|Octave}}==

If a variable is generated without defing a value, e.g. with

the variable is empty, and can be tested with

For numerical values (e.g. vectors or arrays) with a predefined size, often not-a-numbers (NaN's) user used to indicate missing values,

These can be tested with:

<pre>

=={{header|MUMPS}}==

<p>MUMPS does have variables with undefined values, but referencing them usually causes an error. To test for whether a value is undefined use the $Data function. If you are trying to read a value that may be undefined, use $Get as a wrapper. Note that an alternate form of $Get can return a specified value, which must be defined.</p>

=={{header|Nim}}==

So, there are no undefined values except if the user explicitly specified that the variable must not be initialized. This may be useful for instance for a big array which will be initialized later on. To avoid the implicit initialization, we use the pragma <code>{.noInit.}</code>. For instance:

# For a proc, {.noInit.} means that the result is not initialized.

proc p(): array[1000, int] {.noInit.} =

=={{header|OCaml}}==

but if you really need this you can use the built-in "option" type.

It is defined like this: type 'a option = None | Some of 'a *)

inc None;;

=={{header|Oforth}}==

However, variables can be "unbound" or "free". If a program tries to read such a variable, the current thread will be suspended until the variable's value becomes determined.

thread

{Delay 1000}

{System.showInfo "Setting X."}

Explicitly checking the status of a variable with <code>IsFree</code> is discouraged because it can introduce race conditions.

=={{header|PARI/GP}}==

In GP, undefined variables test equal to the monomial in the variable with the name of that variable. So to test if <var>v</var> is undefined, do

In PARI, you can do the same but the function <code>is_entry()</code> is more appropriate:

=={{header|Pascal}}==

=={{header|Perl}}==

use strict;

# Because most of the output is conditional, this serves as

# a clear indicator that the program has run to completion.

Results in:

{{libheader|Phix/basics}}

Phix has a special unassigned value, that can be tested for using object(). (There is no offical way to "un-assign" a variable, though you could fairly easily do so with a bit of inline assembly.)

<span style="color: #004080;">object</span> <span style="color: #000000;">x</span>

<span style="color: #000000;">x</span> <span style="color: #0000FF;">=</span> <span style="color: #000000;">1</span>

<span style="color: #000000;">test</span><span style="color: #0000FF;">()</span>

{{out}}

<pre>

=={{header|PHP}}==

// Check to see whether it is defined

if (!isset($var))

// a clear indicator that the program has run to completion.

echo "Done\n";

Results in:

An internal symbol is initialized to NIL. Depending on the context, this is

interpreted as "undefined". When called as a function, an error is issued:

!? (myfoo 3 4)

myfoo -- Undefined

The function 'default' can be used to initialize a variable if and only

if its current value is NIL:

-> NIL

: MyVar

=={{header|Pike}}==

In Pike variables are always defined. <code>UNDEFINED</code> is only used to indicate the nonexistence of a key or object member. <code>UNDEFINED</code> is not a value and can not be assigned. in such cases it is converted to <math>0</math>. <code>zero_type()</code> is used to test if a key exists or not:

> zero_type(UNDEFINED);

Result: 1

> zero_type(bar->baz);

Result: 0

=={{header|PowerShell}}==

The <code>Get-Variable</code> cmdlet gets the Windows PowerShell variables in the current console. Variables can be filtered by using the <code>-Name</code> parameter.

If a variable doesn't exist an error is returned. Using the <code>-ErrorAction SilentlyContinue</code> parameter suppresses the error message and returns <code>$false</code>.

if (Get-Variable -Name noSuchVariable -ErrorAction SilentlyContinue)

{

$false

}

{{Out}}

<pre>

PowerShell represents things like the file system, registry, functions, variables, etc. with '''providers''' known as PS drives.

One of those PS drives is Variable. <code>Variable:</code> contains all of the variables that are currently stored in memory.

Get-PSProvider

{{Out}}

<pre>

</pre>

To access the Variable PS drive you'd use the same syntax as you would with the file system by specifying <code>Variable:\</code>.

Test-Path Variable:\noSuchVariable

{{Out}}

<pre>

</pre>

If a variable doesn't exist, it technically has a value of <code>$null</code>. <code>$null</code> is an automatic variable that represents "does not exist."

$noSuchVariable -eq $null

{{Out}}

<pre>

=={{header|PureBasic}}==

Variables are defined through a formal declaration or simply upon first use. Each variable is initialized to a value. PureBasic does not allow changing of a variable's status at runtime. It can be tested at compile-time and acted upon, however, it cannot be undefined once it is defined.

CompilerIf Defined(var, #PB_Variable)

CloseConsole()

Sample output:

<pre>var is undefined at first check

In Python names, (variables), can be dynamically created and deleted at run time.

try: name

except NameError: print "name is undefined at first check"

# Because most of the output is conditional, this serves as

# a clear indicator that the program has run to completion.

Results in:

=={{header|R}}==

There are four cases to consider. To test whether a varaible has previously been defined, use <code>exists</code>.

exists("x")

If you want to declare a variable with undefined contents, use <code>NULL</code>.

x <- NULL

If you want to declare a variable with missing values, use <code>NA</code>.

y <- c(1, 4, 9, NA, 25)

z <- c("foo", NA, "baz")

(Note that there are different types of <code>NA</code>, namely <code>NA_integer_</code>, <code>NA_real_</code>, <code>NA_character_</code>, <code>NA_complex_</code> and plain (logical) <code>NA</code>. In practice, you should hardly ever need to explicitly set which type of NA you are using, as it will be done automatically.)

Finally, you test for arguments that haven't been passed into a function with <code>missing</code>.

print_is_missing <- function(x)

{

print_is_missing() # TRUE

print_is_missing(123) # FALSE

=={{header|Racket}}==

recursive definitions. It can be grabbed explicitly with:

-> (letrec ([x x]) x)

#<undefined>

However, it is not used as an implicit value for all (non-existent)

Raku has "interesting" values of undef, but unlike Perl 5, doesn't actually have a value named <tt>undef</tt>. Instead, several very different meanings of undefinedness are distinguished. First, <tt>Nil</tt> represents the absence of a value. The absence of a value cannot be stored. Instead, an attempt to assign <tt>Nil</tt> to a storage location causes that location to revert to its uninitialized state, however that is defined.

This <tt>Any</tt> is an example of another kind of undefined type, which is a typed undef. All reference types have an undefined value representing the type. You can think of it as a sort of "type gluon" that carries a type charge without being a "real" particle. Hence there are undefined values whose names represent types, such as <tt>Int</tt>, <tt>Num</tt>, <tt>Str</tt>, and all the other object types in Raku. As generic objects, such undefined values carry the same metaobject pointer that a real object of the type would have, without being instantiated as a real object. As such, these types are in the type hierarchy. For example, <tt>Int</tt> derives from <tt>Cool</tt>, <tt>Cool</tt> derives from <tt>Any</tt>, and <tt>Any</tt> derives from <tt>Mu</tt>, the most general undefined object (akin to Object in other languages). Since they are real objects of the type, even if undefined, they can be used in reasoning about the type. You don't have to instantiate a <tt>Method</tt> object in order to ask if <tt>Method</tt> is derived from <tt>Routine</tt>, for instance.

Variables default to <tt>Any</tt>, unless declared to be of another type:

my Int $y; say $y.WHAT; # Int()

The user-interface for definedness are [http://design.raku.org/S12.html#Abstract_vs_Concrete_types type smilies] and the <tt>with</tt>-statement.

multi sub foo(Int:D $i where * != 0){ (0..100).roll / $i } # we will never divide by 0

multi sub foo(Int:U $i){ die 'WELP! $i is undefined' } # because undefinedness is deadly

with $i { say 'defined' } # as "if" is looking for Bool::True, "with" is looking for *.defined

with 0 { say '0 may not divide but it is defined' }

There are further some [http://design.raku.org/S03.html operators] for your convenience.

my $ain't-defined = Any;

my $doesn't-matter;

my $s = (f1() orelse f2()); # Please note the parentheses, which are needed because orelse is

# much looser then infix:<=> .

Finally, another major group of undefined values represents failures. Raku is designed with the notion that throwing exceptions is bad for parallel processing, so exceptions are thrown lazily; that is, they tend to be returned in-band instead as data, and are only thrown for real if not properly handled as exception data. (In which case, the exception is required to report the original difficulty accurately.) In order not to disrupt control flow and the synchronicity of event ordering, such failures are returned in-band as a normal values. And in order not to subvert the type system when it comes to return types, failure may be mixed into any reference type which produces an undefined, typed value which is also indicates the nature of the failure.

=={{header|REXX}}==

tlaloc = "rain god of the Aztecs." /*assign a value to the Aztec rain god.*/

/*check if the rain god is defined. */

if symbol(y)=="VAR" then say y ' is defined.'

else say y "isn't defined."

'''output'''

<pre>

=={{header|Ring}}==

# Project : Undefined values

islocal("y") + nl +

islocal("z") + nl

Output:

<pre>

=={{header|Ruby}}==

puts "var is undefined at first check" unless defined? var

# Because most of the output is conditional, this serves as

# a clear indicator that the program has run to completion.

Results in:

A pointer may have a ''null'' value:

let p: *const i32 = ptr::null();

=={{header|Scala}}==

=={{header|Sidef}}==

Sidef variables are initialized with a default ''nil'' value, representing the absence of a value.

x == nil && say "nil value";

defined(x) || say "undefined";

x = nil;

{{out}}

<pre>

However, we can check for the existence of a global binding with:

=={{header|Tcl}}==

Tcl does not have undefined ''values'', but ''variables'' may be undefined by being not set.

# Check to see whether it is defined

if {![info exists var]} {puts "var is undefind at fourth check"}

Yields this output:

<pre>

=={{header|UNIX Shell}}==

echo ${VAR1:-"Not set."}

<pre>VAR1

Not set.</pre>

In conditional expressions, ''null'' as well as the Boolean value ''false'' are considered to be 'false'. All other values (including zero) are considered to be 'true'.

System.print("'f' called.") // note no return value

}

var v = null // explicitly assigned null

{{out}}

=={{header|zkl}}==

Nothing is undefined in zkl (in theory). Numeric values like inf and NaN are "discouraged". "Void" is an object that is used as a "sentinel" for "I have no idea what the proper value should be" but it is just object that can be used like any other object.

1+Void

{{out}}

<pre>