Variables represent storage locations. Every variable has a type that determines what values can be stored in the variable. C# is a type-safe language, and the C# compiler guarantees that values stored in variables are always of the appropriate type. The value of a variable can be changed through assignment or through use of the ++
and --
operators.
A variable shall be definitely assigned (§9.4) before its value can be obtained.
As described in the following subclauses, variables are either initially assigned or initially unassigned. An initially assigned variable has a well-defined initial value and is always considered definitely assigned. An initially unassigned variable has no initial value. For an initially unassigned variable to be considered definitely assigned at a certain location, an assignment to the variable shall occur in every possible execution path leading to that location.
C# defines eight categories of variables: static variables, instance variables, array elements, value parameters, input parameters, reference parameters, output parameters, and local variables. The subclauses that follow describe each of these categories.
Example: In the following code
class A { public static int x; int y; void F(int[] v, int a, ref int b, out int c, in int d) { int i = 1; c = a + b++ + d; } }
x
is a static variable,y
is an instance variable,v[0]
is an array element,a
is a value parameter,b
is a reference parameter,c
is an output parameter,d
is an input parameter, andi
is a local variable. end example
A field declared with the static
modifier is a static variable. A static variable comes into existence before execution of the static
constructor (§15.12) for its containing type, and ceases to exist when the associated application domain ceases to exist.
The initial value of a static variable is the default value (§9.3) of the variable’s type.
For the purposes of definite-assignment checking, a static variable is considered initially assigned.
A field declared without the static
modifier is an instance variable.
An instance variable of a class comes into existence when a new instance of that class is created, and ceases to exist when there are no references to that instance and the instance’s finalizer (if any) has executed.
The initial value of an instance variable of a class is the default value (§9.3) of the variable’s type.
For the purpose of definite-assignment checking, an instance variable of a class is considered initially assigned.
An instance variable of a struct has exactly the same lifetime as the struct variable to which it belongs. In other words, when a variable of a struct type comes into existence or ceases to exist, so too do the instance variables of the struct.
The initial assignment state of an instance variable of a struct is the same as that of the containing struct
variable. In other words, when a struct variable is considered initially assigned, so too are its instance variables, and when a struct variable is considered initially unassigned, its instance variables are likewise unassigned.
The elements of an array come into existence when an array instance is created, and cease to exist when there are no references to that array instance.
The initial value of each of the elements of an array is the default value (§9.3) of the type of the array elements.
For the purpose of definite-assignment checking, an array element is considered initially assigned.
A value parameter comes into existence upon invocation of the function member (method, instance constructor, accessor, or operator) or anonymous function to which the parameter belongs, and is initialized with the value of the argument given in the invocation. A value parameter normally ceases to exist when execution of the function body completes. However, if the value parameter is captured by an anonymous function (§12.19.6.2), its lifetime extends at least until the delegate or expression tree created from that anonymous function is eligible for garbage collection.
For the purpose of definite-assignment checking, a value parameter is considered initially assigned.
Value parameters are discussed further in §15.6.2.2.
A reference parameter is a reference variable (§9.7) which comes into existence upon invocation of the function member, delegate, anonymous function, or local function and its referent is initialized to the variable given as the argument in that invocation. A reference parameter ceases to exist when execution of the function body completes. Unlike value parameters a reference parameter shall not be captured (§9.7.2.9).
The following definite-assignment rules apply to reference parameters.
Note: The rules for output parameters are different, and are described in (§9.2.7). end note
- A variable shall be definitely assigned (§9.4) before it can be passed as a reference parameter in a function member or delegate invocation.
- Within a function member or anonymous function, a reference parameter is considered initially assigned.
Reference parameters are discussed further in §15.6.2.3.3.
An output parameter is a reference variable (§9.7) which comes into existence upon invocation of the function member, delegate, anonymous function, or local function and its referent is initialized to the variable given as the argument in that invocation. An output parameter ceases to exist when execution of the function body completes. Unlike value parameters an output parameter shall not be captured (§9.7.2.9).
The following definite-assignment rules apply to output parameters.
Note: The rules for reference parameters are different, and are described in (§9.2.6). end note
- A variable need not be definitely assigned before it can be passed as an output parameter in a function member or delegate invocation.
- Following the normal completion of a function member or delegate invocation, each variable that was passed as an output parameter is considered assigned in that execution path.
- Within a function member or anonymous function, an output parameter is considered initially unassigned.
- Every output parameter of a function member, anonymous function, or local function shall be definitely assigned (§9.4) before the function member, anonymous function, or local function returns normally.
Output parameters are discussed further in §15.6.2.3.4.
An input parameter is a reference variable (§9.7) which comes into existence upon invocation of the function member, delegate, anonymous function, or local function and its referent is initialized to the variable_reference given as the argument in that invocation. An input parameter ceases to exist when execution of the function body completes. Unlike value parameters an input parameter shall not be captured (§9.7.2.9).
The following definite assignment rules apply to input parameters.
- A variable shall be definitely assigned (§9.4) before it can be passed as an input parameter in a function member or delegate invocation.
- Within a function member, anonymous function, or local function an input parameter is considered initially assigned.
Input parameters are discussed further in §15.6.2.3.2.
A local variable is declared by a local_variable_declaration, declaration_expression, foreach_statement, or specific_catch_clause of a try_statement. A local variable can also be declared by certain kinds of patterns (§11). For a foreach_statement, the local variable is an iteration variable (§13.9.5). For a specific_catch_clause, the local variable is an exception variable (§13.11). A local variable declared by a foreach_statement or specific_catch_clause is considered initially assigned.
A local_variable_declaration can occur in a block, a for_statement, a switch_block, or a using_statement. A declaration_expression can occur as an out
argument_value, and as a tuple_element that is the target of a deconstructing assignment (§12.21.2).
The lifetime of a local variable is the portion of program execution during which storage is guaranteed to be reserved for it. This lifetime extends from entry into the scope with which it is associated, at least until execution of that scope ends in some way. (Entering an enclosed block, calling a method, or yielding a value from an iterator block suspends, but does not end, execution of the current scope.) If the local variable is captured by an anonymous function (§12.19.6.2), its lifetime extends at least until the delegate or expression tree created from the anonymous function, along with any other objects that come to reference the captured variable, are eligible for garbage collection. If the parent scope is entered recursively or iteratively, a new instance of the local variable is created each time, and its initializer, if any, is evaluated each time.
Note: A local variable is instantiated each time its scope is entered. This behavior is visible to user code containing anonymous methods. end note
Note: The lifetime of an iteration variable (§13.9.5) declared by a foreach_statement is a single iteration of that statement. Each iteration creates a new variable. end note
Note: The actual lifetime of a local variable is implementation-dependent. For example, a compiler might statically determine that a local variable in a block is only used for a small portion of that block. Using this analysis, the compiler could generate code that results in the variable’s storage having a shorter lifetime than its containing block.
The storage referred to by a local reference variable is reclaimed independently of the lifetime of that local reference variable (§7.9).
end note
A local variable introduced by a local_variable_declaration or declaration_expression is not automatically initialized and thus has no default value. Such a local variable is considered initially unassigned.
Note: A local_variable_declaration that includes an initializer is still initially unassigned. Execution of the declaration behaves exactly like an assignment to the variable (§9.4.4.5). Using a variable before its initializer has been executed; e.g., within the initializer expression itself or by using a goto_statement which bypasses the initializer; is a compile-time error:
goto L; int x = 1; // never executed L: x += 1; // error: x not definitely assignedWithin the scope of a local variable, it is a compile-time error to refer to that local variable in a textual position that precedes its declarator.
end note
A discard is a local variable that has no name. A discard is introduced by a declaration expression (§12.17) with the identifier _
; and is either implicitly typed (_
or var _
) or explicitly typed (T _
).
Note:
_
is a valid identifier in many forms of declarations. end note
Because a discard has no name, the only reference to the variable it represents is the expression that introduces it.
Note: A discard can however be passed as an output argument, allowing the corresponding output parameter to denote its associated storage location. end note
A discard is not initially assigned, so it is always an error to access its value.
Example:
_ = "Hello".Length; (int, int, int) M(out int i1, out int i2, out int i3) { ... } (int _, var _, _) = M(out int _, out var _, out _);The example assumes that there is no declaration of the name
_
in scope.The assignment to
_
shows a simple pattern for ignoring the result of an expression. The call ofM
shows the different forms of discards available in tuples and as output parameters.end example
The following categories of variables are automatically initialized to their default values:
- Static variables.
- Instance variables of class instances.
- Array elements.
The default value of a variable depends on the type of the variable and is determined as follows:
- For a variable of a value_type, the default value is the same as the value computed by the value_type’s default constructor (§8.3.3).
- For a variable of a reference_type, the default value is
null
.
Note: Initialization to default values is typically done by having the memory manager or garbage collector initialize memory to all-bits-zero before it is allocated for use. For this reason, it is convenient to use all-bits-zero to represent the null reference. end note
At a given location in the executable code of a function member or an anonymous function, a variable is said to be definitely assigned if the compiler can prove, by a particular static flow analysis (§9.4.4), that the variable has been automatically initialized or has been the target of at least one assignment.
Note: Informally stated, the rules of definite assignment are:
- An initially assigned variable (§9.4.2) is always considered definitely assigned.
- An initially unassigned variable (§9.4.3) is considered definitely assigned at a given location if all possible execution paths leading to that location contain at least one of the following:
- A simple assignment (§12.21.2) in which the variable is the left operand.
- An invocation expression (§12.8.10) or object creation expression (§12.8.17.2) that passes the variable as an output parameter.
- For a local variable, a local variable declaration for the variable (§13.6.2) that includes a variable initializer.
The formal specification underlying the above informal rules is described in §9.4.2, §9.4.3, and §9.4.4.
end note
The definite-assignment states of instance variables of a struct_type variable are tracked individually as well as collectively. In additional to the rules described in §9.4.2, §9.4.3, and §9.4.4, the following rules apply to struct_type variables and their instance variables:
- An instance variable is considered definitely assigned if its containing struct_type variable is considered definitely assigned.
- A struct_type variable is considered definitely assigned if each of its instance variables is considered definitely assigned.
Definite assignment is a requirement in the following contexts:
-
A variable shall be definitely assigned at each location where its value is obtained.
Note: This ensures that undefined values never occur. end note
The occurrence of a variable in an expression is considered to obtain the value of the variable, except when
- the variable is the left operand of a simple assignment,
- the variable is passed as an output parameter, or
- the variable is a struct_type variable and occurs as the left operand of a member access.
-
A variable shall be definitely assigned at each location where it is passed as a reference parameter.
Note: This ensures that the function member being invoked can consider the reference parameter initially assigned. end note
-
A variable shall be definitely assigned at each location where it is passed as an input parameter.
Note: This ensures that the function member being invoked can consider the input parameter initially assigned. end note
-
All output parameters of a function member shall be definitely assigned at each location where the function member returns (through a return statement or through execution reaching the end of the function member body).
Note: This ensures that function members do not return undefined values in output parameters, thus enabling the compiler to consider a function member invocation that takes a variable as an output parameter equivalent to an assignment to the variable. end note
-
The
this
variable of a struct_type instance constructor shall be definitely assigned at each location where that instance constructor returns.
The following categories of variables are classified as initially assigned:
- Static variables.
- Instance variables of class instances.
- Instance variables of initially assigned struct variables.
- Array elements.
- Value parameters.
- Reference parameters.
- Input parameters.
- Variables declared in a
catch
clause or aforeach
statement.
The following categories of variables are classified as initially unassigned:
- Instance variables of initially unassigned struct variables.
- Output parameters, including the
this
variable of struct instance constructors without a constructor initializer. - Local variables, except those declared in a
catch
clause or aforeach
statement.
In order to determine that each used variable is definitely assigned, the compiler shall use a process that is equivalent to the one described in this subclause.
The compiler processes the body of each function member that has one or more initially unassigned variables. For each initially unassigned variable v, the compiler determines a definite-assignment state for v at each of the following points in the function member:
- At the beginning of each statement
- At the end point (§13.2) of each statement
- On each arc which transfers control to another statement or to the end point of a statement
- At the beginning of each expression
- At the end of each expression
The definite-assignment state of v can be either:
- Definitely assigned. This indicates that on all possible control flows to this point, v has been assigned a value.
- Not definitely assigned. For the state of a variable at the end of an expression of type
bool
, the state of a variable that isn’t definitely assigned might (but doesn’t necessarily) fall into one of the following sub-states:- Definitely assigned after true expression. This state indicates that v is definitely assigned if the Boolean expression evaluated as true, but is not necessarily assigned if the Boolean expression evaluated as false.
- Definitely assigned after false expression. This state indicates that v is definitely assigned if the Boolean expression evaluated as false, but is not necessarily assigned if the Boolean expression evaluated as true.
The following rules govern how the state of a variable v is determined at each location.
- v is not definitely assigned at the beginning of a function member body.
- The definite-assignment state of v at the beginning of any other statement is determined by checking the definite-assignment state of v on all control flow transfers that target the beginning of that statement. If (and only if) v is definitely assigned on all such control flow transfers, then v is definitely assigned at the beginning of the statement. The set of possible control flow transfers is determined in the same way as for checking statement reachability (§13.2).
- The definite-assignment state of v at the end point of a
block
,checked
,unchecked
,if
,while
,do
,for
,foreach
,lock
,using
, orswitch
statement is determined by checking the definite-assignment state of v on all control flow transfers that target the end point of that statement. If v is definitely assigned on all such control flow transfers, then v is definitely assigned at the end point of the statement. Otherwise, v is not definitely assigned at the end point of the statement. The set of possible control flow transfers is determined in the same way as for checking statement reachability (§13.2).
Note: Because there are no control paths to an unreachable statement, v is definitely assigned at the beginning of any unreachable statement. end note
The definite-assignment state of v on the control transfer to the first statement of the statement list in the block (or to the end point of the block, if the statement list is empty) is the same as the definite-assignment statement of v before the block, checked
, or unchecked
statement.
For an expression statement stmt that consists of the expression expr:
- v has the same definite-assignment state at the beginning of expr as at the beginning of stmt.
- If v if definitely assigned at the end of expr, it is definitely assigned at the end point of stmt; otherwise, it is not definitely assigned at the end point of stmt.
- If stmt is a declaration statement without initializers, then v has the same definite-assignment state at the end point of stmt as at the beginning of stmt.
- If stmt is a declaration statement with initializers, then the definite-assignment state for v is determined as if stmt were a statement list, with one assignment statement for each declaration with an initializer (in the order of declaration).
For a statement stmt of the form:
if ( «expr» ) «then_stmt» else «else_stmt»
- v has the same definite-assignment state at the beginning of expr as at the beginning of stmt.
- If v is definitely assigned at the end of expr, then it is definitely assigned on the control flow transfer to then_stmt and to either else_stmt or to the end-point of stmt if there is no else clause.
- If v has the state “definitely assigned after true expression” at the end of expr, then it is definitely assigned on the control flow transfer to then_stmt, and not definitely assigned on the control flow transfer to either else_stmt or to the end-point of stmt if there is no else clause.
- If v has the state “definitely assigned after false expression” at the end of expr, then it is definitely assigned on the control flow transfer to else_stmt, and not definitely assigned on the control flow transfer to then_stmt. It is definitely assigned at the end-point of stmt if and only if it is definitely assigned at the end-point of then_stmt.
- Otherwise, v is considered not definitely assigned on the control flow transfer to either the then_stmt or else_stmt, or to the end-point of stmt if there is no else clause.
For a switch
statement stmt with a controlling expression expr:
The definite-assignment state of v at the beginning of expr is the same as the state of v at the beginning of stmt.
The definite-assignment state of v at the beginning of a case’s guard clause is
- If v is a pattern variable declared in the switch_label: “definitely assigned”.
- If the switch label containing that guard clause (§13.8.3) is not reachable: “definitely assigned”.
- Otherwise, the state of v is the same as the state of v after expr.
Example: The second rule eliminates the need for the compiler to issue an error if an unassigned variable is accessed in unreachable code. The state of b is “definitely assigned” in the unreachable switch label
case 2 when b
.bool b; switch (1) { case 2 when b: // b is definitely assigned here. break; }end example
The definite-assignment state of v on the control flow transfer to a reachable switch block statement list is
- If the control transfer was due to a ‘goto case’ or ‘goto default’ statement, then the state of v is the same as the state at the beginning of that ‘goto’ statement.
- If the control transfer was due to the
default
label of the switch, then the state of v is the same as the state of v after expr. - If the control transfer was due to an unreachable switch label, then the state of v is “definitely assigned”.
- If the control transfer was due to a reachable switch label with a guard clause, then the state of v is the same as the state of v after the guard clause.
- If the control transfer was due to a reachable switch label without a guard clause, then the state of v is
- If v is a pattern variable declared in the switch_label: “definitely assigned”.
- Otherwise, the state of v is the same as the stat of v after expr.
A consequence of these rules is that a pattern variable declared in a switch_label will be “not definitely assigned” in the statements of its switch section if it is not the only reachable switch label in its section.
Example:
public static double ComputeArea(object shape) { switch (shape) { case Square s when s.Side == 0: case Circle c when c.Radius == 0: case Triangle t when t.Base == 0 || t.Height == 0: case Rectangle r when r.Length == 0 || r.Height == 0: // none of s, c, t, or r is definitely assigned return 0; case Square s: // s is definitely assigned return s.Side * s.Side; case Circle c: // c is definitely assigned return c.Radius * c.Radius * Math.PI; … } }end example
For a statement stmt of the form:
while ( «expr» ) «while_body»
- v has the same definite-assignment state at the beginning of expr as at the beginning of stmt.
- If v is definitely assigned at the end of expr, then it is definitely assigned on the control flow transfer to while_body and to the end point of stmt.
- If v has the state “definitely assigned after true expression” at the end of expr, then it is definitely assigned on the control flow transfer to while_body, but not definitely assigned at the end-point of stmt.
- If v has the state “definitely assigned after false expression” at the end of expr, then it is definitely assigned on the control flow transfer to the end point of stmt, but not definitely assigned on the control flow transfer to while_body.
For a statement stmt of the form:
do «do_body» while ( «expr» ) ;
- v has the same definite-assignment state on the control flow transfer from the beginning of stmt to do_body as at the beginning of stmt.
- v has the same definite-assignment state at the beginning of expr as at the end point of do_body.
- If v is definitely assigned at the end of expr, then it is definitely assigned on the control flow transfer to the end point of stmt.
- If v has the state “definitely assigned after false expression” at the end of expr, then it is definitely assigned on the control flow transfer to the end point of stmt, but not definitely assigned on the control flow transfer to do_body.
For a statement of the form:
for ( «for_initializer» ; «for_condition» ; «for_iterator» )
«embedded_statement»
definite-assignment checking is done as if the statement were written:
{
«for_initializer» ;
while ( «for_condition» )
{
«embedded_statement» ;
LLoop: «for_iterator» ;
}
}
with continue
statements that target the for
statement being translated to goto
statements targeting the label LLoop
. If the for_condition is omitted from the for
statement, then evaluation of definite-assignment proceeds as if for_condition were replaced with true in the above expansion.
The definite-assignment state of v on the control flow transfer caused by a break
, continue
, or goto
statement is the same as the definite-assignment state of v at the beginning of the statement.
For a statement stmt of the form:
throw «expr» ;
the definite-assignment state of v at the beginning of expr is the same as the definite-assignment state of v at the beginning of stmt.
For a statement stmt of the form:
return «expr» ;
- The definite-assignment state of v at the beginning of expr is the same as the definite-assignment state of v at the beginning of stmt.
- If v is an output parameter, then it shall be definitely assigned either:
- after expr
- or at the end of the
finally
block of atry
-finally
ortry
-catch
-finally
that encloses thereturn
statement.
For a statement stmt of the form:
return ;
- If v is an output parameter, then it shall be definitely assigned either:
- before stmt
- or at the end of the
finally
block of atry
-finally
ortry
-catch
-finally
that encloses thereturn
statement.
For a statement stmt of the form:
try «try_block»
catch ( ... ) «catch_block_1»
...
catch ( ... ) «catch_block_n»
- The definite-assignment state of v at the beginning of try_block is the same as the definite-assignment state of v at the beginning of stmt.
- The definite-assignment state of v at the beginning of catch_block_i (for any i) is the same as the definite-assignment state of v at the beginning of stmt.
- The definite-assignment state of v at the end-point of stmt is definitely assigned if (and only if) v is definitely assigned at the end-point of try_block and every catch_block_i (for every i from 1 to n).
For a statement stmt of the form:
try «try_block» finally «finally_block»
- The definite-assignment state of v at the beginning of try_block is the same as the definite-assignment state of v at the beginning of stmt.
- The definite-assignment state of v at the beginning of finally_block is the same as the definite-assignment state of v at the beginning of stmt.
- The definite-assignment state of v at the end-point of stmt is definitely assigned if (and only if) at least one of the following is true:
- v is definitely assigned at the end-point of try_block
- v is definitely assigned at the end-point of finally_block
If a control flow transfer (such as a goto
statement) is made that begins within try_block, and ends outside of try_block, then v is also considered definitely assigned on that control flow transfer if v is definitely assigned at the end-point of finally_block. (This is not an only if—if v is definitely assigned for another reason on this control flow transfer, then it is still considered definitely assigned.)
For a statement of the form:
try «try_block»
catch ( ... ) «catch_block_1»
...
catch ( ... ) «catch_block_n»
finally «finally_block»
definite-assignment analysis is done as if the statement were a try
-finally
statement enclosing a try
-catch
statement:
try
{
try «try_block»
catch ( ... ) «catch_block_1»
...
catch ( ... ) «catch_block_n»
}
finally «finally_block»
Example: The following example demonstrates how the different blocks of a
try
statement (§13.11) affect definite assignment.class A { static void F() { int i, j; try { goto LABEL; // neither i nor j definitely assigned i = 1; // i definitely assigned } catch { // neither i nor j definitely assigned i = 3; // i definitely assigned } finally { // neither i nor j definitely assigned j = 5; // j definitely assigned } // i and j definitely assigned LABEL: ; // j definitely assigned } }end example
For a statement stmt of the form:
foreach ( «type» «identifier» in «expr» ) «embedded_statement»
- The definite-assignment state of v at the beginning of expr is the same as the state of v at the beginning of stmt.
- The definite-assignment state of v on the control flow transfer to embedded_statement or to the end point of stmt is the same as the state of v at the end of expr.
For a statement stmt of the form:
using ( «resource_acquisition» ) «embedded_statement»
- The definite-assignment state of v at the beginning of resource_acquisition is the same as the state of v at the beginning of stmt.
- The definite-assignment state of v on the control flow transfer to embedded_statement is the same as the state of v at the end of resource_acquisition.
For a statement stmt of the form:
lock ( «expr» ) «embedded_statement»
- The definite-assignment state of v at the beginning of expr is the same as the state of v at the beginning of stmt.
- The definite-assignment state of v on the control flow transfer to embedded_statement is the same as the state of v at the end of expr.
For a statement stmt of the form:
yield return «expr» ;
- The definite-assignment state of v at the beginning of expr is the same as the state of v at the beginning of stmt.
- The definite-assignment state of v at the end of stmt is the same as the state of v at the end of expr.
A yield break
statement has no effect on the definite-assignment state.
The following applies to any constant expression, and takes priority over any rules from the following sections that might apply:
For a constant expression with value true
:
- If v is definitely assigned before the expression, then v is definitely assigned after the expression.
- Otherwise v is “definitely assigned after false expression” after the expression.
Example:
int x; if (true) {} else { Console.WriteLine(x); }end example
For a constant expression with value false
:
- If v is definitely assigned before the expression, then v is definitely assigned after the expression.
- Otherwise v is “definitely assigned after true expression” after the expression.
Example:
int x; if (false) { Console.WriteLine(x); }end example
For all other constant expressions, the definite-assignment state of v after the expression is the same as the definite-assignment state of v before the expression.
The following rule applies to these kinds of expressions: literals (§12.8.2), simple names (§12.8.4), member access expressions (§12.8.7), non-indexed base access expressions (§12.8.15), typeof
expressions (§12.8.18), default value expressions (§12.8.21), nameof
expressions (§12.8.23), and declaration expressions (§12.17).
- The definite-assignment state of v at the end of such an expression is the same as the definite-assignment state of v at the beginning of the expression.
The following rules apply to these kinds of expressions: parenthesized expressions (§12.8.5), tuple expressions (§12.8.6), element access expressions (§12.8.12), base access expressions with indexing (§12.8.15), increment and decrement expressions (§12.8.16, §12.9.6), cast expressions (§12.9.7), unary +
, -
, ~
, *
expressions, binary +
, -
, *
, /
, %
, <<
, >>
, <
, <=
, >
, >=
, ==
, !=
, is
, as
, &
, |
, ^
expressions (§12.10, §12.11, §12.12, §12.13), compound assignment expressions (§12.21.4), checked
and unchecked
expressions (§12.8.20), array and delegate creation expressions (§12.8.17) , and await
expressions (§12.9.8).
Each of these expressions has one or more subexpressions that are unconditionally evaluated in a fixed order.
Example: The binary
%
operator evaluates the left hand side of the operator, then the right hand side. An indexing operation evaluates the indexed expression, and then evaluates each of the index expressions, in order from left to right. end example
For an expression expr, which has subexpressions expr₁, expr₂, …, exprₓ, evaluated in that order:
- The definite-assignment state of v at the beginning of expr₁ is the same as the definite-assignment state at the beginning of expr.
- The definite-assignment state of v at the beginning of exprᵢ (i greater than one) is the same as the definite-assignment state at the end of exprᵢ₋₁.
- The definite-assignment state of v at the end of expr is the same as the definite-assignment state at the end of exprₓ.
If the method to be invoked is a partial method that has no implementing partial method declaration, or is a conditional method for which the call is omitted (§22.5.3.2), then the definite-assignment state of v after the invocation is the same as the definite-assignment state of v before the invocation. Otherwise the following rules apply:
For an invocation expression expr of the form:
«primary_expression» ( «arg₁», «arg₂», … , «argₓ» )
or an object-creation expression expr of the form:
new «type» ( «arg₁», «arg₂», … , «argₓ» )
- For an invocation expression, the definite assignment state of v before primary_expression is the same as the state of v before expr.
- For an invocation expression, the definite assignment state of v before arg₁ is the same as the state of v after primary_expression.
- For an object creation expression, the definite assignment state of v before arg₁ is the same as the state of v before expr.
- For each argument argᵢ, the definite assignment state of v after argᵢ is determined by the normal expression rules, ignoring any
in
,out
, orref
modifiers. - For each argument argᵢ for any i greater than one, the definite assignment state of v before argᵢ is the same as the state of v after argᵢ₋₁.
- If the variable v is passed as an
out
argument (i.e., an argument of the form “out v”) in any of the arguments, then the state of v after expr is definitely assigned. Otherwise, the state of v after expr is the same as the state of v after argₓ. - For array initializers (§12.8.17.5), object initializers (§12.8.17.3), collection initializers (§12.8.17.4) and anonymous object initializers (§12.8.17.7), the definite-assignment state is determined by the expansion that these constructs are defined in terms of.
Let the set of assignment targets in an expression e be defined as follows:
- If e is a tuple expression, then the assignment targets in e are the union of the assignment targets of the elements of e.
- Otherwise, the assignment targets in e are e.
For an expression expr of the form:
«expr_lhs» = «expr_rhs»
- The definite-assignment state of v before expr_lhs is the same as the definite-assignment state of v before expr.
- The definite-assignment state of v before expr_rhs is the same as the definite-assignment state of v after expr_lhs.
- If v is an assignment target of expr_lhs, then the definite-assignment state of v after expr is definitely assigned. Otherwise, if the assignment occurs within the instance constructor of a struct type, and v is the hidden backing field of an automatically implemented property P on the instance being constructed, and a property access designating P is an assigment target of expr_lhs, then the definite-assignment state of v after expr is definitely assigned. Otherwise, the definite-assignment state of v after expr is the same as the definite-assignment state of v after expr_rhs.
Example: In the following code
class A { static void F(int[] arr) { int x; arr[x = 1] = x; // ok } }the variable
x
is considered definitely assigned afterarr[x = 1]
is evaluated as the left hand side of the second simple assignment.end example
For an expression expr of the form:
«expr_first» && «expr_second»
- The definite-assignment state of v before expr_first is the same as the definite-assignment state of v before expr.
- The definite-assignment state of v before expr_second is definitely assigned if and only if the state of v after expr_first is either definitely assigned or “definitely assigned after true expression”. Otherwise, it is not definitely assigned.
- The definite-assignment state of v after expr is determined by:
- If the state of v after expr_first is definitely assigned, then the state of v after expr is definitely assigned.
- Otherwise, if the state of v after expr_second is definitely assigned, and the state of v after expr_first is “definitely assigned after false expression”, then the state of v after expr is definitely assigned.
- Otherwise, if the state of v after expr_second is definitely assigned or “definitely assigned after true expression”, then the state of v after expr is “definitely assigned after true expression”.
- Otherwise, if the state of v after expr_first is “definitely assigned after false expression”, and the state of v after expr_second is “definitely assigned after false expression”, then the state of v after expr is “definitely assigned after false expression”.
- Otherwise, the state of v after expr is not definitely assigned.
Example: In the following code
class A { static void F(int x, int y) { int i; if (x >= 0 && (i = y) >= 0) { // i definitely assigned } else { // i not definitely assigned } // i not definitely assigned } }the variable
i
is considered definitely assigned in one of the embedded statements of anif
statement but not in the other. In theif
statement in methodF
, the variablei
is definitely assigned in the first embedded statement because execution of the expression(i = y)
always precedes execution of this embedded statement. In contrast, the variablei
is not definitely assigned in the second embedded statement, sincex >= 0
might have tested false, resulting in the variablei
’s being unassigned.end example
For an expression expr of the form:
«expr_first» || «expr_second»
- The definite-assignment state of v before expr_first is the same as the definite-assignment state of v before expr.
- The definite-assignment state of v before expr_second is definitely assigned if and only if the state of v after expr_first is either definitely assigned or “definitely assigned after true expression”. Otherwise, it is not definitely assigned.
- The definite-assignment statement of v after expr is determined by:
- If the state of v after expr_first is definitely assigned, then the state of v after expr is definitely assigned.
- Otherwise, if the state of v after expr_second is definitely assigned, and the state of v after expr_first is “definitely assigned after true expression”, then the state of v after expr is definitely assigned.
- Otherwise, if the state of v after expr_second is definitely assigned or “definitely assigned after false expression”, then the state of v after expr is “definitely assigned after false expression”.
- Otherwise, if the state of v after expr_first is “definitely assigned after true expression”, and the state of v after expr_ second is “definitely assigned after true expression”, then the state of v after expr is “definitely assigned after true expression”.
- Otherwise, the state of v after expr is not definitely assigned.
Example: In the following code
class A { static void G(int x, int y) { int i; if (x >= 0 || (i = y) >= 0) { // i not definitely assigned } else { // i definitely assigned } // i not definitely assigned } }the variable
i
is considered definitely assigned in one of the embedded statements of anif
statement but not in the other. In theif
statement in methodG
, the variablei
is definitely assigned in the second embedded statement because execution of the expression(i = y)
always precedes execution of this embedded statement. In contrast, the variablei
is not definitely assigned in the first embedded statement, sincex >= 0
might have tested true, resulting in the variablei
’s being unassigned.end example
For an expression expr of the form:
! «expr_operand»
- The definite-assignment state of v before expr_operand is the same as the definite-assignment state of v before expr.
- The definite-assignment state of v after expr is determined by:
- If the state of
v
after expr_operand is definitely assigned, then the state ofv
after expr is definitely assigned. - Otherwise, if the state of
v
after expr_operand is “definitely assigned after false expression”, then the state ofv
after expr is “definitely assigned after true expression”. - Otherwise, if the state of
v
after expr_operand is “definitely assigned after true expression”, then the state of v after expr is “definitely assigned after false expression”. - Otherwise, the state of
v
after expr is not definitely assigned.
- If the state of
For an expression expr of the form:
«expr_first» ?? «expr_second»
- The definite-assignment state of v before expr_first is the same as the definite-assignment state of v before expr.
- The definite-assignment state of v before expr_second is the same as the definite-assignment state of v after expr_first.
- The definite-assignment statement of v after expr is determined by:
- If expr_first is a constant expression (§12.23) with value
null
, then the state of v after expr is the same as the state of v after expr_second. - Otherwise, the state of v after expr is the same as the definite-assignment state of v after expr_first.
- If expr_first is a constant expression (§12.23) with value
For an expression expr of the form:
«expr_cond» ? «expr_true» : «expr_false»
- The definite-assignment state of v before expr_cond is the same as the state of v before expr.
- The definite-assignment state of v before expr_true is definitely assigned if the state of v after expr_cond is definitely assigned or “definitely assigned after true expression”.
- The definite-assignment state of v before expr_false is definitely assigned if the state of v after expr_cond is definitely assigned or “definitely assigned after false expression”.
- The definite-assignment state of v after expr is determined by:
- If expr_cond is a constant expression (§12.23) with value
true
then the state of v after expr is the same as the state of v after expr_true. - Otherwise, if expr_cond is a constant expression (§12.23) with value
false
then the state of v after expr is the same as the state of v after expr_false. - Otherwise, if the state of v after expr_true is definitely assigned and the state of v after expr_false is definitely assigned, then the state of v after expr is definitely assigned.
- Otherwise, the state of v after expr is not definitely assigned.
- If expr_cond is a constant expression (§12.23) with value
For a lambda_expression or anonymous_method_expression expr with a body (either block or expression) body:
- The definite assignment state of a parameter is the same as for a parameter of a named method (§9.2.6, §9.2.7, §9.2.8).
- The definite assignment state of an outer variable v before body is the same as the state of v before expr. That is, definite assignment state of outer variables is inherited from the context of the anonymous function.
- The definite assignment state of an outer variable v after expr is the same as the state of v before expr.
Example: The example
class A { delegate bool Filter(int i); void F() { int max; // Error, max is not definitely assigned Filter f = (int n) => n < max; max = 5; DoWork(f); } void DoWork(Filter f) { ... } }generates a compile-time error since max is not definitely assigned where the anonymous function is declared.
end example
Example: The example
class A { delegate void D(); void F() { int n; D d = () => { n = 1; }; d(); // Error, n is not definitely assigned Console.WriteLine(n); } }also generates a compile-time error since the assignment to
n
in the anonymous function has no affect on the definite-assignment state ofn
outside the anonymous function.end example
For an expression expr of the form:
throw
thrown_expr
- The definite assignment state of v before thrown_expr is the same as the state of v before expr.
- The definite assignment state of v after expr is “definitely assigned”.
Local functions are analyzed in the context of their parent method. There are two control flow paths that matter for local functions: function calls and delegate conversions.
Definite assignment for the body of each local function is defined separately for each call site. At each invocation, variables captured by the local function are considered definitely assigned if they were definitely assigned at the point of call. A control flow path also exists to the local function body at this point and is considered reachable. After a call to the local function, captured variables that were definitely assigned at every control point leaving the function (return
statements, yield
statements, await
expressions) are considered definitely assigned after the call location.
Delegate conversions have a control flow path to the local function body. Captured variables are definitely assigned for the body if they are definitely assigned before the conversion. Variables assigned by the local function are not considered assigned after the conversion.
Note: the above implies that bodies are re-analyzed for definite assignment at every local function invocation or delegate conversion. Compilers are not required to re-analyze the body of a local function at each invocation or delegate conversion. The implementation must produce results equivalent to that description. end note
Example: The following example demonstrates definite assignment for captured variables in local functions. If a local function reads a captured variable before writing it, the captured variable must be definitely assigned before calling the local function. The local function
F1
readss
without assigning it. It is an error ifF1
is called befores
is definitely assigned.F2
assignsi
before reading it. It may be called beforei
is definitely assigned. Furthermore,F3
may be called afterF2
becauses2
is definitely assigned inF2
.void M() { string s; int i; string s2; // Error: Use of unassigned local variable s: F1(); // OK, F2 assigns i before reading it. F2(); // OK, i is definitely assigned in the body of F2: s = i.ToString(); // OK. s is now definitely assigned. F1(); // OK, F3 reads s2, which is definitely assigned in F2. F3(); void F1() { Console.WriteLine(s); } void F2() { i = 5; // OK. i is definitely assigned. Console.WriteLine(i); s2 = i.ToString(); } void F3() { Console.WriteLine(s2); } }end example
For an expression expr of the form:
expr_operand is pattern
- The definite-assignment state of v before expr_operand is the same as the definite-assignment state of v before expr.
- If the variable ‘v’ is declared in pattern, then the definite-assignment state of ‘v’ after expr is “definitely assigned when true”.
- Otherwise the definite assignment state of ‘v’ after expr is the same as the definite assignment state of ‘v’ after expr_operand.
A variable_reference is an expression that is classified as a variable. A variable_reference denotes a storage location that can be accessed both to fetch the current value and to store a new value.
variable_reference
: expression
;
Note: In C and C++, a variable_reference is known as an lvalue. end note
Reads and writes of the following data types shall be atomic: bool
, char
, byte
, sbyte
, short
, ushort
, uint
, int
, float
, and reference types. In addition, reads and writes of enum types with an underlying type in the previous list shall also be atomic. Reads and writes of other types, including long
, ulong
, double
, and decimal
, as well as user-defined types, need not be atomic. Aside from the library functions designed for that purpose, there is no guarantee of atomic read-modify-write, such as in the case of increment or decrement.
A reference variable is a variable that refers to another variable, called the referent (§9.2.6). A reference variable is a local variable declared with the ref
modifier.
A reference variable stores a variable_reference (§9.5) to its referent and not the value of its referent. When a reference variable is used where a value is required its referent’s value is returned; similarly when a reference variable is the target of an assignment it is the referent which is assigned to. The variable to which a reference variable refers, i.e. the stored variable_reference for its referent, can be changed using a ref assignment (= ref
).
Example: The following example demonstrates a local reference variable whose referent is an element of an array:
public class C { public void M() { int[] arr = new int[10]; // element is a reference variable that refers to arr[5] ref int element = ref arr[5]; element += 5; // arr[5] has been incremented by 5 } }end example
A reference return is the variable_reference returned from a returns-by-ref method (§15.6.1). This variable_reference is the referent of the reference return.
Example: The following example demonstrates a reference return whose referent is an element of an array field:
public class C { private int[] arr = new int[10]; public ref readonly int M() { // element is a reference variable that refers to arr[5] ref int element = ref arr[5]; return ref element; // return reference to arr[5]; } }end example
All reference variables obey safety rules that ensure the ref-safe-context of the reference variable is not greater than the ref-safe-context of its referent.
Note: The related notion of a safe-context is defined in (§16.4.12), along with associated constraints. end note
For any variable, the ref-safe-context of that variable is the context where a variable_reference (§9.5) to that variable is valid. The referent of a reference variable shall have a ref-safe-context that is at least as wide as the ref-safe-context of the reference variable itself.
Note: The compiler determines the ref-safe-context through a static analysis of the program text. The ref-safe-context reflects the lifetime of a variable at runtime. end note
There are three ref-safe-contexts:
-
declaration-block: The ref-safe-context of a variable_reference to a local variable (§9.2.9) is that local variable’s scope (§13.6.2), including any nested embedded-statements in that scope.
A variable_reference to a local variable is a valid referent for a reference variable only if the reference variable is declared within the ref-safe-context of that variable.
-
function-member: Within a function a variable_reference to any of the following has a ref-safe-context of function-member:
- Value parameters (§15.6.2.2) on a function member declaration, including the implicit
this
of class member functions; and - The implicit reference (
ref
) parameter (§15.6.2.3.3)this
of a struct member function, along with its fields.
A variable_reference with ref-safe-context of function-member is a valid referent only if the reference variable is declared in the same function member.
- Value parameters (§15.6.2.2) on a function member declaration, including the implicit
-
caller-context: Within a function a variable_reference to any of the following has a ref-safe-context of caller-context:
- Reference parameters (§9.2.6) other than the implicit
this
of a struct member function; - Member fields and elements of such parameters;
- Member fields of parameters of class type; and
- Elements of parameters of array type.
- Reference parameters (§9.2.6) other than the implicit
A variable_reference with ref-safe-context of caller-context can be the referent of a reference return.
These values form a nesting relationship from narrowest (declaration-block) to widest (caller-context). Each nested block represents a different context.
Example: The following code shows examples of the different ref-safe-contexts. The declarations show the ref-safe-context for a referent to be the initializing expression for a
ref
variable. The examples show the ref-safe-context for a reference return:public class C { // ref safe context of arr is "caller-context". // ref safe context of arr[i] is "caller-context". private int[] arr = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 }; // ref safe context is "caller-context" public ref int M1(ref int r1) { return ref r1; // r1 is safe to ref return } // ref safe context is "function-member" public ref int M2(int v1) { return ref v1; // error: v1 isn't safe to ref return } public ref int M3() { int v2 = 5; return ref arr[v2]; // arr[v2] is safe to ref return } public void M4(int p) { int v3 = 6; // context of r2 is declaration-block, // ref safe context of p is function-member ref int r2 = ref p; // context of r3 is declaration-block, // ref safe context of v3 is declaration-block ref int r3 = ref v3; // context of r4 is declaration-block, // ref safe context of arr[v3] is caller-context ref int r4 = ref arr[v3]; } }end example.
Example: For
struct
types, the implicitthis
parameter is passed as a reference parameter. The ref-safe-context of the fields of astruct
type as function-member prevents returning those fields by reference return. This rule prevents the following code:public struct S { private int n; // Disallowed: returning ref of a field. public ref int GetN() => ref n; } class Test { public ref int M() { S s = new S(); ref int numRef = ref s.GetN(); return ref numRef; // reference to local variable 'numRef' returned } }end example.
For a local variable v
:
- If
v
is a reference variable, its ref-safe-context is the same as the ref-safe-context of its initializing expression. - Otherwise its ref-safe-context is declaration-block.
For a parameter p
:
- If
p
is a reference or input parameter, its ref-safe-context is the caller-context. Ifp
is an input parameter, it can’t be returned as a writableref
but can be returned asref readonly
. - If
p
is an output parameter, its ref-safe-context is the caller-context. - Otherwise, if
p
is thethis
parameter of a struct type, its ref-safe-context is the function-member. - Otherwise, the parameter is a value parameter, and its ref-safe-context is the function-member.
For a variable designating a reference to a field, e.F
:
- If
e
is of a reference type, its ref-safe-context is the caller-context. - Otherwise, if
e
is of a value type, its ref-safe-context is the same as the ref-safe-context ofe
.
The conditional operator (§12.18), c ? ref e1 : ref e2
, and reference assignment operator, = ref e
(§12.21.1) have reference variables as operands and yield a reference variable. For those operators, the ref-safe-context of the result is the narrowest context among the ref-safe-contexts of all ref
operands.
For a variable c
resulting from a ref-returning function invocation, its ref-safe-context is the narrowest of the following contexts:
- The caller-context.
- The ref-safe-context of all
ref
,out
, andin
argument expressions (excluding the receiver). - For each input parameter, if there is a corresponding expression that is a variable and there exists an identity conversion between the type of the variable and the type of the parameter, the variable’s ref-safe-context, otherwise the nearest enclosing context.
- The safe-context (§16.4.12) of all argument expressions (including the receiver).
Example: the last bullet is necessary to handle code such as
ref int M2() { int v = 5; // Not valid. // ref safe context of "v" is block. // Therefore, ref safe context of the return value of M() is block. return ref M(ref v); } ref int M(ref int p) { return ref p; }end example
A property invocation and an indexer invocation (either get
or set
) is treated as a function invocation of the underlying accessor by the above rules. A local function invocation is a function invocation.
A value’s ref-safe-context is the nearest enclosing context.
Note: This occurs in an invocation such as
M(ref d.Length)
whered
is of typedynamic
. It is also consistent with arguments corresponding to input parameters. end note
A new
expression that invokes a constructor obeys the same rules as a method invocation (§9.7.2.6) that is considered to return the type being constructed.
- Neither a reference parameter, nor an output parameter, nor an input parameter, nor a
ref
local, nor a parameter or local of aref struct
type shall be captured by lambda expression or local function. - Neither a reference parameter, nor an output parameter, nor an input parameter, nor a parameter of a
ref struct
type shall be an argument for an iterator method or anasync
method. - Neither a
ref
local, nor a local of aref struct
type shall be in context at the point of ayield return
statement or anawait
expression. - For a ref reassignment
e1 = ref e2
, the ref-safe-context ofe2
shall be at least as wide a context as the ref-safe-context ofe1
. - For a ref return statement
return ref e1
, the ref-safe-context ofe1
shall be the caller-context.