Dart Language Specification (0.41, M4+)

# The Dart Programming Language Specification (0.41, M4+)

The Dart Team

The Dart Programming Language Specification (0.41, M4+)

Licensing

Changes

Changes Since Version 0.02

Changes Since Version 0.03

Changes Since Version 0.04

Changes Since Version 0.05

Changes since Version 0.06

Changes since Version 0.07

Changes since Version 0.08

Changes since Version 0.09

Changes since Version 0.10

Changes since Version 0.11

Changes since Version 0.12

Changes since Version 0.20

Changes since Version 0.30

Changes Since Version 0.31

Changes Since Version 0.40

Notation

Overview

Scoping

Privacy

Concurrency

Errors and Warnings

Variables

Evaluation of Implicit  Variable Getters

Functions

Function Declarations

Formal Parameters

Required Formals

Optional Formals

Type of a Function

External Functions

Classes

Instance Methods

Operators

Getters

Setters

Abstract Instance Members

Instance Variables

Constructors

Generative Constructors

Redirecting Constructors

Initializer Lists

Factories

Redirecting Factory Constructors

Constant Constructors

Static Methods

Static Variables

Superclasses

Inheritance and Overriding

Superinterfaces

Interfaces

Superinterfaces

Inheritance and Overriding

Mixins

Mixin Application

Mixin Composition

Generics

Expressions

Constants

Null

Numbers

Booleans

Boolean Conversion

Strings

String Interpolation

Lists

Maps

Throw

Function Expressions

This

Instance Creation

New

Const

Spawning an Isolate

Property Extraction

Function Invocation

Actual Argument List Evaluation

Binding Actuals to Formals

Unqualified Invocation

Function Expression Invocation

Method Invocation

Ordinary Invocation

Static Invocation

Super Invocation

Sending Messages

Getter and Setter Lookup

Getter Invocation

Assignment

Compound Assignment

Conditional

Logical Boolean Expressions

Bitwise Expressions

Equality

Relational Expressions

Shift

Multiplicative Expressions

Unary Expressions

Postfix Expressions

Assignable Expressions

Identifier Reference

Type Test

Type Cast

Statements

Blocks

Expression Statements

Local Variable Declaration

Local Function Declaration

If

For

For Loop

For-in

While

Do

Switch

Rethrow

Try

Return

Labels

Break

Continue

Assert

Libraries and Scripts

Imports

Exports

Parts

Scripts

URIs

Types

Static Types

Dynamic Type System

Type Declarations

Typedef

Interface Types

Function Types

Type dynamic

Type Void

Parameterized Types

Actual Type of a Declaration

Least Upper Bounds

Reference

Lexical Rules

Reserved Words

Operator Precedence

Appendix: Naming Conventions

Notes

This document is synced to version 0.41.

This is a work in progress. Expect the contents and language rules to change over time.

#### Licensing

Except as otherwise noted, the content of this document is licensed under the Creative Commons Attribution 3.0 License, and code samples are licensed under the BSD License.

#### Changes

##### Changes Since Version 0.02

The following changes have been made in version 0.03 since version 0.02. In addition, various typographical errors have been corrected. The changes are listed by section number.

2: Expanded examples of grammar in notation section.

7.10: Removed static warning when imported superinterface of a class contains private members.

8.4: Removed redundant prohibition on default values.

8.5: Removed static warning when imported superinterface of an interface contains private members.

10: Fixed typo in grammar

10.10.1, 10.10.2 : made explicit accessibility requirement for class being constructed.

10.10.2: make clear that referenced constructor must be marked const.

10.14.3: fixed botched sentence where superclass S is introduced.

10.27: qualified definition of v++ so it is clear that v is an identifier.

##### Changes Since Version 0.03

7.1: Added rules prohibiting clashes of inherited variable names or of static and instance methods.

7.1, 8.1: Added missing requirement that overriding methods have same number of required parameters and all optional parameters as overridden method, in same order.

9: Added prohibition against cyclic type hierarchy for type parameters.

10.10:  Clarified requirements on use of parameterized types in instance creation expressions.

10.13.2: Added requirement that qi are distinct.

10.4.2: Static method invocation determines the function (which may involve evaluating a getter) before evaluating the arguments, so that static invocation and top-level function invocation agree.

10:30: Added missing test that type being tested against is in scope and is indeed a type.

11.5.1: Changed for loop to introduce fresh variable for each iteration.

13.8:  Malformed parameterized types generate warnings, not errors(except when used in reified contexts like instance creation and superclasses/interfaces).

##### Changes Since Version 0.04

7.1.1: Removed unary plus operator. Clarified that operator formals must be required.

7.5.3: Filled in a lot of missing detail.

8.4: Allowed factory class to be declared via a qualified name.

10.3: Changed production for Number.

10.10.2: Added requirements that actuals be constant, rules for dealing with inappropriate types of actuals, and examples. Also explicitly prohibit type parameters.

10.13.4: Modified final bullet to keep it inline with similar clauses in other sections. Exact wording of these sections also tweaked slightly.

10.25: Specified ! operator. Eliminated section on prefix expressions and moved contents to section on unary expressions.

14.1: Specified unicode form of Dart source.

##### Changes Since Version 0.05

7:6.1: Clarified how initializing formals can act as optional parameters of generative constructors.

7.6.2:  Treat factories as constructors, so type parameters are implicitly in scope

8.4: Simplify rules for interface factory clauses. Use the keyword default instead of factory.

9: Mention that typedefs can have type parameters.

10.29: Added checked mode test that type arguments match generic type.

13.2: Added definition of malformed types, and requirement on their handling in checked mode.

##### Changes since Version 0.06

5: Library variable initializers must be constant.

7: Added abstract modifier to grammar.

7, 7.7, 7.8, 10.13.3, 10.28:  Superclass static members are not in scope in subclasses, and do not conflict with subclass members.

7.1.1: []= must return void. Operator call added to support function emulation. Removed operator >>>. Made explicit restriction on methods named call  or negate.

10.1: Added !e as constant expression. Clarified what happens if evaluation of a constant fails.

10.7: Map keys need not be constants. However, they are always string literals.

10.9: State restrictions on use of this.

10.10, 10.10.1: Rules for bounds checking of constructor arguments when calling default constructors for interfaces refined.

10.13.4: Revised semantics to account for function emulation.

10.14.1: Revised semantics to account for function emulation.

10.14.2: Revised semantics to account for function emulation.

10.14.3:  Factory constructors cannot contain super invocations. Revised semantics to account for function emulation.

10.16: Specified assignment involving []= operator.

10.16.1: Removed operator >>>.

10.22: Removed operator >>>.

10.26: Postfix -- operator specified. Behavior of postfix operations on subscripted expressions specified.

10:28: Added built-in identifier call.  Banned use of built-in identifiers as types, and made other uses warnings.

10.29:  Moved specification of test that type arguments match generic type to 13.2 .

11.8: Corrected evaluation of case clauses so that case expression is the receiver of ==. Revised specification to correctly deal with blank statements in case clauses.

11:15: Fixed bug in assert specification that could lead to puzzlers.

13.2: Consolidated definition of malformed types.

13.5: Revised semantics to account for function emulation.

##### Changes since Version 0.07

5:  Static variables are lazily initialized, but need not be constants. Orthogonal notion of constant variable introduced.

7.1.1: Added equals operator as part of revised == treatment.

7.6.1: Initializing formals have the same type as the field they correspond to.

7.8: Static variable getter rules revised to deal with lazy initialization.

10: Modified syntax to support cascaded method invocations.

10.1: Removed support for + operator on Strings. Extended string constants to support certain cases of string interpolation. Revised constants to deal with constant variables

10.5: Corrected definition of HEX_DIGIT_SEQUENCE. Support implicit concatenation of adjacent single line strings.

10.13.2: Centralized and corrected type rules for function invocation.

10.14: Moved rules for checking function/method invocations to 10.1.3.2. Added definition of cascaded method invocations.

10.15, 10.16: Updated noSuchMethod() call for getters and setters to conform to planned API.

10.17: Modified syntax to support cascaded method invocations.

10:20: Revised semantics for ==.

10:28: Removed import, library and source from list of built-in identifiers. Revised rules for evaluating identifiers to deal with lazy static variable initialization.

11.13: Fixed bug that allowed continue labeled on non-loops.

12: Revised syntax so no space is permitted between # and directives. Introduced show: combinator. Describe prefix: as a combinator. Added initial discussion of namespaces. Preclude string interpolation in arguments to directives.

##### Changes since Version 0.08

7.1, 7.4: Abstract methods may specify default values.

8, 8.1: Interface methods may specify default values.

10.1: The ~/  operator can apply to doubles.

10.10: Refined rules regarding abstract class instantiation, allowing factories to be used.

11.8: switch  statement specification revised.

11.14: throw may not throw null.

12.2: Imports introduce a scope separate from the library scope. Multiple libraries may share prefix.

13.3.1: Recursive typedefs disallowed.

##### Changes since Version 0.09

3.1: Consolidated discussion of namespaces and scopes. Started to tighten up definitions of scopes.

7: Overriding of fields allowed.

7.1.1: call()  is no longer an operator.

8.3: Added specification of variable declarations in interfaces.

10.1: Static methods and top level functions are compile-time constants.

10.5: Multiline strings can be implicitly concatenated and contain interpolated expressions.

10.30: Type cast expression added.

11.2: Map literals cannot be expression statements.

11.9, 11.14: Clarified type of stack trace.

12.1, 12.2: Added re-export facility.

##### Changes since Version 0.10

3: Discuss reified runtime types.

5: Centralized discussion of implicit getters and setters.

6.4: External functions added.

7: Abstract classes must now be declared explicitly.

7.1.1: Eliminate negate in favor of overloaded unary minus. Eliminate equals in favor of operator ==.

7.3: Setter syntax no longer includes = sign after the name.

7.4: Clarify that getters and setters may be abstract. Eliminate abstract modifier for abstract members. Added static warning if abstract class has abstract member.

7.5: Instance variables can be initialized to non-constants. Moved discussion of implicit getters and setters to section 5.

7.6.1: Clarify that finals can only be set once.

7.6.2: Added redirecting factories.

7.8: Moved discussion of implicit getters and setters to section 5.

8: Eliminated interface declarations.

11.1: Refined definition constant identity/caching.

11.8: Stack traces moved to 12.9.

11.11.1:  Creating an instance via new using an undefined class or constructor is a dynamic error and a static warning, not a compile-time error. Evaluation rules to allow instance variables with non-constant initializers. Instantiating an abstract class via a generative constructor is now a dynamic error.

11.22: Replaced === with built-in function identical(). Equality does not special case identity.

11.30: Type names have meaning as expressions. assert is no longer a built-in identifier, but export, import, library and part are. Removed warning on use of built-in identifiers as variable/function names.

11.32: Type cast expression accepts null.

11.33: Added operation to determine if optional argument was actually passed.

12.4: Added section on local functions.

12.8: Allow switch on compile-time constants of any one type.

12.9: Revised syntax for catch clauses. Specification of stack traces moved from 12.14.

12:10: Made explicit that checked mode tests returns.

12.14: assert is now a reserved word.

13: Revised library syntax and semantics.

14.7: Renamed Dynamic to dynamic.

15.1.1: assert is a reserved word.

##### Changes since Version 0.11

2:  Moved discussion of tokenization to section 15.

11.3: Removed unary plus.

13: Refined definition of show and hide to handle getters and setters in pairs.

14.5: All function types are subtypes of Function.

15:  Added discussion of tokenization.

15.1.1:  Defined meaning of reserved word.

15.1.2: Added initial discussion of dartdoc comments.

##### Changes since Version 0.12

7: Removed abstract modifier from grammar of members.

7.9: Added discussion of new with clause for mixins.

9: Added tentative specification of mixins.

12.1: Revised description of identical() so that it always returns true on numbers with the same class and value.

12.2: Invocations on null throw NoSuchMethodError only.

12.8: throw of null raises NullThrownError.

12.11: Using undefined types as type arguments to a constructor is a dynamic failure.

13: Added discussion of compilation units.  Allow name clause to be optional for libraries, allow export clauses in scripts. Support fully qualified name for libraries. Allow different imports to share same prefix.

15.3.1: Typedefs for mixin applications added.

15.4: Add effects of with clause on subtyping.

15.8.2: Corrected definition of LUBs.

16.1: Added keyword with.

##### Changes since Version 0.20

7.2: Getters cannot return void.

7.10: Added requirement that superclass cannot appear in implements clause.

8.1.1: Clarified status of getter/setter vs. method conflicts in multiple interface inheritance.

12.8: Separated throw into throw e and rethrow.

12.11.1: Arguments to constructor are evaluated before allocating new object.

12.14.3: Calling and undeclared method in a static context is a dynamic error (and a static warning) not a compile-time one.

12.15.1, 12.15.4, 12.17, 12.18: Refined description of InvocationMirror instance passed to noSuchMethod().

16.1: Abandoned requirement for Unicode Normalization Form C.

##### Changes since Version 0.30

7.9, 7.10: Clarified that super types with wrong number of type arguments cause a compilation error.

12.11: Constructors may not be passed type misconstructed type arguments.

15.8: Defined misconstructed parameterized types.

##### Changes Since Version 0.31

7.4: Abstract methods act as pure declarations.

9: Mixin application has forwarding constructors for all superclass constructors.

12.11, 12.11.1: Revised so that wrong generic arity is a runtime error with static warning.

12.31, 12.32:  Eliminated special treatment of malformed types.

13.10: Rethrow is a statement, not an expression.

14: Semicolon required after top-level external declarations.

14.1: Removed requirement that library names be unique within an isolate.

15.2: Revised definition of malformed types. Revised behavior of checked and production modes wrt to malformed and malbounded types. Refined assignability rules wrt new function type subtyping.

15.4: Corrected oversight wrt type variables and Object.

15.5: Liberalized function subtyping rules. Special assignability rules.

15.8: out with misconstructed types, in with malbounded ones.

##### Changes Since Version 0.40

12: Argument definition test construct dropped.

12.7: Map literals can be keyed by arbitrary expressions.

14:  Implicitly named libraries are named ''. Script tags are allowed in all libraries. It is a warning to import or export two libraries with the same name.

### Notation

We distinguish between normative and non-normative text. Normative text defines the rules of Dart. It is given in this font (black Arial 11pt). At this time, non-normative text includes:

1.  Rationale. Discussion of the motivation for language design decisions appears in blue italics.  Distinguishing normative from non-normative helps clarify what part of the text is binding and what part is merely expository.
2. Commentary, given in green (Arial, 11pt) . Comments such as “The careful reader will have noticed that the name Dart  has four characters serve to illustrate or clarify the specification, but are redundant with the normative text.  The difference between commentary and rationale can be subtle. Commentary is more general than rationale, and may include illustrative examples or clarifications.
3. Open questions (in red). Open questions are points that are unsettled in the mind of the author(s) of the specification; expect them  (the questions, not the authors; precision is important in a specification) to be eliminated in the final specification. Should the text at the end of the previous bullet be rationale or commentary?

Reserved words and built-in identifiers appear in this font.

Examples would be switch or class.

Grammar productions are given in a common variant of EBNF. The lefthand side of a production ends with a colon. On the right hand side, alternation is represented by vertical bars, and sequencing by spacing.  As in PEGs, alternation gives priority to the left.  Optional elements of a production are suffixed by a question mark like so: anElephant? . Appending a star to an element of a production means it may be repeated zero or more times. Appending a plus sign to a production means it occurs one or more times. Parentheses are used for grouping. Negation  is represented by prefixing an element of a production with a tilde. Negation is similar to the not combinator of PEGs, but it consumes input if it matches. In the context of a lexical production it consumes a single character if there is one;  otherwise, a single token if there is one.

An example would be:

AProduction:

AnAlternative

| AnotherAlternative

|  OneThing After Another

| ZeroOrMoreThings*

| OneOrMoreThings+

| AnOptionalThing?

| (Some Grouped Things)

| ~NotAThing

| A_LEXICAL_THING

;

Both syntactic and lexical productions are represented this way. Lexical productions are distinguished by their names. The names of lexical productions consist exclusively of upper case characters and underscores. As always, within grammatical productions, whitespace and comments between elements of the production are implicitly ignored unless stated otherwise.

Productions are embedded, as much as possible, in the discussion of the constructs they represent. Punctuation tokens appear in quotes.

A list x1,..., xn denotes any list of n elements of the form xi, 1 <= i <= n. Note that n may be zero, in which case the list is empty. We use such lists extensively throughout this specification.

The notation [x1, ..., xn/y1, ..., yn]E  denotes a copy of E in which all occurrences of yi, 1 <= i <= n have been replaced with xi.

We sometimes abuse list or map literal syntax, writing [o1, ..., on]  (respectively {k1: o1, ..., kn: on}) where the oi and ki may be objects rather than expressions. The intent is to denote a list (respectively map) object whose elements are the oi (respectively, whose keys are the ki and values are the oi).

The specifications of operators often involve statements such as x op y is equivalent to the method invocation x.op(y). Such specifications should be understood as a shorthand for:

x op y is equivalent to the method invocation x.op(y), assuming the class of x actually declared a non-operator method named op’ defining the same function as the operator op. This circumlocution is required because x.op(y), where op is an operator, is not legal syntax. However, it is painfully verbose, and we prefer to state this rule once here, and use a concise and clear notation across the specification.

When the specification refers to the order given in the program, it means the order of the program source code text, scanning left-to-right and top-to-bottom.

References to otherwise unspecified names of program entities (such as classes or functions) are interpreted as the names of members of the Dart core library.

Examples would be the classes Object and Type representing the root of the class hierarchy and the reification of runtime types respectively.

## Overview

Dart is a class-based, single-inheritance, pure object-oriented programming language. Dart is optionally typed and supports reified generics.  The runtime type of every object is represented as an instance of class Type which can be obtained by calling the getter  runtimeType declared in class Object, the root of the Dart class hierarchy.

Dart programs can be statically checked. The static checker will report some violations of the type rules, but such violations do not abort compilation or preclude execution.

Dart programs may be executed in one of two modes: production mode or checked mode. In production mode, static type annotations have absolutely no effect on execution.  In checked mode, assignments are dynamically checked, and certain violations of the type system raise exceptions at run time.

The coexistence between optional typing and reification is based on the following:

1. Reified type information reflects the types of objects at runtime and may always be queried by dynamic typechecking constructs (the analogs of instanceOf, casts, typecase etc. in other languages). Reified type information includes class declarations, the the runtime type (aka class) of of an object, and type arguments to constructors.
2. Static type annotations determine the types of variables and function declarations (including methods and constructors).
3. Production mode respects optional typing. Static type annotations do not affect runtime behavior.
4. Checked mode utilizes static type annotations and dynamic type information aggressively yet selectively to provide early error detection during development.

Dart programs are organized in a modular fashion into units called libraries. Libraries are units of encapsulation and may be mutually recursive.

However they are not first class.  To get multiple copies of a library running simultaneously, one needs to spawn an isolate.

### Scoping

A namespace is a mapping of identifiers to declarations.  Let NS be a namespace. We say that a name n is in NS if n is a key of NS. We say a declaration d is in NS if a key of NS maps to d.

A scope S0 induces a namespace NS0 that maps the simple name of each variable, type or function declaration d declared in S0 to d.  Labels are not included in the induced namespace of a scope; instead they have their own dedicated namespace.

It is therefore impossible, e.g.,  to define a class that declares a method and a field with the same name in Dart. Similarly one cannot declare a top level function with the same name as a library variable or class.

It is a compile-time error if there is more than one entity with the same name declared in the same scope.

In some cases, the name of the declaration differs from the identifier used to declare it.  Setters have names that are distinct from the corresponding getters because they always have an = automatically added at the end, and unary minus has the special name unary-.

Dart is lexically scoped.   Scopes may nest. A name or declaration d  is available in scope S if d  is in the namespace induced by S or if d  is available in the lexically enclosing scope of S. We  say that a name or declaration d  is in scope if d  is available in the current scope.

If a  declaration d named n is in the namespace induced by a scope S, then d hides any declaration named n that is available in the lexically enclosing scope of S.

Names may be introduced into a scope by  declarations within the scope or by other mechanisms such as imports or inheritance.

The interaction of lexical scoping and inheritance is a subtle one.  Ultimately, the question is whether lexical scoping takes precedence over inheritance or vice versa. Dart chooses the former.

Allowing inherited names to take precedence over locally declared names can create unexpected situations as code evolves. Specifically, the behavior of code in a subclass can change without warning  if a new name is introduced in a superclass.  Consider:

library ‘L1’;

class S {}

library ‘L2’;

import ‘L1.dart’;

foo() => 42;

class C extends S{ bar() => foo();}

Now assume a method foo() is added to S.

library ‘l1’;

class S {foo() => 91;}

If inheritance took precedence over the lexical scope, the behavior of C would change in an unexpected way. Neither the author of S nor the author of C are necessarily aware of this. In Dart, if there is a lexically visible method foo(), it will always be called.

Now consider the opposite scenario. We start with a version of S that contains foo(), but do not declare foo() in library L2.  Again, there is a change in behavior - but the author of L2 is the one who introduced the discrepancy that effects their code, and the new code is lexically visible. Both these factors make it more likely that the problem will be detected.

These considerations become even more important if one introduces constructs such as nested classes, which might be considered in future versions of the language.

Good tooling should of course endeavor to inform programmers of such situations (discretely). For example, an identifier that is both inherited and lexically visible could be highlighted (via underlining or colorization). Better yet, tight integration of source control with language aware tools would detect such changes when they occur.

### Privacy

Dart supports two levels of privacy: public and private. A declaration is private if its name begins with an underscore (the _ character) otherwise it is public.

A declaration m is accessible to library L if m is declared in L or if m is public.

Private declarations may only be accessed within the library in which they are declared.

Privacy is, at this point, a static notion tied to a particular piece of code (a library). It is designed to support software engineering concerns rather than security concerns. Untrusted code should always run in an another isolate.  It is possible that libraries will become first class objects and privacy will be a dynamic notion tied to a library instance.

Privacy is indicated by the name of a declaration - hence privacy and naming are not orthogonal. This has the advantage that both humans and machines can recognize access to private declarations at the point of use without knowledge of the context from which the declaration is derived.

### Concurrency

Dart code is always single threaded. There is no shared-state concurrency in Dart. Concurrency is supported via actor-like entities called isolates.

An isolate is a unit of concurrency. It has its own memory and its own thread of control. Isolates communicate by message passing. No mutable state is ever shared between isolates. Isolates are created by spawning.

### Errors and Warnings

This specification distinguishes between several kinds of errors.

Compile-time errors are errors that preclude execution. A compile time error must be reported by a Dart compiler before the erroneous code is executed.

A Dart implementation has considerable freedom as to when compilation takes place. Modern programming language implementations often interleave compilation and execution, so that compilation of a method may be delayed, e.g.,  until it is first invoked. Consequently, compile-time errors in a method m may be reported as late as the time of m’s first invocation.

As a web language, Dart is often loaded directly from source, with no intermediate binary representation. In the interests of rapid loading, Dart implementations may choose to avoid full parsing of method bodies, for example. This can be done by tokenizing the input and checking for balanced curly braces on method body entry. In such an implementation, even syntax errors will be detected only when the method needs to be executed, at which time it will be compiled (JITed).

In a development environment a compiler should of course report compilation errors eagerly so as to  best serve the programmer.

If a compile-time error occurs within the code of a running isolate A, A is immediately suspended.

Typically, A will then be terminated. However, this depends on the overall environment. A Dart engine runs in the context of an embedder, a program that interfaces between the engine and the surrounding computing environment. The embedder will often be a web browser, but need not be; it may be a C++ program on the server for example. When an isolate fails with a compile-time error as described above, control returns to the embedder, along with an exception describing the problem.  This is necessary so that the embedder can clean up resources etc. It is then the embedder’s decision whether to terminate the isolate or not.

Static warnings are those warnings reported by the static checker. They have no effect on execution. Many, but not all, static warnings relate to types, in which case they are known as static type warnings. Static warnings must be provided by Dart compilers used during development.

Dynamic type errors are type errors reported in checked mode.

Run time errors are exceptions raised during execution. Whenever we say that an exception ex is raised or thrown, we mean that  a throw expression of the form: throw ex; was implicitly evaluated. When we say that a C is thrown, where C is an exception class, we mean that an instance of class C is thrown.

If an uncaught exception is thrown by a running isolate A, A is immediately suspended.

## Variables

Variables are storage locations in memory.

variableDeclaration:

declaredIdentifier (',' identifier)*
;

declaredIdentifier:

;

finalConstVarOrType:

final type?

| const type?
| varOrType
;

varOrType:

var
|
type
;

initializedVariableDeclaration:

declaredIdentifier ('=' expression)? (',' initializedIdentifier)*
;

initializedIdentifier:

identifier ('=' expression)?
;

initializedIdentifierList:

initializedIdentifier (',' initializedIdentifier)*
;

A variable that has not been initialized has the initial value null.

A static variable is a variable that is not associated with a particular instance, but rather with an entire library or class.  A variable declared at the top-level of a library is referred to as either a library variable or simply as a top-level variable.

Static variable declarations are initialized lazily. When a static variable v is read, iff it has not yet been assigned, it is set to the result of evaluating its initializer. The precise rules are given in sections 5.1 and 11.30.

The lazy semantics are given because we do not want a language where one tends to define expensive initialization computations, causing long application startup times. This is especially crucial for Dart, which must support the coding of client applications.

A final variable is a variable whose declaration includes the modifier final.

It is a compile-time error if a final instance variable that has been initialized at its point of declaration  is  also initialized in a constructor. It is a compile-time error if a final instance variable that has is initialized by means of an initializing formal of a constructor is  also initialized elsewhere in the same constructor.

It is a compile-time error if a library, static or local variable v is final and v is not initialized at its point of declaration.

Attempting to assign to a final variable elsewhere will cause a NoSuchMethodError to be thrown, because no setter is defined for it. The assignment will also give rise to a static warning for the same reason.

A constant variable is a variable whose declaration includes the modifier const. A constant variable is always implicitly final. A constant variable must be initialized to a compile-time constant or a compile-time error occurs.

If a variable declaration does not explicitly specify a type, the type of the declared variable(s) is dynamic, the unknown type.

A library variable is implicitly static. It is a compile-time error to preface a top level variable declaration with the built-in identifier static.

Variable declarations always induce implicit getters. Mutable variables also induce implicit setters. The scope into which the implicit getters and setters are introduced depends on the kind of variable declaration involved.

A variable declaration of one of the forms T v;, T v = e;, const T v = e;, final T v;   or final T v = e; always induces an implicit getter function with signature

T get v

whose invocation evaluates as described below.

A variable declaration of one of the forms var v;, var v = e;, const  v = e;,  final v;   or final v = e; always induces an implicit getter function with signature

get v

whose invocation evaluates as described below.

A non-final variable declaration of the form T v; or the form T v = e; always induces an implicit setter function with signature

void set v=(T x)

whose execution sets the value of v to the incoming argument x.

A non-final variable declaration of the form var v; or the form var v = e; always induces an implicit setter function with signature

set v=(x)

whose execution sets the value of v to the incoming argument x.

### Evaluation of Implicit  Variable Getters

Let d be the declaration of a variable v.  If d is a local or instance variable, then the invocation of the implicit getter of v evaluates to the value stored in v.

If d is a static or library variable then the implicit getter method of v executes as follows:

1. If d is of one of the forms var v = e; , static T v = e; , static final v = e;  or final T v = e; and no value has yet been stored into v then the initializer expression e is evaluated. If, during the evaluation of e, the getter for v is referenced, a CyclicInitializationError is thrown. If the evaluation succeeded yielding an object o, let r = o, otherwise let r = null. In any case, r is stored into v. The result of executing the getter is r.
2. If d is of one of the forms const v = e;  or const T v = e; the result of the getter is the value of the compile time constant e. Otherwise
3. The result of executing the getter method is the value stored in v.

## Functions

Functions abstract over executable actions.

functionSignature:

metadata returnType? identifier formalParameterList
;

returnType:

void
|
type

;

functionBody:

'=>' expression ';'
|
block
;

block:
'{'
statements '}'
;

Functions include  function declarations, methods, getters, setters and function literals.

All functions have a signature and a body. The signature describes the formal parameters of the function, and possibly its name and return type. The body is a block statement containing the statements executed by the function. A function body of the form  => e is equivalent to a body of the form {return e;}.

If the last statement of a function is not a return statement, the statement return null; is implicitly appended to the function body.

Because Dart is optionally typed, we cannot guarantee that a function that does not return a value will not be used in the context of an expression. Therefore, every function must return a value. See the discussion around the return statement.

### Function Declarations

A function declaration is a function that is not a method, getter, setter or function literal. Function declarations include library functions, which are function declarations at the top level of a library, and local functions, which are functions declarations declared inside other functions.  Library functions are often referred to simply as top-level functions.

A function declaration consists of an identifier indicating the function's name, followed by a signature and body.

The scope of a library function is the scope of the enclosing library. The scope of a local function is  described below. In both cases, the name of the function is in scope in the formal parameters scope of the function.

It is a compile-time error to preface a function declaration with the built-in identifier static.

### Formal Parameters

Every function includes a formal parameter list, which consists of a list of required positional parameters, followed by any optional parameters. The optional parameters may be specified either as a set of named parameters or as a list of positional parameters, but not both.

The formal parameter list of a function introduces a new scope known as the functions formal parameter scope. The formal parameter scope of a function f  is enclosed in the scope where f is declared.

The body of a function introduces a new scope known as the functions body scope. The body of a function f  is enclosed  in the scope introduced by the formal parameter scope of f. Every formal parameter introduces a local variable into the formal parameter scope.

It is a compile-time error if a formal parameter is declared as a constant variable.

formalParameterList:

'(' ')'

| '(' normalFormalParameters ( ‘,’ optionalFormalParameters)? ')'

|  '(' optionalFormalParameters ')'
;

normalFormalParameters:
normalFormalParameter (',' normalFormalParameter)*
;

optionalFormalParameters:
optionalPositionalFormalParameters |

namedFormalParameters
;

optionalPositionalFormalParameters:
'[' defaultFormalParameter (',' defaultFormalParameter)* ']'
;

namedFormalParameters:
'{'
defaultNamedParameter(',' defaultNamedParameter)* '}'
;

#### Required Formals

A required formal parameter is a simple variable declaration.

normalFormalParameter:

functionSignature
|
fieldFormalParameter
|
simpleFormalParameter
;

simpleFormalParameter:

declaredIdentifier
|
;

fieldFormalParameter:

metadata finalConstVarOrType? this '.' identifier
;

#### Optional Formals

Optional parameters may be specified and provided with default values.

defaultFormalParameter:

normalFormalParameter ('=' expression)?
;

defaultNamedParameter:

normalFormalParameter (':' expression)?
;

It is a compile-time error if the default value of an optional parameter is not a compile-time constant.

If no default is explicitly specified for an optional parameter an implicit default of null is provided.

It is a compile-time error if the name of a named optional parameter begins with an ‘_’ character.

The need for this restriction is a direct consequence of the fact that naming and privacy are not orthogonal. If we allowed named parameters to begin with an underscore, they would be considered private and inaccessible to callers from outside the library where it was defined. If a method outside the library overrode a method with a private optional name, it would not be a subtype of the original method. The static checker would of course flag such situations, but the consequence would be that adding a private named formal would break clients outside the library in a way they could not easily correct.

### Type of a Function

If a function does not declare a return type explicitly, its return type is dynamic.

Let F be a function with required formal parameters T1 p1, …, Tn pn., return type T0 and no optional parameters. Then the type of F is

(T1 , …, Tn) → T0.

Let F be a function with required formal parameters T1 p1, …, Tn pn., return type T0 and positional optional parameters Tn+1 pn+1, . . . , Tn+k pn+k. Then the type of F is

(T1 , …, Tn, [Tn+1, …,Tn+k]) → T0.

Let F be a function with required formal parameters T1 p1, …, Tn pn., return type T0 and named optional parameters Tn+1 pn+1, . . . , Tn+k pn+k. Then the type of F is

(T1 , …, Tn, {pn+1:Tn+1, …, pn+k:Tn+k}) → T0.

### External Functions

An external function is a function whose body is provided separately from its declaration. An external function may be a top-level function, a method, a getter, a setter or a non-redirecting constructor. External functions are introduced via the built-in identifier external followed by the function signature.

External functions allow us to introduce  type information  for code that is not statically known to the Dart compiler.

Examples of external functions might be foreign functions (defined in C, or Javascript etc.), primitives of the implementation (as defined by the Dart runtime), or code that was dynamically generated but whose interface is statically known. However, an abstract method is different from an external function, as it has no body.

An external function is connected to its body by an implementation specific mechanism. Attempting to invoke an external function that has not been connected to its body will raise a NoSuchMethodError or some subclass thereof.

The actual syntax is given in the sections on classes and libraries below.

## Classes

A class defines the form and behavior of a set of objects which are its instances.

classDefinition:

metadata abstract? class identifier typeParameters? (superclass mixins?)? interfaces?
'{' (
;

mixins:
with typeList
;

Mixin clauses are not yet supported.

classMemberDefinition:

declaration ';'
|
methodSignature functionBody
;

methodSignature:

| static? setterSignature

| operatorSignature
|
functionSignature
|
static (final | const) type? staticFinalDeclarationList

| final type? initializedIdentifierList
|
static? (var | type) initializedIdentifierList
;

staticFinalDeclarationList:
:
staticFinalDeclaration (',' staticFinalDeclaration)*
;

staticFinalDeclaration:

identifier '=' expression
;

A class has constructors,  instance members and static members. The instance members of a class are its instance methods, getters, setters and instance variables. The static members of a class are its static methods, getters, setters and static variables.  The members of a class are its static and instance members.

Every class has a single superclass except class Object which has no superclass. A class may implement a number of interfaces by declaring them in its implements clause.

An abstract class is a class that is explicitly declared with the abstract modifier, either by means of a class declaration or via a type alias for a mixin application.

We want different behavior for concrete classes and abstract classes. If A is intended to be abstract, we want the static checker to warn about any attempt to instantiate A, and we do not want the checker to complain about unimplemented methods in A. In contrast, if A is intended to be concrete, the checker should warn about all unimplemented methods, but allow clients to instantiate it freely.

The interface of class C is an implicit interface that declares instance members that correspond to the instance members declared by C, and whose direct superinterfaces are the direct superinterfaces of C. When a class name appears as a type or interface, that name denotes the interface of the class.

It is a compile-time error if a class declares two members of the same name.
It is a compile-time error if a class has an instance member and a static member with the same name.

Here are simple examples, that illustrate the difference between “has a member” and “declares a member”. For example, B declares one member named j, but has two such members - one it declares and one it inherits. The rules of inheritance determine what members a class has.

class A {

var i = 0;

var j;

f(x) => 3;

}

class B extends A {

int i = 1; // getter i and setter i= override versions from A

static j; // compile-time error: static getter & setter conflict with instance getter & setter

static f(x) => 3; // compile-time error: static method conflicts with instance method

}

It is a compile time error if a class C declares a member with the same name as C.

### Instance Methods

Instance methods are functions whose declarations are immediately contained within a class declaration and that are not declared static. The instance methods of a class C are those instance methods declared by C and the instance methods inherited by C from its superclass.

It is a compile-time error if an instance method m1 overrides an instance member m2 and  m1 has a different number of required parameters than m2. It is a compile-time error if an instance method m1 overrides  an instance member m2 and  m1 has fewer optional positional parameters than m2. It is a compile-time error if an instance method m1 overrides  an instance member m2 and  m1 does not declare all the named parameters declared by m2.

It is a static warning if an instance method m1 overrides an instance method m2 and the type of m1 is not a subtype of the type of m2. It is a static warning if an instance method m1  overrides an instance member m2,  the signature of m2 explicitly specifies a default value for a formal parameter p and the signature of m1 specifies a different default value for p. It is a static warning if a class C declares an instance method named n and an accessible static member named n is declared in a superclass of C.

#### Operators

Operators are instance methods with special names.

operatorSignature:

returnType? operator operator formalParameterList
;

operator:
'~'
|
binaryOperator
| '[' ']'
| '[' ']' '='

;

binaryOperator:

multiplicativeOperator
|
|
shiftOperator
|
relationalOperator
|’==’
|
bitwiseOperator
;

An operator declaration is identified with built-in identifier operator.

The following names are allowed for user-defined operators: <, >, <=, >=, ==, -, +, /, ~/,  *, %, |, ^, &, <<, >>, []=, [], ~.

It is a compile-time error if the arity of the user-declared operator []= is not 2. It is a compile time error if the arity of a user-declared operator with one of the names:  <, >, <=, >=, ==, +, /, ~/, *, %, |, ^, &, <<, >>, [] is not 1. It is a compile time error if the arity of the user-declared operator - is not 0 or 1.

The - operator is unique in that two overloaded versions are permitted. If the operator has no arguments, it denotes unary minus. If it has an argument, it denotes binary subtraction.

The name of the unary operator - is unary-.

This device allows the two methods to be distinguished for purposes of method lookup, override and reflection.

It is a compile time error if the arity of the user-declared operator ~ is not 0.

It is a compile-time error to declare an optional parameter in an operator.

It is a static warning if the return type of the user-declared operator []= is explicitly declared  and not void.

### Getters

Getters are functions that are used to retrieve the values of object properties.

getterSignature:

type? get identifier

If no return type is specified, the return type of the getter is dynamic.

A getter definition that is prefixed with the static modifier defines a static getter. Otherwise, it defines an instance getter. The name of the getter is given by the identifier in the definition.

The instance getters of a class C are those instance getters declared by C and the instance getters inherited by C from its superclass. The static getters of a class C are those static getters declared by C.

It is a compile-time error if a class has both a getter and a method with the same name. This restriction holds regardless of whether the getter is defined explicitly or implicitly, or whether the getter or the method are inherited or not.

This implies that a getter can never override a method, and a method can never override a getter or field.

It is a static warning if a getter m1 overrides a getter m2 and the type of m1 is not a subtype of the type of m2.  It is a static warning if a class declares a static getter named g and also has a non-static setter named v=. It is a static warning if a class C declares an instance getter named n and an accessible static member named v or v= is declared in a superclass of C.

### Setters

Setters are functions that are used to set the values of object properties.

setterSignature:

returnType? set identifier  formalParameterList

;

If no return type is specified, the return type of the setter is dynamic.

A setter definition that is prefixed with the static modifier defines a static setter. Otherwise, it defines an instance setter. The name of a setter is  obtained by appending the  string ='  to the identifier given in its signature.

Hence, a setter name can never conflict with, override or be overridden by a getter or method.

The instance setters of a class C are those instance setters declared by C and the instance setters inherited by C from its superclass. The static setters of a class C are those static setters declared by C.

It is a compile-time error if a setter’s formal parameter list does not include exactly one required  formal parameter p. We could enforce this via the grammar, but we’d have to specify the evaluation rules in that case.

It is a static warning if a setter declares a return type other than void. It is a static warning if a setter m1 overrides a setter m2 and the type of m1 is not a subtype of the type of m2. It is a static warning if a class has a setter named v= with argument type T and a getter named v with return type S, and T may not be assigned to S. It is a static warning if a class  declares a static setter named v= and also has a non-static member named v. It is a static warning if a class C declares an instance setter named v= and an accessible static member named v= or v is declared in a superclass of C.

#### Abstract Instance Members

An abstract method (respectively abstract getter or abstract setter) is an instance method, getter or setter that is not declared external and does not provide an implementation.

Earlier versions of Dart required that abstract members be identified by prefixing them with an the modifier abstract. The elimination of this requirement is motivated by the desire to use abstract classes as interfaces. Every Dart class induces an implicit interface.

Using an abstract class instead of an interface declaration has important advantages. An abstract class can provide default implementations; it can also provide static methods, obviating the need for service classes such as Collections or Lists, whose entire purpose is to group utilities related to a given type.

Eliminating the requirement for an explicit modifier on members makes abstract classes more concise, making abstract classes an attractive substitute for interface declarations.

Invoking an abstract method, getter or setter always results in an invocation of  noSuchMethod() exactly as if the declaration did not exist, unless a suitable member a is available in a superclass, in which case a is invoked.

The purpose of an abstract method is to provide a declaration for purposes such as type checking and reflection. In classes used as mixins, it is often useful to introduce such declarations for methods that the mixin expects will be provided by the superclass the mixin is applied to.

It is a static warning if an abstract member is declared or inherited in a concrete class.

### Instance Variables

Instance variables are variables whose declarations are immediately contained within a class declaration and that are not declared static. The instance variables of a class C are those instance variables declared by C and the instance variables inherited by C from its superclass.

### Constructors

A constructor is a special member that is used in instance creation expressions (instanceCreation) to produce objects. Constructors may be generative or they may be factories.

A constructor name always begins with the name of its immediately enclosing class, and may optionally be followed by a dot and an identifier id. It is a compile-time error if id is the name of a member declared in the immediately enclosing class. It is a compile-time error if the name of a constructor is not a constructor name.

Iff no constructor is specified for a class C, it implicitly has a default constructor C() : super() {}, , unless C is class Object.

#### Generative Constructors

A generative constructor consists of a constructor name, a constructor parameter list, and either a redirect clause or an initializer list and an optional body.

constructorSignature:

identifier ('.' identifier)? formalParameterList
;

A constructor parameter list is a parenthesized, comma-separated list of formal constructor parameters. A formal constructor parameter is either a formal parameter or an initializing formal. An initializing formal has the form this.id.  It is a compile-time error if id is not the name of an instance variable of the immediately enclosing class. It is a compile-time error if an initializing formal is used by a function other than a non-redirecting generative constructor.

If an explicit type is attached to the initializing formal, that is its static type. Otherwise, the type of an initializing formal named id is Tid, where Tid is the type of the field named id in the immediately enclosing class.  It is a static warning if the static type of id is not assignable to Tid.

Using an initializing formal this.id in a formal parameter list does not introduce a formal parameter name into the scope of the constructor. However, the initializing formal does effect the type of the constructor function exactly as if a formal parameter named id of the same type were introduced in the same position.

Initializing formals are executed during the execution of generative constructors detailed below. Executing an initializing formal this.id causes the field id of the immediately surrounding class to be assigned the value of the corresponding actual parameter.

The above rule allows initializing formals to be used as optional parameters:

class A {

int x;

A([this.x]);

}

is legal, and has the same effect as

class A {

int x;

A([int x]): this.x = x;

}

sc

A fresh instance is an instance whose identity is distinct from any previously allocated instance of its class. A generative constructor always operates on a fresh instance of its immediately enclosing class. The above holds if the constructor is actually run, as it is by new. If a constructor c is referenced by const, c may not be run; instead, a canonical object may be looked up. See the section on instance creation.

If a generative constructor c  is not a redirecting constructor and no body is provided, then c implicitly has an empty body {}.

##### Redirecting Constructors

A generative constructor may be redirecting, in which case its only action is to invoke another generative constructor. A redirecting constructor has no body; instead, it has a redirect clause that specifies which constructor the invocation is redirected to, and with what arguments.

redirection:
':'
this ('.' identifier)? arguments
;

##### Initializer Lists

An initializer list begins with a colon, and consists of a comma-separated list of individual initializers. There are two kinds of initializers.

1. A superinitializer identifies a superconstructor - that is, a specific constructor of the superclass.  Execution of the superinitializer causes the initializer list of the superconstructor to be executed.
2. An instance variable initializer assigns a value to an individual instance variable.

initializers:
':'
superCallOrFieldInitializer (',' superCallOrFieldInitializer)*
;

superCallOrFieldInitializer:

super arguments
|
super '.' identifier arguments
|
fieldInitializer
;

fieldInitializer:
(
this '.')? identifier '=' conditionalExpression cascadeSection*
;

Let k be a generative constructor. Then k may include at most one superinitializer in its initializer list or a compile time error occurs.  If no superinitializer is provided, an implicit superinitializer of the form super() is added at the end of k’s  initializer list, unless the enclosing class is class Object. It is a compile time error if more than one initializer corresponding to a given instance variable appears in k’s list. It is a compile time error if k’s  initializer list contains an initializer for a variable that is initialized by means of an initializing formal of k. It is a compile time error if k’s  initializer list contains an initializer for a final variable f whose declaration includes an initialization expression.

Each final instance variable f declared in the immediately enclosing class must have an initializer in k's initializer list unless it has already been initialized by one of the following means:

1. Initialization at the declaration of f.
2. Initialization by means of an initializing formal of k.

or a compile-time error occurs. It is a compile-time error if k's initializer list contains an initializer for a variable that is not an instance variable declared in the immediately surrounding class.

The  initializer list may of course contain an initializer for any instance variable declared by the immediately surrounding class, even if it is not final.

It is a compile-time error if a generative constructor of class Object includes a superinitializer.

Execution of a generative constructor k is always done with respect to a set of bindings for its formal parameters and with this bound to a fresh instance i and the type parameters of the immediately enclosing class bound to a set of actual type arguments V1, ... , Vm.

These bindings are usually determined by the instance creation expression that invoked the constructor.  However, they may also be determined by a reflective call, or by a call from another (redirecting) constructor.

If k is redirecting, , then its redirect clause has the form this.g(a1, …, an, xn+1: an+1, …, xn+k: an+k) where g identifies another generative constructor of the immediately surrounding class. Then execution of k proceeds by evaluating the argument list (a1, …, an, xn+1: an+1, …, xn+k: an+k), and then executing g with respect to the bindings resulting from the evaluation of (a1, …, an, xn+1: an+1, …, xn+k: an+k) and with this bound to i and the type parameters of the immediately enclosing class bound to V1, ... , Vm.

Otherwise, execution  proceeds as follows:

Any initializing formals declared in k's parameter list are executed in the order they appear in the program.  Then, k's initializers are executed in the order they appear in the program.

We could observe the order by side effecting external routines called. So we need to specify the order.

After all the initializers have completed, the body of k is executed in a scope where this is bound to i. Execution of the body begins with execution of the body of the superconstructor with this bound to i, the type parameters of the immediately enclosing class bound to a set of actual type arguments V1, ... , Vm  and the formal parameter bindings determined by the argument list of the superinitializer of k.

This process ensures that no uninitialized final field is ever seen by code. Note that this  is not in scope on the right hand side of an initializer so no instance method can execute during initialization: an instance method cannot be directly invoked, nor can this be passed into any other code being invoked in the initializer.

Execution of an initializer of the form this.v = e proceeds as follows:

First, the expression e is evaluated to an object o. Then, the instance variable v of the object denoted by this is bound to o. In checked mode, it is a dynamic type error if o is not null and the interface of the class of o is not a subtype of the actual type of the field v.

An initializer of the form v = e is equivalent to an initializer of the form this.v = e.

Execution of a superinitializer of the form super(a1, …, an, xn+1: an+1, …, xn+k: an+k) (respectively super.id(a1, …, an, xn+1: an+1, …, xn+k: an+k)) proceeds as follows:

First, the argument list (a1, …, an, xn+1: an+1, …, xn+k: an+k) is evaluated.

Let C be the class in which the superinitializer appears and let S be the superclass of C.  If S is generic, let U1, ,.., Um be the actual type arguments passed to S in the superclass clause of C.

Then, the initializer list of the constructor S (respectively S.id) is executed with respect to the bindings that resulted from the evaluation of the argument list,  with this bound to the current binding of this, and  the type parameters (if any) of class S bound to the current binding of U1, ,.., Um.

It is a compile-time error if class S does not declare a generative constructor named S (respectively S.id)

#### Factories

A factory is a constructor prefaced by the built-in identifier factory.

factoryConstructorSignature:

factory identifier ('.' identifier)? formalParameterList
;

The return type of a factory whose signature is of the form factory M or the form factory M.id is M if M is not a generic type; otherwise the return type is M <T1, …, Tn>, where T1, …, Tn are the type parameters of the enclosing class.

It is a static warning if M.id is not a constructor name.

It is a compile-time error if M is not the  name  of the immediately enclosing class.

In checked mode, it is a dynamic type error if a factory returns an object whose type is not a subtype of its actual return type.

It seems useless to allow a factory to return null. But it is more uniform to allow it, as the rules currently do.

Factories address classic weaknesses associated with constructors in other languages.

Factories can produce instances that are not freshly allocated: they can come from a cache. Likewise, factories can return instances of different classes.

##### Redirecting Factory Constructors

Redirecting factory constructors are only partially implemented in M3.

A redirecting factory constructor specifies a call to a constructor of another class that is to be used whenever the redirecting constructor is called.

redirectingFactoryConstructorSignature:
const? factory  identifier  ('.' identifier)?  formalParameterList =’ type ('.' identifier)?
;

Calling a redirecting factory constructor k causes the constructor k’ denoted by type (respectively, type.identifier) to be called with the actual arguments passed to k, and returns the result of k’ as the result of k. The resulting constructor call is governed by the same rules as an instance creation expression using new.

It follows that if type or type.id are not defined, or do not refer to a class or constructor, a dynamic error occurs, as with any other undefined constructor call. The same holds if k has more optional positional parameters than k’ or if k  has a named parameter that is not declared by k’.

This implies that the arguments to k are always legal arguments to k’

It is a compile-time error if a redirecting factory constructor redirects to itself, either directly or indirectly via a sequence of redirections.

If a redirecting factory F1 redirects to another redirecting factory F2 and F2 then redirects to F1, then both F1 and F2 are undefined. Such cycles are therefore illegal.

It is a static warning if type does not denote a class accessible in the current scope; if type does denote such a class C it is a static warning if the referenced constructor (be it type or type.id) is not a constructor of C.

Note that it is not possible to modify the arguments being passed to  k’. This is a deliberate decision, so that k’ can easily determine what arguments were actually passed by the caller.

At first glance, one might think that ordinary factory constructors could simply create instances of other classes and return them, and that redirecting factories are unnecessary. However, redirecting factories have several advantages:

1. An abstract class may provide a constant constructor that utilizes the constant constructor of another class.
2. A constructor to which calls are being redirected can determine whether any user arguments were explicitly passed.
3. A redirecting factory constructor avoids the need for forwarders to repeat the default values for formal parameters in their signatures.

It is a compile-time error if k is prefixed with the const modifier but k’ is not a constant constructor.

It is a static warning if the function type of k’ is not a subtype of the type of k.

This implies that the resulting object conforms to the interface of the immediately enclosing class of k.

It is a static type warning if any of the type arguments to k’ are not subtypes of the bounds of the corresponding formal type parameters of type.

#### Constant Constructors

A constant constructor may be used to create compile-time constant objects. A constant constructor is prefixed by the reserved word const.

constantConstructorSignature:

const qualified formalParameterList
;

All the work of a constant constructor must be handled via its initializers.

It is a compile-time error if a constant constructor is declared by a class that has a non-final instance variable.

The above refers to both locally declared and inherited instance variables.

Any expression that appears within the initializer list of a constant constructor must be a potentially constant expression, or a compile-time error occurs.

A potentially constant expression is an expression e that would be a valid constant expression if all formal parameters of e’s immediately enclosing constant constructor were treated as compile-time constants that were guaranteed to evaluate to an integer, boolean or string value as required by their immediately enclosing superexpression.

The difference between a potentially constant expression and a compile-time constant expression deserves some explanation.

The key issue is whether one treats the formal parameters of a constructor as compile-time constants.

If a constant constructor is invoked from a constant object expression, the actual arguments will be required to be compile-time constants. Therefore, if we were assured that constant constructors were always invoked from constant object expressions, we could assume that the formal parameters of a constructor were compile-time constants.

However, constant constructors can also be invoked from ordinary instance creation expressions, and so the above assumption is not generally valid.

Nevertheless, the use of the formal parameters of a constant constructor within the constructor is of considerable utility. The concept of potentially constant expressions is introduced to facilitate limited use of such formal parameters. Specifically, we allow the usage of the formal parameters of a constant constructor for expressions that involve built-in operators, but not for constant objects, lists and maps. This allows for constructors such as:

class C {

final x; final y; final z;

const C(p, q): x = q, y = p + 100, z = p + q;

}

The assignment to x is allowed under the assumption that q is a compile-time constant (even though q is not, in general a compile-time constant).  The assignment to y is similar, but raises additional questions. In this case, the superexpression of p is p + 100, and it requires that p be a numeric compile-time constant for the entire expression to be considered constant.  The wording of the specification allows us to assume that p evaluates to an integer. A similar argument holds for p and q in the assignment to z.

However, the following constructors are disallowed:

class D {

final w;

const D.makeList(p): w = const [p];  // compile-time error

const D.makeMap(p): w = const {“help”: p}; // compile-time error

const D.makeC(p): w = const C(p, 12); // compile-time error

}

The problem is not that the assignments to w are not potentially constant; they are.  However, all these run afoul of the rules for constant lists, maps and objects, all of which independently require their subexpressions to be constant expressions.

All of the illegal constructors of D above could not be sensibly invoked via new, because an expression that must be constant cannot depend on a formal parameter, which may or may not be constant. In contrast, the legal examples make sense regardless of whether the constructor is invoked via const or via new.

Careful readers will of course worry about cases where the actual arguments to C() are constants, but are not of appropriate type. This is precluded by the following rule, combined with the rules for evaluating constant objects.

When invoked from a constant object expression, a constant constructor must throw an exception if any of its actual parameters  is a value that would prevent one of the potentially constant expressions within it to from being a valid compile-time constant.

### Static Methods

Static methods are functions whose declarations are immediately contained within a class declaration and that are declared static. The static methods of a class C are those static methods declared by C.

Inheritance of static methods has little utility in Dart. Static methods cannot be overridden. Any required static function can be obtained from its declaring library, and there is no need to bring it into scope via inheritance. Experience shows that developers are confused by the idea of inherited methods that are not instance methods.

Of course, the entire notion of static methods is debatable, but it is retained here because so many programmers are familiar with it. Dart static methods may be seen as functions of the enclosing library.

### Static Variables

Static variables are variables whose declarations are immediately contained within a class declaration and that are declared static. The static variables of a class C are those static variables declared by C.

### Superclasses

The superclass of a class C that has a with clause with M1, …, Mk and an extends clause extends S is the application of the mixin Mk* .. * M1  to S.  If no with clause is specified then the extends clause of a class C specifies its superclass. If no extends clause is specified, then either:

1. C is Object, which has no superclass. OR
1. Class C is deemed to have an extends clause of the form extends Object, and the rules above apply.

It is a compile-time error to specify an extends clause for class Object.

superclass:

extends type
;

It is a compile-time error if the extends clause of a class C includes a type expression that does not denote a class available in the lexical scope of C.

The type parameters of a generic class are available in the lexical scope of the superclass clause, potentially shadowing classes in the surrounding scope. The following code is therefore illegal and should cause a compile-time error:

class T{}

class G<T> extends T {} // Compilation error: Attempt to subclass a type parameter

A class S is a superclass of a class C iff either:

1.  S is the superclass of C, or
2. S is a superclass of a class S’ and S’ is a superclass of C.

It is a compile-time error if a class C is a superclass of itself.

#### Inheritance and Overriding

A class C inherits any instance members of its superclass that are not overridden by instance members declared in C. It is a compile-time error if the extends clause of a class C includes a parameterized type of the form G<T1, ..., Tn> and G is not a generic class with n type parameters, either directly or as a type argument to another parameterized type.

A class may override instance members that would otherwise have been inherited from its superclass.

Let C be a class declared in library L with superclass S and let C declare an instance member m, and assume S declares an instance member m’ with the same name as m. Then m overrides m’ iff m’ is accessible to L, m has the same name as m’ and neither m nor m’ are fields.

Fields never override each other. The getters and setters induced by fields do.

Whether an override is legal or not is described elsewhere in this specification (see instance methods, getters and setters).

For example getters and setters may not legally override methods and vice versa.   Setters and methods never override each other, because their names always differ.

For convenience, here is a summary of the relevant rules. Remember that this is not normative. The controlling language is in the relevant sections of the specification.

1. There is only one namespace for getters, setters, methods and constructors (3.1). A field f introduces a getter f, and if it is not const or final, a setter f= (7.5, 7.8). When we speak of members here, we mean accessible fields, getters, setters and methods (7).
2. You cannot have two members with the same name in the same class - be they declared or inherited (3.1, 7).
3. Static members are never inherited.
4. It is a warning if you have an static member named m in your class or any superclass (even though it is not inherited) and an instance member of the same name (7.1, 7.2, 7.3).
5. It is a warning if you have a static setter v=, and an instance member v (7.3).
6. It is a warning if you have a static getter v and an instance setter v= (7.2).
7. If you define an instance member named m, and your superclass has an instance member of the same name, they override each other. This may or may not be legal.
8. If two members override each other, it is a static warning if their type signatures are not assignable to each other (7.1, 7.2, 7.3) (and since these are function types, this means the same as ”subtypes of each other”).
9. If two members override each other, it is a compile time error if they have a different number of required parameters (7.1).
10. If two members override each other, it is a compile time error if the overriding member has fewer optional positional parameters than the member being overridden (7.1).
11. If two members override each other, it is a compile time error if the overriding member does not have all the named parameters that the member being overridden has (7.1).
12. Setters, getters and operators never have optional parameters of any kind; it’s a compile-time error (7.1.1, 7.2, 7.3).
13. It is a compile-time error if a member has the same name as its enclosing class (7).
14. A class has an implicit interface (7).
15. Interface members are not inherited by a class, but are inherited by its implicit interface. Interfaces have their own inheritance rules (8.1.1).
16. A member is abstract if it has no body and is not labeled external (7.4, 6.4).
17. A class is abstract iff it is explicitly labeled abstract.
18. It is a static warning a concrete class has an abstract member (declare or inherited).
19. It is a static warning and a dynamic error to call a non-factory constructor of an abstract class (11.10.1).
20. If a class defines an instance member named m, and any of its superinterfaces have a member named m, the interface of the class overrides m.
21. An interface inherits all members of its superinterfaces that are not overridden and not members of multiple superinterfaces.
22. If multiple superinterfaces of an interface define a member with the same name m, then at most one member is inherited. That member (if it exists) is the one whose type is a subtype of all the others. If there is no such member, then:
1. A static warning is given.
2. If possible the interface gets a member named m that has the same number of required parameters as all the members in the superinterfaces, the maximal number of optional positionals, and the superset of named parameters. The types of these are all dynamic. If this is impossible (because the superinterface members differ in their number of required parameters) then no member m appears in the interface.

1. Rule 8 applies to interfaces as well as classes (8.1.1).
2. It is a static warning if a concrete class does not have an implementation for a method in any of its superinterfaces (7.10).
3. The identifier of a named constructor cannot be the same as the name of a member declared (as opposed to inherited) in the same class (7.6).

It is nevertheless convenient to define the override relation between members in this way, so that we can concisely describe the illegal cases.

Note that instance variables do not participate in the override relation, but the getters and setters they induce do. Also, getters don’t override setters and vice versa.  Finally, static members never override anything.

It is a static warning if a non-abstract class inherits an abstract method.

### Superinterfaces

A class has a set of direct superinterfaces. This set includes the interface of its superclass and the interfaces specified in the the implements clause of the class.

interfaces:

implements typeList
;

It is a compile-time error if the implements clause of a class C includes a type expression that does not denote a class available in the lexical scope of C.

In particular, one cannot inherit from a type parameter.

It is a compile-time error if the implements clause of a class includes type dynamic.

It is a compile-time error if a type T appears more than once in the implements clause of  a class. It is a compile-time error if the superclass of a class C appears in the implements clause of C.

One might argue that it is harmless to repeat a type in this way, so why make it an error? The issue is not so much that the situation described in program source is erroneous, but that it is pointless. As such, it is an indication that the programmer may very well have meant to say something else - and that is a mistake that should be called to her or his attention.  Nevertheless, we could simply issue a warning; and perhaps we should and will. That said, problems like these are local and easily corrected on the spot, so we feel justified in taking a harder line.

It is a compile-time error if the interface of a class C is a superinterface of itself.

It is a static warning if the implicit interface of a non-abstract class C includes an instance member m and C does not declare or inherit a corresponding instance member m.

A class does not inherit members from its superinterfaces. However, its implicit interface does.

## Interfaces

An interface defines how one may interact with an object. An interface has methods, getters and setters and a set of superinterfaces.

It is a compile-time error if an interface member m1 overrides an interface member m2 and  m1 has a different number of required parameters than m2. It is a compile-time error if an interface member m1 overrides  an interface member m2 and  m1 does not declare all the named parameters declared by m2 in the same order.

It is a static warning if an interface member m1 overrides an interface member m2 and the type of m1 is not a subtype of the type of m2.  It is a static warning if an interface method m1  overrides an interface method m2,  the signature of m2 explicitly specifies a default value for a formal parameter p and the signature of m1 specifies a different default value for p.

### Superinterfaces

An interface has a set of direct superinterfaces.

An interface J is a superinterface of an interface I iff either J is a direct superinterface of I or J is a superinterface of a direct superinterface of I.

It is a compile-time error if an interface is a superinterface of itself.

#### Inheritance and Overriding

Let I be the implicit interface of a class C.  I inherits any accessible instance members of its superinterfaces that are not overridden by members declared in C.

However, if there are multiple members m1, …,  mk with the same name n that would be inherited (because identically named members existed in several superinterfaces) then at most one member is inherited.

If some but not all of the mi, 1 <= i <= k, are getters, or if some but not all of the mi are setters, none of the mi are inherited, and a static warning is issued.

If the static types T1, …,  Tk of the members m1, …,  mk are not identical, then there must be a member mx such that Tx <: Ti, 1 <= x <= k for all  i, 1 <= i <=  k, or a static type warning occurs. The member that is inherited is mx, if it exists; otherwise:

1. If all of m1, …,  mk have the same number r of required parameters and the same set of named parameters s, then let h = max( numberOfOptionalPositionals( mi ) ), 1 <= i <= k.  I has a method named n, with r required parameters of type dynamic, h optional positional parameters of type dynamic, named parameters s of type dynamic and  return type dynamic.
2. Otherwise none of the members  m1, …,  mk is inherited.

The only situation where the runtime would be concerned with this would be during reflection if a mirror attempted to obtain the signature of an interface member.

The current solution is a tad complex, but is robust in the face of type annotation changes.  Alternatives: (a) No member is inherited in case of conflict. (b) The first m is selected (based on order of superinterface list) (c) Inherited member chosen at random.

(a) means that the presence of an inherited member of an interface varies depending on type signatures.  (b) is sensitive to irrelevant details of the declaration and (c) is liable to give unpredictable results between implementations or even between different compilation sessions.

## Mixins

A mixin describes the difference between a class and its superclass. A mixin may be declared directly or derived from an existing class declaration.

It is a compile-time error if a declared or derived mixin refers to super. It is a compile-time error if a declared or derived mixin explicitly declares a constructor. It is a compile-time error if a mixin is derived from a class whose superclass is not Object.

These restrictions are temporary.  We expect to remove them in later versions of Dart.

The restriction on the use of super avoids the problem of rebinding super when the mixin is bound to different superclasses.

The restriction on constructors simplifies the construction of mixin applications because the process of creating instances is simpler.

The restriction on the superclass means that the type of a class  from which a mixin is derived is always implemented by any class that mixes it in. This allows us to defer the question of whether and how to express the type of the mixin independently of its superclass and super interface types.

Reasonable answers exist for all these issues, but their implementation is non-trivial.

### Mixin Application

A mixin may be applied to a superclass, yielding a new class. Mixin application may occur  when a mixin is mixed into a class declaration via its with clause, or it may occur in the context of a type alias.

mixinApplication:
type mixins  interfaces?

;

A mixin application of the form S with M  defines a class C with superclass S.

A mixin application of the form S with M1, …, Mk  defines a class C whose superclass is the application of the mixin composition  Mk-1 * … * M1 to S.

In both cases above, C declares the same instance members as M (respectively Mk). If any of the instance fields of M (respectively Mk) have initializers, they are executed in the scope of M (respectively Mk) to initialize the corresponding fields of C.

For each generative constructor named qi(Ti1 ai1, ... , Tiki aiki), 1 <= i  <= n of S, C has an implicitly declared constructor named q'i = [C/S]qi of the form

q'i(ai1, ... , aiki):super(ai1, ... , aiki);.

Let K be a class declaration  with the same superclass and interfaces as C,  and the instance members declared by M (respectively M1, ..., Mk). It is a static warning if the declaration of K would cause a static warning.  It is a compile-time error if the declaration of K would cause a compile-time error.

If, for example, M declares an instance member im whose type is at odds with the type of a member of the same name in S, this will result in a static warning just as if we had defined K by means of an ordinary class declaration extending S with a body that included im.

It is a compile-time error if S does not denote a class available in the immediately enclosing scope. It is a compile-time error if M (respectively, any of M1, ..., Mk) does (respectively, do) not denote a class or mixin available in the immediately enclosing scope. It is a compile time error if a well formed mixin cannot be derived from M (respectively, from each of M1, ..., Mk).

Let K be a class declaration  with the same constructors, superclass and interfaces as C,  and the instance members declared by M (respectively M1, ..., Mk). It is a static warning if the declaration of K would cause a static warning.  It is a compile-time error if the declaration of K would cause a compile-time error.

If, for example, M declares an instance member im whose type is at odds with the type of a member of the same name in S, this will result in a static warning just as if we had defined K by means of an ordinary class declaration extending S, with a body that included im.

### Mixin Composition

Dart does not directly support mixin composition, but the concept is useful when defining how the superclass of a class with a mixin clause is created.

The composition of two mixins, M1<T1 … TkM1> and M2<U1 … UkM2>, written M1<T1 … TkM1> * M2<U1 … UkM2> defines an anonymous mixin such that for any class S<V1 … VkS>, the application of M1<T1 … TkM1> * M2<U1 … UkM2>  to S<V1 … VkS> is equivalent to

typedef Id1<T1 … TkM1, U1 … UkM2, V1 … VkS> = abstract Id2<U1 … UkM2, V1 … VkS> with M1 <T1 … TkM1>;

where Id2 denotes

typedef Id2<U1 … UkM2, V1 … VkS> = abstract S<V1 … VkS> with M2<U1 … UkM2>;

and Id1 and Id2 are unique identifiers that do not exist anywhere in the program.

The classes produced by mixin composition are regarded as abstract because they cannot be instantiated independently. They are only introduced as anonymous superclasses of ordinary class declarations and mixin applications. Consequently, no warning is given if a mixin composition includes abstract members, or incompletely implements an interface.

Mixin composition is associative.

Note that any subset of M1, M2 and S may or may not be generic. For any non-generic declaration, the corresponding type parameters may be elided, and if no type parameters remain in the derived declarations Id1 and/or Id2 then the those declarations need not be generic either.

## Generics

A class declaration or type alias G may be generic, that is, G may have formal type parameters declared. A generic declaration induces a family of declarations, one for each set of actual type parameters provided in the program.

typeParameter:
identifier (extends type)?
;

typeParameters:
'<'
typeParameter (',' typeParameter)* '>'
;

A type parameter T may be suffixed with an extends clause that specifies the upper bound for T. If no extends clause is present, the upper bound is Object. It is a static type warning if a type parameter is a supertype of its upper bound. The bounds of type variables are a form of type annotation and have no effect on execution in production mode.

The type parameters of a generic declaration G are in scope in the bounds of all of the type parameters of G. The type parameters of a generic class declaration G are also in scope in the extends and implements clauses of G (if these exist) and in the body of G.  However, a type parameter is considered to be a malformed type when referenced by a static member.

The restriction is necessary since a type variable has no meaning in the context of a static member, because statics are shared among all instantiations of a generic. However, a type variable may be referenced from an instance initializer, even though this is not available.

Because type parameters are in scope in their bounds, we support F-bounded quantification (if you don't know what that is, don't ask). This enables typechecking code such as:

class Ordered<T> {

operator > (T x);

}

class Sorter<T extends Ordered<T>> {

sort(List<T> l) { … l[n] < l[n+1] …}

}

Even where type parameters are in scope there are numerous restrictions at this time:

1. A type parameter cannot be used to name a constructor in an instance creation expression.
2. A type parameter cannot be used as a superclass or superinterface.

The normative versions of these are given in the appropriate sections of this specification. Some of these restrictions may be lifted in the future.

Dart supports metadata which is used to attach user defined annotations to program structures.

('@' qualified (‘.’ identifier)? (arguments)?)*
;

Metadata consists of a series of annotations, each of which begin with the character @, followed by either a reference to a compile-time constant variable, or a call to a constant constructor.

Metadata is associated with the abstract syntax tree of the program construct p that immediately follows the metadata, assuming p is not itself metadata or a comment . Metadata can be retrieved at runtime via a reflective call, provided the annotated program construct p is accessible via reflection.

Reflective access to metadata is not yet implemented as of the M3 release.

Obviously, metadata can also be retrieved statically by parsing the program and evaluating the constants via a suitable interpreter. In fact many if not most uses of metadata are entirely static.

It is important that no runtime overhead be incurred by the introduction of metadata that is not actually used. Because metadata only involves constants, the time at which it is computed is irrelevant so that implementations may skip the metadata during ordinary parsing and execution and evaluate it lazily.

It is possible to associate metadata with constructs that may not be accessible via reflection, such as local variables (though it is conceivable that in the future, richer reflective libraries might provide access to these as well).  This is not as useless as it might seem. As noted above, the data can be retrieved statically if source code is available.

Metadata can appear before a library, class, typedef, type parameter, constructor, factory, function, field, parameter, or variable declaration and before an import or export directive.

## Expressions

An expression is a fragment of Dart code that can be evaluated at run time to yield a value, which is always an object. Every expression has an associated static type. Every value has an associated dynamic type.

|
conditionalExpression

expressionList:

expression (',' expression)*
;

primary:

thisExpression
|
super assignableSelector
|
functionExpression
|
literal
|
identifier
|
newExpression
|
constObjectExpression
| '('
expression ')
;

An expression e may always be enclosed in parentheses, but this never has any semantic effect on e.

Sadly, it may have an effect on the surrounding expression. Given a class C with static method m => 42, C.m() returns 42, but (C).m() produces a NoSuchMethodError.  This anomaly can be corrected by ensuring that every instance of Type has instance members corresponding to its static members. This issue may be addressed in future versions of Dart.

### Constants

A constant expression is an expression whose value can never change, and that can be evaluated entirely at compile time.

A constant expression is one of the following:

1. A literal number.
2. A literal boolean.
3. A literal string  where any interpolated expression is a compile-time constant that evaluates to a numeric, string or boolean value or to null. It would be tempting to allow string interpolation where the interpolated value is any compile-time constant.  However, this would require running the toString() method for constant objects, which could contain arbitrary code.
4. null.
5. A reference to a static constant variable.
6. An identifier expression that denotes a  constant variable or a class.
7. A constant constructor invocation.
8. A constant list literal.
9. A constant map literal.
10. A simple or qualified identifier denoting a top-level function or a static method.
11. A parenthesized expression (e) where e is a constant expression.
12. An expression of the form identical(e1, e2) where e1 and e2 are constant expressions and identical() is statically bound to the predefined dart function identical() discussed below.
13. An expression of one of the forms e1 == e2 or e1 != e2 where e1 and e2 are constant expressions that evaluate to a numeric, string or boolean value or to null.
14. An expression of one of the forms !e,  e1 && e2 or e1 || e2, where e, e1 and e2 are constant expressions that evaluate to a boolean value or to null.
15. An expression of one of the forms ~ e, e1 ^ e2, e1 & e2, e1 | e2, e1 >> e2 or e1 << e2, where e, e1 and e2 are constant expressions that evaluate to an integer value or to null.
16. An expression of one of the forms -e, e1 + e2, e1 - e2, e1 * e2, e1 / e2, e1 ~/ e2, e1 >e2, e1 < e2, e1 >= e2, e1 <= e2 or e1 % e2,  where e, e1 and e2 are constant expressions that evaluate to a numeric value or to null.

It is a compile-time error if evaluation of a compile-time constant would raise an exception.

The above is not dependent on program control-flow. The mere presence of a compile time constant whose evaluation would fail within a program is an error.  This also holds recursively: since compound constants are composed out of constants, if any subpart of a constant is would raise an exception when evaluated, that is an error.

On the other hand, since implementations are free to compile code late, some compile-time errors may manifest quite late.

const x = 1/0;

final y = 1/0;

class K {

m1(){

var z = false;

if (z) {return x;}

else {return 2;}

}

m2() {

if (true) {return y;}

else {return 3;}

}

}

An implementation is free to immediately issue a compilation error for x but it is not required to do so.  It could defer errors if it does not immediately compile the declarations that reference x. For example, it could delay giving the compilation error for x until the first invocation of m1. However, it could not choose to execute m1,  see that the branch that refers to x is not taken and return 2 successfully.

The situation with respect to an invocation m2 is different. Because y is not a compile-time constant (even though its value is), one need not give a compile-time error upon compiling m2. An implementation may run the code, which will cause  the getter for y to be invoked. At that point, the initialization of y must take place, which requires the initializer to be compiled, which will cause a compilation error.

The treatment of null merits some discussion. Consider null + 2.  This expression always causes an error. We could have chosen not to treat it as a constant expression (and in general, not to allow null as a subexpression of numeric or boolean constant expressions).  There are two arguments for including it:

1. It is constant. We can evaluate it at compile-time.
2. It seems more useful to give the error stemming from the evaluation explicitly.

It is a compile-time error if the value of a compile-time constant expression depends on itself.

As an example, consider:

class CircularConsts{ // Illegal program - mutually recursive compile-time constants

static const i = j; // a compile-time constant

static const j = i; // a compile-time constant

}

Let c1 and c2 be a pair of constants. Then identical(c1, c2) iff:

1. c1 evaluates to either null, a constant map, a constant list, an instance of bool, num, String and c1 == c2, OR
2. c1 and c2 are constant objects of the same class C and for each member field of c1 is identical to the corresponding field of c2.

where identical() is the predefined dart function that returns true iff its two arguments are either:

1. The same object.
2. Of type int and have the same numeric value.
3. Of type double, are not NaNs, and have the same numeric value.

### Null

The reserved word null denotes the null object.

nullLiteral:

null

;

The null object is the sole instance of the built-in class Null. Attempting to instantiate Null causes a runtime error. It is a compile-time error for a class to attempt to extend or implement Null. Invoking a method on null yields a NoSuchMethodError unless the method is explicitly implemented by class Null.

The static type of null is bottom.

The decision to use bottom instead of Null allows null to be be assigned everywhere without complaint by the static checker.

### Numbers

A numeric literal is either a decimal or hexadecimal integer of arbitrary size, or a decimal double.

numericLiteral:
NUMBER
| HEX_NUMBER
;

NUMBER:
DIGIT+ ('.' DIGIT+)? EXPONENT?
|  '.' DIGIT+ EXPONENT?
;

EXPONENT:
('e' | 'E') ('+' | '-')? DIGIT+
;

HEX_NUMBER:
'0x' HEX_DIGIT+
| '0X' HEX_DIGIT+
;

HEX_DIGIT:
'a'..'f'
| 'A'..'F'
| DIGIT
;

If a numeric literal begins with the prefix ‘0x’, it is a hexadecimal integer literal, which denotes the hexadecimal integer represented by the part of the literal following ‘0x’. Otherwise, if the numeric literal does not include a decimal point denotes an it is a decimal integer literal, which denotes a decimal integer.  Otherwise, the numeric literal is a literal double which denotes a 64 bit double precision floating point number as specified by the IEEE 754 standard.

Integers are not restricted to a fixed range. Dart integers are true integers, not 32 bit or 64 bit or any other fixed range representation. Their size is limited only by the memory available to the implementation.

It is a compile-time error for a class to attempt to extend or implement int. It is a compile-time error for a class to attempt to extend or implement double. It is a compile-time error for any type other than the types int and double to attempt to extend or implement num.

An integer literal is either a hexadecimal integer literal or a  decimal integer literal.

The static type of an integer literal is int. A literal double is a numeric literal that is not an integer literal. The static type of a literal double is double.

### Booleans

The reserved words true and false denote objects that represent the boolean values true and false respectively. They are the boolean literals.

booleanLiteral:

true
|
false

;

Both  true and false are implement the built-in class bool. They are the only two instances of bool. It is a compile-time error for a class to attempt to extend or implement bool.

It follows that the two boolean literals are the only two instances of bool.

The static type of a boolean literal is bool.

#### Boolean Conversion

Boolean conversion maps any object o into a boolean. Boolean conversion is defined  by the function

(bool v){

assert(v != null);

return identical(v, true);

}(o)

Boolean conversion is used as part of control-flow constructs and boolean expressions.  Ideally, one would simply insist that control-flow decisions be based exclusively on booleans.  This is straightforward in a statically typed setting. In a dynamically typed language, it requires a dynamic check. Sophisticated virtual machines can minimize the penalty involved. Alas, Dart must be compiled into Javascript. Boolean conversion allows this to be done efficiently.

At the same time, this formulation differs radically from Javascript, where most numbers and objects are interpreted as true.  Dart’s approach prevents usages such  if (a-b) … ; because it does not agree with the low level conventions whereby non-null objects or non-zero numbers are treated as true. Indeed, there is no way to derive true from a non-boolean object via boolean conversion, so this kind of low level hackery is nipped in the bud.

Dart also avoids the strange behaviors that can arise due to the interaction of boolean conversion with autoboxing in Javascript. A notorious example is the situation where false can be interpreted as true. In Javascript, booleans are not objects, and instead are autoboxed into objects where “needed”.  If false gets autoboxed into an object, that object can be coerced into true (as it is a non-null object).

### Strings

A string is a sequence of valid UTF-16 code units.

This decision was made for compatibility with web browsers and Javascript. Earlier versions of the specification required a string to be a sequence of valid Unicode code points.  Programmers should not depend on this distinction.

stringLiteral:
multilineString+
| singleLineString
+
;

A string can be either a single line string or a multiline string.

singleLineString:
'  '' ' stringContentDQ* ' " '
| '
' ' StringContentSQ* ' ' '
| 'r' '
' ' (~( ' ' ' | NEWLINE ))* ' ' '
| 'r' ' " ' (~( ' " ' | NEWLINE ))* ' " '
;

A single line string is delimited by either matching single quotes or matching double quotes.

Hence, ‘abc’ and “abc” are both legal strings, as are ‘He said “To be or not to be” did he not?’  and “He said ‘To be or not to be’ didn’t he?”. However “This ‘ is not a valid string, nor is ‘this”.

The grammar ensures that a single line string cannot span more than one line of source code, unless it includes an interpolated expression that spans multiple lines.

Adjacent single line strings are implicitly concatenated to form a single string literal, and so are adjacent multiline strings, but the two forms may not be mixed.

Here is an example

print("A string" "and then another"); // prints: A stringand then another

Early versions of Dart used the operator + for string concatenation. However, this was  dropped, as it leads to puzzlers such as

print("A simple sum: 2 + 2 = " +

2 + 2);

which this prints  'A simple sum: 2 + 2 = 22' rather than 'A simple sum: 2 + 2 = 4'.

Instead, the recommended Dart idiom is to use string interpolation.

print("A simple sum: 2 + 2 =  {2+2}"); String interpolation work well for most cases. The main situation where it is not fully satisfactory is for string literals that are too large to fit on a line. Multiline strings can be useful, but in some cases, we want to visually align the code. This can be expressed by writing smaller strings separated by whitespace, as shown here: 'Imagine this is a very long string that does not fit on a line. What shall we do? ' 'Oh what shall we do? ' 'We shall split it into pieces ' 'like so' multilineString: ' """' stringContentTDQ* '"""' | ' '''' stringContentTSQ* ''''' | ‘r’ '"""' (~("""))* '"""' | ‘r’ ''''' (~('''))* ''''' ; ESCAPE_SEQUENCE: ‘\n’ | ‘\r’ | ‘\f’ | ‘\b’ | ‘\t’ | ‘\v’ | “\x’ HEX_DIGIT HEX_DIGIT | ‘\u’ HEX_DIGIT HEX_DIGIT HEX_DIGIT HEX_DIGIT | ‘\u{‘ HEX_DIGIT_SEQUENCE ‘}’ ; HEX_DIGIT_SEQUENCE: HEX_DIGIT HEX_DIGIT? HEX_DIGIT? HEX_DIGIT? HEX_DIGIT? HEX_DIGIT? ; Multiline strings are delimited by either matching triples of single quotes or matching triples of double quotes. If the first line of a multiline string consists solely of whitespace characters then that line is ignored, including the new line at its end. Strings support escape sequences for special characters. The escapes are: 1. \n for newline, equivalent to \x0A. 2. \r for carriage return, equivalent to \x0D. 3. \f for form feed, equivalent to \x0C. 4. \b for backspace, equivalent to \x08. 5. \t for tab, equivalent to \x09. 6. \v for vertical tab, equivalent to \x0B. 7. \xHEX_DIGIT1 HEX_DIGIT2, equivalent to \u{ HEX_DIGIT1 HEX_DIGIT2}. 8. \uHEX_DIGIT1 HEX_DIGIT2 HEX_DIGIT3 HEX_DIGIT4, equivalent to \u{ HEX_DIGIT1 HEX_DIGIT2 HEX_DIGIT3 HEX_DIGIT4}. 9. \u{HEX_DIGIT_SEQUENCE} is the unicode scalar value represented by the HEX_DIGIT_SEQUENCE. It is a compile-time error if the value of the HEX_DIGIT_SEQUENCE is not a valid unicode scalar value. 10. indicating the beginning of an interpolated expression.
11. Otherwise, \k indicates the character k for any k not in {n, r, f, b, t, v, x, u}.

It is a compile-time error if a non-raw string literal contains a character sequence of the form \x that is not followed by a sequence of two hexadecimal digits. It is a compile-time error if a non-raw string literal  contains a character sequence of the form \u that is not followed by either a sequence of four hexadecimal digits, or by curly brace delimited sequence of hexadecimal digits.

Any string may be prefixed with the character ‘r’, indicating that it is a raw string, in which case no escapes or interpolations are recognized.

stringContentDQ:
~( '
\' | '  "  ' | '$' | NEWLINE ) | '\' ~( NEWLINE ) | stringInterpolation ; stringContentSQ: ~( '\' | '\'' | '$' | NEWLINE )
| '\' ~( NEWLINE )
| stringInterpolation
;

stringContentTDQ:
~( '
\' | '  "  ' | '$' ) | '\' ~( NEWLINE ) | stringInterpolation ; stringContentTSQ: ~( '\' | '\'' | '$' )
| '\' ~( NEWLINE )
|
stringInterpolation
;

NEWLINE:
\n
| \r
;

All string literals implement the built-in class String. It is a compile-time error for a class to attempt to extend or implement String. The static type of a string literal is String.

#### String Interpolation

It is possible to embed expressions within non-raw string literals, such that the these expressions are evaluated, and the resulting values are converted into strings and concatenated with the enclosing string. This process is known as string interpolation.

stringInterpolation:
'$' IDENTIFIER_NO_DOLLAR | '$' '{'
expression '}'
;

The reader will note that the expression inside the interpolation could itself include strings, which could again be interpolated recursively.

An unescaped $character in a string signifies the beginning of an interpolated expression. The$ sign may be followed by either:

1. A single identifier id that must not contain the $character. 2. An expression e delimited by curly braces. The form$id is equivalent to the form ${id}. An interpolated string ‘s1${e}s2  is equivalent to the concatenation of the strings s1, e.toString() and s2’. Likewise an interpolated string “s1${e}s2is equivalent to the concatenation of the strings “s1, e.toString() and “s2. In both cases, it is a runtime error if e.toString() does not return an object of type String. ### Lists A list literal denotes a list, which is an integer indexed collection of objects. listLiteral: const? typeArguments? '[' (expressionList ','?)? ']' ; A list may contain zero or more objects. The number of elements in a list is its size. A list has an associated set of indices. An empty list has an empty set of indices. A non-empty list has the index set {0 … n -1} where n is the size of the list. It is a runtime error to attempt to access a list using an index that is not a member of its set of indices. If a list literal begins with the reserved word const, it is a constant list literal which is a compile-time constant and therefore evaluated at compile-time. Otherwise, it is a runtime list literal and it is evaluated at runtime. Only run-time list literals can be mutated after they are created. Attempting to mutate a constant list literal will result in a dynamic error. It is a compile time error if an element of a constant list literal is not a compile-time constant. It is a compile time error if the type argument of a constant list literal includes a type parameter. The binding of a type parameter is not known at compile-time, so we cannot use type parameters inside compile-time constants. The value of a constant list literal const <E>[e1... en] is an object a that implements the built-in class List<E>. The ith element of a is vi+1, where vi is the value of the compile time expression ei. The value of a constant list literal const [e1... en] is defined as the value of a constant list literal const <dynamic>[e1... en]. It is a run-time error to attempt to modify a constant list literal. Let list1 = const <V>[e11... e1n] and list2 = const <U>[e21... e2n] be two constant list literals and let the elements of list1 and list2 evaluate to o11... o1n and o21... o2n respectively. Iff identical(o1i, o2i) for 1 <= i <= n and V = U then identical(list1, list2). In other words, constant list literals are canonicalized. A runtime list literal <E>[e1... en] is evaluated as follows: 1. First, the expressions e1... en are evaluated in left to right order, yielding objects o1... on. 2. A fresh instance a, of size n, that implements the built-in interface List<E> is allocated. 3. The operator []= is invoked on a with first argument i and second argument oi+1, 0 <= i <= n. 4. The result of the evaluation is a. Note that this specification does not specify an order in which the elements are set. This allows for parallel assignments into the list if an implementation so desires. The order can only be observed in checked mode: if element i is not a subtype of the element type of the list, a dynamic type error will occur when a[i] is assigned oi-1. A runtime list literal [e1... en] is evaluated as <dynamic>[e1... en]. There is no restriction precluding nesting of list literals. It follows from the rules above that <List<int>>[[1, 2, 3], [4, 5, 6]] is a list with type parameter List<int>, containing two lists with type parameter dynamic. The static type of a list literal of the form const <E>[e1... en] or the form <E>[e1... en] is List<E>. The static type a list literal of the form const [e1... en] or the form [e1... en] is List<dynamic>. It is tempting to assume that the type of the list literal would be computed based on the types of its elements. However, for mutable lists this may be unwarranted. Even for constant lists, we found this behavior to be problematic. Since compile-time is often actually runtime, the runtime system must be able to perform a complex least upper bound computation to determine a reasonably precise type. It is better to leave this task to a tool in the IDE. It is also much more uniform (and therefore predictable and understandable) to insist that whenever types are unspecified they are assumed to be the unknown type dynamic. ### Maps A map literal denotes a map object. mapLiteral: const? typeArguments? '{' (mapLiteralEntry (',' mapLiteralEntry)* ','?)? '}' ; mapLiteralEntry: A map literal consists of zero or more entries. Each entry has a key and a value. Each key and each value is denoted by an expression. If a map literal begins with the reserved word const, it is a constant map literal which is a compile-time constant and therefore evaluated at compile-time. Otherwise, it is a run-time map literal and it is evaluated at run-time. Only run-time map literals can be mutated after they are created. Attempting to mutate a constant map literal will result in a dynamic error. It is a compile time error if either a key or a value of an entry in a constant map literal is not a compile-time constant. It is a compile-time error if the key of an entry in a constant map literal not a constant string, number, boolean, list or map, or null. It is a compile time error if the type arguments of a constant map literal include a type parameter. The value of a constant map literal const <K, V>{k1:e1... kn :en} is an object m that implements the built-in class Map<K, V>. The entries of m are ui:vi, 1 <= i <= n, where ui is the value of the compile time expression ki and vi is the value of the compile time expression ei. The value of a constant map literal const {k1:e1... kn :en} is defined as the value of a constant map literal const <dynamic, dynamic>{k1:e1... kn :en}. It is a run-time error to attempt to modify a constant map literal. Let map1 = const <K, V>{k11:e11... k1n :e1n} and map2 = const <J, U>{k21:e21... k2n :e2n} be two constant map literals. Let the keys of map1 and map2 evaluate to s11... s1n and s21... s2n respectively, and let the elements of map1 and map2 evaluate to o11... o1n and o21... o2n respectively. Iff identical(o1i, o2i) and identical(s1i, s2i) for 1 <= i <= n, and V = U then identical(map1, map2). In other words, constant map literals are canonicalized. A runtime map literal <K, V>{k1:e1... kn :en} is evaluated as follows: 1. First, the expression ki is evaluated yielding object ui, then ei is evaluated yielding object oi, for 1 <= i <= n in left to right order, yielding objects u1,o1..., un, on. 2. A fresh instance m that implements the interface of the built-in class Map<K, V> is allocated. 3. The operator []= is invoked on m with first argument ui and second argument oi, 1 <= i <= n. 4. The result of the evaluation is m. A runtime map literal {k1:e1... kn :en} is evaluated as <dynamic, dynamic>{k1:e1... kn :en}. Iff all the keys in a map literal are compile-time constants, it is a static warning if the values of any two keys in the map literal are equal. A map literal is ordered: iterating over the keys and/or values of the maps always happens in the order the keys appeared in the source code. Of course, if a key repeats, the order is defined by first occurrence, but the value is defined by the last. The static type of a map literal of the form const <K, V>{k1:e1... kn :en} or the form <K, V>{k1:e1... kn :en} is Map<K, V>. The static type a map literal of the form const {k1:e1... kn :en} or the form {k1:e1... kn :en} is Map<dynamic, dynamic>. ### Throw The throw expression is used to raise an exception. throwExpression: throw expression ; throwExpressionWithoutCascade: throw expressionWithoutCascade ; The current exception is the last unhandled exception thrown. Evaluation of a throw statement of the form throw e ; proceeds as follows: The expression e is evaluated yielding a value v. If v evaluates to null, then a NullThrownError is thrown. Otherwise, control is transferred to the nearest dynamically enclosing exception handler, with the current exception set to v. There is no requirement that the expression e evaluate to a special kind of exception or error object. The static type of a throw expression is bottom. ### Function Expressions A function literal is an object that encapsulates an executable unit of code. functionExpression: formalParameterList functionExpressionBody ; functionExpressionBody: '=>' expression | block ; A function literal implements the built-in class Function. The static type of a function literal of the form (T1 a1, …, Tn an, [Tn+1 xn+1 = d1, … ,Tn+k xn+k = dk]) => e is (T1, …, Tn, [Tn+1 xn+1, .., Tn+k xn+k]) →T0, where T0 is the static type of e. In any case where Ti ,1 <= i <= n, is not specified, it is considered to have been specified as dynamic. The static type of a function literal of the form (T1 a1, …, Tn an, {Tn+1 xn+1 : d1, … ,Tn+k xn+k : dk}) => e is (T1, …, Tn, {Tn+1 xn+1, .., Tn+k xn+k}) →T0, where T0 is the static type of e. In any case where Ti ,1 <= i <= n, is not specified, it is considered to have been specified as dynamic. The static type of a function literal of the form (T1 a1, …, Tn an, [Tn+1 xn+1 = d1, …, Tn+k xn+k = dk]) {s} is (T1, …, Tn, [Tn+1 xn+1, .., Tn+k xn+k]) → dynamic. In any case where Ti ,1 <= i <= n, is not specified, it is considered to have been specified as dynamic. The static type of a function literal of the form (T1 a1, …, Tn an, {Tn+1 xn+1 : d1, …, Tn+k xn+k : dk}) {s} is (T1, …, Tn, {Tn+1 xn+1, .., Tn+k xn+k}) → dynamic. In any case where Ti ,1 <= i <= n, is not specified, it is considered to have been specified as dynamic. ### This The reserved word this denotes the target of the current instance member invocation. thisExpression: this ; The static type of this is the interface of the immediately enclosing class. We do not support self-types at this point. It is a compile-time error if this appears in a top-level function or variable initializer, in a factory constructor, or in a static method or variable initializer, or in the initializer of an instance variable. ### Instance Creation Instance creation expressions invoke constructors to produce instances. It is a static type warning if the type T in an instance creation expression of one of the forms new T.id(a1, ... , an, xn+1: an+1, ... , xn+k: an+k), new T(a1, ... , an, xn+1: an+1, ... , xn+k: an+k), const T.id(a1, ... , an, xn+1: an+1, ... , xn+k: an+k), const T(a1, ... , an, xn+1: an+1, ... , xn+k: an+k) is malformed or malbounded #### New The new expression invokes a constructor. newExpression: new type ('.' identifier)? arguments ; Let e be a new expression of the form new T.id(a1, .., an, xn+1: an+1, …, xn+k: an+k) or the form new T(a1, .., an, xn+1: an+1, …, xn+k: an+k). If T is a class or parameterized type accessible in the current scope then: 1. If e is of the form new T.id(a1, .., an, xn+1: an+1, …, xn+k: an+k) it is a static warning if T.id is not the name of a constructor declared by the type T. If e of the form new T(a1, .., an, xn+1: an+1, …, xn+k: an+k) it is a static warning if the type T does not declare a constructor with the same name as the declaration of T. If T is a parameterized type S<U1, ,.., Um>, let R = S. It is a compile time error if S is not a generic type with m type parameters. If T is not a parameterized type, let R = T. Furthermore, if e is of the form new T.id(a1, .., an, xn+1: an+1, …, xn+k: an+k) then let q be the constructor T.id, otherwise let q be the constructor T. If R is a generic with l type parameters but T is not a parameterized type, then for 1 <= i <= l, let Vi = dynamic, otherwise let Vi = Ui for 1 <= i <= m. If R is a generic with l != m type parameters, then for 1 <= i <= l, let Vi = dynamic, otherwise let Vi = Ui for 1 <= i <= m. Evaluation of e proceeds as follows: First, the argument list (a1, …, an, xn+1: an+1, …, xn+k: an+k) is evaluated. Then, If q is a non-factory constructor of an abstract class then an AbstractClassInstantiationError is thrown. If T is malformed a dynamic error occurs. In checked mode, if T is malbounded a dynamic error occurs. Otherwise, if q is not defined or not accessible, a NoSuchMethodError is thrown. If q has less than n positional parameters or more than n required parameters, or if q lacks any of the keyword parameters {xn+1, ..., xn+k} a NoSuchMethodError is thrown. Otherwise, if q is a generative or redirecting constructor, then: Note that it this point we are assured that the number of actual type arguments match the number of formal type parameters. A fresh instance, i, of class R is allocated. For each instance variable f of i, if the variable declaration of f has an initializer expression ef, then ef is evaluated to an object of and f is bound to of. Otherwise f is bound to null. Observe that this is not in scope in ef. Hence, the initialization cannot depend on other properties of the object being instantiated. Do we want to say that this is not in scope, or that using this is illegal? Next, q is executed with this bound to i, the type parameters (if any) of R bound to the actual type arguments V1, ..., Vl and the formal parameter bindings that resulted from the evaluation of the argument list. The result of the evaluation of e is i. Otherwise, q is a factory constructor. Then: If q is a redirecting factory constructor of the form T(p1, …, pn+k) = c; or of the form T.id(p1, …, pn+k) = c; then the result of the evaluation of e is equivalent to evaluating the expression [V1, ..., Vm/T1, ..., Tm](new c(a1, …, an, xn+1: an+1, …, xn+k: an+k)). Otherwise, the body of q is executed with respect to the bindings that resulted from the evaluation of the argument list and the type parameters (if any) of q bound to the actual type arguments V1, ,.., Vl resulting in an object i. The result of the evaluation of e is i. It is a static warning if q is a constructor of an abstract class and q is not a factory constructor. The above gives precise meaning to the idea that instantiating an abstract class leads to a warning at compile time. A similar clause applies to constant object creation in the next section. In particular, a factory constructor can be declared in an abstract class and used safely, as it will either produce a valid instance or lead to a warning inside its own declaration. The static type of a new expression of either the form new T.id(a1, .., an) or the form new T(a1, .., an) is T. It is a static warning if the static type of ai, 1 <= i <= n+ k may not be assigned to the type of the corresponding formal parameter of the constructor T.id (respectively T). #### Const A constant object expression invokes a constant constructor. constObjectExpression: const type ('.' identifier)? arguments ; Let e be a constant object expression of the form const T.id(a1, .., an, xn+1: an+1, …, xn+k: an+k) or the form const T(a1, .., an, xn+1: an+1, …, xn+k: an+k). It is a compile-time error if T is not a class accessible in the current scope. It is a compile-time error if T is a parameterized type. If e is of the form const T.id(a1, .., an, xn+1: an+1, …, xn+k: an+k) it is a compile-time error is a compile-time error if T.id is not the name of a constant constructor declared by the type T. If e is of the form const T(a1, .., an, xn+1: an+1, …, xn+k: an+k) it is a compile-time error if the type T does not declare a constant constructor with the same name as the declaration of T. In all of the above cases, it is a compile-time error if ai, 1 < = i <= n + k, is not a compile-time constant expression. If T is a parameterized type S<U1, ,.., Um>, let R = S; It is a compile time error if S is not a generic type with m type parameters. If T is not a parameterized type, let R = T. Finally, if R is generic but T is not a parameterized type, then for 1 <= i <= m, let Vi = dynamic, otherwise let Vi = Ui. Evaluation of e proceeds as follows: First, if e is of the form const T.id(a1, .., an, xn+1: an+1, …, xn+k: an+k) then let i be the value of the expression new T.id(a1, .., an, xn+1: an+1, …, xn+k: an+k). Otherwise, e must be of the form const T(a1, .., an, xn+1: an+1, …, xn+k: an+k), in which case let i be the result of evaluating new T(a1, .., an, xn+1: an+1, …, xn+k: an+k) . Then: 1. If during execution of the program, a constant object expression has already evaluated to an instance j of class R with type arguments Vi 1 <= i <= m, then: 1. For each instance variable f of i, let vif be the value of the field f in i, and let vjf be the value of the field f in j. If identical(vif , vjf) for all fields f in i, then the value of e is j, otherwise the value of e is i. 1. Otherwise the value of e is i. In other words, constant objects are canonicalized. In order to determine if an object is actually new, one has to compute it; then it can be compared to any cached instances. If an equivalent object exists in the cache, we throw away the newly created object and use the cached one. Objects are equivalent if they have identical fields and identical type arguments. Since the constructor cannot induce any side effects, the execution of the constructor is unobservable. The constructor need only be executed once per call site, at compile-time. The static type of a constant object expression of either the form const T.id(a1, .., an) or the form const T(a1, .., an) is T. It is a static warning if the static type of ai, 1 <= i <= n+ k may not be assigned to the type of the corresponding formal parameter of the constructor T.id (respectively T). It is a compile-time error if evaluation of a constant object results in an uncaught exception being thrown. To see how such situations might arise, consider the following examples: class A { static final x; const A(var p): p = x * 10; } const A(“x”); //compile-time error const A(5); // legal class IntPair { const IntPair(this.x, this.y); final int x; final int y; operator *(v) => new IntPair(x*v, y*v); } const A(const IntPair(1, 2)); // compile-time error: illegal in a subtler way Due to the rules governing constant constructors, evaluating the constructor A() with the argument “x” or the argument const IntPair(1, 2) would cause it to throw an exception, resulting in a compile-time error. Given an instance creation expression of the form const q(a1, .., an) it is a static warning if q is the constructor of an abstract class but q is not a factory constructor. ### Spawning an Isolate Spawning an isolate is accomplished via what is syntactically an ordinary library call, invoking one of the functions spawnUri() or spawnFunction() defined in the dart:isolate library. However, such calls have the semantic effect of creating a new isolate with its own memory and thread of control. An isolate's memory is finite, as is the space available to its thread's call stack. It is possible for a running isolate to exhaust its memory or stack, resulting in a run-time error that cannot be effectively caught, which will force the isolate to be suspended. As discussed above, the handling of a suspended isolate is the responsibility of the embedder. ### Property Extraction Property extraction allows for a member of an object to be concisely extracted from the object. If o is an object, and if m is the name of a method member of o, then o.m is defined to be equivalent to: 1. (r1, .., rn, {p1 : d1, …, pk : dk}){return o.m(r1, .., rn, p1: p1, .., pk: pk);} if m has required parameters r1, …, rn, and named parameters p1 .. pk with defaults d1, …, dk. 2. (r1, .., rn, [p1 = d1, …, pk = dk]){return o.m(r1, .., rn, p1, .., pk);} if m has required parameters r1, …, rn, and optional positional parameters p1 .. pk with defaults d1, …, dk. Otherwise, if m is the name of a getter member of o (declared implicitly or explicitly) then o.m evaluates to the result of invoking the getter. Observations: 1. One cannot extract a getter or a setter. 2. One can tell whether one implemented a property via a method or via field/getter, which means that one has to plan ahead as to what construct to use, and that choice is reflected in the interface of the class. ### Function Invocation Function invocation occurs in the following cases: when a function expression is invoked, when a method, getter or setter is invoked or when a constructor is invoked (either via instance creation , constructor redirection or super initialization). The various kinds of function invocation differ as to how the function to be invoked, f, is determined as well as whether this is bound. Once f has been determined, the formal parameters of f are bound to the corresponding actual arguments. The body of f is then executed with the aforementioned bindings. Execution of the body terminates when the first of the following occurs; 1. An uncaught exception is thrown 2. A return statement immediately nested in the body of f is executed. 3. The last statement of the body completes execution. #### Actual Argument List Evaluation Function invocation involve evaluation of the list of actual arguments to the function and binding of the results to the function’s formal parameters. arguments: '(' argumentList? ')' ; argumentList: namedArgument (',' namedArgument)* | expressionList (',' namedArgument)* ; namedArgument: label expression ; Evaluation of an actual argument list of the form (a1 .. am, q1: am+1, …, ql: am+l) proceeds as follows: The arguments a1, …, am+l are evaluated in the order they appear in the program, yielding objects o1 .. om+l. Simply stated, an argument list consisting of m positional arguments and l named arguments is evaluated from left to right. #### Binding Actuals to Formals Let f be the function, let p1, …, pn be the positional parameters of f and let pn+1, …, pn+k be the optional parameters declared by f. An evaluated actual argument list (o1, …, om+l) derived from an actual argument list of the form (a1 .. am, q1: am+1, …, ql: am+l) is bound to the formal parameters of f as follows: Again, we have an argument list consisting of m positional arguments and l named arguments. We have a function with h required parameters and k optional parameters. The number of positional arguments must be at least as large as the number of required parameters, and no larger than the number of positional parameters. All named arguments must have a corresponding named parameter. You may not provide a given named argument more than once. If an optional parameter has no corresponding argument, it gets its default value. In checked mode, all arguments must belong to subtypes of the type of their corresponding formal. If m < h or m > n, a run-time error occurs. Furthermore, each qi, 1 <= i <= l, must have a corresponding named parameter in the set {pn+1, …, pn+k} or a run time error occurs. Then pi is bound to the value of oi, 1 <= i <= m, and qj is bound to the value of om+j, 1 <= j <= l. All remaining formal parameters of f are bound to their default values. All of these remaining parameters are necessarily optional and thus have default values. If l > 0, then it is necessarily the case that n = h, because a method cannot have both optional positional parameters and named parameters. In checked mode, it is a dynamic type error if oi is not null and the actual type of pi is not a supertype of the type of oi, 1 <= i <= m. It is a dynamic type error if, in checked mode, om+j is not null and the actual type of qj is not a supertype of the type of om+j, 1 <= j <= l. It is a compile-time error if qi = qj for any i != j. Let Ti be the static type of ai, let Si be the type of pi, 1 <= i <= n+k and let Sq be the type of the named parameter q of f. It is a static warning if Tj may not be assigned to Sj, 1 <= j <= m. It is a static warning if m < h or if m > n. Furthermore, each qi, 1 <= i <= l, must have a corresponding named parameter in the set {pn+1, …, pn+k} or a static warning occurs. It is a static warning if Tm+j may not be assigned to Sr, where r = qj, 1 <= j <= l. #### Unqualified Invocation An unqualified function invocation i has the form id(a1, …, an, xn+1: an+1, …, xn+k: an+k), where id is an identifier. What about library prefixes? If there exists a lexically visible declaration named id, let fid be the innermost such declaration. Then: 1. If fid is a local function, a library function, a library or static getter or a variable then i is interpreted as a function expression invocation. 2. Otherwise, If fid is a static method of the enclosing class C, i is equivalent to the static method invocation C.id(a1, …, an, xn+1: an+1, …, xn+k: an+k). 3. Otherwise, if i occurs inside a top level or static function (be it function, method, getter, or setter) or variable initializer, evaluation of i causes a NoSuchMethodError to be thrown. Otherwise, i is equivalent to the ordinary method invocation this.id(a1, …, an, xn+1: an+1, …, xn+k: an+k). #### Function Expression Invocation A function expression invocation i has the form ef(a1, …, an, xn+1: an+1, …, xn+k: an+k), where ef is an expression. If ef is an identifier id, then id must necessarily denote a local function, a library function, a library or static getter or a variable as described above, or i is not considered a function expression invocation. If ef is a property extraction expression, then i is not a function expression invocation and is instead recognized as an ordinary method invocation. a.b(x) is parsed as a method invocation of method m() on object a, not as an invocation of getter b on a followed by a function call (a.b)(x). If a method or getter b exists, the two will be equivalent. However, if b is not defined on a, the resulting invocation of noSuchMethod() would differ. The InvocationMirror passed to noSuchMethod() would describe a call to a method b with argument x in the former case, and a call to a getter b (with no arguments) in the latter. Otherwise: A function expression invocation i of the form ef(a1, …, an, xn+1: an+1, …, xn+k: an+k) is equivalent to the ordinary method invocation ef.call(a1, …, an, xn+1: an+1, …, xn+k: an+k). The implication of this definition, and the other definitions involving the method call(), is that user defined types can be used as function values provided iff they define a call() method. The method call() is special in this regard. The signature of the call() method determines the appropriate signature used when using the object via the built-in invocation syntax. It is a static type warning if the static type F of ef may not be assigned to a function type. If F is not a function type, the static type of i is dynamic. Otherwise the static type of i is the declared return type of F. ### Method Invocation Method invocation can take several forms as specified below. #### Ordinary Invocation An ordinary method invocation i has the form o.m(a1, …, an, xn+1: an+1, …, xn+k: an+k). Method invocation involves method lookup, defined next. The result of looking up a method m in object o with respect to library L is the result of looking up method m in class C with respect to library L, where C is the class of o. The result of looking up method m in class C with respect to library L is: 1. If C declares an instance method named m that is accessible to L, then that method is the result of the lookup. Otherwise, if C has a superclass S, then the result of the lookup is the result of looking up method m in S with respect to L. Otherwise, we say that the lookup has failed. Evaluation of an ordinary method invocation i of the form o.m(a1, …, an, xn+1: an+1, …, xn+k: an+k) proceeds as follows: First, the expression o is evaluated to a value vo. Next, the argument list (a1, …, an, xn+1: an+1, …, xn+k: an+k) is evaluated yielding actual objects o1, …, on+k. Let f be the result of looking up method m in vo with respect to the current library L. If the method lookup succeeded, the body of f is executed with respect to the bindings that resulted from the evaluation of the argument list, and with this bound to vo. The value of i is the value returned after f is executed. If the method lookup has failed, then let g be the result of looking up getter m in vo with respect to L. If the getter lookup succeeded, let vg be the value of the getter invocation o.m. Then the value of i is the result of invoking the static method Function.apply() with arguments vg, [o1, …, on], {xn+1: on+1, …, xn+k: on+k}. If getter lookup has also failed, then a new instance im of the predefined class InvocationMirror is created, such that : 1. im.isMethod evaluates to true. 2. im.memberName evaluates to ‘m’. 3. im.positionalArguments evaluates to an immutable list containing the same values as [o1, …, on]. 4. im.namedArguments evaluates to an immutable map with the same keys and values as {xn+1:on+1, …, xn+k : on+k}. 5. Then the method noSuchMethod() is looked up in o and invoked with argument im, and the result of this invocation is the result of evaluating i. Notice that the wording carefully avoids re-evaluating the receiver o and the arguments ai. Let T be the static type of o. It is a static type warning if T does not have an accessible instance member named m. If T.m exists, it is a static warning if the type F of T.m may not be assigned to a function type. If T.m does not exist, or if F is not a function type, the static type of i is dynamic. Otherwise the static type of i is the declared return type of F. #### Cascaded Invocations A cascaded method invocation has the form e..suffix where suffix is a sequence of operator, method, getter or setter invocations. cascadeSection: '..' (cascadeSelector arguments*)(assignableSelector arguments*)* (assignmentOperator expressionWithoutCascade)? ; cascadeSelector: '[' expression ']' | identifier ; A cascaded method invocation expression of the form e..suffix is equivalent to the expression (t){t.suffix; return t;}(e). #### Static Invocation A static method invocation i has the form C.m(a1, …, an, xn+1: an+1, …, xn+k: an+k). It is a static warning if C does not denote a class in the current scope. It is a static warning if C does not declare a static method or getter m. Note that the absence of C or of C.m is statically detectable. Nevertheless, we choose not to define this situation as an error. The goal is to allow coding to proceed in the order that suits the developer rather than eagerly insisting on consistency. The warnings are given statically at compile-time to help developers catch errors. However, developers need not correct these problems immediately in order to make progress. As of this writing, the above situations are treated as compile time errors by the implementations. This will change over time. Note the requirement that C declare the method. This means that static methods declared in superclasses of C cannot be invoked via C. Evaluation of i proceeds as follows: If C does not denote a class in the current scope, or if C does not declare a static method or getter m then the argument list (a1, …, an, xn+1: an+1, …, xn+k: an+k) is evaluated after which a NoSuchMethodError is thrown. Otherwise, evaluation proceeds as follows: 1. If the member m declared by C is a getter, then i is equivalent to the expression (C.m)(a1, …, an, xn+1: an+1, …, xn+k: an+k). 2. Otherwise, let f be the method m declared in class C. Next, the argument list (a1, …, an, xn+1: an+1, …, xn+k: an+k) is evaluated. The body of f is then executed with respect to the bindings that resulted from the evaluation of the argument list. The value of i is the value returned after the body of f is executed. It is a static type warning if the type F of C.m may not be assigned to a function type. If F is not a function type, or if C.m does not exist, the static type of i is dynamic. Otherwise the static type of i is the declared return type of F. #### Super Invocation A super method invocation i has the form super.m(a1, …, an, xn+1: an+1, …, xn+k: an+k). Evaluation of i proceeds as follows: First, the argument list (a1, …, an, xn+1: an+1, …, xn+k: an+k) is evaluated yielding actual objects o1, …, on+k. Let S be the superclass of the immediately enclosing class and let f be the result of looking up method m in S with respect to the current library L. If the method lookup succeeded, the body of f is executed with respect to the bindings that resulted from the evaluation of the argument list, and with this bound to the current value of this. The value of i is the value returned after f is executed. If the method lookup has failed, then let g be the result of looking up getter m in S with respect to L. If the getter lookup succeeded, let vg be the value of the getter invocation super.m. Then the value of i is the result of invoking the static method Function.apply() with arguments vg, [o1, …, on], {xn+1: on+1, …, xn+k: on+k}. If the getter lookup has also failed, then a new instance im of the predefined class InvocationMirror is created, such that : 1. im.isMethod evaluates to true. 2. im.memberName evaluates to ‘m’. 3. im.positionalArguments evaluates to to an immutable list containing the same values as [o1, …, on]. 4. im.namedArguments evaluates to an immutable map with the same keys and values as {xn+1:on+1, …, xn+k : on+k}. Then the method noSuchMethod() is looked up in S and invoked on this with argument im, and the result of this invocation is the result of evaluating i. It is a compile-time error if a super method invocation occurs in a top-level function or variable initializer, in an instance variable initializer or initializer list, in class Object, in a factory constructor, or in a static method or variable initializer. It is a static type warning if S does not have an accessible instance member named m. If S.m exists, it is a static warning if the type F of S.m may not be assigned to a function type. If S.m does not exist, or if F is not a function type, the static type of i is dynamic. Otherwise the static type of i is the declared return type of F. #### Sending Messages Messages are the sole means of communication among isolates. Messages are sent by invoking specific methods in the Dart libraries; there is no specific syntax for sending a message. In other words, the methods supporting sending messages embody primitives of Dart that are not accessible to ordinary code, much like the methods that spawn isolates. ### Getter and Setter Lookup The result of looking up a getter (respectively setter) m in object o with respect to library L is the result of looking up getter (respectively setter) m in class C with respect to L, where C is the class of o. The result of looking up getter (respectively setter) m in class C with respect to library L is: 1. If C declares an instance getter (respectively setter) named m that is accessible to L, then that getter (respectively setter) is the result of the lookup. Otherwise, if C has a superclass S, then the result of the lookup is the result of looking up getter (respectively setter) m in S with respect to L. Otherwise, we say that the lookup has failed. ### Getter Invocation A getter invocation provides access to the value of a property. Evaluation of a getter invocation i of the form e.m proceeds as follows: First, the expression e is evaluated to an object o. Then, the getter function m is looked up in o with respect to the current library, and its body is executed with this bound to o. The value of the getter invocation expression is the result returned by the call to the getter function. If the getter lookup has failed, then a new instance im of the predefined class InvocationMirror is created, such that : 1. im.isGetter evaluates to true. 2. im.memberName evaluates to ‘m’. 3. im.positionalArguments evaluates to the value of const []. 4. im.namedArguments evaluates to the value of const {}. Then the method noSuchMethod() is looked up in o and invoked with argument im, and the result of this invocation is the result of evaluating i. Let T be the static type of e. It is a static type warning if T does not have a getter named m. The static type of i is the declared return type of T.m, if T.m exists; otherwise the static type of i is dynamic. Evaluation of a getter invocation i of the form C.m proceeds as follows: If there is no class C in the enclosing lexical scope of i, or if C does not declare, implicitly or explicitly, a getter named m, a NoSuchMethodError is thrown. Otherwise, the getter function C.m is invoked. The value of i is the result returned by the call to the getter function. It is a static warning if there is no class C in the enclosing lexical scope of i, or if C does not declare, implicitly or explicitly, a getter named m. The static type of i is the declared return type of C.m if it exists or dynamic otherwise. As of this writing, the above situations are treated as compile time errors by the implementations. Evaluation of a top-level getter invocation i of the form m, where m is an identifier, proceeds as follows: The getter function m is invoked. The value of i is the result returned by the call to the getter function. Note that the invocation is always defined. Per the rules for identifier references, an identifier will not be treated as a top level getter invocation unless the getter i is defined. The static type of i is the declared return type of m. ### Assignment An assignment changes the value associated with a mutable variable or property. assignmentOperator: '=' | compoundAssignmentOperator ; Evaluation of an assignment a of the form v = e proceeds as follows: If there is no declaration d with name v= in the lexical scope enclosing the assignment, then: 1. If a occurs inside a top level or static function (be it function, method, getter, or setter) or variable initializer, evaluation of a causes e to be evaluated, after which a NoSuchMethodError is thrown. 2. Otherwise, the assignment is equivalent to the assignment this.v = e. Otherwise, let d be the innermost declaration whose name is v, if it exists. If d is the declaration of a local or library variable, the expression e is evaluated to an object o. Then, the variable v is bound to o. The value of the assignment expression is o. Otherwise, if d is the declaration of a static variable in class C, then the assignment is equivalent to the assignment C.v = e. Otherwise, the assignment is equivalent to the assignment this.v = e. In checked mode, it is a dynamic type error if o is not null and the interface of the class of o is not a subtype of the actual type of v. It is a static type warning if the static type of e may not be assigned to the static type of v. The static type of the expression v = e is the static type of e. Evaluation of an assignment of the form C.v = e proceeds as follows: If there is no class C in the enclosing lexical scope of the assignment, or if C does not declare, implicitly or explicitly, a setter v=, then a NoSuchMethodError is thrown. Otherwise, the expression e is evaluated to an object o. Then, the setter C.v= is invoked with its formal parameter bound to o. The value of the assignment expression is o. It is a static warning if there is no class C in the enclosing lexical scope of the assignment, or if C does not declare, implicitly or explicitly, a setter v=. As of this writing, the above situations are treated as compile time errors by the implementations. In checked mode, it is a dynamic type error if o is not null and the interface of the class of o is not a subtype of the static type of C.v. It is a static type warning if the static type of e may not be assigned to the static type of C.v. The static type of the expression C.v = e is the static type of e. Evaluation of an assignment of the form e1.v = e2 proceeds as follows: The expression e1 is evaluated to an object o1. Then, the expression e2 is evaluated to an object o2. Then, the setter v= is looked up in o1 with respect to the current library, and its body is executed with its formal parameter bound to o2 and this bound to o2. If the setter lookup has failed, then a new instance im of the predefined class InvocationMirror is created, such that : 1. im.isSetter evaluates to true. 2. im.memberName evaluates to ‘v=’. 3. im.positionalArguments evaluates to to an immutable list containing the same values as [o2]. 4. im.namedArguments evaluates to the value of const {}. Then the method noSuchMethod() is looked up in o1 with argument im. The value of the assignment expression is o2 irrespective of whether setter lookup has failed or succeeded. In checked mode, it is a dynamic type error if o2 is not null and the interface of the class of o2 is not a subtype of the actual type of e1.v. Let T be the static type of e1. It is a static type warning if T does not have an accessible instance setter named v=. It is a static type warning if the static type of e2 may not be assigned to T. The static type of the expression e1.v = e2 is the static type of e2. Evaluation of an assignment of the form e1[e2] = e3 is equivalent to the evaluation of the expression (a, i, e){a.[]=(i, e); return e; } (e1, e2, e3). The static type of the expression e1[e2] = e3 is the static type of e3. It is as static warning if an assignment of the form v = e occurs inside a top level or static function (be it function, method, getter, or setter) or variable initializer and there is no declaration d with name v= in the lexical scope enclosing the assignment. #### Compound Assignment A compound assignment of the form v op= e is equivalent to v = v op e. A compound assignment of the form C.v op= e is equivalent to C.v = C.v op e. A compound assignment of the form e1.v op= e2 is equivalent to ((x) => x.v = x.v op e2)(e1) where x is a variable that is not used in e2. A compound assignment of the form e1[e2] op=e3 is equivalent to ((a, i) => a[i] = a[i] op e3)(e1, e2) where a and i are a variables that are not used in e3. compoundAssignmentOperator: ‘*=' | '/=' | '~/=' | '%=' | '+=' | '-=' | '<<=' | '>>=' | '&=' | '^=' | '|=' ; ### Conditional A conditional expression evaluates one of two expressions based on a boolean condition. conditionalExpression: logicalOrExpression ('?' expressionWithoutCascade ':' expressionWithoutCascade)? ; Evaluation of a conditional expression c of the form e1 ? e2 : e3 proceeds as follows: First, e1 is evaluated to an object o1. In checked mode, it is a dynamic type error if o1 is not of type bool. Otherwise o1 is then subjected to boolean conversion producing an object r. If r is true, then the value of c is the result of evaluating the expression e2. Otherwise the value of c is the result of evaluating the expression e3. It is a static type warning if the type of e1 may not be assigned to bool. The static type of c is the least upper bound of the static type of e2 and the static type of e3. ### Logical Boolean Expressions The logical boolean expressions combine boolean objects using the boolean conjunction and disjunction operators. logicalOrExpression: logicalAndExpression ('||' logicalAndExpression)* ; logicalAndExpression: bitwiseOrExpression ('&&' bitwiseOrExpression)* ; A logical boolean expression is either a bitwise expression, or an invocation of a logical boolean operator on an expression e1 with argument e2. Evaluation of a logical boolean expression b of the form e1 || e2 causes the evaluation of e1; if e1 evaluates to true, the result of evaluating b is true, otherwise e2 is evaluated to an object o, which is then subjected to boolean conversion producing a an object r, which is the value of b. Evaluation a logical boolean expression b of the form e1 && e2 causes the evaluation e1; if e1 does not evaluate to true, the result of evaluating b is false, otherwise e2 is evaluated to an object o, which is then subjected to boolean conversion producing an object r, which is the value of b. The static type of a logical boolean expression is bool. ### Bitwise Expressions Bitwise expressions invoke the bitwise operators on objects. bitwiseOrExpression: bitwiseXorExpression ('|' bitwiseXorExpression)* | super ('|' bitwiseXorExpression)+ ; bitwiseXorExpression: bitwiseAndExpression ('^' bitwiseAndExpression)* | super ('^' bitwiseAndExpression)+ ; bitwiseAndExpression: equalityExpression ('&' equalityExpression)* | super ('&' equalityExpression)+ ; bitwiseOperator: '&' | '^' | '|' ; A bitwise expression is either an equality expression, or an invocation of a bitwise operator on either super or an expression e1, with argument e2. A bitwise expression of the form e1 op e2 is equivalent to the method invocation e1.op(e2). A bitwise expression of the form super op e2 is equivalent to the method invocation super.op(e2). It should be obvious that the static type rules for these expressions are defined by the equivalence above - ergo, by the type rules for method invocation and the signatures of the operators on the type e1. The same holds in similar situations throughout this specification. ### Equality Equality expressions test objects for identity or equality. equalityExpression: relationalExpression ( equalityOperator relationalExpression)? | super equalityOperator relationalExpression ; equalityOperator: '==' | '!=' ; An equality expression is either a relational expression, or an invocation of an equality operator on either super or an expression e1, with argument e2. Evaluation of an equality expression ee of the form e1 == e2 proceeds as follows: 1. The expression e1 is evaluated to an object o1. 2. The expression e2 is evaluated to an object o2. 1. If either o1 or o2 is null, then ee evaluates to identical(o1,o2). Otherwise, 1. ee is equivalent to the method invocation o1.==(o2). Evaluation of an equality expression ee of the form super == e proceeds as follows: 1. The expression e is evaluated to an object o. 2. If either this or o is null, then ee evaluates to identical(this,o). Otherwise, 3. ee is equivalent to the method invocation super.==(o). As a result of the above definition, user defined == methods can assume that their argument is non-null, and avoid the standard boiler-plate prelude: if (identical(null, arg)) return false; Another implication is that there is never a need to use identical to test against null, nor should anyone ever worry about whether to write identical(null, e) or identical(e, null). An equality expression of the form e1 != e2 is equivalent to the expression !(e1 == e2 ). An equality expression of the form super != e is equivalent to the expression !(super == e)). The static type of an equality expression is bool. ### Relational Expressions Relational expressions invoke the relational operators on objects. relationalExpression: shiftExpression ( typeTest | typeCast | relationalOperator shiftExpression)? | super relationalOperator shiftExpression ; relationalOperator: '>=' | '>' | '<=' | '<' ; A relational expression is either a shift expression, or an invocation of a relational operator on either super or an expression e1, with argument e2. A relational expression of the form e1 op e2 is equivalent to the method invocation e1.op(e2). A relational expression of the form super op e2 is equivalent to the method invocation super.op(e2). ### Shift Shift expressions invoke the shift operators on objects. shiftExpression: additiveExpression (shiftOperator additiveExpression)* | super (shiftOperator additiveExpression)+ ; shiftOperator: '<<' | '>>' ; A shift expression is either an additive expression, or an invocation of a shift operator on either super or an expression e1, with argument e2. A shift expression of the form e1 op e2 is equivalent to the method invocation e1.op(e2). A shift expression of the form super op e2 is equivalent to the method invocation super.op(e2). Note that this definition implies left-to-right evaluation order among shift expressions: e1 << e2 << e3 is evaluated as (e1 << e2 ).<< (e3) which is equivalent to (e1 << e2) << e3. The same holds for additive and multiplicative expressions. ### Additive Expressions Additive expressions invoke the addition operators on objects. additiveExpression: multiplicativeExpression (additiveOperator multiplicativeExpression)* | super (additiveOperator multiplicativeExpression)+ ; additiveOperator: '+' | '-' ; An additive expression is either a multiplicative expression, or an invocation of an additive operator on on either super or an expression e1, with argument e2. An additive expression of the form e1 op e2 is equivalent to the method invocation e1.op(e2). An additive expression of the form super op e2 is equivalent to the method invocation super.op(e2). ### Multiplicative Expressions Multiplicative expressions invoke the multiplication operators on objects. multiplicativeExpression: unaryExpression (multiplicativeOperator unaryExpression)* | super (multiplicativeOperator unaryExpression)+ ; multiplicativeOperator: '*' | ' /' | ' %' | ' ~/' ; A multiplicative expression is either a unary expression, or an invocation of a multiplicative operator on either super or an expression e1, with argument e2. A multiplicative expression of the form e1 op e2 is equivalent to the method invocation e1.op(e2). A multiplicative expression of the form super op e2 is equivalent to the method invocation super.op(e2). ### Unary Expressions Unary expressions invoke unary operators on objects. unaryExpression: prefixOperator unaryExpression | postfixExpression | prefix Operator super | incrementOperator assignableExpression ; prefixOperator: '-' unaryOperator: '!' | '~' ; A unary expression is either a postfix expression, an invocation of a prefix operator on an expression e, or an invocation of a unary operator on either super or an expression e. The expression !e is equivalent to the expression e? false: true. Evaluation of an expression of the form ++e is equivalent to e += 1. Evaluation of an expression of the form --e is equivalent to e -= 1. A unary expression u of the form op e is equivalent to a method invocation expression e.op(). An expression of the form op super is equivalent to the method invocation super.op(). ### Postfix Expressions Postfix expressions invoke the postfix operators on objects. postfixExpression: assignableExpression postfixOperator | primary selector* ; postfixOperator: incrementOperator ; selector: assignableSelector | arguments ; incrementOperator: '++' | '--' ; A postfix expression is either a primary expression, a function, method or getter invocation, or an invocation of a postfix operator on an expression e. A postfix expression of the form v++, where v is an identifier, is equivalent to (){var r = v; v = r + 1; return r}(). The above ensures that if v is a field, the getter gets called exactly once. Likewise in the cases below. A postfix expression of the form C.v ++ is equivalent to (){var r = C.v; C.v = r + 1; return r}(). A postfix expression of the form e1.v++ is equivalent to (x){var r = x.v; x.v = r + 1; return r}(e1). A postfix expression of the form e1[e2]++ is equivalent to (a, i){var r = a[i]; a[i] = r + 1; return r}(e1, e2) A postfix expression of the form v--, where v is an identifier, is equivalent to (){var r = v; v = r - 1; return r}(). A postfix expression of the form C.v-- is equivalent to (){var r = C.v; C.v = r - 1; return r}(). A postfix expression of the form e1.v-- is equivalent to (x){var r = x.v; x.v = r - 1; return r}(e1). A postfix expression of the form e1[e2]-- is equivalent to (a, i){var r = a[i]; a[i] = r - 1; return r}(e1, e2) ### Assignable Expressions Assignable expressions are expressions that can appear on the left hand side of an assignment. This section describes how to evaluate these expressions when they do not constitute the complete left hand side of an assignment. Of course, if assignable expressions always appeared as the left hand side, one would have no need for their value, and the rules for evaluating them would be unnecessary. However, assignable expressions can be subexpressions of other expressions and therefore must be evaluated. assignableExpression: primary (arguments* assignableSelector)+ | super assignableSelector | identifier ; assignableSelector: '[' expression ']' | '.' identifier ; An assignable expression is either: 1. An identifier. 2. An invocation of a method, getter or list access operator on an expression e. 3. An invocation of a getter or list access operator on super. An assignable expression of the form id is evaluated as an identifier expression. An assignable expression of the form e.id(a1, …, an) is evaluated as a method invocation. An assignable expression of the form e.id is evaluated as a getter invocation. An assignable expression of the form e1[e2] is evaluated as a method invocation of the operator method [] on e1 with argument e2. An assignable expression of the form super.id is evaluated as a getter invocation. An assignable expression of the form super[e2] is equivalent to the method invocation super.[e2]. ### Identifier Reference An identifier expression consists of a single identifier; it provides access to an object via an unqualified name. identifier: IDENTIFIER ; IDENTIFIER_NO_DOLLAR: IDENTIFIER_START_NO_DOLLAR IDENTIFIER_PART_NO_DOLLAR* ; IDENTIFIER: IDENTIFIER_START IDENTIFIER_PART* ; ; BUILT_IN_IDENTIFIER: abstract | as | dynamic | export | external | factory | get | implements | import | library | operator | part | set | static | typedef ; IDENTIFIER_START: IDENTIFIER_START_NO_DOLLAR | '$'
;

IDENTIFIER_START_NO_DOLLAR:
LETTER
| '_'
;

IDENTIFIER_PART_NO_DOLLAR:
IDENTIFIER_START_NO_DOLLAR
| DIGIT
;

IDENTIFIER_PART:
IDENTIFIER_START
| DIGIT
;

qualified:

identifier ('.' identifier)?
;

A built-in identifier is one of the identifiers produced by the production BUILT_IN_IDENTIFIER. It is a compile-time error if a built-in identifier is used as the declared name of a class, type parameter or type alias. It is a compile-time error to use a built-in identifier other than dynamic as a type annotation.

Built-in identifiers are identifiers that are used as keywords in Dart, but are not reserved words in Javascript. To minimize incompatibilities when porting Javascript code to Dart, we do not make these into reserved words. A built-in identifier may not be used to name a class or type.  In other words, they are treated as reserved words when used as types. This eliminates many confusing situations without causing compatibility problems. After all, a Javascript program has no type declarations or annotations so no clash can occur.  Furthermore, types should begin with an uppercase letter (see the appendix) and so no clash should occur in any Dart user program anyway.

Evaluation of an identifier expression e of the form id proceeds as follows:

Let d be the innermost declaration in the enclosing lexical scope whose name is id.  If no such declaration exists in the lexical scope, let d be the declaration of the inherited member named id if it exists.

1. If d is a class or type alias T, the value of e is an instance of class Type reifying T.
2. If d is a type parameter T, then the value of e is the value of the actual type argument corresponding to T that was  passed to the generative constructor that created the current binding of this. We are assured that this is well defined, because if we were in a static member the reference to T would be a compile-time error.
1. If d is a library variable then:
1. If d is of one of the forms var v = ei; , T v = ei; , final v = ei; , final T v = ei; , and no value has yet been stored into v then the initializer expression ei is evaluated. If, during the evaluation of ei, the getter for v is referenced, a CyclicInitializationError is thrown. If the evaluation succeeded yielding an object o, let r = o, otherwise let r = null. In any case, r is stored into v. The value of e is r.
2. If d is of one of the forms const v = e; or const T v = e; the result of the getter is the value of the compile time constant e. Otherwise
3. e evaluates to the current binding of id.
1. If d is a local variable or formal parameter then e evaluates to the current binding of id.
2. If d is a static method, top level function or local function then e evaluates to the function defined by d.
1. If d is the declaration of a static variable or static getter declared in class C, then e is equivalent to the getter invocation C.id.
2. If d is the declaration of a top level getter, then e is equivalent to the getter invocation id.
3. Otherwise, if e occurs inside a top level or static function (be it function, method, getter,  or setter) or variable initializer, evaluation of e causes a NoSuchMethodError to be thrown.
4. Otherwise e is equivalent to the property extraction  this.id.

It is as static warning if an identifier expression of the form id occurs inside a top level or static function (be it function, method, getter, or setter) or variable initializer and there is no declaration d with name id in the lexical scope enclosing the expression.

### Type Test

The is-expression tests if an object is a member of a type.

typeTest:

isOperator type

;

isOperator:
is '!'?
;

Evaluation of the is-expression e is T proceeds as follows:

The expression e is evaluated to a value v. Then, if the interface of the class of v is a subtype of T, the is-expression evaluates to true. Otherwise it evaluates to false.

It follows that e is Object is always true. This makes sense in a language where everything is an object.

Also note that null is T is false unless T = Object, T = dynamic or T = Null. Since the class Null is not exported by the core library, the latter will not occur in user code. The former two are useless, as is anything of the form e is Object or e is dynamic.  Users should test for a null value directly rather than via a type test on Object.

The is-expression e is! T is equivalent to the expression !(e is T).

If T is malformed the test always succeeds. This is a consequence of the rule that malformed types are treated as dynamic.

The static type of an is-expression is bool.

### Type Cast

The cast-expression ensures that an object is a member of a type.

typeCast:

asOperator type

;

asOperator:
as
;

Evaluation of the cast expression e as T proceeds as follows:

The expression e is evaluated to a value v. Then, if the interface of the class of v is a subtype of T, the cast-expression evaluates to v. Otherwise, if v is null, the cast-expression evaluates to v.  In all other cases,  a CastError is thrown.

If T is malformed the cast always succeeds. This is a consequence of the rule that malformed types are treated as dynamic.

The static type of a cast expression e as T is T.

## Statements

statements:
statement*
;

statement:
label* nonLabelledStatement
;

nonLabelledStatement:

block

| localVariableDeclaration ';'
|
forStatement

### Blocks

A block statement supports sequencing of code.

Execution of a block statement {s1 …  sn} proceeds as follows:

For i = 1 .. n, si is executed.

### Expression Statements

An expression statement consists of an expression that does not begin with a non-constant map literal that has no explicit type arguments.

The restriction on maps is designed to resolve an ambiguity in the grammar, when a statement begins with {.

expressionStatement:

expression? ';'

Execution of an expression statement e; proceeds by evaluating e.

It is a compile-time error if a non-constant map literal that has no explicit type arguments appears in a place where a statement is expected.

### Local Variable Declaration

A variable declaration statement declares a new local variable.

localVariableDeclaration:

initializedVariableDeclaration ';'

Executing a variable declaration statement of one of the forms var v = e;, T v = e;, const v = e; , const T v = e;, final v = e; or final T v = e; proceeds as follows:

The expression e is evaluated to an object o. Then, the variable v is set to o.

A variable declaration statement of the form var v; is equivalent to var v = null;.

A variable declaration statement of the form T v; is equivalent to T v = null;.

This holds regardless of the type T. For example, int i; does not cause i to be initialized to zero. Instead, i is initialized to null, just as if we had written var i; or Object i; or Collection<String> i;

To do otherwise would undermine the optionally typed nature of Dart, causing type annotations to modify program behavior.

It is a compile-time error if e refers to the name v or the name v=.

This rule may seem redundant in light of the scoping rules above. After all:

f() { // a top level function

var v = v* 2;

}

is already an error since v is not yet in scope in its own initializer. However, consider the more insidious:

var v = 0; // top level variable

f() { // a top level function

var v = v* 2; // compile-time error

}

### Local Function Declaration

A function declaration statement declares a new local function.

localFunctionDeclaration:

functionSignature functionBody
;

A function declaration statement of one of the forms id signature {statements} or T id signature {statements} causes a new function named id to be added to the innermost enclosing scope at the point immediately following the function declaration statement.

As for all function declarations, the function is also made available under its name in the function’s formal parameters scope.

This implies that local functions can be directly recursive, but not mutually recursive. Consider these examples:

top() { // a top level function

f(x) => x > 0? x*f(x-1): 1; // recursion is legal

g1(x) => h(x, 1); // error: h is not in scope yet

h(x, n) => x > 1? h(x-1, n*x): n; // again, recursion is fine

g2(x) => h(x, 1); // legal

p1(x) => q(x,x); // illegal

q1(a, b) => a > 0 ? p1(a-1): b; // fine

q2(a, b) => a > 0 ? p2(a-1): b; // illegal

p1(x) => q2(x,x); // fine

}

There is no way to write a pair of mutually recursive local functions, because one always has to come before the other is in scope. These cases are quite rare, and can always be managed by defining a pair of variables first, then assigning them appropriate closures:

top2() { // a top level function

var p, q;

p = (x) => q(x,x); // illegal

q = (a, b) => a > 0 ? p(a-1): b; // fine

}

### If

The if statement allows for conditional execution of statements.

ifStatement:

if '(' expression ')' statement (else statement)?
;

Execution of an if statement of the form if(b) s1 else s2 proceeds as follows:

First, the expression b is evaluated to an object o. In checked mode, it is a dynamic type error if o is not of type bool. Otherwise, o is then subjected to boolean conversion, producing an object r.  If r is true, then the statement s1 is executed, otherwise statement s2 is executed.

It is a static type warning if the type of the expression b may not be assigned to bool.

An if statement of the form  if (b) s1 is equivalent to the if statement if(b) s1 else {}.

### For

The for statement supports iteration.

forStatement:

for '(' forLoopParts ')' statement
;

forLoopParts:
forInitializerStatement
expression? ';' expressionList?
|
declaredIdentifier in expression
|
identifier in expression
;

forInitializerStatement:

localVariableDeclaration ';'
|
expression? ';'
;

The for statement has two forms - the traditional for loop and the foreach statement.

#### For Loop

Execution of a for statement of the form  for (var v = e0 ; c e) s proceeds as follows:

If c is empty let c’ be true, otherwise let c’ be c.

First the variable declaration statement var v = e0  is executed. Then:

1. If this is the first iteration of the for loop, let v’ be v. Otherwise,  let v’ be the variable v’’ created in the previous execution of step 4.
2. The expression [v’/v]c is evaluated and subjected to boolean conversion. If the result is false, the for loop completes. Otherwise, execution continues at step 3.
3. The statement [v’/v]s is executed.
4. Let v’’ be a fresh variable.  v’’ is bound to the value of v’.
5. The expression [v’’/v]e is evaluated, and the process recurses at step 1.

The definition above is intended to prevent the common error where users create a closure inside a for loop, intending to close over the current binding of the loop variable, and find (usually after a painful process of debugging and learning) that all the created closures have captured the same value - the one current in the last iteration executed.

Instead, each iteration has its own distinct variable.  The first iteration uses the variable created by the initial declaration. The expression executed at the end of each iteration uses a fresh variable v’’, bound to the value of the current iteration variable, and then modifies \$v’’ as required for the next iteration.

#### For-in

A for statement of the form  for (varOrType? id in e) s is equivalent to the following code:

var n0 = e.iterator;

while (n0.moveNext()) {

varOrType? id = n0.current;

s

}

where n0 is an identifier that does not occur anywhere in the program.

### While

The while statement supports conditional iteration, where the condition is evaluated prior to the loop.

whileStatement:

while '(' expression ')' statement

;

Execution of a while statement of the form while (e) s; proceeds as follows:

The expression e is evaluated to an object o. In checked mode, it is a dynamic type error if o is not of type bool. Otherwise, o is subjected to boolean conversion, producing an object r.  If r is true, then s is executed and then the while statement is re-executed recursively. If r is false, execution of the while statement is complete.

It is a static type warning if the type of e may not be assigned to bool.

### Do

The do statement supports conditional iteration, where the condition is evaluated after the loop.

doStatement:

do statement while '(' expression ')' ';'
;

Execution of a do statement of the form do s while (e); proceeds as follows:

The statement s is executed. Then, the expression e is evaluated to an object o. In checked mode, it is a dynamic type error if o is not of type bool. Otherwise, o is then subjected to boolean conversion, producing an object r. If r is false, execution of the do statement is complete. If r is true, then the do statement is re-executed recursively.

It is a static type warning if the type of e cannot be assigned to bool.

### Switch

The switch statement supports dispatching control among a large number of cases.

switchStatement:

switch '(' expression ')' '{' switchCase* defaultCase? '}'
;

switchCase:
label* (
case expression ':') statements
;

defaultCase:
label*
default ':' statements
;

Given a switch statement of the form switch (e) { label11 … label1j1 case e1: s1 … labeln1 ..labelnjn case en: sn default: sn+1} or the form switch (e) { label11 … label1j1 case e1: s1 … labeln1 ..labelnjn case en: sn}, it is a compile-time error if the expressions ek are not compile-time constants, for all 1 <= k <= n. It is a compile-time error if values of the expressions ek are not instances of the same class C, for all 1 <= k <= n, or of a class that implements int or String.

In other words, all the expressions in the cases evaluate to constants of the exact same user defined class, or are of certain known types.  Note that the values of the expressions are known at compile-time, and are independent of any static type annotations.

It is a compile-time error if the class C implements the operator == unless the value of the expression is a string or integer.

The prohibition on user defined equality allows us to implement the switch efficiently for user defined types. We could formulate matching in terms of identity instead with the same efficiency. However, if a type defines an equality operator, programmers would find it quite surprising that equal objects did not match.

The switch statement should only be used in very limited situations (e.g., interpreters or scanners).

Execution of a switch statement of the form  switch (e) {label11 ..label1j1 case e1: s1 … labeln1 ..labelnjn case en: sn default: sn+1} or the form switch (e) { label11 … label1j1 case e1: s1 … labeln1 ..labelnjn case en: sn} proceeds as follows:

The statement var id = e; is evaluated, where id is a variable whose name is distinct from any other variable in the program.  In checked mode, it is a run time error if the value of e is not an instance of the same class as the constants e1 … en.

Next, the case clause case e1: s1 is executed if it exists. If case e1: s1 does not exist, then the default clause is executed by executing sn+1.

Note that if there are no case clauses (n = 0), the type of e does not matter.

A case clause introduces a new scope, nested in the lexically surrounding scope. The scope of a case clause ends immediately after the case clause’s statement.

Execution of a case clause case ek: sk of a switch statement switch (e) {label11 ..label1j1 case e1: s1 … labeln1 ..labelnjn case en: sn default: sn+1} proceeds as follows:

The expression ek == id  is evaluated  to an object o which is then subjected to boolean conversion yielding a value v.

If v is not true, the following case,  case ek+1: sk+1 is executed if it exists. If case ek+1: sk+1 does not exist, then the default clause is executed by executing sn+1.

If v is true, let h be the smallest integer such that h >= k and sh is non-empty.  If no such h exists, let h = n + 1. The sequence of statements sh is then executed.  If execution reaches the point after sh  then a runtime error occurs, unless h = n + 1.

Execution of a case clause case ek: sk of a switch statement switch (e) {label11 ..label1j1 case e1: s1 … labeln1 ..labelnjn case en: sn} proceeds as follows:

The expression ek == id  is evaluated  to an object o which is then subjected to boolean conversion yielding a value v.

If v is not true, the following case,  case ek+1: sk+1 is executed if it exists.

If v is true, let h be the smallest integer such that h >= k and sh is non-empty.  The sequence of statements sh is executed if it exists.  If execution reaches the point after sh  then a runtime error occurs, unless h = n.

In other words, there is no implicit fall-through between cases. The last case in a switch (default or otherwise) can ‘fall-through’ to the end of the statement.

It is a static warning if the type of e may not be assigned to the type of ek. It is a static warning if the last statement of the statement sequence sk is not a break, continue, return or throw statement.

The behavior of switch cases intentionally differs from the C tradition.  Implicit fall through is a known cause of programming errors and therefore disallowed.  Why not simply break the flow implicitly at the end of every case, rather than requiring explicit code to do so?  This would indeed be cleaner.  It would also be cleaner to insist that each case have a single (possibly compound) statement.  We have chosen not to do so in order to facilitate porting of switch statements from other languages.  Implicitly breaking the control flow at the end of a case would silently alter the meaning of ported code that relied on fall-through, potentially forcing the programmer to deal with subtle bugs. Our design ensures that the difference is immediately brought to the coder’s attention.  The programmer will be notified at compile-time if they forget to end a case with a statement that terminates the straight-line control flow. We could make this warning a compile-time error, but refrain from doing so because do not wish to force the programmer to deal with this issue immediately while porting code.  If developers ignore the warning and run their code, a run time error will prevent the program from misbehaving in hard-to-debug ways (at least with respect to this issue).

The sophistication of the analysis of fall-through is another issue. For now, we have opted for a very straightforward syntactic requirement. There are obviously situations where code does not fall through, and yet does not conform to these simple rules, e.g.:

switch (x) {

case 1: try { … return;} finally { … return;}

}

Very elaborate code in a case clause is probably bad style in any case, and such code can always be refactored.

### Rethrow

The rethrow statement is used to re-raise an exception.

rethrowStatement:

rethrow
;

Execution of a rethrow statement proceeds as follows:

Control is transferred to the nearest dynamically enclosing exception handler.

No change is made to the current exception.

It is a compile-time error if a rethrow statement is not enclosed within a on-catch clause.

### Try

The try statement supports the definition of exception handling code in a structured way.

tryStatement:

try block (onPart+ finallyPart? | finallyPart)
;

onPart:
catchPart block

| on type catchPart? block

;

catchPart:

catch '(' identifier (',' identifier)? ')'
;

finallyPart:

finally block
;

A try statement consists of a block statement, followed by at least one of:

1. A set of on-catch clauses, each of which specifies (either explicitly or implicitly) the type of exception object to be handled, one or two exception parameters and a block statement.
2. A finally clause, which consists of a block statement.

The syntax is designed to be upward compatible with existing Javascript programs. The on clause can be omitted, leaving what looks like a Javascript catch clause.

An on-catch clause of the form on T catch (p1, p2) s or on T s matches an object o if the type of o is a subtype of T.  It is a static warning if T does not denote a type available in the lexical scope of the catch clause.

An on-catch clause of the form on T catch (p1) s is equivalent to an on-catch clause on T catch (p1, p2) s, where p2 is an identifier that does not occur anywhere else in the program.

An on-catch clause of the form catch (p) s is equivalent to an an on-catch clause on Object catch (p) s. An on-catch clause of the form catch (p1, p2) s is equivalent to an an on-catch clause on Object catch (p1, p2) s.

The active stack trace is an object whose toString() method produces a string that is a record of exactly those function activations within the current isolate that had not completed execution at the point where the current exception was thrown.

This implies that no synthetic function activations may be added to the trace, nor may any source level activations be omitted. This means, for example, that any inlining of functions done as an optimization must not be visible in the trace. Similarly, any synthetic routines used by the implementation must not appear in the trace.

Nothing is said about how any native function calls may be represented in the trace.

For each such function activation, the active stack trace includes the name of the function, the bindings of all its formal parameters, local variables and this, and the position at which the function was executing.

The term position should not be interpreted as a line number, but rather as a precise position - the exact character index of the expression that raised the exception.

The definition below is an attempt to characterize exception handling without resorting to a normal/abrupt completion formulation. It has the advantage that one need not specify abrupt completion behavior for every compound statement.  On the other hand, it is new different and needs more thought.

A try statement try s1 on-catch1 ... on-catchn finally sf  defines an exception handler h that executes as follows:

The on-catch clauses are examined in order, starting with on-catch1, until either an on-catch clause that matches the current exception is found, or the list of on-catch clauses has been exhausted. If an on-catch clause on-catchk is found, it is executed. If no on-catch clause is found, the finally clause is executed. Then, execution resumes at the end of the try statement.

A finally clause finally s defines an exception handler h that executes by executing the finally clause. Then, execution resumes at the end of the try statement.

Execution of an on-catch clause on T catch (p1, p2) s of a try statement t proceeds as follows:

Variables p1 of static type T, and p2 of static type StackTrace are implicitly declared, with a scope comprising s.  The variable p1 is bound to the current exception, and p2 is bound to the current stack trace. Next, the statement s is executed in the dynamic scope of the exception handler defined by the finally clause of t. Then, the current exception and active stack trace both become undefined.

Execution of a finally clause finally s of a try statement proceeds as follows:

The statement s is executed. Then, if the current exception is defined, control is transferred to the nearest dynamically enclosing exception handler.

Execution of a try statement of the form try s1 on-catch1 ... on-catchn finally sf  proceeds as follows:

The statement s1 is executed in the dynamic scope of the exception handler defined by the try statement. Then, the finally clause is executed.

Whether any of the on-catch clauses is executed depends on whether a matching exception has been raised by s1 (see the specification of the throw statement).

If s1 has raised an exception, it will transfer control to the try statement’s handler, which will examine the on-catch clauses in order for a match as specified above. If no matches are found, the handler will execute the finally clause.

If a matching handler was found, it will execute first, and then the finally clause will be executed.

If an exception is raised during execution of an on-catch clause, this will transfer control to the handler for the finally clause, causing the finally clause to execute in this case as well.

If no exception was raised, the finally clause is also executed. Execution of the finally clause could also raise an exception, which will cause transfer of control to the next enclosing handler.

A try statement of the form try s1 on-catch1 ... on-catchn  is equivalent to the statement  try s1 on-catch1 ... on-catchn finally {}

### Return

The return statement returns a result to the caller of a function.

returnStatement:

return expression? ';'
;

Executing a return statement

return e;

first causes evaluation of the expression e, producing an object o. Next, control is transferred to the caller of the current function activation, and the object o is provided to the caller as the result of the function call.

It is a static type warning if the type of e may not be assigned to the declared return type of the immediately enclosing function.

In checked mode, it is a dynamic type error if o is not null and the runtime type of o is not a subtype of the actual return type of the immediately enclosing function.

It is a compile-time error if a return statement of the form return e; appears in a generative constructor.

It is quite easy to forget to add the factory prefix for a constructor, accidentally converting a factory into a generative constructor. The static checker may detect a type mismatch in some, but not all, of these cases. The rule above helps catch such errors, which can otherwise be very hard to recognize. There is no real downside to it, as returning a value from a generative constructor is meaningless.

Let f be the function immediately enclosing a return statement of the form return; It is a static warning if both of the following conditions hold:

1.  f is not a generative constructor.
2. The return type of f may not be assigned to void.

Hence, a static warning will not be issued if f has no declared return type, since the return type would be dynamic and dynamic may be assigned to void. However, any function that declares a return type must return an expression explicitly.

This helps catch situations where users forget to return a value in a return statement.

A return statement with no expression, return; is executed by executing the statement return null; if it occurs inside a method, getter, setter or factory; otherwise, the return statement necessarily occurs inside a generative constructor, in which case it is executed by executing return this;.

Despite the fact that return; is executed as if by a return e;, it is important to understand that it is not a static warning to include a statement of the form return; in a generative constructor. The rules relate only to the specific syntactic form return e;.

The motivation for formulating return; in this way stems from the basic requirement that all function invocations indeed return a value. Function invocations are expressions, and we cannot rely on a mandatory typechecker to always prohibit use of void functions in expressions. Hence, a return statement must always return a value, even if no expression is specified.

The question then becomes, what value should a return statement return when no return expression is given. In a generative constructor, it is obviously the object being constructed (this). In void functions we use null. A void function is not expected to participate in an expression, which is why it is marked void in the first place. Hence, this situation is a mistake which should be detected as soon as possible. The static rules help here, but if the code is executed, using null leads to fast failure, which is desirable in this case. The same rationale applies for function bodies that do not contain a return statement at all.

### Labels

A label is an identifier followed by a colon. A labeled statement is a statement prefixed by a label L.   A labeled case clause is a case clause within a switch statement prefixed by a label L.

The sole role of labels is to provide targets for the break and continue statements.

label:

identifier ':'
;

The semantics of a labeled statement L: s are identical to those of the statement s. The namespace of labels is distinct from the one used for types, functions and variables.

The scope of a label that labels a statement s is s. The scope of a label that labels a case clause of a switch statement s is s.

Labels should be avoided by programmers at all costs. The motivation for including labels in the language is primarily making Dart a better target for code generation.

### Break

The break statement consists of the reserved word break and an optional label.

breakStatement:

break identifier? ';'
;

Let sb be a break statement. If sb is of the form break L; then let sE be the the innermost labeled statement with label L enclosing sb. If sb is of the form break; then let sE be the the innermost  do, for, switch or while statement enclosing sb. It is a compile-time error if no such statement sE exists within the innermost function in which sb occurs.  Furthermore, let s1... sn be those try statements that are both enclosed in sE and that enclose sb, and that have a finally clause. Lastly, let fj be the finally clause of sj, 1 <= j <= n.   Executing sb first executes f1 ... fn in innermost-clause-first order and then terminates sE.

### Continue

The continue statement consists of the reserved word continue and an optional label.

continueStatement:

continue identifier? ';'
;

Let sc be a continue statement. If sc is of the form continue L; then let sE be the the innermost labeled do, for or while statement or case clause with label L that encloses sc. If sc is of the form continue; then let sE be the the innermost  do, for or while statement enclosing sc. It is a compile-time error if no such statement or case clause sE exists within the innermost function in which sc occurs.  Furthermore, let s1... sn be those try statements that are both enclosed in sE and that enclose sc, and that have a finally clause. Lastly, let fj be the finally clause of sj, 1 <= j <= n.   Executing sc first executes f1 ... fn in innermost-clause-first order. Then, if sE is a case  clause control is transferred to the case clause; otherwise, sE is necessarily a loop and execution resumes after the last statement in the loop body.

In a while loop, that would be the boolean expression before the body. In a do loop, it would be the boolean expression after the body. In a for loop, it would be the increment clause.  In other words, execution continues to the next iteration of the loop.

### Assert

An assert statement is used to disrupt normal execution if a given boolean condition does not hold.

assertStatement:

assert '(' conditionalExpression ')' ';'
;

The assert statement has no effect in production mode. In checked mode, execution of an assert statement assert(e); proceeds as follows:

The conditional expression e is evaluated to an object o. If the class of o is a subtype of Function then let r be the result of invoking o with no arguments. Otherwise, let r be o.  It is a dynamic type error if o is not of type bool or of type Function, or if r is not of type bool.  If r is false, we say that the assertion failed. If r is true, we say that the assertion succeeded. If the assertion succeeded, execution of the assert statement is complete. If the assertion failed, an AssertionError is thrown.

It is a static type warning if the type of e may not be assigned to either bool or () → bool

Why is this a statement, not a built in function call? Because it is handled magically so it has no effect and no overhead in production mode. Also, in the absence of final methods. one could not prevent it being overridden (though there is no real harm in that). Overall, perhaps it could be defined as a function, and the overhead issue could be viewed as an optimization.

## Libraries and Scripts

A Dart program consists of one or more libraries, and may be built out of one or more {\em compilation units}. A compilation unit may be a library or a part.

A library consists of (a possibly empty) set of imports, and a set of top level declarations. A top level declaration is either a class, a type alias declaration, a function or a variable declaration.

topLevelDefinition:

classDefinition
|
typeAlias

| external functionSignature ';'

| external getterSignature ';'

| external setterSignature ';'
|
functionSignature functionBody
|
returnType? getOrSet identifier formalParameterList functionBody
|
(final | const) type? staticFinalDeclarationList ';'
|
variableDeclaration ';'
;

getOrSet:

get

| set

;

libraryDefinition:

libraryName? importOrExport* partDirective*  topLevelDefinition*

;

scriptTag:

“#!” (~NEWLINE)* NEWLINE

;

libraryName:

metadata library identifier (“.” identifier)* “;”   ;

importOrExport:

libraryImport

| libraryExport

;

Libraries may be explicitly named or implicitly named. An explicitly named library begins with the word library (possibly prefaced by any applicable metadata annotations), followed by a fully qualified identifier that gives the name of the library.

Technically, each dot and identifier is a separate token and so spaces between them are acceptable. However, the actual library name is the concatenation of the simple identifiers and dots and contains no spaces.

An implicitly named library has the empty string as its name.

The name of a library is used to tie it to separately compiled parts of the library (called parts) and  can be used for printing and, more generally, reflection. The name may be relevant for further language evolution.

Libraries intended for widespread use should avoid name collisions.  Dart's pub package management system provides a mechanism for doing so.  Each pub package is guaranteed a unique name, effectively enforcing a global namespace.

A library may optionally begin with a script tag. Script tags are intended for use with scripts. Script tags can be used to identify the interpreter of the script to whatever computing environment the script is embedded in. The script tag must appear before any whitespace or comments.  A script  tag begins with the characters #! and ends at the end of the line. Any characters that follow #! in the script tag are ignored by the Dart implementation.

Libraries are units of privacy. A private declaration declared within a library L can only be accessed by code within L. Any attempt to access a private member declaration from outside L will cause a run-time error.

Since top level privates are not imported, using the top level privates of another library is never possible.

The public namespace of library L is the mapping that maps the simple name of each public top level member m of L to m.

The scope of a library L consists of the names introduced of all top level declarations declared in L, and the names added by L's imports.

### Imports

An import directive (often abbreviated as import) specifies how a namespace derived from one

library is to be used in the scope of another library.

libraryImport:

metadata import  uri (as identifier)? combinator* “;”

;

combinator:

show identifierList

| hide identifierList

;

identifierList:

identifier (, identifier)*

;

An import specifies a URI x where the declaration of an imported library is to be found. It is a compile-time error if x is not a compile-time constant, or if x involves string interpolation. It is a compile-time error if the compilation unit found at the specified URI is not a library declaration. The interpretation of URIs is described in section 13.5 below.

The current library is the library currently being compiled. The import modifies the import namespace of the current library in a manner that is determined by the imported library and by the optional elements of the import.

Imports assume a global namespace of libraries (at least per isolate). They also assume the library is in control, rather than the other way around.

An import directive I may optionally include:

1. A prefix clause of the form as Id used to prefix names imported by I.
2. Namespace combinator clauses used to restrict the set of names imported by I. Currently, two namespace combinators are supported: hide and show.

Let I be an import directive that refers to a URI via the string s1. Evaluation of I  proceeds as follows:

First,

1. If the URI that is the value of s1 has not yet been accessed by an import or export directive in the current isolate then the contents of the URI are compiled to yield a library B. Because libraries may have mutually recursive imports, care must be taken to avoid an infinite regress.

1. Otherwise, the contents of the URI denoted by s1 have been compiled into a library B within the current isolate.

Let NS0 be the exported namespace of B. Then, for each combinator clause Ci, 1 <= i <= n,  in I:

1. If Ci is of the form  show id1, …, idk , then let NSi = show([id1, …, idk], NSi-1) where show(l,n) takes a list of identifiers l and a namespace n, and produces a namespace that  maps each name in l to the same element that n does. Furthermore, for each name x in l, if n defines the name  x= then the new namespace maps x= to the same element that n does. Otherwise the resulting mapping is undefined.
2. If Ci is of the form  hide id1, …, idk , then let NSi = hide([id1, …, idk], NSi-1) where hide(l, n) takes a list of identifiers l and a namespace n, and produces a namespace that is identical to n except that for each name k in l, k and k= are undefined.

Next, if I includes a prefix clause of the form as p, let NS =  prefix(p, NSn) where prefix(id, n), takes an identifier id and produces a namespace that has, for each entry mapping key k to declaration d in n,  an entry mapping id.k to d. Otherwise, let NS = NSn.

It is a compile-time error if the current library declares a top-level member named p.

Then, for each entry mapping key k to declaration d in NS, d is made available in the top level scope of L under the name k unless either:

1. a declaration with the name k exists in L, OR
2. a prefix clause of the form  as k is used in L.

The greatly increases the chance that a member can be added to a library without breaking its importers.

If a name N is referenced by a library L and N is introduced into the top level scope L by more than one import then:

1. It is a static warning if N is used as a type annotation.
2. In checked mode, it is a dynamic error if N is used as a type annotation and referenced during a subtype test.
3. Otherwise, it is a compile-time error.

We say that the namespace NS has been imported into L.

It is neither an error nor a warning if N is introduced by two or more imports but never referred to.

The policy above makes libraries more robust in the face of additions made to their imports.

A clear distinction needs to be made between this approach, and seemingly similar policies with respect to classes or interfaces.  The use of a class or interface, and of its members, is separate from its declaration. The usage and declaration may occur in widely separated places in the code, and may in fact be authored by different people or organizations.  It is important that errors are given at the offending declaration so that the party that receives the error can respond to it a meaningful way.

In contrast a library comprises both imports and their usage; the library is under the control of a single party and so any problem stemming from the import can be resolved even if it is reported at the use site.

On a related note, the provenance of the conflicting elements is not considered. An  element that is imported via distinct paths may conflict with itself. This avoids variants of the well known "diamond" problem.

It is a static warning to import two different libraries with the same name.

A widely disseminated library should be given a name that will not conflict with other such libraries. The preferred mechanism for this is using pub, the Dart package manager, which provides a global namespace for libraries, and conventions that leverage that namespace.

Note that no errors or warnings are given if one hides or shows a name that is not in a namespace.  This prevents situations where removing a name from a library would cause breakage of a client library.

The dart core library is implicitly imported into every dart library other than itself via an  import clause of the form

import  “dart:core”;

unless the importing library explicitly imports dart:core.

Any import of dart:core, even if restricted via show, hide or as, preempts the automatic import.

It would be nice if there was nothing special about dart:core. However, its use is pervasive, which leads to the decision to import it automatically.  However, some library L  may wish to define entities with names used by dart:core (which it can easily do, as the names declared by a library take precedence). Other libraries may wish to use L and may want to use members of L that conflict with the core library without having to use a prefix. The above rule makes this possible, essentially canceling dart:core's special treatment by means of yet another special rule.

### Exports

A library L exports a namespace, meaning that the declarations in the namespace are made available to other libraries if they choose to import L.  The namespace that L exports is known as its exported namespace.

libraryExport:

metadata export  uri combinator* “;”

An export specifies a URI x where the declaration of an exported library is to be found. It is a compile-time error if the compilation unit found at the specified URI is not a library declaration.

We say that a name is exported by a library (or equivalently, that a library exports a name) if the name is in the library’s exported namespace. We say that a declaration is exported by a library (or equivalently, that a library exports a declaration) if the declaration is in the library’s exported namespace.

A library always exports all names and all declarations in its public namespace. In addition, a library may choose to re-export additional libraries via export directives, often referred to simply as exports.

Let E be an export directive that refers to a URI via the string s1. Evaluation of E  proceeds as follows:

First,

1. If the URI that is the value of s1 has not yet been accessed by an import or export directive in the current isolate then the contents of the URI are compiled to yield a library B.
1. Otherwise, the contents of the URI denoted by s1 have been compiled into a library B within the current isolate.

Let NS0 be the exported namespace of B. Then, for each combinator clause Ci, 1 <= i <= n,  in E:

1. If Ci is of the form  show id1, …, idk , then let NSi = show([id1, …, idk], NSi-1).
2. If Ci is of the form  hide id1, …, idk , then let NSi = hide([id1, …, idk], NSi-1).

For each entry mapping key k to declaration d  in NSn  an entry mapping k to d is added to the exported namespace of L unless a  top-level declaration with the name k exists in L.  We say that L re-exports library B, and also that L re-exports namespace ImportNamespace. When no confusion can arise, we may simply state that L re-exports B, or that L re-exports ImportNamespace.

It is a compile-time error if a name N is re-exported by a library L and N is introduced into the export namespace of L by more than one export.

### Parts

A library may be divided into parts, each of which can be stored in a separate location. A library identifies its parts by listing them via part directives.

A part directive specifies a URI where a Dart compilation unit that should be incorporated into the current library may be found.

partDirective:

metadata part  stringLiteral “;”

;

metadata part of identifier ( “.” identifier)* “;”
;

partDeclaration:
;

A part header begins with  part of  followed by the name of the library the part belongs to.  A part declaration consists of a part header followed by a sequence of top level declarations.

Compiling a part directive of the form part s; causes the Dart system to attempt to compile the contents of the URI that is the value of s. The top level declarations at that URI are then compiled by the Dart compiler in the scope of the current library. It is a compile time error if the contents of the URI are not a valid part declaration. It is a static warning if the referenced part declaration p names a library other than the current library as the library to which p belongs.

It is a compile-time error if s is not a compile-time constant, or if s involves string interpolation.

### Scripts

A script is a library with a top level function main(). A script S may be executed as follows:

First, S is compiled as a library as specified above. Then, the top level function main() that is in scope in S is invoked with no arguments. It is a run time error if S does not declare or import a top level function main().

The names of scripts are optional, in the interests of interactive, informal use. However, any script of long term value should be given a name as a matter of good practice.

### URIs

URIs are specified by means of string literals:

uri:

stringLiteral

;

It is a compile-time error if the string literal x that describes a URI is not a compile-time constant, or if x involves string interpolation.

This specification does not discuss the interpretation of URIs, with the following exceptions.

The interpretation of URIs is mostly left to the surrounding computing environment. For example, if Dart is running in a web browser, that browser will likely interpret some URIs. While it might seem attractive to specify, say, that URIs are interpreted with respect to a standard such as IETF RFC 3986, in practice this will usually depend on the browser and cannot be relied upon.

A URI of the form dart:s is interpreted as a reference to a library s that is part of the Dart implementation.

A URI of the form package:s is interpreted as a URI of the form packages/s relative to an implementation specified location.

This location will often be the location of the root library presented to the Dart compiler. However, implementations may supply means to override or replace this choice.

The intent is that, during development, Dart programmers can rely on a package manager to find elements of their program. Such package managers may provide a directory structure starting at a directory packages where they place the required dart code (or links thereto).

Otherwise, any relative URI is interpreted as relative to the the location of the current library. All further interpretation of URIs is implementation dependent.

This means it is dependent on the embedder.

## Types

Dart supports optional typing based on interface types.

The type system is unsound, due to the covariance of generic types. This is a deliberate choice (and undoubtedly controversial).  Experience has shown that sound type rules for generics fly in the face of programmer intuition. It is easy for tools to provide a sound type analysis if they choose, which may be useful for tasks like refactoring.

### Static Types

Static type annotations are used in variable declarations (including formal parameters) and in the return types of functions, and in the bounds of type variables.  Static type annotations are used during static checking and when running programs in checked mode. They have no effect whatsoever in production mode.

type:

typeName typeArguments?
;

typeName:

qualified
;

typeArguments:
'<'
typeList '>'
;

typeList:

type (',' type)*
;

A Dart implementation must provide a static checker that detects and reports exactly those situations this specification identifies as static warnings and only those situations. However:

1. Running  the static checker on a program P is not required for compiling and running P.
2. Running the static checker on a program P must not prevent successful compilation of P nor may it prevent the execution of P, regardless of whether any static warnings occur.

Nothing precludes additional tools that implement alternative static analyses (e.g., interpreting the existing type annotations in a sound manner such as either non-variant generics, or inferring declaration based variance from the actual declarations). However, using these tools must not preclude successful compilation and execution of Dart code.

A type T is malformed iff:

1. T has the form id or the form  prefix.id, and in the enclosing lexical scope, the name id (respectively prefix.id) does not denote a type.
2. T denotes a type variable in the enclosing lexical scope, but occurs in the signature or body of a static member.
3. T is a parameterized type of the form G<S1, .., Sn>, and G is malformed.

Any use of a malformed  type gives rise to a static warning. A malformed type is then interpreted as dynamic by the static type checker and the runtime.

This ensures that the developer is spared a series of cascading warnings as the malformed type interacts with other types.

### Dynamic Type System

A Dart implementation must support execution in both production mode and checked mode.  Those dynamic checks specified as occurring specifically in checked mode must be performed iff the code is executed in checked mode.

In checked mode, it is a dynamic type error if  a malbounded type is used in a subtype test.

Consider the following program

typedef F(bool x);

f(foo x) => x;

main() {

if (f is F) {

print("yoyoma");

}

}

The type of the formal parameter of f is foo, which is undeclared in the lexical scope. This will lead to a static type warning. At runtime the program will print yoyoma because foo is treated as dynamic.

As another example take

var i;

i  j; //  a variable j of type i (supposedly)

main() {

j = new Object(); // fails in checked mode

}

Since i is not a type, a static warning will be issue at the declaration of j. However, the program can be executed without incident. The undeclared type i is treated as dynamic so even in checked mode, the  implicit subtype test at the assignment succeeds.

One have chosen to treat malformed types as type dynamic. Earlier versions of this specification did so in some cases but not in others. We found the rules to be too complex, and have opted to harmonize the specification. Given that a static warning is issued, there is no need for the runtime to deal with extra complexity by treating malformed types specially.

Here is an example involving malbounded types:

class I<T extends num> {}

class J{}

class A<T> implements J,  I<T> // type warning: T is not a subtype of num

{ ...

}

Given the declarations above, the following

I x = new A<String>();

will cause a dynamic type error in checked mode, because the assignment requires a subtype test A<String> <: I. To show that this holds, we need to show that A<String>I<String>, but I<String> is a malbounded type, causing the dynamic error.  No error is thrown in production mode. Note that

J x = new A<String>();

does not cause a dynamic error, as there is no need to test against I<String> in this case.

Similarly, in production mode

A x = new A<String>();

bool b = x is I;

b is bound to true, but in checked mode the second line causes a dynamic type error.

### Type Declarations

#### Typedef

A type alias declares a name for a type expression or mixin application.

typeAlias:

;

typeAliasBody:

identifier typeParameters?=’ abstract? mixinApplication ‘;’

;

functionTypeAlias:

functionPrefix typeParameters? formalParameterList ';'
;

functionPrefix:

returnType? identifier
;

The effect of a type alias of the form  typedef T id (T1 p1, .., Tn pn, [Tn+1 pn+1, …, Tn+k pn+k]) declared in a library L is is to introduce the name id into the scope of L, bound to the function type (T1, .., Tn, [ Tn+1 pn+1, …,  Tn+k pn+k])  → T.  The effect of a type alias of the form  typedef T id (T1 p1, .., Tn pn, {Tn+1 pn+1, …, Tn+k pn+k}) declared in a library L is is to introduce the name id into the scope of L, bound to the function type (T1, .., Tn, { Tn+1 pn+1, …,  Tn+k pn+k]})  → T. In either case, if no return type is specified, it is taken to be dynamic. Likewise, if a type annotation is omitted on a formal parameter, it is taken to be dynamic.

The effect of a type alias of the form typedef C = M; or the form typedef C<T1, .., Tn> = M;  is to introduce the name C into the scope of L, bound to the class defined by the mixin application M. The name of the class is also set to C. Iff the type alias body includes the built-in identifier abstract, the class being defined is an abstract class.

It is a compile-time error if any default values are specified in the signature of a function type alias. It is a compile-time error if a typedef refers to itself via a chain of references that does not include a class declaration.

Hence

typedef F F(F f);

is illegal, as are

typedef B A();

typedef A B();

but

typedef D C();

class D { C foo(){}}

is legal, because the references goes through a class declaration.

### Interface Types

The implicit interface of class I is a direct supertype of the implicit interface of class J iff:

1. If I is Object, and J has no extends clause.
2. If I is listed in the extends clause of J.
3. If I is listed in the implements clause of J.
1. If I is listed in the with clause of J.
1. If J is a mixin application of the mixin of I.

A type T is more specific than a type S, written T ≪ S,  if one of the following conditions is met:

1. Reflexivity: T is S.
2. T is bottom.
3. S is dynamic.
4. Direct supertype: S is a direct supertype of T.
5. T is a type parameter and S is the upper bound of T.
6. T is a type parameter and S is Object.
7. Covariance: T is of the form I<T1, ..., Tn> and S is of the form I<S1, ..., Sn> and Ti  ≪ Si , 1 <= i <= n.
8. Transitivity: T ≪ U and U ≪ S.

≪ is a partial order on types.

T is a subtype of S, written T <: S, iff [bottom/dynamic]T ≪ S.

Note that <: is not a partial order on types, it is only binary relation on types. This is because <: is not transitive. If it was, the subtype rule would have a cycle. For example:

List <: List<String> and List<int> <: List, but List<int> is not a subtype of List<String>.

Although <: is not a partial order on types, it does contain a partial order, namely ≪. This means that, barring raw types, intuition about classical subtype rules does apply.

S is a supertype of T, written S :> T, iff T is a subtype of S.

The supertypes of an interface are its direct supertypes and their supertypes.

An interface type T may be assigned to a type S, written  T ⇔ S, iff either T <: S or S <: T.

This rule may surprise readers accustomed to conventional typechecking. The intent of the ⇔ relation is not to ensure that an assignment is correct. Instead, it aims to only flag assignments that are almost certain to be erroneous, without precluding assignments that may work.

For example, assigning a value of static type Object to a variable with static type String, while not guaranteed to be correct, might be fine if the runtime value happens to be a string.

### Function Types

Function types come in two variants:

1. The types of functions that only have positional parameters. These have the general form (T1, ..., Tn, [Tn+1, …, Tn+k]) → T.
2. The types of functions with named parameters. These have the general form (T1, ..., Tn, {Tx1 x1, …, Txk xk}) → T.

A function type (T1, ..., Tk, [Tk+1, …, Tn+m]) → T is a subtype of the function type (S1, ..., Sk+j, [Sk+j+1, …, Sn]) → S, if all of the following conditions are met:

1. Either
1. S is void,
2. or T ⇔ S.
1. For all i , 1 <= i <= n, Ti ⇔ Si.

A function type (T1, ..., Tn, {Tx1 x1, …, Txk xk}) → T is a subtype of the function type (S1, ..., Sn, {Sy1 y1, …, Sym ym}) → S, if all of the following conditions are met:

1. Either
1. S is void,
2. or T ⇔ S.
1. For all i , 1 <= i <=  n, Ti ⇔ Si.
2. k >= m and yi in {x1, …, xk}, 1 <= i <= m.
3. For all yi in {y1, …, ym}, yi = xj => Tj ⇔ Si.

In addition, the following subtype rules apply:

(T1, ..., Tn, []) → T <: (T1, ..., Tn) → T.

(T1, ..., Tn) → T <: (T1, ..., Tn, {}) → T.

(T1, ..., Tn, {}) → T <: (T1, ..., Tn) → T.

(T1, ..., Tn) → T <: (T1, ..., Tn, []) → T.

The naive reader might conclude that, since it is not legal to declare a function with an empty optional parameter list, these rules are pointless. However, the induce  useful relationships between function types that declare no optional parameters and those that do.

A function type T may be assigned to a type function type S, written T ⇔ S, iff T <: S.

A functions is always an instance of some class that implements the class Function. All function types are subtypes of Function. If a type I includes an instance method named call(), and the type of call() is the function type F, then I is considered to be a subtype of F.

### Type dynamic

The built-in identifier dynamic denotes the unknown type.

If no static type annotation has been provided the type system assumes the declaration has the type dynamic. If a generic type is used but type arguments are not provided, then the type arguments default to the unknown type.

This means that given a generic declaration G<T1, …, Tn>, the type G is equivalent to G<dynamic, …, dynamic>.

Type dynamic has methods for every possible identifier and arity, with every possible combination of named parameters. These methods all have dynamic as their return type, and their formal parameters all have type dynamic.

Type dynamic has properties for every possible identifier. These properties all have type dynamic.

From a usability perspective, we want to ensure that the checker does not issue errors everywhere an unknown type is used. The definitions above ensure that no secondary errors are reported when accessing an unknown type.

The current rules say that missing type arguments are treated as if they were the type dynamic. An alternative is to consider them as meaning Object.  This would lead to earlier error detection in checked mode, and more aggressive errors during static typechecking. For example:

(1) typedAPI(G<String> g){...}

(2) typedAPI(new G());

Under the alternative rules, (2) would cause a runtime error in checked mode. This seems desirable from the perspective of error localization. However, when a dynamic error is raised at (2), the only way to keep running is rewriting (2) into

(3) typedAPI(new G<String>());

This forces users to write type information in their client code just because they are calling a typed API.  We do not want to impose this on Dart programmers, some of which may be blissfully unaware of types in general, and genericity in particular.

What of static checking? Surely we would want to flag (2) when users have explicitly asked for static typechecking? Yes, but the reality is that the Dart static checker is likely to be running in the background by default. Engineering teams typically desire a “clean build” free of warnings and so the checker is designed to be extremely charitable. Other tools can interpret the type information more aggressively and warn about violations of conventional (and sound) static type discipline.

### Type Void

The special type void may only be used as the return type of a function: it is a compile-time error to use void in any other context.

For example, as a type argument, or as the type of a variable or parameter

Void is not an interface type.

The only subtype relations that pertain to void are therefore:

1. void <: void(by reflexivity)
2. bottom <: void (as bottom is a subtype of all types).
3. void <: dynamic (as dynamic is a supertype of all types)

Hence, the static checker will issue warnings if one attempts to access a member of the result of a void method invocation (even for members of null, such as ==).  Likewise, passing the result of a void method as a parameter or assigning it to a variable will cause a warning unless the variable/formal parameter has type dynamic.

On the other hand, it is possible to return the result of a void method from within a void method. One can also return null; or a value of type dynamic. Returning any other result will cause a type warning. In checked mode, a dynamic type error would arise if a non-null object was returned from a void method (since no object has runtime type dynamic).

### Parameterized Types

A parameterized type is an invocation of a generic type declaration.

Let T be a parameterized type G<S1, …, Sn>. If G is not a generic type, the type arguments S1, …, Sn are discarded. If G has m != n type parameters, T is treated as as a parameterized type with m arguments, all of which are dynamic.

In short, any arity mismatch results in all type arguments being dropped, and replaced with the correct number of type arguments, all set to dynamic. Of course, a static warning will be issued. This behavior is not yet implemented.

Otherwise, let Ti be the type parameters of G and let Bi be the bound of Ti, 1 <= i <= n. T is malbounded iff either Si is malbounded  or Si is not a subtype of [S1,  …, Sn/T1, ..., Tn]Bi,   1 <= i <= n.

Note, that, in checked mode, it is a dynamic type error if a malbounded type is used in a type test as specified above.

Any use of a malbounded type gives rise to a static warning.

If S is the static type of a member m of G, then the static type of the member m of G<A1, …, An> is [A1, …, An/T1, …, Tn]S where T1, …, Tn are the formal type parameters of G. Let Bi be the bounds of Ti, 1 <= i <= n. It is a static type warning if Ai is not a subtype of [A1, …, An/T1, …, Tn]Bi, 1 <= i <= n. It is a static type warning if G is not a generic type with exactly n type parameters.

#### Actual Type of a Declaration

A type T depends on a type parameter U iff:

1. T is U.
2. T is a parameterized type, and one of the type arguments of T depends on U.

Let T be the declared type of a declaration d, as it appears in the program source. The actual type of d is

1. [A1, ..., An/U1, ..., Un]T if T depends on type parameters U1, ..., Un, and Ai is the actual value of Ui, 1 <= i <= n.
2. T otherwise.

#### Least Upper Bounds

Given two interfaces I and J, let SI be the set of superinterfaces of I,  let SJ be the set of superinterfaces of J and let S = (I SI ) (J  SJ ). Furthermore, we define Sn = {T | T  S   depth(T) =n} for any finite n, where depth(T) is the number of steps in the longest inheritance path from T to Object. Let q be the largest number such that Sq has cardinality one. The least upper bound of I and J is the sole element of Sq.

## Reference

### Lexical Rules

Dart source text is represented as a sequence of Unicode code points.  This sequence is first converted into a sequence of tokens according to the lexical rules given in this specification.  At any point in the tokenization process, the longest possible token is recognized.

#### Reserved Words

A reserved word may not be used as an identifier; it is a compile-time error if a reserved word is used where an identifier is expected.

assert, break, case, catch, class, const, continue, default, do, else, enum, extends, false, final, finally, for, if, in, is, new, null, rethrow, return, super, switch, this, throw, true, try, var, void, while, with.

LETTER:
'a'..'z'
| 'A'..'Z'
;

DIGIT:
'0'..'9'
;

WHITESPACE:
('\t' | ' ' | NEWLINE)+
;

Comments are sections of program text that are used for documentation.

SINGLE_LINE_COMMENT:
'//' ~(NEWLINE)* (NEWLINE)?
;

MULTI_LINE_COMMENT:
'/*'
(MULTI_LINE_COMMENT | ~ '*/')* '*/'
;

Dart supports both single-line and multi-line comments. A single line comment begins with the token //. Everything between // and the end of line must be ignored by the Dart compiler unless the comment is a documentation comment.

A multi-line comment begins with the token /* and ends with the token */.  Everything between /* and */ must be ignored by the Dart compiler unless the comment is a documentation comment. Comments may nest.

Documentation comments are comments that begin with the tokens /// or  /**. Documentation comments are intended to be processed by a tool that produces human readable documentation. Below, we refer to any such tool as DartDoc, and owe refer to the output of DartDoc as the formatted output''.

The scope of a documentation comment  always excludes the imported namespace of the enclosing library. Only names declared in the enclosing library are considered in scope within a documentation comment.

The scope of a documentation comment immediately preceding the declaration of a class C is the instance scope of C, excluding any names introduced via the import namespace of the enclosing library.

The scope of a documentation comment immediately preceding the declaration of a function f  is the  scope in force at the very beginning of the body of f,  excluding any names introduced via the import namespace of the enclosing library.

Within a documentation comment, text is tokenized. Most tokens have no significance, and will be reproduced verbatim in the formatted output produced by the DartDoc tool.  A number of tokens have special significance.

A blank line indicates that the following text will begin a new paragraph in the formatted output.

A token of the form [id] will be replaced by a link in the formatted output. The link will point at the declaration named id. The title of the link will be id. It is a static warning if id does not denote a declaration that available in the  scope of the documentation comment. If id is the name of a class, the generated link will point at the class, even if the class has a constructor of the same name.

A token of the form [new c] will be replaced by a link in the formatted output. The link will point at the declaration of a constructor named c. The title of the link will be c.  It is a static warning if c does not denote a constructor that available in the  scope of the documentation comment.

A token of the form [:code:] or code` will render code as code in the formatted output.

A token of the form *foo1 … fook* or _foo1 … fook_ will cause foo1 … fook to be emphasized in the formatted output.

A token of the form **foo1 … fook** or __foo1 … fook__ will cause foo1 … fook to be strongly emphasized in the formatted output.

A token of the form [id](uri)  will be replaced by a link in the formatted output. The link will point at the declaration named id in L. The title of the link will be id. It is a static warning if uri is not the URI of a dart library L, or if id is not a name declared in the exported namespace of L.  If id is the name of a class, the generated link will point at the class, even if the class has a constructor of the same name.

A token of the form [new c](uri) will be replaced by a link in the formatted output. The link will point at the constructor named c in L. The title of the link will be c. It is a static warning if uri is not the URI of a dart library L, or if c is not the name of a constructor of a class declared in the exported namespace of L.

A token of the form [s](uri) where s is a string will be replaced by a link in the formatted output. The link will point at the URI given by uri. The title of the link will be the value of the string s.

A token of the form <uri> will be replaced by a link in the formatted output. The link will point at the location specified by uri.

A token of the form ## indicates that the following text, up to the nearest newline, will appear as a heading in the formatted output. If the line in question ends with ##, those two characters are elided in the formatted output.

A token of the form * indicates that the following text, up to the nearest blank line, will appear as a bulleted item in the formatted output.

A token of the form n.  where n is an integer indicates that the following text, up to the nearest blank line, will appear as a numbered item in the formatted output. The number used will be n.

### Operator Precedence

Operator precedence is given implicitly by the grammar.

The following non-normative table may be helpful:

 Description Operator Associativity Precedence Unary postfix . , ?id, e++, e--, e1[e2], e1(), () None 15 Unary prefix -e, !e, ~e, ++e, --e None 14 Multiplicative *, /, ~/, % Left 13 Additive +, - Left 12 Shift <<, >> Left 11 Relational <, >, <=, >=, as, is None 10 Equality ==, != None 9 Bitwise AND & Left 8 Bitwise XOR ^ Left 7 Bitwise Or | Left 6 Logical And && Left 5 Logical Or || Left 4 Conditional e1? e2 : e3 None 3 Cascade .. Left 2 Assignment =, *=, /=, +=, =+, ~=, %=, <<=, >>>=, >>=, &=, ^= Right 1

## Appendix: Naming Conventions

The following naming conventions are customary in Dart programs.

1. The names of compile time constant variables never use lower case letters. If they consist of multiple words, those words are separated by underscores. Examples: PI, I_AM_A_CONSTANT.
2. The names of functions (including getters, setters, methods and local or library functions) and non-constant variables begin with a lowercase letter. If the name consists of multiple words, each word (except the first) begins with an uppercase letter. No other uppercase letters are used. Examples: camlCase, , dart4TheWorld.
3. The names of types (including classes, type variables and type aliases) begin with an uppercase letter. If the name consists of multiple words, each word begins with an uppercase letter. No other uppercase letters are used. Examples: CamlCase, Dart4TheWorld.
4. The names of type variables are short (preferably single letter). Examples: T, S, K, V , E.
5. The names of libraries or library prefixes never use upper case letters. If they consist of multiple words, those words are separated by underscores. Example: my_favorite_library.