Dart 速查表

Dart 语言旨在让从其他编程语言转来的开发者们能够轻松学习，但也有它的独特之处。本篇将为你介绍一些最重要的语言特性。

这篇教程中的嵌入式编辑器已经完成了部分代码片段。你可以在这些编辑器上将代码补充完整，然后点击 Run (运行) 按钮进行测试。这些编辑器上还包含了健全的测试代码；你可以随时研究这些代码来学习测试方面的知识，但 不要编辑测试代码。

为了将表达式的值放在字符串中，请使用 ${expression}。若表达式为单个标识符，则可以省略 {}。

下面是一些使用字符串插值的例子：

下面的方法接收两个整型变量作为参数，然后让它返回一个包含以空格分隔的整数的字符串。例如，stringify(2, 3) 应该返回 '2 3'。

String stringify(int x, int y) {
  TODO('Return a formatted string here');
}


// Tests your solution (Don't edit!): 
void main() {
  assert(stringify(2, 3) == '2 3',
      "Your stringify method returned '${stringify(2, 3)}' instead of '2 3'");
  print('Success!');
}

String stringify(int x, int y) {
  return '$x $y';
}

Dart 要求使用健全的空安全，这意味着除非变量显式声明为可空类型，否则它们将不能为空。换句话说，类型默认是不可为空的。

举个例子，下面的代码在空安全下是有错误的，因为 int 类型的变量不能为 null：

int a = null; // INVALID.

你可以通过在类型后添加 ? 来表示该类型可空：

int? a = null; // Valid.

在所有 Dart 版本中，null 在未初始化的变量里都是默认值，所以你可以这样简化你的代码：

int? a; // The initial value of a is null.

想了解更多有关 Dart 的空安全的内容，请阅读 健全的空安全。

在此 DartPad 中声明以下两种变量：

• 一个可空的 String，名为 name，值为 'Jane'。

• 一个可空的 String，名为 address，值为 null。

可以忽略以下代码一开始在 DartPad 中的错误。

// TODO: Declare the two variables here


// Tests your solution (Don't edit!): 
void main() {
  try {
    if (name == 'Jane' && address == null) {
      // verify that "name" is nullable
      name = null;
      print('Success!');
    } else {
      print('Not quite right, try again!');
    }
  } catch (e) {
    print('Exception: ${e.runtimeType}');
  }
}

String? name = 'Jane';
String? address;

Dart 提供了一系列方便的运算符用于处理可能会为空值的变量。其中一个是 ??= 赋值运算符，仅当该变量为空值时才为其赋值：

int? a; // = null
a ??= 3;
print(a); // <-- Prints 3.

a ??= 5;
print(a); // <-- Still prints 3.

另外一个避空运算符是 ??，如果该运算符左边的表达式返回的是空值，则会计算并返回右边的表达式。

print(1 ?? 3); // <-- Prints 1.
print(null ?? 12); // <-- Prints 12.

尝试在下面的代码片段中交替使用 ??= 和 ?? 操作符，实现期望的需求。

可以忽略以下代码一开始在 DartPad 中的错误。

String? foo = 'a string';
String? bar; // = null

// Substitute an operator that makes 'a string' be assigned to baz.
String? baz = foo /* TODO */ bar;

void updateSomeVars() {
  // Substitute an operator that makes 'a string' be assigned to bar.
  bar /* TODO */ 'a string';
}


// Tests your solution (Don't edit!):
void main() {
  try {
    updateSomeVars();
    
    if (foo != 'a string') {
      print('Looks like foo somehow ended up with the wrong value.');
    } else if (bar != 'a string') {
      print('Looks like bar ended up with the wrong value.');
    } else if (baz != 'a string') {
      print('Looks like baz ended up with the wrong value.');
    } else {
      print('Success!');
    }
  } catch (e) {
    print('Exception: ${e.runtimeType}.');
  }
  
}

// Substitute an operator that makes 'a string' be assigned to baz.
String? baz = foo ?? bar;

void updateSomeVars() {
  // Substitute an operator that makes 'a string' be assigned to bar.
  bar ??= 'a string';
}

要保护可能会为空的属性的正常访问，请在点（.）之前加一个问号（?）。

myObject?.someProperty

上述代码等效于以下内容：

(myObject != null) ? myObject.someProperty : null

你可以在一个表达式中连续使用多个 ?.：

myObject?.someProperty?.someMethod()

如果 myObject 或 myObject.someProperty 为空，则前面的代码返回 null（并不再调用 someMethod）。

下面的函数将一个可空的字符串作为参数。请尝试使用条件属性访问来让它返回 str 的大写形式，如果 str 为 null 则返回 null。

String? upperCaseIt(String? str) {
  // TODO: Try conditionally accessing the `toUpperCase` method here.
}


// Tests your solution (Don't edit!):
void main() {
  try {
    String? one = upperCaseIt(null);
    if (one != null) {
      print('Looks like you\'re not returning null for null inputs.');
    } else {
      print('Success when str is null!');
    }
  } catch (e) {
    print('Tried calling upperCaseIt(null) and got an exception: \n ${e.runtimeType}.');
  }
  
  try {
    String? two = upperCaseIt('a string');
    if (two == null) {
      print('Looks like you\'re returning null even when str has a value.');
    } else if (two != 'A STRING') {
      print('Tried upperCaseIt(\'a string\'), but didn\'t get \'A STRING\' in response.');
    } else {
      print('Success when str is not null!');
    }
  } catch (e) {
    print('Tried calling upperCaseIt(\'a string\') and got an exception: \n ${e.runtimeType}.');
  }
}

String? upperCaseIt(String? str) {
  return str?.toUpperCase();
}

Dart 内置了对 list、map 以及 set 的支持。你可以通过字面量直接创建它们：

final aListOfStrings = ['one', 'two', 'three'];
final aSetOfStrings = {'one', 'two', 'three'};
final aMapOfStringsToInts = {
  'one': 1,
  'two': 2,
  'three': 3,
};

Dart 的类型推断可以自动帮你分配这些变量的类型。在这个例子中，推断类型是 List<String>、Set<String>和 Map<String, int>。

你也可以手动指定类型：

final aListOfInts = <int>[];
final aSetOfInts = <int>{};
final aMapOfIntToDouble = <int, double>{};

在使用子类型的内容初始化列表，但仍希望列表为 List <BaseType> 时，指定其类型很方便：

final aListOfBaseType = <BaseType>[SubType(), SubType()];

尝试将以下变量设定为指定的值。替换当前的 null 值。

// Assign this a list containing 'a', 'b', and 'c' in that order:
final aListOfStrings = null;

// Assign this a set containing 3, 4, and 5:
final aSetOfInts = null;

// Assign this a map of String to int so that aMapOfStringsToInts['myKey'] returns 12:
final aMapOfStringsToInts = null;

// Assign this an empty List<double>:
final anEmptyListOfDouble = null;

// Assign this an empty Set<String>:
final anEmptySetOfString = null;

// Assign this an empty Map of double to int:
final anEmptyMapOfDoublesToInts = null;


// Tests your solution (Don't edit!):
void main() {
  final errs = <String>[];
  
  if (aListOfStrings is! List<String>) {
    errs.add('aListOfStrings should have the type List<String>.');
  } else if (aListOfStrings.length != 3) {
    errs.add('aListOfStrings has ${aListOfStrings.length} items in it, \n rather than the expected 3.');
  } else if (aListOfStrings[0] != 'a' || aListOfStrings[1] != 'b' || aListOfStrings[2] != 'c') {
    errs.add('aListOfStrings doesn\'t contain the correct values (\'a\', \'b\', \'c\').');
  }

  if (aSetOfInts is! Set<int>) {
    errs.add('aSetOfInts should have the type Set<int>.');
  } else if (aSetOfInts.length != 3) {
    errs.add('aSetOfInts has ${aSetOfInts.length} items in it, \n rather than the expected 3.');
  } else if (!aSetOfInts.contains(3) || !aSetOfInts.contains(4) || !aSetOfInts.contains(5)) {
    errs.add('aSetOfInts doesn\'t contain the correct values (3, 4, 5).');
  }

  if (aMapOfStringsToInts is! Map<String, int>) {
    errs.add('aMapOfStringsToInts should have the type Map<String, int>.');
  } else if (aMapOfStringsToInts['myKey'] != 12) {
    errs.add('aMapOfStringsToInts doesn\'t contain the correct values (\'myKey\': 12).');
  }

  if (anEmptyListOfDouble is! List<double>) {
    errs.add('anEmptyListOfDouble should have the type List<double>.');
  } else if (anEmptyListOfDouble.isNotEmpty) {
    errs.add('anEmptyListOfDouble should be empty.');
  }

  if (anEmptySetOfString is! Set<String>) {
    errs.add('anEmptySetOfString should have the type Set<String>.');
  } else if (anEmptySetOfString.isNotEmpty) {
    errs.add('anEmptySetOfString should be empty.');
  }

  if (anEmptyMapOfDoublesToInts is! Map<double, int>) {
    errs.add('anEmptyMapOfDoublesToInts should have the type Map<double, int>.');
  } else if (anEmptyMapOfDoublesToInts.isNotEmpty) {
    errs.add('anEmptyMapOfDoublesToInts should be empty.');
  }

  if (errs.isEmpty) {
    print('Success!');
  } else {
    errs.forEach(print);
  }

  // ignore_for_file: unnecessary_type_check
}

// Assign this a list containing 'a', 'b', and 'c' in that order:
final aListOfStrings = ['a', 'b', 'c'];

// Assign this a set containing 3, 4, and 5:
final aSetOfInts = {3, 4, 5};

// Assign this a map of String to int so that aMapOfStringsToInts['myKey'] returns 12:
final aMapOfStringsToInts = {'myKey': 12};

// Assign this an empty List<double>:
final anEmptyListOfDouble = <double>[];

// Assign this an empty Set<String>:
final anEmptySetOfString = <String>{};

// Assign this an empty Map of double to int:
final anEmptyMapOfDoublesToInts = <double, int>{};

你也许已经在 Dart 代码中见到过 => 符号。这种箭头语法是一种定义函数的方法，该函数将在其右侧执行表达式并返回其值。

例如，考虑调用这个 List 类中的 any 方法：

bool hasEmpty = aListOfStrings.any((s) {
  return s.isEmpty;
});

这里是一个更简单的代码实现：

bool hasEmpty = aListOfStrings.any((s) => s.isEmpty);

尝试使用箭头语法完成下面语句：

class MyClass {
  int value1 = 2;
  int value2 = 3;
  int value3 = 5;
  
  // Returns the product of the above values:
  int get product => TODO();
  
  // Adds 1 to value1:
  void incrementValue1() => TODO();
  
  // Returns a string containing each item in the
  // list, separated by commas (e.g. 'a,b,c'): 
  String joinWithCommas(List<String> strings) => TODO();
}


// Tests your solution (Don't edit!):
void main() {
  final obj = MyClass();
  final errs = <String>[];
  
  try {
    final product = obj.product;
    
    if (product != 30) {
      errs.add('The product property returned $product \n instead of the expected value (30).'); 
    } 
  } catch (e) {
    print('Tried to use MyClass.product, but encountered an exception: \n ${e.runtimeType}.');
    return;
  }

  try {
    obj.incrementValue1();
    
    if (obj.value1 != 3) {
      errs.add('After calling incrementValue, value1 was ${obj.value1} \n instead of the expected value (3).'); 
    } 
  } catch (e) {
    print('Tried to use MyClass.incrementValue1, but encountered an exception: \n ${e.runtimeType}.');
    return;
  }

  try {
    final joined = obj.joinWithCommas(['one', 'two', 'three']);
    
    if (joined != 'one,two,three') {
      errs.add('Tried calling joinWithCommas([\'one\', \'two\', \'three\']) \n and received $joined instead of the expected value (\'one,two,three\').'); 
    } 
  } catch (e) {
    print('Tried to use MyClass.joinWithCommas, but encountered an exception: \n ${e.runtimeType}.');
    return;
  }

  if (errs.isEmpty) {
    print('Success!');
  } else {
    errs.forEach(print);
  }
}

class MyClass {
  int value1 = 2;
  int value2 = 3;
  int value3 = 5;

  // Returns the product of the above values:
  int get product => value1 * value2 * value3;
  
  // Adds 1 to value1:
  void incrementValue1() => value1++; 
  
  // Returns a string containing each item in the
  // list, separated by commas (e.g. 'a,b,c'): 
  String joinWithCommas(List<String> strings) => strings.join(',');
}

要对同一对象执行一系列操作，请使用级联（..）。我们都看到过这样的表达式：

myObject.someMethod()

它在 myObject 上调用 someMethod 方法，而表达式的结果是 someMethod 的返回值。

下面是一个使用级连语法的相同表达式：

myObject..someMethod()

Although it still invokes someMethod() on myObject, the result of the expression isn't the return value—it's a reference to myObject!

虽然它仍然在 myObject 上调用了 someMethod，但表达式的结果却不是该方法返回值，而是是 myObject 对象的引用！使用级联，你可以将需要单独操作的语句链接在一起。例如，下方的代码使用了空判断调用符 (?.) 在 button 不为 null 时获取属性：

var button = querySelector('#confirm');
button?.text = 'Confirm';
button?.classes.add('important');
button?.onClick.listen((e) => window.alert('Confirmed!'));
button?.scrollIntoView();

现在你可以在第一个级联位置，使用 空判断 级联操作符 (?..)，它可以确保级联操作均在实例不为 null 时执行。使用空判断级联后，你也不再需要 button 变量了：

querySelector('#confirm')
  ?..text = 'Confirm'
  ..classes.add('important')
  ..onClick.listen((e) => window.alert('Confirmed!'))
  ..scrollIntoView();

使用级联创建一个语句，分别将 BigObject 的 anInt 属性设为 1、aString 属性设为 String!、aList 属性设置为 [3.0] 然后调用 allDone()。

class BigObject {
  int anInt = 0;
  String aString = '';
  List<double> aList = [];
  bool _done = false;
  
  void allDone() {
    _done = true;
  }
}

BigObject fillBigObject(BigObject obj) {
  // Create a single statement that will update and return obj:
  return TODO('obj..');
}


// Tests your solution (Don't edit!):
void main() {
  BigObject obj;

  try {
    obj = fillBigObject(BigObject());
  } catch (e) {
    print('Caught an exception of type ${e.runtimeType} \n while running fillBigObject');
    return;
  }

  final errs = <String>[];

  if (obj.anInt != 1) {
    errs.add(
        'The value of anInt was ${obj.anInt} \n rather than the expected (1).');
  }

  if (obj.aString != 'String!') {
    errs.add(
        'The value of aString was \'${obj.aString}\' \n rather than the expected (\'String!\').');
  }

  if (obj.aList.length != 1) {
    errs.add(
        'The length of aList was ${obj.aList.length} \n rather than the expected value (1).');
  } else {
    if (obj.aList[0] != 3.0) {
      errs.add(
          'The value found in aList was ${obj.aList[0]} \n rather than the expected (3.0).');
    }
  }
  
  if (!obj._done) {
    errs.add('It looks like allDone() wasn\'t called.');
  }

  if (errs.isEmpty) {
    print('Success!');
  } else {
    errs.forEach(print);
  }
}

BigObject fillBigObject(BigObject obj) {
  return obj
    ..anInt = 1
    ..aString = 'String!'
    ..aList.add(3)
    ..allDone();
}

任何需要对属性进行更多控制而不是允许简单字段访问的时候，你都可以自定义 getter 和 setter。

例如，你可以用来确保属性值合法：

class MyClass {
  int _aProperty = 0;

  int get aProperty => _aProperty;

  set aProperty(int value) {
    if (value >= 0) {
      _aProperty = value;
    }
  }
}

你还可以使用 getter 来定义计算属性：

class MyClass {
  final List<int> _values = [];

  void addValue(int value) {
    _values.add(value);
  }

  // A computed property.
  int get count {
    return _values.length;
  }
}

想象你有一个购物车类，其中有一个私有的 List<double> 类型的 prices 属性。添加以下内容：

• 一个名为 total 的 getter，用于返回总价格。

• 只要新列表不包含任何负价格， setter 就会用新的列表替换列表（在这种情况下，setter 应该抛出 InvalidPriceException）。

可以忽略以下代码一开始在 DartPad 中的错误。

class InvalidPriceException {}

class ShoppingCart {
  List<double> _prices = [];
  
  // TODO: Add a "total" getter here:

  // TODO: Add a "prices" setter here:
}


// Tests your solution (Don't edit!):
void main() {
  var foundException = false;
  
  try {
    final cart = ShoppingCart();
    cart.prices = [12.0, 12.0, -23.0];
  } on InvalidPriceException {
    foundException = true;
  } catch (e) {
    print('Tried setting a negative price and received a ${e.runtimeType} \n instead of an InvalidPriceException.');
    return;
  }
  
  if (!foundException) {
    print('Tried setting a negative price \n and didn\'t get an InvalidPriceException.');
    return;
  }
  
  final secondCart = ShoppingCart();
  
  try {
    secondCart.prices = [1.0, 2.0, 3.0];
  } catch(e) {
    print('Tried setting prices with a valid list, \n but received an exception: ${e.runtimeType}.');
    return;
  }
  
  if (secondCart._prices.length != 3) {
    print('Tried setting prices with a list of three values, \n but _prices ended up having length ${secondCart._prices.length}.');
    return;
  }

  if (secondCart._prices[0] != 1.0 || secondCart._prices[1] != 2.0 || secondCart._prices[2] != 3.0) {
    final vals = secondCart._prices.map((p) => p.toString()).join(', ');
    print('Tried setting prices with a list of three values (1, 2, 3), \n but incorrect ones ended up in the price list ($vals) .');
    return;
  }
  
  var sum = 0.0;
  
  try {
    sum = secondCart.total;
  } catch (e) {
    print('Tried to get total, but received an exception: ${e.runtimeType}.');
    return;
  }
  
  if (sum != 6.0) {
    print('After setting prices to (1, 2, 3), total returned $sum instead of 6.');
    return;
  }
  
  print('Success!');
}

// Add a "total" getter here:
double get total => _prices.fold(0, (e, t) => e + t);

// Add a "prices" setter here:
set prices(List<double> value) {
  if (value.any((p) => p < 0)) {
    throw InvalidPriceException();
  }
  
  _prices = value;
}

Dart 有两种传参方法：位置参数和命名参数。位置参数你可能会比较熟悉：

int sumUp(int a, int b, int c) {
  return a + b + c;
}
  // ···
  int total = sumUp(1, 2, 3);

在 Dart 里，你可以将这些参数包裹在方括号中，使其变成可选位置参数：

int sumUpToFive(int a, [int? b, int? c, int? d, int? e]) {
  int sum = a;
  if (b != null) sum += b;
  if (c != null) sum += c;
  if (d != null) sum += d;
  if (e != null) sum += e;
  return sum;
}
  // ···
  int total = sumUpToFive(1, 2);
  int otherTotal = sumUpToFive(1, 2, 3, 4, 5);

可选位置参数永远放在方法参数列表的最后。除非你给它们提供一个默认值，否则默认为 null:

int sumUpToFive(int a, [int b = 2, int c = 3, int d = 4, int e = 5]) {
  // ···
}

void main() {
  int newTotal = sumUpToFive(1);
  print(newTotal); // <-- prints 15
}

实现一个名为 joinWithCommas 的方法，它接收一至五个整数，然后返回由逗号分隔的包含这些数字的字符串。以下是方法调用和返回值的一些示例：

String joinWithCommas(int a, [int? b, int? c, int? d, int? e]) {
  return TODO();
}


// Tests your solution (Don't edit!):
void main() {
  final errs = <String>[];
  
  try {
    final value = joinWithCommas(1);
    
    if (value != '1') {
      errs.add('Tried calling joinWithCommas(1) \n and got $value instead of the expected (\'1\').'); 
    } 
  } on UnimplementedError {
    print('Tried to call joinWithCommas but failed. \n Did you implement the method?');
    return;
  } catch (e) {
    print('Tried calling joinWithCommas(1), \n but encountered an exception: ${e.runtimeType}.');
    return;
  }

  try {
    final value = joinWithCommas(1, 2, 3);
    
    if (value != '1,2,3') {
      errs.add('Tried calling joinWithCommas(1, 2, 3) \n and got $value instead of the expected (\'1,2,3\').'); 
    } 
  } on UnimplementedError {
    print('Tried to call joinWithCommas but failed. \n Did you implement the method?');
    return;
  } catch (e) {
    print('Tried calling joinWithCommas(1, 2 ,3), \n but encountered an exception: ${e.runtimeType}.');
    return;
  }

  try {
    final value = joinWithCommas(1, 2, 3, 4, 5);
    
    if (value != '1,2,3,4,5') {
      errs.add('Tried calling joinWithCommas(1, 2, 3, 4, 5) \n and got $value instead of the expected (\'1,2,3,4,5\').'); 
    } 
  } on UnimplementedError {
    print('Tried to call joinWithCommas but failed. \n Did you implement the method?');
    return;
  } catch (e) {
    print('Tried calling stringify(1, 2, 3, 4 ,5), \n but encountered an exception: ${e.runtimeType}.');
    return;
  }

  if (errs.isEmpty) {
    print('Success!');
  } else {
    errs.forEach(print);
  }
}

String joinWithCommas(int a, [int? b, int? c, int? d, int? e]) {
  var total = '$a';
  if (b != null) total = '$total,$b';
  if (c != null) total = '$total,$c';
  if (d != null) total = '$total,$d';
  if (e != null) total = '$total,$e';
  return total;
}

你可以在参数列表的靠后位置使用花括号 ({}) 来定义命名参数。

除非显式使用 required 进行标记，否则命名参数默认是可选的。

void printName(String firstName, String lastName, {String? middleName}) {
  print('$firstName ${middleName ?? ''} $lastName');
}

void main() {
  printName('Dash', 'Dartisan');
  printName('John', 'Smith', middleName: 'Who');
  // Named arguments can be placed anywhere in the argument list
  printName('John', middleName: 'Who', 'Smith');
}

正如你所料，这些参数默认为 null，但你也可以为其提供默认值。

如果一个参数的类型是非空的，那么你必须要提供一个默认值（如下方代码所示），或者将其标记为 required（如 构造部分所示）。

void printName(String firstName, String lastName, {String middleName = ''}) {
  print('$firstName $middleName $lastName');
}

一个方法不能同时使用可选位置参数和可选命名参数。

向 MyDataObject 类添加一个 copyWith() 实例方法，它应该包含三个可空的命名参数。

• int? newInt
• String? newString
• double? newDouble

copyWith 方法应该根据当前实例返回一个新的 MyDataObject 并将前面参数（如果有的话）的数据复制到对象的属性中。例如，如果 newInt 不为空，则将其值复制到 anInt 中。

可以忽略以下代码一开始在 DartPad 中的错误。

class MyDataObject {
  final int anInt;
  final String aString;
  final double aDouble;

  MyDataObject({
     this.anInt = 1,
     this.aString = 'Old!',
     this.aDouble = 2.0,
  });

  // TODO: Add your copyWith method here:
}


// Tests your solution (Don't edit!):
void main() {
  final source = MyDataObject();
  final errs = <String>[];
  
  try {
    final copy = source.copyWith(newInt: 12, newString: 'New!', newDouble: 3.0);
    
    if (copy.anInt != 12) {
      errs.add('Called copyWith(newInt: 12, newString: \'New!\', newDouble: 3.0), \n and the new object\'s anInt was ${copy.anInt} rather than the expected value (12).');
    }
    
    if (copy.aString != 'New!') {
      errs.add('Called copyWith(newInt: 12, newString: \'New!\', newDouble: 3.0), \n and the new object\'s aString was ${copy.aString} rather than the expected value (\'New!\').');
    }
    
    if (copy.aDouble != 3) {
      errs.add('Called copyWith(newInt: 12, newString: \'New!\', newDouble: 3.0), \n and the new object\'s aDouble was ${copy.aDouble} rather than the expected value (3).');
    }
  } catch (e) {
    print('Called copyWith(newInt: 12, newString: \'New!\', newDouble: 3.0) \n and got an exception: ${e.runtimeType}');
  }
  
  try {
    final copy = source.copyWith();
    
    if (copy.anInt != 1) {
      errs.add('Called copyWith(), and the new object\'s anInt was ${copy.anInt} \n rather than the expected value (1).');
    }
    
    if (copy.aString != 'Old!') {
      errs.add('Called copyWith(), and the new object\'s aString was ${copy.aString} \n rather than the expected value (\'Old!\').');
    }
    
    if (copy.aDouble != 2) {
      errs.add('Called copyWith(), and the new object\'s aDouble was ${copy.aDouble} \n rather than the expected value (2).');
    }
  } catch (e) {
    print('Called copyWith() and got an exception: ${e.runtimeType}');
  }
  
  if (errs.isEmpty) {
    print('Success!');
  } else {
    errs.forEach(print);
  }
}

  MyDataObject copyWith({int? newInt, String? newString, double? newDouble}) {
    return MyDataObject(
      anInt: newInt ?? this.anInt,
      aString: newString ?? this.aString,
      aDouble: newDouble ?? this.aDouble,
    );
  }

Dart 代码可以抛出和捕获异常。与 Java 不同的是，Dart 的所有异常都是 unchecked exception。方法不会声明它们可能抛出的异常，你也不需要捕获任何异常。

虽然 Dart 提供了 Exception 和 Error 类型，但是你可以抛出任何非空对象：

throw Exception('Something bad happened.');
throw 'Waaaaaaah!';

使用 try、on 以及 catch 关键字来处理异常：

try {
  breedMoreLlamas();
} on OutOfLlamasException {
  // A specific exception
  buyMoreLlamas();
} on Exception catch (e) {
  // Anything else that is an exception
  print('Unknown exception: $e');
} catch (e) {
  // No specified type, handles all
  print('Something really unknown: $e');
}

The try keyword works as it does in most other languages. Use the on keyword to filter for specific exceptions by type, and the catch keyword to get a reference to the exception object.

如果你无法完全处理该异常，请使用 rethrow 关键字再次抛出异常：

try {
  breedMoreLlamas();
} catch (e) {
  print('I was just trying to breed llamas!');
  rethrow;
}

要执行一段无论是否抛出异常都会执行的代码，请使用 finally：

try {
  breedMoreLlamas();
} catch (e) {
  // ... handle exception ...
} finally {
  // Always clean up, even if an exception is thrown.
  cleanLlamaStalls();
}

在下面实现 tryFunction() 方法。它应该会执行一个不可靠的方法，然后做以下操作：

• 如果 untrustworthy() 抛出了 ExceptionWithMessage，则调用 logger.logException 并传入使用异常类型和消息（尝试使用 on 和 catch）。

• 如果 untrustworthy() 抛出了一个 Exception，则调用 logger.logException 并传入使用异常类型（这次请尝试使用 on）。

• 如果 untrustworthy() 抛出了其他对象，请不要捕获该异常。

• 捕获并处理完所有内容后，调用 logger.doneLogging（尝试使用 finally）。

typedef VoidFunction = void Function();

class ExceptionWithMessage {
  final String message;
  const ExceptionWithMessage(this.message);
}

// Call logException to log an exception, and doneLogging when finished.
abstract class Logger {
  void logException(Type t, [String? msg]);
  void doneLogging();
}

void tryFunction(VoidFunction untrustworthy, Logger logger) {
  try {
    untrustworthy();
  } on ExceptionWithMessage catch (e) {
    logger.logException(e.runtimeType, e.message); 
  } on Exception catch (e) {
    logger.logException(e.runtimeType);
  } finally {
    logger.doneLogging();
  }
}

// Tests your solution (Don't edit!):
class MyLogger extends Logger {
  Type? lastType;
  String lastMessage = '';
  bool done = false;
  
  void logException(Type t, [String? message]) {
    lastType = t;
    lastMessage = message ?? lastMessage;
  }
  
  void doneLogging() => done = true;  
}

void main() {
  final errs = <String>[];
  var logger = MyLogger();
  
  try {
    tryFunction(() => throw Exception(), logger);
  
    if ('${logger.lastType}' != 'Exception' && '${logger.lastType}' != '_Exception') {
      errs.add('Untrustworthy threw an Exception, but a different type was logged: \n ${logger.lastType}.');
    }
    
    if (logger.lastMessage != '') {
      errs.add('Untrustworthy threw an Exception with no message, but a message \n was logged anyway: \'${logger.lastMessage}\'.');
    }
    
    if (!logger.done) {
      errs.add('Untrustworthy threw an Exception, \n and doneLogging() wasn\'t called afterward.');
    }
  } catch (e) {
    print('Untrustworthy threw an exception, and an exception of type \n ${e.runtimeType} was unhandled by tryFunction.');
  }
  
  logger = MyLogger();
  
  try {
    tryFunction(() => throw ExceptionWithMessage('Hey!'), logger);
  
    if (logger.lastType != ExceptionWithMessage) {
      errs.add('Untrustworthy threw an ExceptionWithMessage(\'Hey!\'), but a \n different type was logged: ${logger.lastType}.');
    }
    
    if (logger.lastMessage != 'Hey!') {
      errs.add('Untrustworthy threw an ExceptionWithMessage(\'Hey!\'), but a \n different message was logged: \'${logger.lastMessage}\'.');
    }
    
    if (!logger.done) {
      errs.add('Untrustworthy threw an ExceptionWithMessage(\'Hey!\'), \n and doneLogging() wasn\'t called afterward.');
    }
  } catch (e) {
    print('Untrustworthy threw an ExceptionWithMessage(\'Hey!\'), \n and an exception of type ${e.runtimeType} was unhandled by tryFunction.');
  }
  
  logger = MyLogger();
  bool caughtStringException = false;

  try {
    tryFunction(() => throw 'A String', logger);
  } on String {
    caughtStringException = true;
  }

  if (!caughtStringException) {
    errs.add('Untrustworthy threw a string, and it was incorrectly handled inside tryFunction().');
  }
  
  logger = MyLogger();
  
  try {
    tryFunction(() {}, logger);
  
    if (logger.lastType != null) {
      errs.add('Untrustworthy didn\'t throw an Exception, \n but one was logged anyway: ${logger.lastType}.');
    }
    
    if (logger.lastMessage != '') {
      errs.add('Untrustworthy didn\'t throw an Exception with no message, \n but a message was logged anyway: \'${logger.lastMessage}\'.');
    }
    
    if (!logger.done) {
      errs.add('Untrustworthy didn\'t throw an Exception, \n but doneLogging() wasn\'t called afterward.');
    }
  } catch (e) {
    print('Untrustworthy didn\'t throw an exception, \n but an exception of type ${e.runtimeType} was unhandled by tryFunction anyway.');
  }
  
  if (errs.isEmpty) {
    print('Success!');
  } else {
    errs.forEach(print);
  }
}

void tryFunction(VoidFunction untrustworthy, Logger logger) {
  try {
    untrustworthy();
  } on ExceptionWithMessage catch (e) {
    logger.logException(e.runtimeType, e.message);
  } on Exception {
    logger.logException(Exception);
  } finally {
    logger.doneLogging();
  }
}

Dart 提供了一个方便的快捷方式，用于为构造方法中的属性赋值：在声明构造方法时使用 this.propertyName。

class MyColor {
  int red;
  int green;
  int blue;

  MyColor(this.red, this.green, this.blue);
}

final color = MyColor(80, 80, 128);

此技巧同样也适用于命名参数。属性名为参数的名称：

class MyColor {
  ...

  MyColor({required this.red, required this.green, required this.blue});
}

final color = MyColor(red: 80, green: 80, blue: 80);

在上面的代码中，red、green 和 blue 被标记为 required，因为这些 int 数值不能为空。如果你指定了默认值，你可以忽略 required。

对于可选参数，默认值为期望值：

MyColor([this.red = 0, this.green = 0, this.blue = 0]);
// or
MyColor({this.red = 0, this.green = 0, this.blue = 0});

使用 this 语法向 MyClass 添加一行构造方法，并接收和分配全部（三个）属性。

可以忽略以下代码一开始在 DartPad 中的错误。

class MyClass {
  final int anInt;
  final String aString;
  final double aDouble;
  
  // TODO: Create the constructor here.
}


// Tests your solution (Don't edit!):
void main() {
  final errs = <String>[];
  
  try {
    final obj = MyClass(1, 'two', 3);
    
    if (obj.anInt != 1) {
      errs.add('Called MyClass(1, \'two\', 3) and got an object with anInt of ${obj.anInt} \n instead of the expected value (1).');
    }

    if (obj.anInt != 1) {
      errs.add('Called MyClass(1, \'two\', 3) and got an object with aString of \'${obj.aString}\' \n instead of the expected value (\'two\').');
    }

    if (obj.anInt != 1) {
      errs.add('Called MyClass(1, \'two\', 3) and got an object with aDouble of ${obj.aDouble} \n instead of the expected value (3).');
    }
  } catch (e) {
    print('Called MyClass(1, \'two\', 3) and got an exception \n of type ${e.runtimeType}.');
  }
  
  if (errs.isEmpty) {
    print('Success!');
  } else {
    errs.forEach(print);
  }
}

MyClass(this.anInt, this.aString, this.aDouble);

有时，当你在实现构造函数时，你需要在构造函数体执行之前进行一些初始化。例如，final 修饰的字段必须在构造函数体执行之前赋值。在初始化列表中执行此操作，该列表位于构造函数的签名与其函数体之间：

Point.fromJson(Map<String, double> json)
    : x = json['x']!,
      y = json['y']! {
  print('In Point.fromJson(): ($x, $y)');
}

初始化列表也是放置断言的便利位置，它仅会在开发期间运行：

NonNegativePoint(this.x, this.y)
    : assert(x >= 0),
      assert(y >= 0) {
  print('I just made a NonNegativePoint: ($x, $y)');
}

完成下面的 FirstTwoLetters 的构造函数。使用的初始化列表将 word 的前两个字符分配给 letterOne 和 LetterTwo 属性。要获得额外的信用，请添加一个 断言 以捕获少于两个字符的单词。

可以忽略以下代码一开始在 DartPad 中的错误。

class FirstTwoLetters {
  final String letterOne;
  final String letterTwo;

  // TODO: Create a constructor with an initializer list here:
  FirstTwoLetters(String word)

}


// Tests your solution (Don't edit!):
void main() {
  final errs = <String>[];

  try {
    final result = FirstTwoLetters('My String');
    
    if (result.letterOne != 'M') {
      errs.add('Called FirstTwoLetters(\'My String\') and got an object with \n letterOne equal to \'${result.letterOne}\' instead of the expected value (\'M\').');
    }

    if (result.letterTwo != 'y') {
      errs.add('Called FirstTwoLetters(\'My String\') and got an object with \n letterTwo equal to \'${result.letterTwo}\' instead of the expected value (\'y\').');
    }
  } catch (e) {
    errs.add('Called FirstTwoLetters(\'My String\') and got an exception \n of type ${e.runtimeType}.');
  }

  bool caughtException = false;
  
  try {
    FirstTwoLetters('');
  } catch (e) {
    caughtException = true;
  }
  
  if (!caughtException) {
    errs.add('Called FirstTwoLetters(\'\') and didn\'t get an exception \n from the failed assertion.');
  }
  
  if (errs.isEmpty) {
    print('Success!');
  } else {
    errs.forEach(print);
  }
}

  FirstTwoLetters(String word)
      : assert(word.length >= 2),
        letterOne = word[0],
        letterTwo = word[1];

为了允许一个类具有多个构造方法， Dart 支持命名构造方法：

class Point {
  double x, y;

  Point(this.x, this.y);

  Point.origin()
      : x = 0,
        y = 0;
}

为了使用命名构造方法，请使用全名调用它：

final myPoint = Point.origin();

给 Color 类添加一个叫做 Color.black 的方法，它将会把三个属性的值都设为 0。

可以忽略以下代码一开始在 DartPad 中的错误。

class Color {
  int red;
  int green;
  int blue;
  
  Color(this.red, this.green, this.blue);

  // TODO: Create a named constructor called "Color.black" here:

}


// Tests your solution (Don't edit!):
void main() {
  final errs = <String>[];

  try {
    final result = Color.black();
    
    if (result.red != 0) {
      errs.add('Called Color.black() and got a Color with red equal to \n ${result.red} instead of the expected value (0).');
    }

    if (result.green != 0) {
      errs.add('Called Color.black() and got a Color with green equal to \n ${result.green} instead of the expected value (0).');
    }

    if (result.blue != 0) {
  errs.add('Called Color.black() and got a Color with blue equal to \n ${result.blue} instead of the expected value (0).');
    }
  } catch (e) {
    print('Called Color.black() and got an exception of type \n ${e.runtimeType}.');
    return;
  }

  if (errs.isEmpty) {
    print('Success!');
  } else {
    errs.forEach(print);
  }
}

  Color.black()
      : red = 0,
        green = 0,
        blue = 0;

Dart 支持工厂构造方法。它能够返回其子类甚至 null 对象。要创建一个工厂构造方法，请使用 factory 关键字。

class Square extends Shape {}

class Circle extends Shape {}

class Shape {
  Shape();

  factory Shape.fromTypeName(String typeName) {
    if (typeName == 'square') return Square();
    if (typeName == 'circle') return Circle();

    throw ArgumentError('Unrecognized $typeName');
  }
}

替换名为 IntegerHolder.fromList 工厂构造函数中的 TODO(); 行，使其执行以下操作：

• 如果列表只有一个值，那么就用它来创建一个 IntegerSingle 实例。

• 如果这个列表有两个值，那么按其顺序创建一个 IntegerDouble 实例。

• 如果这个列表有三个值，那么按其顺序创建一个 IntegerTriple 实例。

• 否则，抛出一个 Error。

如果成功，控制台应显示 Success!。

class IntegerHolder {
  IntegerHolder();
  
  // Implement this factory constructor.
  factory IntegerHolder.fromList(List<int> list) {
    TODO();
  }
}

class IntegerSingle extends IntegerHolder {
  final int a;

  IntegerSingle(this.a);
}

class IntegerDouble extends IntegerHolder {
  final int a;
  final int b;

  IntegerDouble(this.a, this.b);
}

class IntegerTriple extends IntegerHolder {
  final int a;
  final int b;
  final int c;

  IntegerTriple(this.a, this.b, this.c);
}

// Tests your solution (Don't edit from this point to end of file):
void main() {
  final errs = <String>[];

  // Run 5 tests to see which values have valid integer holders
  for (var tests = 0; tests < 5; tests++) {
    if (!testNumberOfArgs(errs, tests)) return;
  }

  // The goal is no errors with values 1 to 3,
  // but have errors with values 0 and 4.
  // The testNumberOfArgs method adds to the errs array if
  // the values 1 to 3 have an error and
  // the values 0 and 4 don't have an error
  if (errs.isEmpty) {
    print('Success!');
  } else {
    errs.forEach(print);
  }
}

bool testNumberOfArgs(List<String> errs, int count) {
  bool _threw = false;
  final ex = List.generate(count, (index) => index + 1);
  final callTxt = "IntegerHolder.fromList(${ex})";
  try {
    final obj = IntegerHolder.fromList(ex);
    final String vals = count == 1 ? "value" : "values";
    // Uncomment the next line if you want to see the results realtime
    // print("Testing with ${count} ${vals} using ${obj.runtimeType}.");
    testValues(errs, ex, obj, callTxt);
  } on Error {
    _threw = true;
  } catch (e) {
    switch (count) {
      case (< 1 && > 3):
        if (!_threw) {
          errs.add('Called ${callTxt} and it didn\'t throw an Error.');
        }
      default:
        errs.add('Called $callTxt and received an Error.');
    }
  }
  return true;
}

void testValues(List<String> errs, List<int> expectedValues, IntegerHolder obj,
    String callText) {
  for (var i = 0; i < expectedValues.length; i++) {
    int found;
    if (obj is IntegerSingle) {
      found = obj.a;
    } else if (obj is IntegerDouble) {
      found = i == 0 ? obj.a : obj.b;
    } else if (obj is IntegerTriple) {
      found = i == 0
          ? obj.a
          : i == 1
              ? obj.b
              : obj.c;
    } else {
      throw ArgumentError(
          "This IntegerHolder type (${obj.runtimeType}) is unsupported.");
    }

    if (found != expectedValues[i]) {
      errs.add(
          "Called $callText and got a ${obj.runtimeType} " + 
          "with a property at index $i value of $found " +
          "instead of the expected (${expectedValues[i]}).");
    }
  }
}

  switch (list.length) {
    case 1:
      return IntegerSingle(list[0]);
    case 2:
      return IntegerDouble(list[0], list[1]);
    case 3:
      return IntegerTriple(list[0], list[1], list[2]);
    default:
      throw ArgumentError("List must between 1 and 3 items. This list was ${list.length} items.");
  }

有时一个构造方法仅仅用来重定向到该类的另一个构造方法。重定向方法没有主体，它在冒号（:）之后调用另一个构造方法。

class Automobile {
  String make;
  String model;
  int mpg;

  // The main constructor for this class.
  Automobile(this.make, this.model, this.mpg);

  // Delegates to the main constructor.
  Automobile.hybrid(String make, String model) : this(make, model, 60);

  // Delegates to a named constructor
  Automobile.fancyHybrid() : this.hybrid('Futurecar', 'Mark 2');
}

还记得我们之前提到的 Color 类吗？创建一个叫做 black 的命名构造方法，但这次我们不要手动分配属性，而是将 0 作为参数，重定向到默认的构造方法。

可以忽略以下代码一开始在 DartPad 中的错误。

class Color {
  int red;
  int green;
  int blue;
  
  Color(this.red, this.green, this.blue);

  // TODO: Create a named constructor called "black" here
  // and redirect it to call the existing constructor
}


// Tests your solution (Don't edit!):
void main() {
  final errs = <String>[];

  try {
    final result = Color.black();
    
    if (result.red != 0) {
      errs.add('Called Color.black() and got a Color with red equal to \n ${result.red} instead of the expected value (0).');
    }

    if (result.green != 0) {
      errs.add('Called Color.black() and got a Color with green equal to \n ${result.green} instead of the expected value (0).');
    }

    if (result.blue != 0) {
  errs.add('Called Color.black() and got a Color with blue equal to \n ${result.blue} instead of the expected value (0).');
    }
  } catch (e) {
    print('Called Color.black() and got an exception of type ${e.runtimeType}.');
    return;
  }

  if (errs.isEmpty) {
    print('Success!');
  } else {
    errs.forEach(print);
  }
}

  Color.black() : this(0, 0, 0);

如果你的类生成的对象永远都不会更改，则可以让这些对象成为编译时常量。为此，请定义 const 构造方法并确保所有实例变量都是 final 的。

class ImmutablePoint {
  static const ImmutablePoint origin = ImmutablePoint(0, 0);

  final int x;
  final int y;

  const ImmutablePoint(this.x, this.y);
}

修改 Recipe 类，使其实例成为常量，并创建一个执行以下操作的常量构造方法：

• 该方法有三个参数：ingredients、calories 和 milligramsOfSodium。（按照此顺序）

• 使用 this 语法自动将参数值分配给同名的对象属性。

• 在 Recipe 的构造方法声明之前，用 const 关键字使其成为常量。

可以忽略以下代码一开始在 DartPad 中的错误。

class Recipe {
  List<String> ingredients;
  int calories;
  double milligramsOfSodium;

  // TODO: Create a const constructor here"

}


// Tests your solution (Don't edit!):
void main() {
  final errs = <String>[];

  try {
    const obj = Recipe(['1 egg', 'Pat of butter', 'Pinch salt'], 120, 200);
    
    if (obj.ingredients.length != 3) {
      errs.add('Called Recipe([\'1 egg\', \'Pat of butter\', \'Pinch salt\'], 120, 200) \n and got an object with ingredient list of length ${obj.ingredients.length} rather than the expected length (3).');
    }
    
    if (obj.calories != 120) {
      errs.add('Called Recipe([\'1 egg\', \'Pat of butter\', \'Pinch salt\'], 120, 200) \n and got an object with a calorie value of ${obj.calories} rather than the expected value (120).');
    }
    
    if (obj.milligramsOfSodium != 200) {
      errs.add('Called Recipe([\'1 egg\', \'Pat of butter\', \'Pinch salt\'], 120, 200) \n and got an object with a milligramsOfSodium value of ${obj.milligramsOfSodium} rather than the expected value (200).');
    }
  } catch (e) {
    print('Tried calling Recipe([\'1 egg\', \'Pat of butter\', \'Pinch salt\'], 120, 200) \n and received a null.');
  }

  if (errs.isEmpty) {
    print('Success!');
  } else {
    errs.forEach(print);
  }
}

class Recipe {
  final List<String> ingredients;
  final int calories;
  final double milligramsOfSodium;

  const Recipe(this.ingredients, this.calories, this.milligramsOfSodium);
}

我们希望你能够喜欢这个教程来学习或测试你对 Dart 语言中一些有趣特性的了解。

接下来你可以尝试：

• 尝试其他 Dart 教程。

• 阅读 Dart 语言之旅。

• 在 DartPad 上进行练习。

• 获取 Dart SDK。

除非另有说明，文档之所提及适用于 Dart 3.5.3 版本，本页面最后更新时间: 2024-08-07。 查看文档源码 或者 报告页面问题。