Contents

Dart cheatsheet

The Dart language is designed to be easy to learn for coders coming from other languages, but it has a few unique features. This tutorial walks you through the most important of these language features.

The embedded editors in this tutorial have partially completed code snippets. You can use these editors to test your knowledge by completing the code and clicking the Run button. The editors also contain thorough test code; don't edit the test code, but feel free to study it to learn about testing.

If you need help, expand the Solution for... dropdown beneath each DartPad for an explanation and the answer.

String interpolation

#

To put the value of an expression inside a string, use ${expression}. If the expression is an identifier, you can omit the {}.

Here are some examples of using string interpolation:

StringResult
'${3 + 2}''5'
'${"word".toUpperCase()}''WORD'
'$myObject'The value of myObject.toString()

Code example

#

The following function takes two integers as parameters. Make it return a string containing both integers separated by a space. For example, stringify(2, 3) should return '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!');
}
Solution for string interpolation example

Both x and y are simple values, and Dart's string interpolation will handle converting them to string representations. All you need to do is use the $ operator to reference them inside single quotes, with a space in between:

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

Nullable variables

#

Dart enforces sound null safety. This means values can't be null unless you say they can be. In other words, types default to non-nullable.

For example, consider the following code. With null safety, this code returns an error. A variable of type int can't have the value null:

dart
int a = null; // INVALID.

When creating a variable, add ? to the type to indicate that the variable can be null:

dart
int? a = null; // Valid.

You can simplify that code a bit because, in all versions of Dart, null is the default value for uninitialized variables:

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

To learn more about null safety in Dart, read the sound null safety guide.

Code example

#

Declare two variables in this DartPad:

  • A nullable String named name with the value 'Jane'.
  • A nullable String named address with the value null.

Ignore all initial errors in the DartPad.

// 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}');
  }
}
Solution for nullable variables example

Declare the two variables as String followed by ?. Then, assign 'Jane' to name and leave address uninitialized:

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

Null-aware operators

#

Dart offers some handy operators for dealing with values that might be null. One is the ??= assignment operator, which assigns a value to a variable only if that variable is currently null:

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

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

Another null-aware operator is ??, which returns the expression on its left unless that expression's value is null, in which case it evaluates and returns the expression on its right:

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

Code example

#

Try substituting in the ??= and ?? operators to implement the described behavior in the following snippet.

Ignore all initial errors in the DartPad.

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}.');
  }
  
}
Solution for null-aware operators example

All you need to do in this exercise is replace the TODO comments with either ?? or ??=. Read the text above to make sure you understand both, and then give it a try:

dart
// 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';
}

Conditional property access

#

To guard access to a property or method of an object that might be null, put a question mark (?) before the dot (.):

dart
myObject?.someProperty

The preceding code is equivalent to the following:

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

You can chain multiple uses of ?. together in a single expression:

dart
myObject?.someProperty?.someMethod()

The preceding code returns null (and never calls someMethod()) if either myObject or myObject.someProperty is null.

Code example

#

The following function takes a nullable string as a parameter. Try using conditional property access to make it return the uppercase version of str, or null if str is 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}.');
  }
}
Solution for conditional property access example

If this exercise wanted you to conditionally lowercase a string, you could do it like this: str?.toLowerCase(). Use the equivalent method to uppercase a string!

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

Collection literals

#

Dart has built-in support for lists, maps, and sets. You can create them using literals:

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

Dart's type inference can assign types to these variables for you. In this case, the inferred types are List<String>, Set<String>, and Map<String, int>.

Or you can specify the type yourself:

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

Specifying types is handy when you initialize a list with contents of a subtype, but still want the list to be List<BaseType>:

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

Code example

#

Try setting the following variables to the indicated values. Replace the existing null values.

// 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
}
Solution for collection literals example

Add a list, set, or map literal after each equals sign. Remember to specify the types for the empty declarations, since they can't be inferred.

dart
// 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>{};

Arrow syntax

#

You might have seen the => symbol in Dart code. This arrow syntax is a way to define a function that executes the expression to its right and returns its value.

For example, consider this call to the List class's any() method:

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

Here's a simpler way to write that code:

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

Code example

#

Try finishing the following statements, which use arrow syntax.

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);
  }
}
Solution for arrow syntax example

For the product, you can use * to multiply the three values together. For incrementValue1, you can use the increment operator (++). For joinWithCommas, use the join method found in the List class.

dart
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(',');
}

Cascades

#

To perform a sequence of operations on the same object, use cascades (..). We've all seen an expression like this:

dart
myObject.someMethod()

It invokes someMethod() on myObject, and the result of the expression is the return value of someMethod().

Here's the same expression with a cascade:

dart
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!

Using cascades, you can chain together operations that would otherwise require separate statements. For example, consider the following code, which uses the conditional member access operator (?.) to read properties of button if it isn't null:

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

To instead use cascades, you can start with the null-shorting cascade (?..), which guarantees that none of the cascade operations are attempted on a null object. Using cascades shortens the code and makes the button variable unnecessary:

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

Code example

#

Use cascades to create a single statement that sets the anInt, aString, and aList properties of a BigObject to 1, 'String!', and [3.0] (respectively) and then calls 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);
  }
}
Solution for cascades example

The best solution for this exercise starts with obj.. and has four assignment operations chained together. Start with return obj..anInt = 1, then add another cascade (..) and start the next assignment.

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

Getters and setters

#

You can define getters and setters whenever you need more control over a property than a simple field allows.

For example, you can make sure a property's value is valid:

dart
class MyClass {
  int _aProperty = 0;

  int get aProperty => _aProperty;

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

You can also use a getter to define a computed property:

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

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

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

Code example

#

Imagine you have a shopping cart class that keeps a private List<double> of prices. Add the following:

  • A getter called total that returns the sum of the prices
  • A setter that replaces the list with a new one, as long as the new list doesn't contain any negative prices (in which case the setter should throw an InvalidPriceException).

Ignore all initial errors in the DartPad.

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!');
}
Solution for getters and setters example

Two functions are handy for this exercise. One is fold, which can reduce a list to a single value (use it to calculate the total). The other is any, which can check each item in a list with a function you give it (use it to check if there are any negative prices in the prices setter).

dart
// 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;
}

Optional positional parameters

#

Dart has two kinds of function parameters: positional and named. Positional parameters are the kind you're likely familiar with:

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

With Dart, you can make these positional parameters optional by wrapping them in brackets:

dart
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);

Optional positional parameters are always last in a function's parameter list. Their default value is null unless you provide another default value:

dart
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
}

Code example

#

Implement a function called joinWithCommas() that accepts one to five integers, then returns a string of those numbers separated by commas. Here are some examples of function calls and returned values:

Function callReturned value
joinWithCommas(1)'1'
joinWithCommas(1, 2, 3)'1,2,3'
joinWithCommas(1, 1, 1, 1, 1)'1,1,1,1,1'

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);
  }
}
Solution for positional parameters example

The b, c, d, and e parameters are null if they aren't provided by the caller. The important thing, then, is to check whether those arguments are null before you add them to the final string.

dart
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;
}

Named parameters

#

Using a curly brace syntax at the end of the parameter list, you can define parameters that have names.

Named parameters are optional unless they're explicitly marked as required.

dart
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');
}

As you might expect, the default value of a nullable named parameter is null, but you can provide a custom default value.

If the type of a parameter is non-nullable, then you must either provide a default value (as shown in the following code) or mark the parameter as required (as shown in the constructor section).

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

A function can't have both optional positional and named parameters.

Code example

#

Add a copyWith() instance method to the MyDataObject class. It should take three named, nullable parameters:

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

Your copyWith() method should return a new MyDataObject based on the current instance, with data from the preceding parameters (if any) copied into the object's properties. For example, if newInt is non-null, then copy its value into anInt.

Ignore all initial errors in the DartPad.

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);
  }
}
Solution for named parameters example

The copyWith method shows up in a lot of classes and libraries. Yours should do a few things: use optional named parameters, create a new instance of MyDataObject, and use the data from the parameters to fill it (or the data from the current instance if the parameters are null). This is a chance to get more practice with the ?? operator!

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

Exceptions

#

Dart code can throw and catch exceptions. In contrast to Java, all of Dart's exceptions are unchecked. Methods don't declare which exceptions they might throw and you aren't required to catch any exceptions.

Dart provides Exception and Error types, but you're allowed to throw any non-null object:

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

Use the try, on, and catch keywords when handling exceptions:

dart
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.

If you can't completely handle the exception, use the rethrow keyword to propagate the exception:

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

To execute code whether or not an exception is thrown, use finally:

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

Code example

#

Implement tryFunction() below. It should execute an untrustworthy method and then do the following:

  • If untrustworthy() throws an ExceptionWithMessage, call logger.logException with the exception type and message (try using on and catch).
  • If untrustworthy() throws an Exception, call logger.logException with the exception type (try using on for this one).
  • If untrustworthy() throws any other object, don't catch the exception.
  • After everything's caught and handled, call logger.doneLogging (try using 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);
  }
}
Solution for exceptions example

This exercise looks tricky, but it's really one big try statement. Call untrustworthy inside the try, and then use on, catch, and finally to catch exceptions and call methods on the logger.

dart
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();
  }
}

Using this in a constructor

#

Dart provides a handy shortcut for assigning values to properties in a constructor: use this.propertyName when declaring the constructor:

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

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

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

This technique works for named parameters, too. Property names become the names of the parameters:

dart
class MyColor {
  ...

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

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

In the preceding code, red, green, and blue are marked as required because these int values can't be null. If you add default values, you can omit required:

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

Code example

#

Add a one-line constructor to MyClass that uses this. syntax to receive and assign values for all three properties of the class.

Ignore all initial errors in the DartPad.

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);
  }
}
Solution for `this` example

This exercise has a one-line solution. Declare the constructor with this.anInt, this.aString, and this.aDouble as its parameters in that order.

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

Initializer lists

#

Sometimes when you implement a constructor, you need to do some setup before the constructor body executes. For example, final fields must have values before the constructor body executes. Do this work in an initializer list, which goes between the constructor's signature and its body:

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

The initializer list is also a handy place to put asserts, which run only during development:

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

Code example

#

Complete the FirstTwoLetters constructor below. Use an initializer list to assign the first two characters in word to the letterOne and LetterTwo properties. For extra credit, add an assert to catch words of less than two characters.

Ignore all initial errors in the DartPad.

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);
  }
}
Solution for initializer lists example

Two assignments need to happen: letterOne should be assigned word[0], and letterTwo should be assigned word[1].

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

Named constructors

#

To allow classes to have multiple constructors, Dart supports named constructors:

dart
class Point {
  double x, y;

  Point(this.x, this.y);

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

To use a named constructor, invoke it using its full name:

dart
final myPoint = Point.origin();

Code example

#

Give the Color class a constructor named Color.black that sets all three properties to zero.

Ignore all initial errors in the DartPad.

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);
  }
}
Solution for named constructors example

The declaration for your constructor should begin with Color.black():. In the initializer list (after the colon), set red, green, and blue to 0.

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

Factory constructors

#

Dart supports factory constructors, which can return subtypes or even null. To create a factory constructor, use the factory keyword:

dart
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');
  }
}

Code example

#

Replace the line TODO(); in the factory constructor named IntegerHolder.fromList to return the following:

  • If the list has one value, create an IntegerSingle instance using that value.
  • If the list has two values, create an IntegerDouble instance using the values in order.
  • If the list has three values, create an IntegerTriple instance using the values in order.
  • Otherwise, throw an Error.

If you succeed, the console should display 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]}).");
    }
  }
}

Solution for factory constructors example

Inside the factory constructor, check the length of the list, then create and return an IntegerSingle, IntegerDouble, or IntegerTriple as appropriate.

Replace TODO(); with the following code block.

dart
  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.");
  }

Redirecting constructors

#

Sometimes a constructor's only purpose is to redirect to another constructor in the same class. A redirecting constructor's body is empty, with the constructor call appearing after a colon (:).

dart
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');
}

Code example

#

Remember the Color class from above? Create a named constructor called black, but rather than manually assigning the properties, redirect it to the default constructor with zeros as the arguments.

Ignore all initial errors in the DartPad.

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);
  }
}
Solution for redirecting constructors example

Your constructor should redirect to this(0, 0, 0).

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

Const constructors

#

If your class produces objects that never change, you can make these objects compile-time constants. To do this, define a const constructor and make sure that all instance variables are final.

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

  final int x;
  final int y;

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

Code example

#

Modify the Recipe class so its instances can be constants, and create a constant constructor that does the following:

  • Has three parameters: ingredients, calories, and milligramsOfSodium (in that order).
  • Uses this. syntax to automatically assign the parameter values to the object properties of the same name.
  • Is constant, with the const keyword just before Recipe in the constructor declaration.

Ignore all initial errors in the DartPad.

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);
  }
}
Solution for const constructors example

To make the constructor const, you'll need to make all the properties final.

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

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

What's next?

#

We hope you enjoyed using this tutorial to learn or test your knowledge of some of the most interesting features of the Dart language.

What you can try next includes: