Fixing type promotion failures

Contents

Only local variables can be promoted
Other causes and workarounds

This page has information to help you understand why type promotion failures occur, with tips on how to fix them. For background information, see Working with nullable fields, a section in Understanding null safety.

Only local variables can be promoted

The cause: You’re trying to promote a property or this, but only local variables can be promoted.

Example:

class C {
  int? i;                  // (1)
  void f() {
    if (i == null) return;
    print(i.isEven);       // (2) ERROR
  }
}

The Dart compiler produces an error message for (2) that points to (1) and explains that i can’t be promoted to a non-nullable type because it’s a field.

The usual fix is either to use i! or to create a local variable of type int that holds the value of i.

Here’s an example of using i!:

print(i!.isEven);

And here’s an example of creating a local variable (which can be named i) that holds the value of i:

class C {
  int? i;
  void f() {
    final i = this.i;
    if (i == null) return;
    print(i.isEven);
  }
}

This example features an instance field, but it could instead use an instance getter, a static field or getter, a top-level variable or getter, or this. (Although promoting this would be sound, implementing it would be difficult and not very useful.)

Other causes and workarounds

This section covers other common causes of type promotion failure. The workarounds for many of these are one or more of the following:

Create a local variable that has the type you need.
Add an explicit null check.
Use ! or as if you’re sure an expression can’t be null.

Possibly written after promotion

The cause: Trying to promote a variable that might have been written to since it was promoted.

Example:

void f(bool b, int? i, int? j) {
  if (i == null) return;
  if (b) {
    i = j;           // (1)
  }
  if (!b) {
    print(i.isEven); // (2) ERROR
  }
}

In this example, when flow analysis hits (1), it demotes i from non-nullable int back to nullable int?. A human can tell that the access at (2) is safe because there’s no code path that includes both (1) and (2), but flow analysis isn’t smart enough to see that, because it doesn’t track correlations between conditions in separate if statements.

You might fix the problem by combining the two if statements:

void f(bool b, int? i, int? j) {
  if (i == null) return;
  if (b) {
    i = j;
  } else {
    print(i.isEven);
  }
}

In straight-line control flow cases like these (no loops), flow analysis takes into account the right hand side of the assignment when deciding whether to demote. As a result, another way to fix this code is to change the type of j to int.

void f(bool b, int? i, int j) {
  if (i == null) return;
  if (b) {
    i = j;
  }
  if (!b) {
    print(i.isEven);
  }
}

Possibly written in a previous loop iteration

The cause: You’re trying to promote something that might have been written to in a previous iteration of a loop, and so the promotion was invalidated.

Example:

void f(Link? p) {
  if (p != null) return;
  while (true) {    // (1)
    print(p.value); // (2) ERROR
    var next = p.next;
    if (next == null) break;
    p = next;       // (3)
  }
}

When flow analysis reaches (1), it looks ahead and sees the write to p at (3). But because it’s looking ahead, it hasn’t yet figured out the type of the right-hand side of the assignment, so it doesn’t know whether it’s safe to retain the promotion. To be safe, it invalidates the promotion.

You might fix this problem by moving the null check to the top of the loop:

void f(Link? p) {
  while (p != null) {
    print(p.value);
    p = p.next;
  }
}

This situation can also arise in switch statements if a case block has a label, because you can use labeled switch statements to construct loops:

void f(int i, int? j, int? k) {
  if (j == null) return;
  switch (i) {
    label:
    case 0:
      print(j.isEven); // ERROR
      j = k;
      continue label;
  }
}

Again, you can fix the problem by moving the null check to the top of the loop:

void f(int i, int? j, int? k) {
  switch (i) {
    label:
    case 0:
      if (j == null) return;
      print(j.isEven);
      j = k;
      continue label;
  }
}

In catch after possible write in try

The cause: The variable might have been written to in a try block, and execution is now in a catch block.

Example:

void f(int? i, int? j) {
  if (i == null) return;
  try {
    i = j;                 // (1)
    // ... Additional code ...
    if (i == null) return; // (2)
    // ... Additional code ...
  } catch (e) {
    print(i.isEven);       // (3) ERROR
  }
}

In this case, flow analysis doesn’t consider i.isEven (3) safe, because it has no way of knowing when in the try block the exception might have occurred, so it conservatively assumes that it might have happened between (1) and (2), when i was potentially null.

Similar situations can occur between try and finally blocks, and between catch and finally blocks. Because of a historical artifact of how the implementation was done, these try/catch/finally situations don’t take into account the right-hand side of the assignment, similar to what happens in loops.

To fix the problem, make sure that the catch block doesn’t rely on assumptions about the state of variables that get changed inside the try block. Remember, the exception might occur at any time during the try block, possibly when i is null.

The safest solution is to add a null check inside the catch block:

// ···
} catch (e) {
  if (i != null) {
    print(i.isEven); // (3) OK due to the null check in the line above.
  } else {
    // Handle the case where i is null.
  }
}

Or, if you’re sure that an exception can’t occur while i is null, just use the ! operator:

// ···
} catch (e) {
  print(i!.isEven); // (3) OK because of the `!`.
}

Subtype mismatch

The cause: The type you’re trying to promote to isn’t a subtype of the variable’s current promoted type (or wasn’t a subtype at the time of the promotion attempt).

Example:

void f(Object o) {
  if (o is Comparable /* (1) */) {
    if (o is Pattern /* (2) */) {
      print(o.matchAsPrefix('foo')); // (3) ERROR
    }
  }
}

In this example, o is promoted to Comparable at (1), but it isn’t promoted to Pattern at (2), because Pattern isn’t a subtype of Comparable. (The rationale is that if it did promote, then you wouldn’t be able to use methods on Comparable.) Note that just because Pattern isn’t a subtype of Comparable doesn’t mean the code at (3) is dead; o might have a type—like String—that implements both Comparable and Pattern.

One possible solution is to create a new local variable so that the original variable is promoted to Comparable, and the new variable is promoted to Pattern:

void f(Object o) {
  if (o is Comparable /* (1) */) {
    Object o2 = o;
    if (o2 is Pattern /* (2) */) {
      print(
          o2.matchAsPrefix('foo')); // (3) OK; o2 was promoted to `Pattern`.
    }
  }
}

However, someone who edits the code later might be tempted to change Object o2 to var o2. That change gives o2 a type of Comparable, which brings back the problem of the object not being promotable to Pattern.

A redundant type check might be a better solution:

void f(Object o) {
  if (o is Comparable /* (1) */) {
    if (o is Pattern /* (2) */) {
      print((o as Pattern).matchAsPrefix('foo')); // (3) OK
    }
  }
}

Another solution that sometimes works is when you can use a more precise type. If line 3 cares only about strings, then you can use String in your type check. Because String is a subtype of Comparable, the promotion works:

void f(Object o) {
  if (o is Comparable /* (1) */) {
    if (o is String /* (2) */) {
      print(o.matchAsPrefix('foo')); // (3) OK
    }
  }
}

Write captured by a local function

The cause: The variable has been write captured by a local function or function expression.

Example:

void f(int? i, int? j) {
  var foo = () {
    i = j;
  };
  // ... Use foo ... 
  if (i == null) return; // (1)
  // ... Additional code ...
  print(i.isEven);       // (2) ERROR
}

Flow analysis reasons that as soon as the definition of foo is reached, it might get called at any time, therefore it’s no longer safe to promote i at all. As with loops, this demotion happens regardless of the type of the right hand side of the assignment.

Sometimes it’s possible to restructure the logic so that the promotion is before the write capture:

void f(int? i, int? j) {
  if (i == null) return; // (1)
  // ... Additional code ...
  print(i.isEven); // (2) OK
  var foo = () {
    i = j;
  };
  // ... Use foo ...
}

Another option is to create a local variable, so it isn’t write captured:

void f(int? i, int? j) {
  var foo = () {
    i = j;
  };
  // ... Use foo ...
  var i2 = i;
  if (i2 == null) return; // (1)
  // ... Additional code ...
  print(i2.isEven); // (2) OK because `i2` isn't write captured.
}

Or you can do a redundant check:

void f(int? i, int? j) {
  var foo = () {
    i = j;
  };
  // ... Use foo ...
  if (i == null) return; // (1)
  // ... Additional code ...
  print(i!.isEven); // (2) OK due to `!` check.
}

Written outside of the current closure or function expression

The cause: The variable is written to outside of a closure or function expression, and the type promotion location is inside the closure or function expression.

Example:

void f(int? i, int? j) {
  if (i == null) return;
  var foo = () {
    print(i.isEven); // (1) ERROR
  };
  i = j;             // (2)
}

Flow analysis reasons that there’s no way to determine when foo might get called, so it might get called after the assignment at (2), and thus the promotion might no longer be valid. As with loops, this demotion happens regardless of the type of the right hand side of the assignment.

A solution is to create a local variable:

void f(int? i, int? j) {
  if (i == null) return;
  var i2 = i;
  var foo = () {
    print(i2.isEven); // (1) OK because `i2` isn't changed later.
  };
  i = j; // (2)
}

A particularly nasty case looks like this:

void f(int? i) {
  i ??= 0;
  var foo = () {
    print(i.isEven); // ERROR
  };
}

In this case, a human can see that the promotion is safe because the only write to i uses a non-null value and happens before foo is ever created. But flow analysis isn’t that smart.

Again, a solution is to create a local variable:

void f(int? i) {
  var j = i ?? 0;
  var foo = () {
    print(j.isEven); // OK
  };
}

This solution works because j is inferred to have a non-nullable type (int) due to its initial value (i ?? 0). Because j has a non-nullable type, whether or not it’s assigned later, j can never have a non-null value.

Write captured outside of the current closure or function expression

The cause: The variable you’re trying to promote is write captured outside of a closure or function expression, but this use of the variable is inside of the closure or function expression that’s trying to promote it.

Example:

void f(int? i, int? j) {
  var foo = () {
    if (i == null) return;
    print(i.isEven); // ERROR
  };
  var bar = () {
    i = j;
  };
}

Flow analysis reasons that there’s no way of telling what order foo and bar might be executed in; in fact, bar might even get executed halfway through executing foo (due to foo calling something that calls bar). So it isn’t safe to promote i at all inside foo.

The best solution is probably to create a local variable:

void f(int? i, int? j) {
  var foo = () {
    var i2 = i;
    if (i2 == null) return;
    print(i2.isEven); // OK because i2 is local to this closure.
  };
  var bar = () {
    i = j;
  };
}