Add null to array_map() callback parameters of shorter arrays by SanderMuller · Pull Request #6874 · phpstan/phpstan-src

SanderMuller · 2026-10-28T18:47:03Z

With more than one array, array_map() calls the callback once per index of the longest array. Each shorter array gives null for the missing indexes:

php -r 'var_dump(array_map(fn ($a, $b) => [$a, $b], [1], [3, 4]));'

The second call gets $a = null, so the first array is padded too, not only the 2nd+ arrays.

PHPStan typed each callback parameter as the value type of its array, without null. This PR adds null to the parameter of an array whose largest size is smaller than the largest size of another array. It does this only when the size of every array is known, for example constant arrays or array{1, 2}|array{1}.

It compares the largest sizes and not the smallest, because array_map($cb, $a, $a) never pads, whatever sizes $a can have. An arg with ... is skipped where it is not expanded into its arrays.

The logic is in ArrayMapCallbackParameterTypes. Four places use it: ParametersAcceptorSelector, ContextualClosureParameterResolver, ArrayMapFunctionReturnTypeExtension and ArrayMapParameterClosureTypeExtension. The two shadowed classes call the PHP helper from their C++ twins.

A typed callback such as function (int $a, int $b) over [1, 2] and [3] now gets an argument error. PHP throws a TypeError for that call.

General arrays stay as they are. I tried to make them nullable too, but three existing regression tests then report errors:

array_map($cb, $arr, $arr) (bug-12317)
array_map($callback, $this->items, array_keys($this->items)) (array_map_multiple.php)
array_map(fn (string $a, string $b) => ..., ...$tuple) with array{list<string>, list<string>} (bug-10527)

The first is provably the same length, the second is the same length in practice, and the third can differ. Should general arrays become nullable, maybe in bleeding edge, with an exception for the same expression?

Closes phpstan/phpstan#13217

🤖 Generated with Claude Code

With more than one array, array_map() passes null for the missing indexes of each shorter array. When the size of every array is known, the callback parameter of an array that can be shorter than another one now gets null. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>

staabm · 2026-10-30T08:14:02Z

+		// the closure gets one parameter per arg here, not one per unpacked array
+		array_map(function ($a, $b) {
+			assertType('1|2|3', $a);
+			assertType('array{4, 5, 6}', $b);


I don't think $b is an array? https://3v4l.org/QsHlT#v

You are right, $b gets 4, 5 and 6. PHPStan 2.3.4 also infers array{4, 5, 6} there, because an unpacked arg stays one parameter and is not expanded into its arrays. This PR does not change that, but the test locked the wrong type in. I removed that case. I can fix the unpacked args in a separate PR.

yes, please report a new phpstan issue with expectations

Opened phpstan/phpstan#15620 with the expected types. I have a fix ready and will open the PR after this one is merged, because both change the same code.

The test asserted array{4, 5, 6} for the parameter of an unpacked arg. At runtime the parameter gets 4, 5 and 6. PHPStan already inferred this wrong type before this change, so the test must not lock it in. Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>

staabm · 2026-10-30T09:07:08Z

+			assertType('1|2', $a);
+			assertType('3|null', $b);


unrelated side-question:

do we differentiate between types beeing infered from native vs. phpdoc?
so would assertType and assertNativeType lead different results when we feed a phpdoc array-shape or native values like [1, 2] into array_map?

Yes, they can differ. With native arrays like [1, 2] and [3], assertType and assertNativeType both give 3|null for $b. With a PHPDoc shape, the native type of the array is array, so assertNativeType gives mixed for its value. I added both cases to the test.

staabm · 2026-10-30T09:09:30Z

+				$callbackParameters = [];
+				foreach (ArrayMapCallbackParameterTypes::addPaddingNull($callbackValueTypes, $callbackArrayTypes) as $valueType) {
+					$callbackParameters[] = new DummyParameter('item', $valueType, optional: false, passedByReference: PassedByReference::createNo(), variadic: false, defaultValue: null);
+				}
+				$nativeCallbackParameters = [];
+				foreach (ArrayMapCallbackParameterTypes::addPaddingNull($nativeCallbackValueTypes, $nativeCallbackArrayTypes) as $valueType) {
+					$nativeCallbackParameters[] = new DummyParameter('item', $valueType, optional: false, passedByReference: PassedByReference::createNo(), variadic: false, defaultValue: null);
+				}


since we have 2 separate loops for native-types vs. non-native-types, I think we should also have assertType and assertNativeType assertions.

said differently: do we have test-coverage for both these loops here?

just double checked: if I comment out line 251, no tests fail

I added assertNativeType lines, but none of them fail when line 251 is commented out. An inline closure gets its native types from ContextualClosureParameterResolver, not from this callable type. I also tried a closure in a ternary, a closure in a variable and a typed arrow function, and the output did not change. I found no case where the native callable type here has an effect. It was built the same way before this PR, so I kept the two loops in line.

staabm · 2026-10-30T09:12:22Z

 			$parameters[] = new DummyParameter('item', $type, optional: false, passedByReference: PassedByReference::createNo(), variadic: false, defaultValue: null);
+		}
+		$nativeParameters = [];
+		foreach ($nativeTypes as $nativeType) {


same here.. might need assertNativeType to cover this code

Covered now. The new assertNativeType lines fail when this line or the native padding above it is removed. They also pass with Turbo loaded.

staabm · 2026-10-31T09:25:12Z

ping in case you missed the latest feedback

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>

SanderMuller force-pushed the array-map-null-padding branch from 04e7c92 to d231b6d Compare October 29, 2026 12:43

staabm reviewed Oct 30, 2026

View reviewed changes

Comment thread src/Reflection/ArrayMapCallbackParameterTypes.php Resolved

staabm approved these changes Oct 30, 2026

View reviewed changes

staabm requested a review from VincentLanglet October 30, 2026 08:49

staabm reviewed Oct 30, 2026

View reviewed changes

Assert the native types of array_map() callback parameters

87c6971

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>

VincentLanglet approved these changes Oct 31, 2026

View reviewed changes

This was referenced Oct 31, 2026

Unpacked args to array_map() and immediately invoked closures are typed as one value phpstan/phpstan#15619

Closed

Unpacked args to array_map() and immediately invoked closures are typed as one value phpstan/phpstan#15620

Open

Navigation Menu

Repository navigation

Add null to array_map() callback parameters of shorter arrays - #6874

Add null to array_map() callback parameters of shorter arrays#6874
SanderMuller wants to merge 3 commits into
phpstan:2.3.xfrom
SanderMuller:array-map-null-padding

SanderMuller commented Oct 28, 2026

Uh oh!

staabm Oct 30, 2026

Uh oh!

SanderMuller Oct 30, 2026

Uh oh!

staabm Oct 30, 2026

Uh oh!

SanderMuller Oct 31, 2026 •

edited

Loading

Uh oh!

Uh oh!

staabm Oct 30, 2026 •

edited

Loading

Uh oh!

SanderMuller Oct 31, 2026

Uh oh!

staabm Oct 30, 2026 •

edited

Loading

Uh oh!

staabm Oct 30, 2026

Uh oh!

SanderMuller Oct 31, 2026

Uh oh!

staabm Oct 30, 2026

Uh oh!

SanderMuller Oct 31, 2026

Uh oh!

staabm commented Oct 31, 2026

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

SanderMuller commented Oct 28, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

SanderMuller Oct 31, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

staabm Oct 30, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

staabm Oct 30, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

staabm commented Oct 31, 2026

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

SanderMuller Oct 31, 2026 •

edited

Loading

staabm Oct 30, 2026 •

edited

Loading

staabm Oct 30, 2026 •

edited

Loading