Skip to content

Treat lookahead/lookbehind as zero-width in RegexGroupParser and look through non-capturing groups in getRootAlternation #5603

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
phpstan-bot wants to merge 1 commit into
base: 2.1.x
Choose a base branch
from
+98 −4
Open

Treat lookahead/lookbehind as zero-width in RegexGroupParser and look through non-capturing groups in getRootAlternation #5603

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
55 changes: 51 additions & 4 deletions src/Type/Regex/RegexGroupParser.php
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -448,18 +448,38 @@ private function getRootAlternation(TreeNode $group): ?TreeNode
if (
$group->getId() === '#capturing'
&& count($group->getChildren()) === 1
&& $group->getChild(0)->getId() === '#alternation'
) {
return $group->getChild(0);
$child = $group->getChild(0);
if ($child->getId() === '#alternation') {
return $child;
}

if (
$child->getId() === '#noncapturing'
&& count($child->getChildren()) === 1
&& $child->getChild(0)->getId() === '#alternation'
) {
return $child->getChild(0);
}
}

// 1st token within a named capturing group is a token holding the group-name
if (
$group->getId() === '#namedcapturing'
&& count($group->getChildren()) === 2
&& $group->getChild(1)->getId() === '#alternation'
) {
return $group->getChild(1);
$child = $group->getChild(1);
if ($child->getId() === '#alternation') {
return $child;
}

if (
$child->getId() === '#noncapturing'
&& count($child->getChildren()) === 1
&& $child->getChild(0)->getId() === '#alternation'
) {
return $child->getChild(0);
}
}

return null;
Expand Down Expand Up @@ -606,6 +626,10 @@ private function walkGroupAst(
$walkResult = $walkResult->numeric(TrinaryLogic::createNo());
}

if (in_array($ast->getId(), ['#lookahead', '#negativelookahead', '#lookbehind', '#negativelookbehind'], true)) {
return $walkResult;
}

foreach ($children as $child) {
$walkResult = $this->walkGroupAst(
$child,
Expand All @@ -620,6 +644,10 @@ private function walkGroupAst(

private function isMaybeEmptyNode(TreeNode $node, string $patternModifiers, bool &$isNonFalsy): bool
{
if (in_array($node->getId(), ['#lookahead', '#negativelookahead', '#lookbehind', '#negativelookbehind'], true)) {
return true;
}

if ($node->getId() === '#quantification') {
[$min] = $this->getQuantificationRange($node);

Expand All @@ -640,6 +668,25 @@ private function isMaybeEmptyNode(TreeNode $node, string $patternModifiers, bool
return $literal === '';
}

if ($node->getId() === '#alternation') {
$allNonFalsy = true;
foreach ($node->getChildren() as $child) {
$childNonFalsy = false;
if ($this->isMaybeEmptyNode($child, $patternModifiers, $childNonFalsy)) {
return true;
}
if ($childNonFalsy) {
continue;
}

$allNonFalsy = false;
}
if ($allNonFalsy) {
$isNonFalsy = true;
}
return false;
}

foreach ($node->getChildren() as $child) {
if (!$this->isMaybeEmptyNode($child, $patternModifiers, $isNonFalsy)) {
return false;
Expand Down
47 changes: 47 additions & 0 deletions tests/PHPStan/Analyser/nsrt/preg_match_shapes.php
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1074,6 +1074,53 @@ function bug12792(string $string): void {
}
}

function bug12840(string $s): void {
// Empty alternation in non-capturing group
if (preg_match('~^((?:|\d+))x$~s', $s, $matches) === 1) {
assertType("array{non-falsy-string, ''|numeric-string}", $matches);
}

// Lookahead in alternation - lookahead is zero-width
if (preg_match('~^(a?(?:(?=x)|y))x$~s', $s, $matches) === 1) {
assertType("array{non-falsy-string, string}", $matches);
}

// Lookbehind in alternation - lookbehind is zero-width
if (preg_match('~^(a?(?:(?<=x)|y))x$~s', $s, $matches) === 1) {
assertType("array{non-falsy-string, string}", $matches);
}

// Negative lookahead - zero-width
if (preg_match('~^(a?(?:(?!z)|y))x$~s', $s, $matches) === 1) {
assertType("array{non-falsy-string, string}", $matches);
}

// Negative lookbehind - zero-width
if (preg_match('~^(a?(?:(?<!z)|y))x$~s', $s, $matches) === 1) {
assertType("array{non-falsy-string, string}", $matches);
}

// Lookahead alone in group should not produce non-empty-string (it's zero-width)
if (preg_match('~^((?=x))x$~s', $s, $matches) === 1) {
assertType("array{non-falsy-string, string}", $matches);
}

// Two non-capturing groups with empty alternation
if (preg_match('~^((?:|\d+)(?:|\w+))x$~s', $s, $matches) === 1) {
assertType("array{non-falsy-string, string}", $matches);
}

// Named capturing group with non-capturing wrapper alternation
if (preg_match('~^(?P<g>(?:|\d+))x$~s', $s, $matches) === 1) {
assertType("array{0: non-falsy-string, g: ''|numeric-string, 1: ''|numeric-string}", $matches);
}

// Lookahead in concatenation with literal - group still captures the literal
if (preg_match('~^((?=x)a)x$~s', $s, $matches) === 1) {
assertType("array{non-falsy-string, non-falsy-string}", $matches);
}
}

function testExtendedSyntaxEscapedHash(string $string): void {
// \# in extended mode should be treated as literal hash, not comment
if (preg_match('/^ ([\#.]) $/x', $string, $matches)) {
Expand Down
Loading