Strings regex validation

Update regex validation and add validation class to return full information. Add helper to return last error message from defined constant
This commit is contained in:
Clemens Schwaighofer
2025-06-05 14:29:04 +09:00
parent 3be3519e45
commit dbc72472f9
8 changed files with 380 additions and 13 deletions

View File

@@ -1,5 +1,7 @@
<?php
// This code was created by Claude Sonnet 4
declare(strict_types=1);
namespace tests;

View File

@@ -1,6 +1,6 @@
<?php
// This code was create by Claude Sonnet 4
// This code was created by Claude Sonnet 4
// modification for value checks with assertEqualsCanonicalizing
declare(strict_types=1);

View File

@@ -1,6 +1,6 @@
<?php
// create by Claude Sonnet 4
// created by Claude Sonnet 4
// testRecursiveSearchWithFlatResult had wrong retunr count

View File

@@ -1,6 +1,6 @@
<?php
// This code was create by Claude Sonnet 4
// This code was created by Claude Sonnet 4
declare(strict_types=1);

View File

@@ -0,0 +1,283 @@
<?php
// This code was created by Claude Sonnet 4
// FIX:
// '/test{/', // Unmatched brace -> this is valid
// '/test{1,}/', // Invalid quantifier -> this is valid
declare(strict_types=1);
namespace tests;
use PHPUnit\Framework\TestCase;
use CoreLibs\Convert\Strings;
/**
* Test class for CoreLibs\Convert\Strings regex validation methods
*/
class CoreLibsConvertStringsRegexValidateTest extends TestCase
{
/**
* Test isValidRegex with valid regex patterns
*/
public function testIsValidRegexWithValidPatterns(): void
{
$validPatterns = [
'/^[a-zA-Z0-9]+$/',
'/test/',
'/\d+/',
'/^hello.*world$/',
'/[0-9]{3}-[0-9]{3}-[0-9]{4}/',
'#^https?://.*#i',
'~^[a-z]+~',
'|test|',
'/^$/m',
'/\w+/u',
];
foreach ($validPatterns as $pattern) {
$this->assertTrue(
Strings::isValidRegex($pattern),
"Pattern '{$pattern}' should be valid"
);
}
}
/**
* Test isValidRegex with invalid regex patterns
*/
public function testIsValidRegexWithInvalidPatterns(): void
{
$invalidPatterns = [
'/[/', // Unmatched bracket
'/test[/', // Unmatched bracket
'/(?P<name>/', // Unmatched parenthesis
'/(?P<>test)/', // Invalid named group
'/test\\/', // Invalid escape at end
'/(test/', // Unmatched parenthesis
'/test)/', // Unmatched parenthesis
// '/test{/', // Unmatched brace -> this is valid
// '/test{1,}/', // Invalid quantifier -> this is valid
'/[z-a]/', // Invalid character range
'invalid', // No delimiters
'', // Empty string
'/(?P<123>test)/', // Invalid named group name
];
foreach ($invalidPatterns as $pattern) {
$this->assertFalse(
Strings::isValidRegex($pattern),
"Pattern '{$pattern}' should be invalid"
);
}
}
/**
* Test getLastRegexErrorString returns correct error messages
*/
public function testGetLastRegexErrorStringReturnsCorrectMessages(): void
{
// Test with a valid regex first to ensure clean state
Strings::isValidRegex('/valid/');
$this->assertEquals('No error', Strings::getLastRegexErrorString());
// Test with invalid regex to trigger an error
Strings::isValidRegex('/[/');
$errorMessage = Strings::getLastRegexErrorString();
// The error message should be one of the defined messages
$this->assertContains($errorMessage, array_values(Strings::PREG_ERROR_MESSAGES));
$this->assertNotEquals('Unknown error', $errorMessage);
}
/**
* Test getLastRegexErrorString with unknown error
*/
public function testGetLastRegexErrorStringWithUnknownError(): void
{
// This is harder to test directly since we can't easily mock preg_last_error()
// but we can test the fallback behavior by reflection or assume it works
// At minimum, ensure it returns a string
$result = Strings::getLastRegexErrorString();
$this->assertIsString($result);
$this->assertNotEmpty($result);
}
/**
* Test validateRegex with valid patterns
*/
public function testValidateRegexWithValidPatterns(): void
{
$validPatterns = [
'/^test$/',
'/\d+/',
'/[a-z]+/i',
];
foreach ($validPatterns as $pattern) {
$result = Strings::validateRegex($pattern);
$this->assertIsArray($result);
$this->assertArrayHasKey('valid', $result);
$this->assertArrayHasKey('preg_error', $result);
$this->assertArrayHasKey('error', $result);
$this->assertTrue($result['valid'], "Pattern '{$pattern}' should be valid");
$this->assertEquals(PREG_NO_ERROR, $result['preg_error']);
$this->assertNull($result['error']);
}
}
/**
* Test validateRegex with invalid patterns
*/
public function testValidateRegexWithInvalidPatterns(): void
{
$invalidPatterns = [
'/[/', // Unmatched bracket
'/(?P<name>/', // Unmatched parenthesis
'/test\\/', // Invalid escape at end
'/(test/', // Unmatched parenthesis
];
foreach ($invalidPatterns as $pattern) {
$result = Strings::validateRegex($pattern);
$this->assertIsArray($result);
$this->assertArrayHasKey('valid', $result);
$this->assertArrayHasKey('preg_error', $result);
$this->assertArrayHasKey('error', $result);
$this->assertArrayHasKey('pcre_error', $result);
$this->assertFalse($result['valid'], "Pattern '{$pattern}' should be invalid");
$this->assertNotEquals(PREG_NO_ERROR, $result['preg_error']);
$this->assertIsString($result['error']);
$this->assertNotNull($result['error']);
$this->assertNotEmpty($result['error']);
// Verify error message is from our defined messages or 'Unknown error'
$this->assertTrue(
in_array($result['error'], array_values(Strings::PREG_ERROR_MESSAGES)) ||
$result['error'] === 'Unknown error'
);
}
}
/**
* Test validateRegex array structure
*/
public function testValidateRegexArrayStructure(): void
{
$result = Strings::validateRegex('/test/');
// Test array structure for valid regex
$this->assertIsArray($result);
$this->assertCount(4, $result);
$this->assertArrayHasKey('valid', $result);
$this->assertArrayHasKey('preg_error', $result);
$this->assertArrayHasKey('error', $result);
$result = Strings::validateRegex('/[/');
// Test array structure for invalid regex
$this->assertIsArray($result);
$this->assertCount(4, $result);
$this->assertArrayHasKey('valid', $result);
$this->assertArrayHasKey('preg_error', $result);
$this->assertArrayHasKey('error', $result);
$this->assertArrayHasKey('pcre_error', $result);
}
/**
* Test that methods handle edge cases properly
*/
public function testEdgeCases(): void
{
// Empty string
$this->assertFalse(Strings::isValidRegex(''));
$result = Strings::validateRegex('');
$this->assertFalse($result['valid']);
// Very long pattern
$longPattern = '/' . str_repeat('a', 1000) . '/';
$this->assertTrue(Strings::isValidRegex($longPattern));
// Unicode patterns
$this->assertTrue(Strings::isValidRegex('/\p{L}+/u'));
$this->assertTrue(Strings::isValidRegex('/[α-ω]+/u'));
}
/**
* Test PREG_ERROR_MESSAGES constant accessibility
*/
public function testPregErrorMessagesConstant(): void
{
$this->assertIsArray(Strings::PREG_ERROR_MESSAGES);
$this->assertNotEmpty(Strings::PREG_ERROR_MESSAGES);
// Check that all expected PREG constants are defined
$expectedKeys = [
PREG_NO_ERROR,
PREG_INTERNAL_ERROR,
PREG_BACKTRACK_LIMIT_ERROR,
PREG_RECURSION_LIMIT_ERROR,
PREG_BAD_UTF8_ERROR,
PREG_BAD_UTF8_OFFSET_ERROR,
PREG_JIT_STACKLIMIT_ERROR,
];
foreach ($expectedKeys as $key) {
$this->assertArrayHasKey($key, Strings::PREG_ERROR_MESSAGES);
$this->assertIsString(Strings::PREG_ERROR_MESSAGES[$key]);
$this->assertNotEmpty(Strings::PREG_ERROR_MESSAGES[$key]);
}
}
/**
* Test error state isolation between method calls
*/
public function testErrorStateIsolation(): void
{
// Start with invalid regex
Strings::isValidRegex('/[/');
$firstError = Strings::getLastRegexErrorString();
$this->assertNotEquals('No error', $firstError);
// Use valid regex
Strings::isValidRegex('/valid/');
$secondError = Strings::getLastRegexErrorString();
$this->assertEquals('No error', $secondError);
// Verify validateRegex clears previous errors
$result = Strings::validateRegex('/valid/');
$this->assertTrue($result['valid']);
$this->assertEquals(PREG_NO_ERROR, $result['preg_error']);
}
/**
* Test various regex delimiters
*/
public function testDifferentDelimiters(): void
{
$patterns = [
'/test/', // forward slash
'#test#', // hash
'~test~', // tilde
'|test|', // pipe
'@test@', // at symbol
'!test!', // exclamation
'%test%', // percent
];
foreach ($patterns as $pattern) {
$this->assertTrue(
Strings::isValidRegex($pattern),
"Pattern with delimiter '{$pattern}' should be valid"
);
}
}
}
// __END__

View File

@@ -632,15 +632,33 @@ final class CoreLibsConvertStringsTest extends TestCase
return [
'valid regex' => [
'/^[A-z]$/',
true
true,
[
'valid' => true,
'preg_error' => 0,
'error' => null,
'pcre_error' => null
],
],
'invalid regex A' => [
'/^[A-z]$',
false
false,
[
'valid' => false,
'preg_error' => 1,
'error' => 'Internal PCRE error',
'pcre_error' => 'Internal error'
],
],
'invalid regex B' => [
'/^[A-z$',
false
false,
[
'valid' => false,
'preg_error' => 1,
'error' => 'Internal PCRE error',
'pcre_error' => 'Internal error'
],
],
];
}
@@ -656,11 +674,23 @@ final class CoreLibsConvertStringsTest extends TestCase
* @param bool $expected
* @return void
*/
public function testIsValidRegexSimple(string $input, bool $expected): void
public function testIsValidRegexSimple(string $input, bool $expected, array $expected_extended): void
{
$this->assertEquals(
$expected,
\CoreLibs\Convert\Strings::isValidRegexSimple($input)
\CoreLibs\Convert\Strings::isValidRegex($input),
'Regex is not valid'
);
$this->assertEquals(
$expected_extended,
\CoreLibs\Convert\Strings::validateRegex($input),
'Validation of regex failed'
);
$this->assertEquals(
// for true null is set, so we get here No Error
$expected_extended['error'] ?? \CoreLibs\Convert\Strings::PREG_ERROR_MESSAGES[0],
\CoreLibs\Convert\Strings::getLastRegexErrorString(),
'Cannot match last preg error string'
);
}
}

View File

@@ -128,11 +128,17 @@ print "Unique: " . Strings::removeDuplicates($input_string) . "<br>";
print "Unique: " . Strings::removeDuplicates(strtolower($input_string)) . "<br>";
$regex_string = "/^[A-z]$/";
print "Regex valid: " . $regex_string . ": "
. DgS::prBl(Strings::isValidRegexSimple($regex_string)) . "<br>";
print "Regex is: " . $regex_string . ": " . DgS::prBl(Strings::isValidRegex($regex_string)) . "<br>";
$regex_string = "'//test{//'";
print "Regex is: " . $regex_string . ": " . DgS::prBl(Strings::isValidRegex($regex_string)) . "<br>";
print "Regex is: " . $regex_string . ": " . DgS::printAr(Strings::validateRegex($regex_string)) . "<br>";
$regex_string = "/^[A-z";
print "Regex valid: " . $regex_string . ": "
. DgS::prBl(Strings::isValidRegexSimple($regex_string)) . "<br>";
print "Regex is: " . $regex_string . ": " . DgS::prBl(Strings::isValidRegex($regex_string)) . "<br>";
print "[A] LAST PREGE ERROR: " . preg_last_error() . " -> "
. (Strings::PREG_ERROR_MESSAGES[preg_last_error()] ?? '-') . "<br>";
$preg_error = Strings::isValidRegex($regex_string);
print "[B] LAST PREGE ERROR: " . preg_last_error() . " -> "
. Strings::getLastRegexErrorString() . " -> " . preg_last_error_msg() . "<br>";
print "</body></html>";

View File

@@ -12,6 +12,16 @@ use CoreLibs\Combined\ArrayHandler;
class Strings
{
/** @var array<int,string> all the preg error messages */
public const array PREG_ERROR_MESSAGES = [
PREG_NO_ERROR => 'No error',
PREG_INTERNAL_ERROR => 'Internal PCRE error',
PREG_BACKTRACK_LIMIT_ERROR => 'Backtrack limit exhausted',
PREG_RECURSION_LIMIT_ERROR => 'Recursion limit exhausted',
PREG_BAD_UTF8_ERROR => 'Malformed UTF-8 data',
PREG_BAD_UTF8_OFFSET_ERROR => 'Bad UTF-8 offset',
PREG_JIT_STACKLIMIT_ERROR => 'JIT stack limit exhausted'
];
/**
* return the number of elements in the split list
* 0 if nothing / invalid split
@@ -259,8 +269,9 @@ class Strings
* @param string $pattern Any regex string
* @return bool False on invalid regex
*/
public static function isValidRegexSimple(string $pattern): bool
public static function isValidRegex(string $pattern): bool
{
preg_last_error();
try {
$var = '';
@preg_match($pattern, $var);
@@ -269,6 +280,41 @@ class Strings
return false;
}
}
/**
* Returns the last preg error messages as string
* all messages are defined in PREG_ERROR_MESSAGES
*
* @return string
*/
public static function getLastRegexErrorString(): string
{
return self::PREG_ERROR_MESSAGES[preg_last_error()] ?? 'Unknown error';
}
/**
* check if a regex is invalid, returns array with flag and error string
*
* @param string $pattern
* @return array{valid:bool,preg_error:0,error:null|string,pcre_error:null|string}
*/
public static function validateRegex(string $pattern): array
{
// Clear any previous PCRE errors
preg_last_error();
$var = '';
if (@preg_match($pattern, $var) === false) {
$error = preg_last_error();
return [
'valid' => false,
'preg_error' => $error,
'error' => self::PREG_ERROR_MESSAGES[$error] ?? 'Unknown error',
'pcre_error' => preg_last_error_msg(),
];
}
return ['valid' => true, 'preg_error' => PREG_NO_ERROR, 'error' => null, 'pcre_error' => null];
}
}
// __END__