Strings::matchAll(): added option 'lazy'

Author: dg

src/Utils/Strings.php

@@ -589,6 +589,7 @@ public static function match(
 	/**
 	 * Searches the string for all occurrences matching the regular expression and
 	 * returns an array of arrays containing the found expression and each subexpression.
+	 * @return ($lazy is true ? \Generator<int, array> : array[])
 	 */
 	public static function matchAll(
 		string $subject,
@@ -599,21 +600,41 @@ public static function matchAll(
 		bool $unmatchedAsNull = false,
 		bool $patternOrder = false,
 		bool $utf8 = false,
-	): array
+		bool $lazy = false,
+	): array|\Generator
 	{
-		$flags = is_int($captureOffset) // back compatibility
-			? $captureOffset
-			: ($captureOffset ? PREG_OFFSET_CAPTURE : 0) | ($unmatchedAsNull ? PREG_UNMATCHED_AS_NULL : 0) | ($patternOrder ? PREG_PATTERN_ORDER : 0);
-
 		if ($utf8) {
 			$offset = strlen(self::substring($subject, 0, $offset));
 			$pattern .= 'u';
 		}
 
+		if ($lazy) {
+			$flags = PREG_OFFSET_CAPTURE | ($unmatchedAsNull ? PREG_UNMATCHED_AS_NULL : 0);
+			return (function () use ($utf8, $captureOffset, $flags, $subject, $pattern, $offset) {
+				$counter = 0;
+				while (
+					$offset <= strlen($subject) - ($counter ? 1 : 0)
+					&& self::pcre('preg_match', [$pattern, $subject, &$m, $flags, $offset])
+				) {
+					$offset = $m[0][1] + max(1, strlen($m[0][0]));
+					if (!$captureOffset) {
+						$m = array_map(fn($item) => $item[0], $m);
+					} elseif ($utf8) {
+						$m = self::bytesToChars($subject, [$m])[0];
+					}
+					yield $counter++ => $m;
+				}
+			})();
+		}
+
 		if ($offset > strlen($subject)) {
 			return [];
 		}
 
+		$flags = is_int($captureOffset) // back compatibility
+			? $captureOffset
+			: ($captureOffset ? PREG_OFFSET_CAPTURE : 0) | ($unmatchedAsNull ? PREG_UNMATCHED_AS_NULL : 0) | ($patternOrder ? PREG_PATTERN_ORDER : 0);
+
 		self::pcre('preg_match_all', [
 			$pattern, $subject, &$m,
 			($flags & PREG_PATTERN_ORDER) ? $flags : ($flags | PREG_SET_ORDER),
@@ -622,7 +643,6 @@ public static function matchAll(
 		return $utf8 && $captureOffset
 			? self::bytesToChars($subject, $m)
 			: $m;
-
 	}
 
 

tests/Utils/Strings.match().phpt

@@ -41,5 +41,6 @@ Assert::same([['k', 7]], Strings::match('žluťoučký kůň', '#[e-l]+#u', capt
 
 
 // right edge
-Assert::null(Strings::match('hello world!', '', offset: 50));
-Assert::null(Strings::match('', '', offset: 1));
+Assert::same([''], Strings::match('he', '#(?<=e)#', offset: 2));
+Assert::same(null, Strings::match('he', '#(?<=x)#', offset: 2));
+Assert::same(null, Strings::match('he', '##', offset: 3));

tests/Utils/Strings.matchAll().phpt

@@ -16,6 +16,12 @@ require __DIR__ . '/../bootstrap.php';
 Assert::same([], Strings::matchAll('hello world!', '#([E-L])+#'));
 
 
+// sentinel
+Assert::same([
+	[''], [''], [''],
+], Strings::matchAll('he', '##'));
+
+
 // capturing
 Assert::same([
 	['hell', 'l'],
@@ -81,4 +87,6 @@ Assert::same([['e', null]], Strings::matchAll('hello world!', '#e(x)*#', unmatch
 
 
 // right edge
-Assert::same([], Strings::matchAll('hello world!', '', offset: 50));
+Assert::same([['']], Strings::matchAll('he', '#(?<=e)#', offset: 2));
+Assert::same([], Strings::matchAll('he', '#(?<=x)#', offset: 2));
+Assert::same([], Strings::matchAll('he', '##', offset: 3));

tests/Utils/Strings.matchAll.lazy().phpt

@@ -0,0 +1,89 @@
+<?php
+
+/**
+ * Test: Nette\Utils\Strings::matchAll()
+ */
+
+declare(strict_types=1);
+
+use Nette\Utils\Strings;
+use Tester\Assert;
+
+require __DIR__ . '/../bootstrap.php';
+
+
+// not matched
+Assert::type(Generator::class, Strings::matchAll('hello world!', '#([E-L])+#', lazy: true));
+Assert::same(0, iterator_count(Strings::matchAll('hello world!', '#([E-L])+#', lazy: true)));
+
+
+// sentinel
+Assert::same(
+	[['h'], ['e']],
+	iterator_to_array(Strings::matchAll('he', '#.#', lazy: true)),
+);
+
+Assert::same(
+	[[''], ['']],
+	iterator_to_array(Strings::matchAll('he', '##', lazy: true)),
+);
+
+
+// right edge
+Assert::same(
+	[['']],
+	iterator_to_array(Strings::matchAll('he', '#(?<=e)#', offset: 2, lazy: true)),
+);
+
+Assert::same(
+	[],
+	iterator_to_array(Strings::matchAll('he', '#(?<=x)#', offset: 2, lazy: true)),
+);
+
+Assert::same(
+	[],
+	iterator_to_array(Strings::matchAll('he', '##', offset: 3, lazy: true)),
+);
+
+
+// capturing
+Assert::same([
+	['hell', 'l'],
+	['l', 'l'],
+], iterator_to_array(Strings::matchAll('hello world!', '#([e-l])+#', lazy: true)));
+
+Assert::same([
+	['hell'],
+	['l'],
+], iterator_to_array(Strings::matchAll('hello world!', '#[e-l]+#', lazy: true)));
+
+
+// options
+Assert::same([
+	[['lu', 2], ['l', 2], ['u', 3]],
+	[['ou', 6], ['o', 6], ['u', 7]],
+	[['k', 10], ['k', 10], ['', 11]],
+	[['k', 14], ['k', 14], ['', 15]],
+], iterator_to_array(Strings::matchAll('žluťoučký kůň!', '#([a-z])([a-z]*)#u', captureOffset: true, lazy: true)));
+
+Assert::same([
+	[['lu', 1], ['l', 1], ['u', 2]],
+	[['ou', 4], ['o', 4], ['u', 5]],
+	[['k', 7], ['k', 7], ['', 8]],
+	[['k', 10], ['k', 10], ['', 11]],
+], iterator_to_array(Strings::matchAll('žluťoučký kůň!', '#([a-z])([a-z]*)#u', captureOffset: true, utf8: true, lazy: true)));
+
+Assert::same(
+	[['l'], ['k'], ['k']],
+	iterator_to_array(Strings::matchAll('žluťoučký kůň', '#[e-l]+#u', offset: 2, lazy: true)),
+);
+
+Assert::same(
+	[['k'], ['k']],
+	iterator_to_array(Strings::matchAll('žluťoučký kůň', '#[e-l]+#u', offset: 2, utf8: true, lazy: true)),
+);
+
+Assert::same(
+	[['e', null]],
+	iterator_to_array(Strings::matchAll('hello world!', '#e(x)*#', unmatchedAsNull: true, lazy: true)),
+);