Fixed bug #79257

Replace an existing entry for a given name only if we have a match.

3 files changed, 237 insertions, 5 deletions

diff --git a/NEWS b/NEWS
index 1c1b11382a..d8467ac018 100644
--- a/NEWS
+++ b/NEWS
@@ -21,6 +21,8 @@ PHP                                                                        NEWS
   . Fixed bug #79188 (Memory corruption in preg_replace/preg_replace_callback
     and unicode). (Nikita)
   . Fixed bug #79241 (Segmentation fault on preg_match()). (Nikita)
+  . Fixed bug #79257 (Duplicate named groups (?J) prefer last alternative even
+    if not matched). (Nikita)
 
 - Standard:
   . Fixed bug #79254 (getenv() w/o arguments not showing changes). (cmb)
diff --git a/ext/pcre/php_pcre.c b/ext/pcre/php_pcre.c
index c50bd2fba2..39896bb07b 100644
--- a/ext/pcre/php_pcre.c
+++ b/ext/pcre/php_pcre.c
@@ -995,6 +995,20 @@ static inline void populate_match_value(
 	}
 }
 
+static inline void add_named(
+		zval *subpats, zend_string *name, zval *val, zend_bool unmatched) {
+	/* If the DUPNAMES option is used, multiple subpatterns might have the same name.
+	 * In this case we want to preserve the one that actually has a value. */
+	if (!unmatched) {
+		zend_hash_update(Z_ARRVAL_P(subpats), name, val);
+	} else {
+		if (!zend_hash_add(Z_ARRVAL_P(subpats), name, val)) {
+			return;
+		}
+	}
+	Z_TRY_ADDREF_P(val);
+}
+
 /* {{{ add_offset_pair */
 static inline void add_offset_pair(
 		zval *result, const char *subject, PCRE2_SIZE start_offset, PCRE2_SIZE end_offset,
@@ -1023,8 +1037,7 @@ static inline void add_offset_pair(
 	}
 
 	if (name) {
-		Z_ADDREF(match_pair);
-		zend_hash_update(Z_ARRVAL_P(result), name, &match_pair);
+		add_named(result, name, &match_pair, start_offset == PCRE2_UNSET);
 	}
 	zend_hash_next_index_insert(Z_ARRVAL_P(result), &match_pair);
 }
@@ -1054,8 +1067,7 @@ static void populate_subpat_array(
 				populate_match_value(
 					&val, subject, offsets[2*i], offsets[2*i+1], unmatched_as_null);
 				if (subpat_names[i]) {
-					Z_TRY_ADDREF(val);
-					zend_hash_update(Z_ARRVAL_P(subpats), subpat_names[i], &val);
+					add_named(subpats, subpat_names[i], &val, offsets[2*i] == PCRE2_UNSET);
 				}
 				zend_hash_next_index_insert(Z_ARRVAL_P(subpats), &val);
 			}
@@ -1063,7 +1075,7 @@ static void populate_subpat_array(
 				for (i = count; i < num_subpats; i++) {
 					ZVAL_NULL(&val);
 					if (subpat_names[i]) {
-						zend_hash_update(Z_ARRVAL_P(subpats), subpat_names[i], &val);
+						zend_hash_add(Z_ARRVAL_P(subpats), subpat_names[i], &val);
 					}
 					zend_hash_next_index_insert(Z_ARRVAL_P(subpats), &val);
 				}
diff --git a/ext/pcre/tests/bug79257.phpt b/ext/pcre/tests/bug79257.phpt
new file mode 100644
index 0000000000..b52a0d6574
--- /dev/null
+++ b/ext/pcre/tests/bug79257.phpt
@@ -0,0 +1,218 @@
+--TEST--
+Bug #79257: Duplicate named groups (?J) prefer last alternative even if not matched
+--FILE--
+<?php
+
+preg_match('/(?J)(?:(?<g>foo)|(?<g>bar))/', 'foo', $matches);
+var_dump($matches);
+preg_match('/(?J)(?:(?<g>foo)|(?<g>bar))/', 'foo', $matches,
+    PREG_UNMATCHED_AS_NULL);
+var_dump($matches);
+preg_match('/(?J)(?:(?<g>foo)|(?<g>bar))/', 'foo', $matches,
+    PREG_OFFSET_CAPTURE);
+var_dump($matches);
+preg_match('/(?J)(?:(?<g>foo)|(?<g>bar))/', 'foo', $matches,
+    PREG_UNMATCHED_AS_NULL|PREG_OFFSET_CAPTURE);
+var_dump($matches);
+
+preg_match('/(?J)(?:(?<g>foo)|(?<g>bar))(?<h>baz)/', 'foobaz', $matches);
+var_dump($matches);
+preg_match('/(?J)(?:(?<g>foo)|(?<g>bar))(?<h>baz)/', 'foobaz', $matches,
+    PREG_UNMATCHED_AS_NULL);
+var_dump($matches);
+preg_match('/(?J)(?:(?<g>foo)|(?<g>bar))(?<h>baz)/', 'foobaz', $matches,
+    PREG_OFFSET_CAPTURE);
+var_dump($matches);
+preg_match('/(?J)(?:(?<g>foo)|(?<g>bar))(?<h>baz)/', 'foobaz', $matches,
+    PREG_UNMATCHED_AS_NULL|PREG_OFFSET_CAPTURE);
+var_dump($matches);
+
+?>
+--EXPECT--
+array(3) {
+  [0]=>
+  string(3) "foo"
+  ["g"]=>
+  string(3) "foo"
+  [1]=>
+  string(3) "foo"
+}
+array(4) {
+  [0]=>
+  string(3) "foo"
+  ["g"]=>
+  string(3) "foo"
+  [1]=>
+  string(3) "foo"
+  [2]=>
+  NULL
+}
+array(3) {
+  [0]=>
+  array(2) {
+    [0]=>
+    string(3) "foo"
+    [1]=>
+    int(0)
+  }
+  ["g"]=>
+  array(2) {
+    [0]=>
+    string(3) "foo"
+    [1]=>
+    int(0)
+  }
+  [1]=>
+  array(2) {
+    [0]=>
+    string(3) "foo"
+    [1]=>
+    int(0)
+  }
+}
+array(4) {
+  [0]=>
+  array(2) {
+    [0]=>
+    string(3) "foo"
+    [1]=>
+    int(0)
+  }
+  ["g"]=>
+  array(2) {
+    [0]=>
+    string(3) "foo"
+    [1]=>
+    int(0)
+  }
+  [1]=>
+  array(2) {
+    [0]=>
+    string(3) "foo"
+    [1]=>
+    int(0)
+  }
+  [2]=>
+  array(2) {
+    [0]=>
+    NULL
+    [1]=>
+    int(-1)
+  }
+}
+array(6) {
+  [0]=>
+  string(6) "foobaz"
+  ["g"]=>
+  string(3) "foo"
+  [1]=>
+  string(3) "foo"
+  [2]=>
+  string(0) ""
+  ["h"]=>
+  string(3) "baz"
+  [3]=>
+  string(3) "baz"
+}
+array(6) {
+  [0]=>
+  string(6) "foobaz"
+  ["g"]=>
+  string(3) "foo"
+  [1]=>
+  string(3) "foo"
+  [2]=>
+  NULL
+  ["h"]=>
+  string(3) "baz"
+  [3]=>
+  string(3) "baz"
+}
+array(6) {
+  [0]=>
+  array(2) {
+    [0]=>
+    string(6) "foobaz"
+    [1]=>
+    int(0)
+  }
+  ["g"]=>
+  array(2) {
+    [0]=>
+    string(3) "foo"
+    [1]=>
+    int(0)
+  }
+  [1]=>
+  array(2) {
+    [0]=>
+    string(3) "foo"
+    [1]=>
+    int(0)
+  }
+  [2]=>
+  array(2) {
+    [0]=>
+    string(0) ""
+    [1]=>
+    int(-1)
+  }
+  ["h"]=>
+  array(2) {
+    [0]=>
+    string(3) "baz"
+    [1]=>
+    int(3)
+  }
+  [3]=>
+  array(2) {
+    [0]=>
+    string(3) "baz"
+    [1]=>
+    int(3)
+  }
+}
+array(6) {
+  [0]=>
+  array(2) {
+    [0]=>
+    string(6) "foobaz"
+    [1]=>
+    int(0)
+  }
+  ["g"]=>
+  array(2) {
+    [0]=>
+    string(3) "foo"
+    [1]=>
+    int(0)
+  }
+  [1]=>
+  array(2) {
+    [0]=>
+    string(3) "foo"
+    [1]=>
+    int(0)
+  }
+  [2]=>
+  array(2) {
+    [0]=>
+    NULL
+    [1]=>
+    int(-1)
+  }
+  ["h"]=>
+  array(2) {
+    [0]=>
+    string(3) "baz"
+    [1]=>
+    int(3)
+  }
+  [3]=>
+  array(2) {
+    [0]=>
+    string(3) "baz"
+    [1]=>
+    int(3)
+  }
+}