pdo_odbc: allocate sufficient space for retrieving unicode data

Closes GH-9498

Author: jnahmias

ext/pdo_odbc/odbc_stmt.c

@@ -607,6 +607,19 @@ static int odbc_stmt_describe(pdo_stmt_t *stmt, int colno)
 		}
 	}
 	colsize = displaysize;
+	if ( S->cols[colno].coltype == SQL_WCHAR
+#ifdef SQL_WVARCHAR
+		|| S->cols[colno].coltype == SQL_WVARCHAR
+#endif
+#ifdef SQL_WLONGVARCHAR
+		|| S->cols[colno].coltype == SQL_WLONGVARCHAR
+#endif
+	) {
+		/* displaysize is counted by characters;
+		for unicode, each could take up to 4 bytes in UTF-8;
+		see https://www.rfc-editor.org/rfc/rfc3629 */
+		colsize = displaysize * 4;
+	}
 
 	col->maxlen = S->cols[colno].datalen = colsize;
 	col->name = zend_string_init(S->cols[colno].colname, colnamelen, 0);

ext/pdo_odbc/tests/gh9498_1.phpt

@@ -0,0 +1,46 @@
+--TEST--
+Bug GH-9498 (unicode string corruption during SELECT)
+--EXTENSIONS--
+pdo_odbc
+--SKIPIF--
+<?php
+require 'ext/pdo/tests/pdo_test.inc';
+PDOTest::skip();
+?>
+--FILE--
+<?php
+require 'ext/pdo/tests/pdo_test.inc';
+$db = PDOTest::test_factory(dirname(__FILE__) . '/common.phpt');
+$db->setAttribute(\PDO::ATTR_ERRMODE, \PDO::ERRMODE_EXCEPTION);
+
+$select_test = "SELECT
+    CAST(0x72006f017e016f007600fd00 AS nvarchar(6)) AS pink_in_czech_binary_encoded_utf16le
+  , CAST(0x6b26 AS nvarchar(1)) AS beamed_eighth_notes_binary_encoded_utf16le
+  , NCHAR(0x266b) AS beamed_eighth_notes_nchar
+  , N'♫' AS \"beamed_eighth_notes_u+266b_constant\"
+  , N'růžový' AS pink_in_czech_unicode_string_constant
+  , NCHAR(0xfb01) AS ligature_fi_nchar
+  , N'𝄋' AS \"segno_u+1d10b_constant\"
+;";
+var_dump($db->query($select_test, \PDO::FETCH_ASSOC)->fetchAll());
+?>
+--EXPECT--
+array(1) {
+  [0]=>
+  array(7) {
+    ["pink_in_czech_binary_encoded_utf16le"]=>
+    string(9) "růžový"
+    ["beamed_eighth_notes_binary_encoded_utf16le"]=>
+    string(3) "♫"
+    ["beamed_eighth_notes_nchar"]=>
+    string(3) "♫"
+    ["beamed_eighth_notes_u+266b_constant"]=>
+    string(3) "♫"
+    ["pink_in_czech_unicode_string_constant"]=>
+    string(9) "růžový"
+    ["ligature_fi_nchar"]=>
+    string(3) "fi"
+    ["segno_u+1d10b_constant"]=>
+    string(4) "𝄋"
+  }
+}