Add test to ensure invalid UTF-8 strings are not marked as valid

Author: Girgias

ext/zend_test/tests/strings_not_marked_as_utf8.phpt

@@ -0,0 +1,131 @@
+--TEST--
+Check that invalid UTF-8 strings are NOT marked as valid UTF-8
+--EXTENSIONS--
+zend_test
+--FILE--
+<?php
+// Invalid 2 Octet Sequence
+$non_utf8 = "\xc3\x28";
+
+echo "Integer cast to string concatenated to invalid UTF-8:\n";
+$i = 2563;
+$s = (string) $i;
+$s .= "\xc3\x28";
+var_dump(zend_test_is_string_marked_as_valid_utf8($s));
+
+echo "Float cast to string concatenated to invalid UTF-8:\n";
+$f = 26.7;
+$s = (string) $f;
+$s .= "\xc3\x28";
+var_dump(zend_test_is_string_marked_as_valid_utf8($s));
+$f = 2e100;
+$s = (string) $f;
+$s .= "\xc3\x28";
+var_dump(zend_test_is_string_marked_as_valid_utf8($s));
+
+echo "Concatenation known valid UTF-8 strings in variables, followed by concatenation of invalid UTF-8:\n";
+$s1 = "f";
+$s2 = "o";
+$s = $s1 . $s2;
+$s = $s . $non_utf8;
+var_dump(zend_test_is_string_marked_as_valid_utf8($s));
+
+echo "Multiple concatenation known valid UTF-8 strings in variables, followed by concatenation of invalid UTF-8:\n";
+$s1 = "f";
+$s2 = "o";
+$s = $s1 . $s2 . $non_utf8;
+var_dump(zend_test_is_string_marked_as_valid_utf8($s));
+
+echo "Concatenation known valid UTF-8 with invalid UTF-8 in assignment:\n";
+$s = "f" . "\xc3\x28";
+var_dump(zend_test_is_string_marked_as_valid_utf8($s));
+
+// The "foo" string matches with a "Foo" class which is registered by the zend_test extension.
+// That class name does not have the "valid UTF-8" flag because class names in general
+// don't have to be UTF-8. As the "foo" string here goes through the interning logic,
+// the string gets replaced by the "foo" string from the class, which does
+// not have the "valid UTF-8" flag. We therefore choose a different test case: "fxo".
+// The previous "foo" test case works because it is not interned.
+echo "Multiple concatenation known valid UTF-8 and invalid UTF-8 in assignment:\n";
+$s = "f" . "o" . "\xc3\x28";
+var_dump(zend_test_is_string_marked_as_valid_utf8($s));
+
+echo "Concatenation known valid UTF-8 string with empty string in variables, followed by concatenation of invalid UTF-8:\n";
+$s1 = "f";
+$s2 = "";
+$s = $s1 . $s2;
+$s = $s . $non_utf8;
+var_dump(zend_test_is_string_marked_as_valid_utf8($s));
+$s1 = "f";
+$s2 = "";
+$s = $s2 . $s1;
+$s = $s . $non_utf8;
+var_dump(zend_test_is_string_marked_as_valid_utf8($s));
+
+echo "Concatenation known valid UTF-8 string with empty string in assignment, followed by concatenation of invalid UTF-8:\n";
+$s = "f" . "";
+$s = $s . $non_utf8;
+var_dump(zend_test_is_string_marked_as_valid_utf8($s));
+$s = "" . "f";
+$s = $s . $non_utf8;
+var_dump(zend_test_is_string_marked_as_valid_utf8($s));
+
+echo "Concatenation in loop:\n";
+const COPY_TIMES = 10_000;
+$string = "a";
+
+$string_concat = $string;
+for ($i = 1; $i < COPY_TIMES; $i++) {
+    $string_concat = $string_concat . $string;
+}
+$string_concat = $string_concat . $non_utf8;
+var_dump(zend_test_is_string_marked_as_valid_utf8($string_concat));
+
+echo "Concatenation in loop (compound assignment):\n";
+$string = "a";
+
+$string_concat = $string;
+for ($i = 1; $i < COPY_TIMES; $i++) {
+    $string_concat .= $string;
+}
+$string_concat = $string_concat . $non_utf8;
+var_dump(zend_test_is_string_marked_as_valid_utf8($string_concat));
+
+echo "Concatenation of objects:\n";
+class ToString {
+    public function __toString() : string{
+        return "z";
+    }
+}
+$o = new ToString();
+$s = $o . $o;
+$s = $s . $non_utf8;
+var_dump(zend_test_is_string_marked_as_valid_utf8($s));
+
+?>
+--EXPECT--
+Integer cast to string concatenated to invalid UTF-8:
+bool(false)
+Float cast to string concatenated to invalid UTF-8:
+bool(false)
+bool(false)
+Concatenation known valid UTF-8 strings in variables, followed by concatenation of invalid UTF-8:
+bool(false)
+Multiple concatenation known valid UTF-8 strings in variables, followed by concatenation of invalid UTF-8:
+bool(false)
+Concatenation known valid UTF-8 with invalid UTF-8 in assignment:
+bool(false)
+Multiple concatenation known valid UTF-8 and invalid UTF-8 in assignment:
+bool(false)
+Concatenation known valid UTF-8 string with empty string in variables, followed by concatenation of invalid UTF-8:
+bool(false)
+bool(false)
+Concatenation known valid UTF-8 string with empty string in assignment, followed by concatenation of invalid UTF-8:
+bool(false)
+bool(false)
+Concatenation in loop:
+bool(false)
+Concatenation in loop (compound assignment):
+bool(false)
+Concatenation of objects:
+bool(false)