Introduce convenience macros for copying flags that hold when concatenating two strings

This abstracts away, and cleans up, the flag handling for properties of
strings that hold when concatenating two strings if they both hold that
property. (These macros also work with simply copies of strings because
a copy of a string can be considered a concatenation with the empty
string.) This gets rid of some branches and some repetitive code, and
leaves room for adding more flags like these in the future.

Author: nielsdos

Zend/zend_operators.c

@@ -1981,11 +1981,7 @@ ZEND_API zend_result ZEND_FASTCALL concat_function(zval *result, zval *op1, zval
 		size_t op2_len = Z_STRLEN_P(op2);
 		size_t result_len = op1_len + op2_len;
 		zend_string *result_str;
-		uint32_t flags = 0;
-
-		if (ZSTR_IS_VALID_UTF8(Z_STR_P(op1)) && ZSTR_IS_VALID_UTF8(Z_STR_P(op2))) {
-			flags = IS_STR_VALID_UTF8;
-		}
+		uint32_t flags = ZSTR_GET_COPYABLE_CONCAT_PROPERTIES_BOTH(Z_STR_P(op1), Z_STR_P(op2));
 
 		if (UNEXPECTED(op1_len > ZSTR_MAX_LEN - op2_len)) {
 			zend_throw_error(NULL, "String size overflow");

Zend/zend_string.c

@@ -191,11 +191,9 @@ ZEND_API zend_string* ZEND_FASTCALL zend_interned_string_find_permanent(zend_str
 	return zend_interned_string_ht_lookup(str, &interned_strings_permanent);
 }
 
-static zend_string* ZEND_FASTCALL zend_init_string_for_interning(zend_string *str, bool persistent) {
-	uint32_t flags = 0;
-	if (ZSTR_IS_VALID_UTF8(str)) {
-		flags = IS_STR_VALID_UTF8;
-	}
+static zend_string* ZEND_FASTCALL zend_init_string_for_interning(zend_string *str, bool persistent)
+{
+	uint32_t flags = ZSTR_GET_COPYABLE_CONCAT_PROPERTIES(str);
 	zend_ulong h = ZSTR_H(str);
 	zend_string_delref(str);
 	str = zend_string_init(ZSTR_VAL(str), ZSTR_LEN(str), persistent);

Zend/zend_string.h

@@ -82,6 +82,27 @@ END_EXTERN_C()
 #define ZSTR_IS_INTERNED(s)					(GC_FLAGS(s) & IS_STR_INTERNED)
 #define ZSTR_IS_VALID_UTF8(s)				(GC_FLAGS(s) & IS_STR_VALID_UTF8)
 
+/* These are properties, encoded as flags, that will hold on the resulting string
+ * after concatenating two strings that have these property.
+ * Example: concatenating two UTF-8 strings yields another UTF-8 string. */
+#define ZSTR_COPYABLE_CONCAT_PROPERTIES		(IS_STR_VALID_UTF8)
+
+#define ZSTR_GET_COPYABLE_CONCAT_PROPERTIES(s) 				(GC_FLAGS(s) & ZSTR_COPYABLE_CONCAT_PROPERTIES)
+/* This macro returns the copyable concat properties which hold on both strings. */
+#define ZSTR_GET_COPYABLE_CONCAT_PROPERTIES_BOTH(s1, s2)	(GC_FLAGS(s1) & GC_FLAGS(s2) & ZSTR_COPYABLE_CONCAT_PROPERTIES)
+
+#define ZSTR_COPY_CONCAT_PROPERTIES(out, in) do { \
+	zend_string *_out = (out); \
+	uint32_t properties = ZSTR_GET_COPYABLE_CONCAT_PROPERTIES((in)); \
+	GC_ADD_FLAGS(_out, properties); \
+} while (0)
+
+#define ZSTR_COPY_CONCAT_PROPERTIES_BOTH(out, in1, in2) do { \
+	zend_string *_out = (out); \
+	uint32_t properties = ZSTR_GET_COPYABLE_CONCAT_PROPERTIES_BOTH((in1), (in2)); \
+	GC_ADD_FLAGS(_out, properties); \
+} while (0)
+
 #define ZSTR_EMPTY_ALLOC() zend_empty_string
 #define ZSTR_CHAR(c) zend_one_char_string[c]
 #define ZSTR_KNOWN(idx) zend_known_strings[idx]

Zend/zend_vm_def.h

@@ -384,11 +384,7 @@ ZEND_VM_HANDLER(8, ZEND_CONCAT, CONST|TMPVAR|CV, CONST|TMPVAR|CV, SPEC(NO_CONST_
 		zend_string *op1_str = Z_STR_P(op1);
 		zend_string *op2_str = Z_STR_P(op2);
 		zend_string *str;
-		uint32_t flags = 0;
-
-		if (ZSTR_IS_VALID_UTF8(op1_str) && ZSTR_IS_VALID_UTF8(op2_str)) {
-			flags = IS_STR_VALID_UTF8;
-		}
+		uint32_t flags = ZSTR_GET_COPYABLE_CONCAT_PROPERTIES_BOTH(op1_str, op2_str);
 
 		if (OP1_TYPE != IS_CONST && UNEXPECTED(ZSTR_LEN(op1_str) == 0)) {
 			if (OP2_TYPE == IS_CONST || OP2_TYPE == IS_CV) {
@@ -3147,11 +3143,7 @@ ZEND_VM_COLD_CONSTCONST_HANDLER(53, ZEND_FAST_CONCAT, CONST|TMPVAR|CV, CONST|TMP
 		zend_string *op1_str = Z_STR_P(op1);
 		zend_string *op2_str = Z_STR_P(op2);
 		zend_string *str;
-		uint32_t flags = 0;
-
-		if (ZSTR_IS_VALID_UTF8(op1_str) && ZSTR_IS_VALID_UTF8(op2_str)) {
-			flags = IS_STR_VALID_UTF8;
-		}
+		uint32_t flags = ZSTR_GET_COPYABLE_CONCAT_PROPERTIES_BOTH(op1_str, op2_str);
 
 		if (OP1_TYPE != IS_CONST && UNEXPECTED(ZSTR_LEN(op1_str) == 0)) {
 			if (OP2_TYPE == IS_CONST || OP2_TYPE == IS_CV) {
@@ -3248,9 +3240,7 @@ ZEND_VM_COLD_CONSTCONST_HANDLER(53, ZEND_FAST_CONCAT, CONST|TMPVAR|CV, CONST|TMP
 		memcpy(ZSTR_VAL(str), ZSTR_VAL(op1_str), ZSTR_LEN(op1_str));
 		memcpy(ZSTR_VAL(str) + ZSTR_LEN(op1_str), ZSTR_VAL(op2_str), ZSTR_LEN(op2_str)+1);
 
-		if (ZSTR_IS_VALID_UTF8(op1_str) && ZSTR_IS_VALID_UTF8(op2_str)) {
-			GC_ADD_FLAGS(str, IS_STR_VALID_UTF8);
-		}
+		ZSTR_COPY_CONCAT_PROPERTIES_BOTH(str, op1_str, op2_str);
 		ZVAL_NEW_STR(EX_VAR(opline->result.var), str);
 		if (OP1_TYPE != IS_CONST) {
 			zend_string_release_ex(op1_str, 0);