Merge pull request #64321 from KoBeWi/s_p_l_i_t

Add support for empty delimiter in `String.split()`

core/string/ustring.cpp

@@ -1180,9 +1180,14 @@ Vector<String> String::split(const String &p_splitter, bool p_allow_empty, int p
 	int len = length();
 
 	while (true) {
-		int end = find(p_splitter, from);
-		if (end < 0) {
-			end = len;
+		int end;
+		if (p_splitter.is_empty()) {
+			end = from + 1;
+		} else {
+			end = find(p_splitter, from);
+			if (end < 0) {
+				end = len;
+			}
 		}
 		if (p_allow_empty || (end > from)) {
 			if (p_maxsplit <= 0) {
@@ -1223,7 +1228,15 @@ Vector<String> String::rsplit(const String &p_splitter, bool p_allow_empty, int
 			break;
 		}
 
-		int left_edge = rfind(p_splitter, remaining_len - p_splitter.length());
+		int left_edge;
+		if (p_splitter.is_empty()) {
+			left_edge = remaining_len - 1;
+			if (left_edge == 0) {
+				left_edge--; // Skip to the < 0 condition.
+			}
+		} else {
+			left_edge = rfind(p_splitter, remaining_len - p_splitter.length());
+		}
 
 		if (left_edge < 0) {
 			// no more splitters, we're done

core/string/ustring.h

@@ -345,8 +345,8 @@ public:
 	String get_slice(String p_splitter, int p_slice) const;
 	String get_slicec(char32_t p_splitter, int p_slice) const;
 
-	Vector<String> split(const String &p_splitter, bool p_allow_empty = true, int p_maxsplit = 0) const;
-	Vector<String> rsplit(const String &p_splitter, bool p_allow_empty = true, int p_maxsplit = 0) const;
+	Vector<String> split(const String &p_splitter = "", bool p_allow_empty = true, int p_maxsplit = 0) const;
+	Vector<String> rsplit(const String &p_splitter = "", bool p_allow_empty = true, int p_maxsplit = 0) const;
 	Vector<String> split_spaces() const;
 	Vector<float> split_floats(const String &p_splitter, bool p_allow_empty = true) const;
 	Vector<float> split_floats_mk(const Vector<String> &p_splitters, bool p_allow_empty = true) const;

core/variant/variant_call.cpp

@@ -1509,8 +1509,8 @@ static void _register_variant_builtin_methods() {
 	bind_method(String, to_camel_case, sarray(), varray());
 	bind_method(String, to_pascal_case, sarray(), varray());
 	bind_method(String, to_snake_case, sarray(), varray());
-	bind_method(String, split, sarray("delimiter", "allow_empty", "maxsplit"), varray(true, 0));
-	bind_method(String, rsplit, sarray("delimiter", "allow_empty", "maxsplit"), varray(true, 0));
+	bind_method(String, split, sarray("delimiter", "allow_empty", "maxsplit"), varray("", true, 0));
+	bind_method(String, rsplit, sarray("delimiter", "allow_empty", "maxsplit"), varray("", true, 0));
 	bind_method(String, split_floats, sarray("delimiter", "allow_empty"), varray(true));
 	bind_method(String, join, sarray("parts"), varray());
 

doc/classes/String.xml

@@ -634,11 +634,11 @@
 		</method>
 		<method name="rsplit" qualifiers="const">
 			<return type="PackedStringArray" />
-			<param index="0" name="delimiter" type="String" />
+			<param index="0" name="delimiter" type="String" default="&quot;&quot;" />
 			<param index="1" name="allow_empty" type="bool" default="true" />
 			<param index="2" name="maxsplit" type="int" default="0" />
 			<description>
-				Splits the string by a [param delimiter] string and returns an array of the substrings, starting from right.
+				Splits the string by a [param delimiter] string and returns an array of the substrings, starting from right. If [param delimiter] is an empty string, each substring will be a single character.
 				The splits in the returned array are sorted in the same order as the original string, from left to right.
 				If [param allow_empty] is [code]true[/code], and there are two adjacent delimiters in the string, it will add an empty string to the array of substrings at this position.
 				If [param maxsplit] is specified, it defines the number of splits to do from the right up to [param maxsplit]. The default value of 0 means that all items are split, thus giving the same result as [method split].
@@ -710,11 +710,11 @@
 		</method>
 		<method name="split" qualifiers="const">
 			<return type="PackedStringArray" />
-			<param index="0" name="delimiter" type="String" />
+			<param index="0" name="delimiter" type="String" default="&quot;&quot;" />
 			<param index="1" name="allow_empty" type="bool" default="true" />
 			<param index="2" name="maxsplit" type="int" default="0" />
 			<description>
-				Splits the string by a [param delimiter] string and returns an array of the substrings. The [param delimiter] can be of any length.
+				Splits the string by a [param delimiter] string and returns an array of the substrings. The [param delimiter] can be of any length. If [param delimiter] is an empty string, each substring will be a single character.
 				If [param allow_empty] is [code]true[/code], and there are two adjacent delimiters in the string, it will add an empty string to the array of substrings at this position.
 				If [param maxsplit] is specified, it defines the number of splits to do from the left up to [param maxsplit]. The default value of [code]0[/code] means that all items are split.
 				If you need only one element from the array at a specific index, [method get_slice] is a more performant option.

tests/core/string/test_string.h

@@ -485,6 +485,7 @@ TEST_CASE("[String] Splitting") {
 
 	const char *slices_l[3] = { "Mars", "Jupiter", "Saturn,Uranus" };
 	const char *slices_r[3] = { "Mars,Jupiter", "Saturn", "Uranus" };
+	const char *slices_3[4] = { "t", "e", "s", "t" };
 
 	l = s.split(",", true, 2);
 	CHECK(l.size() == 3);
@@ -498,6 +499,13 @@ TEST_CASE("[String] Splitting") {
 		CHECK(l[i] == slices_r[i]);
 	}
 
+	s = "test";
+	l = s.split();
+	CHECK(l.size() == 4);
+	for (int i = 0; i < l.size(); i++) {
+		CHECK(l[i] == slices_3[i]);
+	}
+
 	s = "Mars Jupiter Saturn Uranus";
 	const char *slices_s[4] = { "Mars", "Jupiter", "Saturn", "Uranus" };
 	l = s.split_spaces();