DOMXPath::quote(string $str): string

ext/dom/php_dom.stub.php

@@ -934,6 +934,8 @@ public function registerNamespace(string $prefix, string $namespace): bool {}
         public function registerPhpFunctions(string|array|null $restrict = null): void {}
 
         public function registerPhpFunctionNS(string $namespaceURI, string $name, callable $callable): void {}
+
+        public static function quote(string $str): string {}
     }
 #endif
 

ext/dom/tests/DOMXPath_quote.phpt

@@ -0,0 +1,84 @@
+--TEST--
+Test DOMXPath::quote with various inputs
+--SKIPIF--
+<?php if (!class_exists('DOMXPath')) die('skip DOMXPath not available.'); ?>
+--FILE--
+<?php
+$dom = new DOMDocument();
+$xpath = new DOMXPath($dom);
+
+
+/**
+ * Quote a string for use in an XPath expression.
+ * 
+ * Example: $xp->query("//span[contains(text()," . $xp->quote($string) . ")]")
+ * 
+ * @param string $string string to quote.
+ * @return string quoted string.
+ */
+function UserlandDOMXPathQuote(string $string): string
+{
+    if (false === \strpos($string, '\'')) {
+        return '\'' . $string . '\'';
+    }
+    if (false === \strpos($string, '"')) {
+        return '"' . $string . '"';
+    }
+    // if the string contains both single and double quotes, construct an
+    // expression that concatenates all non-double-quote substrings with
+    // the quotes, e.g.:
+    //  'foo'"bar => concat("'foo'", '"bar")
+    $sb = [];
+    while ($string !== '') {
+        $bytesUntilSingleQuote = \strcspn($string, '\'');
+        $bytesUntilDoubleQuote = \strcspn($string, '"');
+        $quoteMethod = ($bytesUntilSingleQuote > $bytesUntilDoubleQuote) ? "'" : '"';
+        $bytesUntilQuote = \max($bytesUntilSingleQuote, $bytesUntilDoubleQuote);
+        $sb[] = $quoteMethod . \substr($string, 0, $bytesUntilQuote) . $quoteMethod;
+        $string = \substr($string, $bytesUntilQuote);
+    }
+    $sb = \implode(',', $sb);
+    return 'concat(' . $sb . ')';
+}
+
+
+
+$tests = [
+    'foo' => "'foo'", // no quotes
+    '"foo' => '\'"foo\'', // double quotes only
+    '\'foo' => '"\'foo"', // single quotes only
+    '\'foo"bar' => 'concat("\'foo",\'"bar\')', // both; double quotes in mid-string
+    '\'foo"bar"baz' => 'concat("\'foo",\'"bar"baz\')', // multiple double quotes in mid-string
+    '\'foo"' => 'concat("\'foo",\'"\')', // string ends with double quotes
+    '\'foo""' => 'concat("\'foo",\'""\')', // string ends with run of double quotes
+    '"\'foo' => 'concat(\'"\',"\'foo")', // string begins with double quotes
+    '""\'foo' => 'concat(\'""\',"\'foo")', // string begins with run of double quotes
+    '\'foo""bar' => 'concat("\'foo",\'""bar\')', // run of double quotes in mid-string
+];
+
+foreach ($tests as $input => $expected) {
+    $result = $xpath->quote($input);
+    if ($result === $expected) {
+        echo "Pass: {$input} => {$result}\n";
+    } else {
+        echo 'Fail: ';
+        var_dump([
+            'input' => $input,
+            'expected' => $expected,
+            'result' => $result,
+            'userland_implementation_result' => UserlandDOMXPathQuote($input),
+        ]);
+    }
+}
+?>
+--EXPECT--
+Pass: foo => 'foo'
+Pass: "foo => '"foo'
+Pass: 'foo => "'foo"
+Pass: 'foo"bar => concat("'foo",'"bar')
+Pass: 'foo"bar"baz => concat("'foo",'"bar"baz')
+Pass: 'foo" => concat("'foo",'"')
+Pass: 'foo"" => concat("'foo",'""')
+Pass: "'foo => concat('"',"'foo")
+Pass: ""'foo => concat('""',"'foo")
+Pass: 'foo""bar => concat("'foo",'""bar')

ext/dom/xpath.c

@@ -446,6 +446,93 @@ PHP_METHOD(DOMXPath, registerPhpFunctionNS)
 	);
 }
 
+/* {{{ */
+PHP_METHOD(DOMXPath, quote) {
+	char *input;
+	size_t input_len;
+	char *output;
+	size_t output_len = 0;
+
+	if (zend_parse_parameters(ZEND_NUM_ARGS(), "s", &input, &input_len) ==
+			FAILURE) {
+		RETURN_THROWS();
+	}
+	if (memchr(input, '\'', input_len) == NULL) {
+		output_len = input_len + 2;
+		output = emalloc(output_len);
+		output[0] = '\'';
+		memcpy(output + 1, input, input_len);
+		output[output_len - 1] = '\'';
+	} else if (memchr(input, '"', input_len) == NULL) {
+		output_len = input_len + 2;
+		output = emalloc(output_len);
+		output[0] = '"';
+		memcpy(output + 1, input, input_len);
+		output[output_len - 1] = '"';
+	} else {
+		// need to do the concat() trick published by Robert Rossney at https://stackoverflow.com/a/1352556/1067003
+		// first lets calculate the length (probably faster than repeated reallocs)
+		output_len = strlen("concat(");
+		for (size_t i = 0; i < input_len; ++i) {
+			uintptr_t bytesUntilSingleQuote =
+					(uintptr_t)memchr(input + i, '\'', input_len - i);
+			if (bytesUntilSingleQuote == 0) {
+				bytesUntilSingleQuote = input_len - i;
+			} else {
+				bytesUntilSingleQuote = bytesUntilSingleQuote - (uintptr_t)(input + i);
+			}
+			uintptr_t bytesUntilDoubleQuote =
+					(uintptr_t)memchr(input + i, '"', input_len - i);
+			if (bytesUntilDoubleQuote == 0) {
+				bytesUntilDoubleQuote = input_len - i;
+			} else {
+				bytesUntilDoubleQuote = bytesUntilDoubleQuote - (uintptr_t)(input + i);
+			}
+			const size_t bytesUntilQuote =
+					(bytesUntilSingleQuote > bytesUntilDoubleQuote)
+							? bytesUntilSingleQuote
+							: bytesUntilDoubleQuote;
+			i += bytesUntilQuote - 1;
+			output_len += 1 + bytesUntilQuote + 1 + 1; // "bytesUntilQuote"[,)]
+		}
+		output = emalloc(output_len);
+		size_t outputPos = strlen("concat(");
+		memcpy(output, "concat(", outputPos);
+		for (size_t i = 0; i < input_len; ++i) {
+			uintptr_t bytesUntilSingleQuote =
+					(uintptr_t)memchr(input + i, '\'', input_len - i);
+			if (bytesUntilSingleQuote == 0) {
+				bytesUntilSingleQuote = input_len - i;
+			} else {
+				bytesUntilSingleQuote = bytesUntilSingleQuote - (uintptr_t)(input + i);
+			}
+			uintptr_t bytesUntilDoubleQuote =
+					(uintptr_t)memchr(input + i, '"', input_len - i);
+			if (bytesUntilDoubleQuote == 0) {
+				bytesUntilDoubleQuote = input_len - i;
+			} else {
+				bytesUntilDoubleQuote = bytesUntilDoubleQuote - (uintptr_t)(input + i);
+			}
+			const size_t bytesUntilQuote =
+					(bytesUntilSingleQuote > bytesUntilDoubleQuote)
+							? bytesUntilSingleQuote
+							: bytesUntilDoubleQuote;
+			const char quoteMethod =
+					(bytesUntilSingleQuote > bytesUntilDoubleQuote) ? '\'' : '"';
+			output[outputPos++] = quoteMethod;
+			memcpy(output + outputPos, input + i, bytesUntilQuote);
+			outputPos += bytesUntilQuote;
+			output[outputPos++] = quoteMethod;
+			i += bytesUntilQuote - 1;
+			output[outputPos++] = ',';
+		}
+		output[outputPos - 1] = ')';
+	}
+	RETVAL_STRINGL(output, output_len);
+	efree(output); // todo: directly return a string and avoid the copy, probably possible but idk how (reviewers may want to check this)
+}
+/* }}} */
+
 #endif /* LIBXML_XPATH_ENABLED */
 
 #endif