encoding.html: add escape() function (#16666)

vlib/encoding/html/escape.v

@@ -0,0 +1,20 @@
+module html
+
+[params]
+pub struct EscapeConfig {
+	quote bool = true
+}
+
+// escape converts special characters in the input, specifically "<", ">", and "&"
+// to HTML-safe sequences. If `quote` is set to true (which is default), quotes in
+// HTML will also be translated. Both double and single quotes will be affected.
+// **Note:** escape() supports funky accents by doing nothing about them. V's UTF-8
+// support through `string` is robust enough to deal with these cases.
+pub fn escape(input string, config EscapeConfig) string {
+	tag_free_input := input.replace_each(['&', '&amp;', '<', '&lt;', '>', '&gt;'])
+	return if config.quote {
+		tag_free_input.replace_each(['"', '&quot;', "'", '&#x27;'])
+	} else {
+		tag_free_input
+	}
+}

vlib/encoding/html/escape_test.v

@@ -0,0 +1,22 @@
+import encoding.html
+
+fn test_escape_html() {
+	assert html.escape('<>&') == '&lt;&gt;&amp;'
+	assert html.escape('No change') == 'No change'
+	assert html.escape('<b>Bold text</b>') == '&lt;b&gt;Bold text&lt;/b&gt;'
+	assert html.escape('<img />') == '&lt;img /&gt;'
+	assert html.escape("' onmouseover='alert(1)'") == '&#x27; onmouseover=&#x27;alert(1)&#x27;'
+	assert html.escape("<a href='http://www.example.com'>link</a>") == '&lt;a href=&#x27;http://www.example.com&#x27;&gt;link&lt;/a&gt;'
+	assert html.escape("<script>alert('hello');</script>") == '&lt;script&gt;alert(&#x27;hello&#x27;);&lt;/script&gt;'
+	// Cases obtained from:
+	// https://github.com/apache/commons-lang/blob/master/src/test/java/org/apache/commons/lang3/StringEscapeUtilsTest.java
+	assert html.escape('plain text') == 'plain text'
+	assert html.escape('') == ''
+	assert html.escape('bread & butter') == 'bread &amp; butter'
+	assert html.escape('"bread" & butter') == '&quot;bread&quot; &amp; butter'
+	assert html.escape('greater than >') == 'greater than &gt;'
+	assert html.escape('< less than') == '&lt; less than'
+	// Leave accents as-is
+	assert html.escape('café') == 'café'
+	assert html.escape('<p>façade</p>') == '&lt;p&gt;façade&lt;/p&gt;'
+}

vlib/vweb/vweb.v

@@ -10,6 +10,7 @@ import net.http
 import net.urllib
 import time
 import json
+import encoding.html
 
 // A type which don't get filtered inside templates
 pub type RawHtml = string
@@ -725,12 +726,5 @@ fn send_string(mut conn net.TcpConn, s string) ! {
 // It used by `vlib/v/gen/c/str_intp.v:130` for string interpolation inside vweb templates
 // TODO: move it to template render
 fn filter(s string) string {
-	return s.replace_each([
-		'<',
-		'&lt;',
-		'"',
-		'&quot;',
-		'&',
-		'&amp;',
-	])
+	return html.escape(s)
 }