Comparison of EncodeForURL, URLEncodedFormat and EncodeRFC3986 (UDF)

EncodeRFC3986.cfm

<cfscript>
	/**
	* URI encoding per RFC 3986, which has treats the following as unreserved: ALPHA / DIGIT / "-" / "." / "_" / "~"
	* @text Text to encode
	* @returns URI encoded text
	*/
	public function encodeRFC3986(required string text) {
		// Requires CF10+
		Local.encoded = encodeForURL(arguments.text);
		// Reverse encoding of tilde "~"
		Local.encoded = replace( Local.encoded, "%7E", "~", "all" );
		// Change space encoding from "+" to "%20"
		Local.encoded = replace( Local.encoded, "+", "%20", "all" );
		// Asterisk "*" should be encoded
		Local.encoded = replace( Local.encoded, "*", "%2A", "all" );
		
		return Local.encoded;
	}
	
	/**
	* URI encoding per RFC 3986, which has treats the following as unreserved: ALPHA / DIGIT / "-" / "." / "_" / "~"
	* @text Text to encode
	* @encoding UTF-8 or ISO-8859-1
	* @returns URI encoded text
	*/
	public function encodeRFC3986JAVA(required string text, required string encoding) {
		Local.oEncoder = createObject("java", "java.net.URLEncoder");
		<!---
		Local.oCharset = createObject("java","java.nio.charset.StandardCharsets");	
		oCharset.UTF_8.toString()
		oCharset.ISO_8859_1.toString()
		--->		

		// Requires CF10+
		Local.encoded = oEncoder.encode( arguments.text, encoding );
		// Reverse encoding of tilde "~"
		Local.encoded = replace( Local.encoded, "%7E", "~", "all" );
		// Change space encoding from "+" to "%20"
		Local.encoded = replace( Local.encoded, "+", "%20", "all" );
		// Asterisk "*" should be encoded
		Local.encoded = replace( Local.encoded, "*", "%2A", "all" );
		
		return Local.encoded;
	}
	
	// Non-reserved characters
	writeOutput("<hr><strong>Non-reserved characters (excluding letters and numbers):</strong><br>");
	variables.nonReservedChars = ["-","_",".","~"," "];
	for (variables.original in variables.nonReservedChars) {
		writeOutput("<br> Original #variables.original#  || encodeForURL = [ #encodeForURL( variables.original )# ]");
		writeOutput(" urlEncodedFormat = [ #urlEncodedFormat( variables.original )# ]");
		writeOutput(" encodeRFC3986 = [ #encodeRFC3986( variables.original )# ]");
		writeOutput(" encodeRFC3986JAVA = [ #encodeRFC3986JAVA( variables.original, "UTF-8" )# ]");
		writeOutput(" encodeRFC3986JAVA = [ #encodeRFC3986JAVA( variables.original, "ISO-8859-1" )# ]");
	}
	

	// Reserved characters
	writeOutput("<hr><strong>Reserved characters:</strong><br>");
	variables.reservedChars = listToArray(": / ? ## [ ] @ ! $ & ' ( ) * + , ; = ", " ");
	for (variables.original in variables.reservedChars) {
		writeOutput("<br> Original #variables.original#  || encodeForURL = [ #encodeForURL( variables.original )# ]");
		writeOutput(" urlEncodedFormat = [ #urlEncodedFormat( variables.original )# ]");
		writeOutput(" encodeRFC3986 = [ #encodeRFC3986( variables.original )# ]");
		writeOutput(" encodeRFC3986JAVA = [ #encodeRFC3986JAVA( variables.original, "UTF-8" )# ]");
		writeOutput(" encodeRFC3986JAVA = [ #encodeRFC3986JAVA( variables.original, "ISO-8859-1" )# ]");
	}
	
	// Unicode
	variables.unicodeText = ["An preost wes on leoden, Laȝamon was ihoten"
					, "He wes Leovenaðes sone -- liðe him be Drihten."
					, "He wonede at Ernleȝe at æðelen are chirechen,"
					, "Uppen Sevarne staþe, sel þar him þuhte,"
					, "Onfest Radestone, þer he bock radde."
				];
				
	for (variables.line in variables.unicodeText) {
		writeOutput("<hr> <strong>#variables.line#</strong>");
		writeOutput("<br><br> 1. [ #encodeForURL( variables.line )# ]");
		writeOutput("<br> 2. [ #urlEncodedFormat( variables.line )# ]");
		writeOutput("<br> 3. [ #encodeRFC3986( variables.line )# ]");
		writeOutput("<br> 4. [ #encodeRFC3986JAVA( variables.line, "UTF-8" )# ]");
		writeOutput("<br> 4. [ #encodeRFC3986JAVA( variables.line, "ISO-8859-1" )# ]");
	}
</cfscript>