LibTextCodec: Don't allocate Strings on encoding normalisation

This ripples down to LibWeb's HTML and XHR decoders, which therefore
become less allocation heavy.

Userland/Libraries/LibTextCodec/Decoder.h

@@ -80,7 +80,7 @@ public:
 };
 
 Decoder* decoder_for(String const& encoding);
-Optional<String> get_standardized_encoding(const String& encoding);
+Optional<StringView> get_standardized_encoding(StringView encoding);
 
 // This returns the appropriate Unicode decoder for the sniffed BOM or nullptr if there is no appropriate decoder.
 Decoder* bom_sniff_to_decoder(StringView);