Perl: Passing Unicode string to chr()

I'm pretty new to Perl so pardon me for what will probably turn into a most obvious answer.

I'm trying to pass a Unicode markup to the chr() function. Here's a redacted example of my script.

#!/usr/bin/perl
$unicode = "/u0026amp;";
print chr("0x".substr($unicode, 4, 2))."\n";

This properly extracts the 26 from the $unicode variable. However the problem as I can tell is that the chr() function doesn't like quoted strings, but if I remove the quotes the x in 0x gets removed and becomes an invalid 026 rather than a valid 0x26.

Anyways, 开发者_开发百科this really gets down to. How can I keep the x in

chr("0x".substr($unicode, 4, 2))

from disappearing and send the proper 0x26

You do not use "0x" strings with chr, the way you can with ord and hex. So you want:

 chr(hex($hexstring))