Added: setting for unescaping special chars (#46)

[signebedi]

Based on discussion in #46. I added a default setting called "unescape_special_chars".

DEFAULT_SETTINGS = {
    ...
    "unescape_special_chars": False, # see discussion https://github.com/matthiask/html-sanitizer/issues/46
}

When true, special characters are added back to the HTML immediately before returning:

        # see discussion https://github.com/matthiask/html-sanitizer/issues/46
        if self.unescape_special_chars:
            html = ( html
                    .replace("&", "&")
                    .replace("&lt;", "<")
                    .replace("&gt;", ">")
            )

I welcome feedback. This is a minimalist change intended only to give end users the ability to retain some of the special characters that are stripped out during the sanitation process. If I have broken assumptions, or if I have missed other special characters, please feel free to make note of these and I am happy to make further changes before the code is merged.