Conform encoding-label matching to Encoding spec

This change makes the parser’s encoding-name matching conform to the current
Encoding spec at https://encoding.spec.whatwg.org/#concept-encoding-get —
which requires that only leading and trailing whitespace be removed from
a string before checking if it matches any valid encoding names.

Otherwise, without this change, the parser instead implements
https://www.unicode.org/reports/tr22/tr22-8.html#Charset_Alias_Matching —
which requires deleting “all characters except a-z, A-Z, and 0-9” from
a string before checking if it matches any valid encoding names. That
difference makes us fail two html5-tests cases.

Author: sideshowbarker

src/nu/validator/htmlparser/io/Encoding.java

@@ -254,9 +254,7 @@ public static String toNameKey(String str) {
             if (c >= 'A' && c <= 'Z') {
                 c += 0x20;
             }
-            if (!((c >= '\t' && c <= '\r') || (c >= '\u0020' && c <= '\u002F')
-                    || (c >= '\u003A' && c <= '\u0040')
-                    || (c >= '\u005B' && c <= '\u0060') || (c >= '\u007B' && c <= '\u007E'))) {
+            if (!(c == ' ' || c == '\t' || c == '\n' || c == '\f' || c == '\r')) {
                 buf[j] = c;
                 j++;
             }