Document attribute charset and its aliases to return UTF-8

LayoutTests/fast/dom/document-attribute-js-null.html

Index: LayoutTests/fast/dom/document-attribute-js-null.html
diff --git a/LayoutTests/fast/dom/document-attribute-js-null.html b/LayoutTests/fast/dom/document-attribute-js-null.html
index 7d99616ad1d45ca7051ef209798c45ae327ecb75..de379ef30fa730f5dd4c1f48842afaca2692a053 100644
--- a/LayoutTests/fast/dom/document-attribute-js-null.html
+++ b/LayoutTests/fast/dom/document-attribute-js-null.html
@@ -66,7 +66,10 @@
                     attributes: [
                         {name: 'xmlVersion', expectedExceptionCode: 9},
                         {name: 'documentURI', expectedNull: null},
-                        {name: 'charset', expectedNull: undefined},
+                        {name: 'charset', expectedNull: 'UTF-8'},

[dominicc (has gone to gerrit), 2015/06/22 01:23:31]

I think the case of this is wrong per
https://encoding.spec.whatwg.org/#names-and-labels

[Habib Virji, 2015/06/22 14:07:30]

Yes it does not match it as utf-8, but these test was to show it does not return
nullvalue. I can work on following CL to make UTF8Encoding() handle possible
different values as per the specification.

[philipj_slow, 2015/06/23 08:32:56]

Actually https://dom.spec.whatwg.org/#dom-document-characterset special-cases a
bunch of encoding to use uppercase labels instead. I'm pretty sure Anne would be
happy to always return a lowercase value if there's implementor interest to make
that change. Not in this CL, though.

+                        {name: 'defaultCharset', expectedNull: undefined},
+                        {name: 'characterSet', expectedNull: 'UTF-8'},
+                        {name: 'inputEncoding', expectedNull: 'UTF-8'},
                         {name: 'selectedStylesheetSet', expectedNull: null}
                     ]
                 },