Doctype parser error when casing is wrong

<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
10: from /home/web/.gem/gems/oga-2.16/lib/oga/oga.rb:23:in `parse_html'
 9: from /home/web/.gem/gems/oga-2.16/lib/oga/oga.rb:23:in `parse'
 8: from /home/web/.gem/gems/oga-2.16/lib/oga/xml/parser.rb:223:in `each_token'
 7: from /home/web/.gem/gems/oga-2.16/lib/oga/xml/lexer.rb:175:in `advance'
 6: from /home/web/.gem/gems/oga-2.16/lib/oga/xml/lexer.rb:130:in `read_data'
 5: from /home/web/.gem/gems/oga-2.16/lib/oga/xml/lexer.rb:176:in `block in advance'
 4: from /home/web/.gem/gems/oga-2.16/lib/oga/xml/lexer.rb:176:in `advance_native'
 3: from /home/web/.gem/gems/oga-2.16/lib/oga/xml/lexer.rb:266:in `on_doctype_name'
 2: from /home/web/.gem/gems/oga-2.16/lib/oga/xml/lexer.rb:223:in `add_token'
 1: from /home/web/.gem/gems/oga-2.16/lib/oga/xml/parser.rb:226:in `block in each_token'
/home/web/.gem/gems/oga-2.16/lib/oga/xml/parser.rb:251:in `parser_error': Unexpected doctype name, expected doctype closing tag instead on line 1 (LL::ParserError)

It seems to be throwing the error when the word "public" is not capitalized in a doctype. Not sure if this is a WONTFIX or not since PUBLIC is probably correct, but it does seem a bit unforgiving. I'm hotfixing it for now with a REGEX to replace with the right casing.

For the record I'm using Oga to parse out millions of html files from the geocities archive. It is less brutal about re-writing the output document than nokogiri is. The fairly weird site I got this exception with is: https://geocities.gallery/Heartland/7792/