Companion

The reasons.

Properties

Link copied to clipboard

The start of the document decodes two ways, so a DOCTYPE might be hidden: an unpaired surrogate in UTF-16, or a non-ASCII byte before the root element under EUC-JP or GB18030, whose decoders can drop the character that follows.

Link copied to clipboard

The document carries a DOCTYPE, or another markup declaration (<!ENTITY, <!ELEMENT, …), before the root element. A web page's <!DOCTYPE html> is NotAFeed(Html) instead.

Link copied to clipboard

What comes before the root element (byte-order marks, white space, comments, processing instructions) does not end within the first 64 KiB.

Link copied to clipboard

An encoding the parser will not read safely: UCS-4, EBCDIC, or a charset named in the Content-Type header or the XML declaration that is outside the set the parser reads. RomeFeedParser reads UTF-8, UTF-16, US-ASCII, ISO-8859-n, windows-125n, KOI8-R, KOI8-U, Shift_JIS, EUC-KR, GBK, GB2312, Big5, EUC-JP and GB18030, plus only these aliases: utf8, ascii, latin1, sjis, windows-31j, iso8859-n, iso_8859-n and cp125n.