Subject:Compressed content in CADATA Author:Tony Lavinio Date:11 Nov 2009 07:57 AM
That's not valid XML.
If the file had an XML header that stated version 1.1, then you
could hold the characters with Unicode values < 32 (except null).
And relatively few parsers handle XML 1.1; that standard was a
But CDATA does not mean "can hold binary content". It's just a
wrapper that lets you avoid having to escape certain characters
such as < and &. The "C" means "character".
Clicking on the "Char" link gets us to a list of valid Unicode values,
 Char ::= #x9 | #xA | #xD | [#x20-#xD7FF] | [#xE000-#xFFFD] | [#x10000-#x10FFFF] /* any Unicode character, excluding the surrogate blocks, FFFE, and FFFF. */
The only valid values with a Unicode value >=0 and <=31 are TAB,
CR and LF.