dfdl:initiator="Message-ID Message-Id" i.e., it's a whitespace separated list of alternatives. There's also a property called dfdl:ignoreCase="yes", but that would allow "mEssAge-iD" as well, which you probably don't want. On Wed, Feb 27, 2013 at 12:48 PM, Garriss Jr., James P. <jgarriss@mitre.org>wrote:
Consider the following IMF headers:****
** **
Message-ID: < CACQ76nh2VDypuwm1aHc1-s8ZmmyXg8XEosS6af6yurkVdT+8QQ@mail.gmail.com>****
Message-Id: <208E8372-A9D9-468A-9B4B-BD97ED28F64C@garriss.org>****
** **
Notice the difference: ** **
** **
Message-ID: ****
vs ****
Message-Id:****
** **
Per the RFC, only the first is legal, but both are used in practice, and we’d like to allow either. In our DFDL schema, “Message-ID” is an initiator, but suddenly we have two alternate initiators for the same element.****
** **
How we model this in DFDL?****
** **
TIA****
-- dfdl-wg mailing list dfdl-wg@ogf.org https://www.ogf.org/mailman/listinfo/dfdl-wg
-- Mike Beckerle | OGF DFDL Workgroup Co-Chair | Tresys Technology | www.tresys.com