It would be nonsensical for @type on <href> to be overridden by @type on <keydef> for two reasons: 1) They are two different domains of addressing. Key definitions are strictly limited to maps and topics, xref is to any element that takes @id, so the two @type domains are necessarily distinct. 2) The referencing element is saying what it expects the the ultimate target to be. That can only be known by the referencing element--it wouldn't be knowable or sensible or, in the case of non-topic or non-map elements, possible to specify on <keydef>. But I understand the precedence rules to be that the initial reference governs, just as for conref. Cheers, E. On 9/19/11 8:50 AM, "Nitchie, Chris" <
cnitchie@ptc.com> wrote: > We¹ve recently had a customer issue crop up related to handling of the @type > attribute when resolving key references. I¹m pretty sure we need to change our > handling here, but I also think we implemented things according to the spirit > of the spec, so I wanted to run this by the TC and get your thoughts. > > The situation is that when you have an xref with a keyref that specifies a > sub-topic element ID, should the effective @type attribute come from the key > reference or the key definition? > > <keydef keys=²footnotes² href=²footnotes.dita² type=²concept²/> > > <xref keyref=²footnotes/fn1² type=²fn²/> > > We¹ve always taken the position that the key referencing element¹s attributes > take precedence over the key definition¹s except for those related to the > reference, specifically @href, @scope, @format, and @type. It¹s pretty clear > that the effective @scope and @format should always come from the key > definition, but for @type it¹s not quite so obvious. > > I couldn¹t find any specific language in the spec for anything besides @href, > so I think this is something that we¹ll need to clarify somehow. I think we¹re > going to modify our processing to have the key reference¹s @type take > precedence when a sub-topic ID is specified on the keyref. In the meantime I¹d > appreciate any thoughts. > > Chris > > <
http://www.ptc.com/ > > > Chris Nitchie > Senior Software Engineer > > T 734.352.2879 F 734.997.0201 > E
cnitchie@ptc.com < mailto:
cnitchie@ptc.com > > > PTC.com <
http://www.ptc.com/ > > > -- Eliot Kimber Senior Solutions Architect "Bringing Strategy, Content, and Technology Together" Main: 512.554.9368
www.reallysi.com www.rsuitecms.com