[osis-core] startMilestone/endMilestone

[Patrick Durusau]

Troy,

Troy A. Griffitts wrote:

>> 4. Note that the startMileston/endMilestone with a type attribute 
>> should solve the quotation problem that Troy has raised.
>
>
> So we are suggesting using something like:
>
> <startMilestone type="q" ID="1234" />
> this is a quote
> <endMilestone IDREF="1234" />
>
> ? 

Yes/

>
>
> Not sure if I like this better than:
>
> <q mStart="1234" />
> this is a quote
> <q mEnd="1234" />
>
Hmmm, Steve, if my oft cited <q/> equivalency to <q></q> is correct, is 
Troy also correct that if I have optional content for the <q> model, 
that use of optional milestone attributes could work this way? (Does get 
me back to 39 main elements so I am not an impartial judge on this one.)

Not wild about the implication of linking via data values in mStart and 
mEnd. I suppose we could declare a processing rule that all mStart 
values must have matching mEnd values. Easy enough to script and could 
enforce order in document stream.

>
> At least with the latter, when we're processing the quote tag <q> 
> we'll get both the container and the milestone usages.
>
>
> The downside of using the former is that when looking for any tag, 
> I'll need to search for <tag> AND also for <startMilestone type="tag">
>
> The downside of using the latter is that when processing any tag with 
> attributeGroup 'milestone' (which would contain attributes 'mStart' 
> and 'mEnd'), I have to check for attribute 'mStart' to see if I'm a 
> container or a milestone type of tag.

Not sure that is much more than an if-else statement. If mStart do, else 
it is regular element. Not sure I see much overhead there.

Probably need to get Steve to evaluate for any hidden gotchas.

Patrick


>
> Still better like the way the latter looks.
>
>     -Troy.
>

-- 
Patrick Durusau
Director of Research and Development
Society of Biblical Literature
pdurusau@emory.edu