RE: [dita] Nested index terms

["Grosso, Paul" <pgrosso@ptc.com>]

I believe there are several examples in this email thread.

In my mind, it's anytime the content of an indexterm 
contains what the DOM would call a text chunk (other
than all whitespace) other than as the first child of 
the indexterm.  So in

<indexterm>Top level
  <indexterm>Nested 1</indexterm>
  <indexterm>Nested 2</indexterm>
  with additional text
</indexterm>

the additional text is "with additional text" because
it comes after a couple indexterm children of the
overall indexterm.

paul

> -----Original Message-----
> From: JoAnn Hackos [mailto:joann.hackos@comtech-serv.com] 
> Sent: Friday, 2005 September 30 14:36
> To: Paul Prescod; Robert D Anderson; Grosso, Paul
> Cc: dita@lists.oasis-open.org
> Subject: RE: [dita] Nested index terms
> 
> I would like to see an example of what you mean by "with additional
> text" for an index item.
> JoAnn 
> 
> -----Original Message-----
> From: Paul Prescod [mailto:paul.prescod@blastradius.com] 
> Sent: Friday, September 30, 2005 12:59 PM
> To: Robert D Anderson; Grosso, Paul
> Cc: dita@lists.oasis-open.org
> Subject: RE: [dita] Nested index terms
> 
> We should agree on the appropriate behaviour for the "with additional
> text" and put something in the DITA spec. Either it means something
> useful or it is buggy extra stuff that should be warned about.
> 
> Thanks for raising this Paul G. 
> 
> It is very analogous to my old favorite:
> 
> <section>
> This is <title>a section</title>. With a few of
> <title>titles</title><p>And a paragraph</p> and another
> <title>title</title></section>
> 
> What does this mean? What is the section title? What should be output?
> Should XMetaL warn that this is meaningless?
> 
> -----Original Message-----
> From: Robert D Anderson [mailto:robander@us.ibm.com] 
> Sent: Friday, September 30, 2005 10:26 AM
> To: Grosso, Paul
> Cc: dita@lists.oasis-open.org
> Subject: RE: [dita] Nested index terms
> 
> In the IBM internal implementation, we convert this markup:
> <indexterm>Top level
>   <indexterm>Nested 1</indexterm>
>   <indexterm>Nested 2</indexterm>
>   with additional text
> </indexterm>
> 
> to this in the index:
> Top level with additional text
>    Nested 1
>    Nested 2
> 
> So, the sample above is equivalent to this, and many other possible
> combinations:
> <indexterm>Top level with additional text
>   <indexterm>Nested 1</indexterm>
> </indexterm>
> <indexterm><indexterm>Nested 2</indexterm>
> Top level with additional text
> </indexterm>
> 
> The DITA open toolkit has this same behavior as its goal, but is not
> there
> yet - the indexing code could not be completely reused, due to
> differences
> between the toolkit architecture and our internal toolset's
> architecture.
> 
> I understand that with the Blast Radius implementation, the "with
> additional text" part of my original sample would be dropped, with a
> warning. That seems equally acceptable to me, as long as there is a
> warning
> - this does not come up often with our users, though somebody did
> request
> that we support it way-back-when.
> 
> Robert D Anderson
> IBM Authoring Tools Development
> Chief Architect, DITA Open Toolkit
> 
> 
>  
> 
>              "Grosso, Paul"
> 
>              <pgrosso@ptc.com>
> 
>  
> To 
>              09/30/2005 11:49          <dita@lists.oasis-open.org>
> 
>              AM
> cc 
>  
> 
>  
> Subject 
>                                        RE: [dita] Nested index terms
> 
>  
> 
>  
> 
>  
> 
>  
> 
>  
> 
>  
> 
> 
> 
> 
> 
> Thanks for answering, Paul.
> 
> I'd be interested in hearing what others think too,
> both in terms of what users would do/expect and in
> terms of what other current implementors do.
> 
> More comments below.
> 
> > -----Original Message-----
> > From: Paul Prescod [mailto:paul.prescod@blastradius.com]
> > Sent: Friday, 2005 September 30 10:43
> > To: Grosso, Paul; dita@lists.oasis-open.org
> > Subject: [dita] Nested index terms
> >
> > I don't think that Paul got an answer to his question. Your
> > code in the
> > second example below should generate:
> >
> > Top level
> >            Nested 1
> >            Nested 2
> >
> > That's how Blast Radius implemented it anyhow.
> 
> So, am I correct to understand that you're saying that:
> 
> <indexterm>Top level
>   <indexterm>Nested 1</indexterm>
>   <indexterm>Nested 2</indexterm>
> </indexterm>
> 
> is effectively equivalent to:
> 
> <indexterm>Top level
>   <indexterm>Nested 1</indexterm>
> </indexterm>
> <indexterm>Top level
>   <indexterm>Nested 2</indexterm>
> </indexterm>
> 
> > The second example should probably be treated as an error. It
> 
> Do you mean "first example" here?
> 
> paul
> 
> > is a flaw
> > of XML schemas and DTDs (inherited from SGML!) that does not make it
> > possible for us to specify that explicitly in the DTD/schema.
> >
> > -----Original Message-----
> > From: Grosso, Paul [mailto:pgrosso@ptc.com]
> > Sent: Wednesday, September 28, 2005 11:08 AM
> > To: dita@lists.oasis-open.org
> > Subject: RE: [dita] Groups - DITA 1.1 Issue #45: Add See, See Also
> > indexing elements (IssueNumber45.html) uploaded
> >
> >
> > There is something about indexterm (irrespective of
> > this current proposal) that has always concerned me:
> > its mixed content model.  Is something like:
> >
> > <indexterm>Top level
> >   <indexterm>Nested</indexterm>
> >   index term content.
> > </indexterm>
> >
> > allowed (the DTD allows it)?  If so, what are the
> > processing expectations?
> >
> > Also, what are the processing expectations of
> >
> > <indexterm>Top level
> >   <indexterm>Nested 1</indexterm>
> >   <indexterm>Nested 2</indexterm>
> > </indexterm>
> >
> > (the DTD allows this too)?
> >
> >
> 
> 
> 
>