Re: RFC proposed change for data bindings alpha 5 dist

[Yuzo Kanomata <yuzok@ics.uci.edu>]

Hi Susan,

Oops, I saw Enumeration once a few weeks ago in XML Schema definitions, but 
I missed it when I was going back through. I think forwarding the question 
would be fine, but it should be edited a bit as this was aimed in scope as 
a group discussion point. I am stumped as to what GXL expects.

I have been inferring my questions from some things I recall seeing in 
PCDATA from other sources and thinking about how people interpret PCDATA. 
Err, maybe  more of a thinking out loud exercise. So I am recalling old 
import data problems/headaches I've had. For example in a dtd where one 
sends a list of email addresses for export, there are may be defined in a 
dtd:
<!ELEMENT addresslist (#PCDATA)>

but different email clients do the following:
1. <addresslist>alice@uci.edu, bob@uci.edu, chuck@uci.edu</addresslist>

2. <addresslist>alice@uci.edu; bob@uci.edu; chuck@uci.edu</addresslist>

Each works fine IFF you are importing to and from clients of the same 
"type", but as to a seemingly easy export between different applications, 
well you'd have to do some hacking.

Having found 4 examples in the GXL Examples which had enum in them, and 
those being single values: <enum>from</enum>. I made the alpha-4-dist 
simple strings (the same punt that is in the xmlschema on the gupro site), 
but I felt a bit of discomfort looking at the def in the dtd which states 
values of enum.

It could be interpreted as each enum value exists as a separate enum 
element, but since you can have many attrs under a element you could have 
the enumeration list a, b, c listed as well:
<attr id="me"><enum>a
<attr id="next"><enum>b
<attr id="last"><enum>c
In which case you could easily and legally say
<attr id="a"><enum>a,b,c
these examples meaning I am "a" in a group of a, b, and c.

So I am wondering if any tools or set of tools follow a certain convention 
in using enum. Since GXL calls it enum with PCDATA then XML Schema 
"enumeration" and my NMTOKENS proposals can be seen as logical successors, 
but I just don't know ...

Sort of a rehash of the original post, but a question in my mind.

Yuzo

--On Thursday, August 19, 2004 4:41 PM -0700 Susan Elliott Sim 
<ses@ics.uci.edu> wrote:

> Dear Yuzo,
>
> I've looked over your proposal and I think you've made a reasonable
> choice from the available options that XML Schema provides.  Moreover,
> it replicates the structure of the <enumeration> tag in XML Schema.
>
> My question at this point is how does GXL expect <enum> data to
> be encoded?  I've looked through my files and I can't find a
> definitive answer.  (Rob: the example in the user manual is
> incorrect; my fault not yours.) There is some documentation on
> constraints in GXL, but it's in German and I don't have a copy
> of it.
>
> We have a couple options.
> a) I can forward your question to GXL mailing list.
> b) You can join the GXL mailing list and post the question yourself.
> http://mailhost.uni-koblenz.de/mailman/listinfo/gxl
>
> Susan