On Monday, April 22, 2002, at 09:18 AM, [EMAIL PROTECTED] wrote: > I can certainly see the benefit of exposing "actual values", especially > for > elements/attributes, so that applications don't need to parse the values > again, but it needs both time and a good design to support it. One of > the > issues would be how to represent actual values. For example, what would > be > an actual value for type "gYearMonth"? How about "duration"?
I actually like your current internal representation, it is space-efficient and it matches well the ISO 8601 thing. What I would do is to rationalize (add appropriate interfaces, etc.) the current DV classes (there are many useful functionalities inside there) and make shallow wrapper objects to expose to the user. From what I've seen in your code you implement very closely the W3C recommendation in your internal representation. It would seem only natural to me to expose this stuff to the user via appropriate interfaces. > We (some committers of Xerces project) are trying to provide more > information via PSVI. As the first step, we want to expose information > required by the schema spec (3.X.5 sections of the structure spec): > schema > components, element/attribute declarations, type definitions, etc. > > So IMO, exposing "actual values" won't be the highest priority in the > very > near future. But it would definitely be helpful if you want to share > some > thoughts on how it can be implemented, and we'll consider including it > if > it's easy enough. I have to build this stuff for my own usage anyway and I'd rather do it right. Can I help with the PSVI design/implementation? Maybe with more manpower you can give more priority to the PSVI :) TIA, ciao, - Fabio --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
