Subject: Re: Discussion Topics Re-Revisited
On Wed, 24 May 2000, Petit, John wrote: 2) Unique IDs. There was also a great deal of talk today about what >unique IDs would be used for and how they would be applied. It seems that >the consensus was that unique IDs would be applied at the schema level, >not the document level. This means that two XML documents which are from >the same schema would have unique IDs for a specific tag, but not each >content instance. As shown below, the IDs are unique for the tag <FRUIT> >but the same in each document instance. (Be gentle now, these documents >are only here to illustrate a point, not be full ebXML doc examples!) > > Document 1 ><FOOD id=559849JUE98> > <FRUIT id=39408JEX98> apple</FRUIT> ></FOOD> > > Document 2 > ><FOOD id=559849JUE98> > <FRUIT id=39408JEX98> orange</FRUIT> ></FOOD> > I don't think I understand the rationale for a unique ID for each element type and not the content. By virtue of the unique name each element is unique. Is the rationale to eliminate the need for XML Namespaces, i.e., <ECC:FOOD xmlns="xmlns:ECC="http://www.eccnet.com/namespaces/ecc-food.xsd"> If it is, I understand the rationale for this approach. However, I would argue against using an XML ID for this methodology. Also, it wouldn't pass a parse for a unique ID. If I were looking for a <FRUIT> and I don't know what the fruit is but I have use the Unique ID in XPATH, I would get the first food - which may be spinach. I recommend that you call the ID above something more meaningful to eliminate confusion. Just a suggestion. Betty /\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/ Betty Harvey | Phone: 410-787-9200 Electronic Commerce Connection, Inc. | FAX: 410-787-9830 harvey@eccnet.com | Washington,DC SGML/XML Users Grp URL: http://www.eccnet.com | http://www.eccnet.com/xmlug/ /\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\\/\/
Powered by
eList eXpress LLC