Felix Sasaki scripsit: > I am wondering whether this concerns only the infoset. I could imagine to > define a data type with RELAX NG, which is "case normalized". RELAX NG > uses the XML Schema datatypes, but does not create an post-validation > infoset. This made me realize what the real problem is with XML Schema as currently designed. Imagine a schema for papers to be published by a certain agency or conference that has multiple working languages. It might be reasonable for the root element of such documents to have a required xml:lang attribute whose value represents, say, English, French, or Russian. In XML Schema or RELAX NG, one would therefore specify this attribute as of type "language" and having possible values "en", "fr", and "ru". Unfortunately, a document with an xml:lang value of "EN" would not pass validation, even though BCP 47 defines "en" and "EN" as synonymous. It gets even worse with longer language tags. This is unfortunate and should be fixed. -- John Cowan cowan at ccil.org www.reutershealth.com www.ccil.org/~cowan If a traveler were informed that such a man [as Lord John Russell] was leader of the House of Commons, he may well begin to comprehend how the Egyptians worshiped an insect. --Benjamin Disraeli _______________________________________________ Ltru mailing list Ltru at ietf.org https://www1.ietf.org/mailman/listinfo/ltru
Note Well: Messages sent to this mailing list are the opinions of the senders and do not imply endorsement by the IETF.