|
Not quite sure I follow you. Are you suggesting that the Serialization format in fact could be identical to the XQuery syntax that would produce that format ? Much like say the JSON format is equivalent to the JavaScript code that produces that data ? That could work ... but there's some issues offhand. 1) You have to have an XQuery parser or atleast part of one --> One of the use cases is to be able to parse this data into NON-XDM Aware programs 2) typed data may be more difficult ( is "1" an xs:integer or xs:decimal ? If the actual XDM data had this type we want to serialize it) --> One of the use cases is to have access to the type of all items 3) May not be canonical. That is there may be more then one way to represent the same sequence --> One of the use cases is a canonical format. although that is only needed once you reconstitute the sequence. All these could possibly be solved with some "rules" about what subset of XQuery is used. e.g. But it may take a LOT of rules. David A. Lee dlee@calldei.com http://www.calldei.com http://www.xmlsh.org 812-482-5224 Kurt Cagle wrote: 6fa681b10909191951v221b11cerb0b8e9e9ae444383@mail.gmail.com" type="cite">I'm going to ask what may be an obvious question, but wouldn't it make sense for a serialization of a sequence to correspond on the output to the serialization on the input? That is to say, if you had a structure: |