- From: Mark Seaborne <m_seaborne@mac.com>
- Date: Wed, 31 May 2006 11:56:49 +0100
- To: www-forms <www-forms@w3.org>
Has anyone actually implemented this? I tried it a while ago and didn't have a good experience. Consequently I resort to using generate-id(), which although not part of XForms is actually available in a good range of XForms processors. All the best Mark On 31 May 2006, at 11:37, Allan Beaufour wrote: > > The specification of the instance element includes this sentence: > "In this process, nodes of type xsd:ID are modified to remain as > unique values in the instance data." > [http://www.w3.org/TR/2006/REC-xforms-20060314/slice9.html#action- > insert] > > To that, I have a few questions. > > 1) When exactly is this modification taking place? > > It matters since nodes might have types associated through <xf:bind > type=""/>, so these needs to be processed before the nodes actually > are of type xsd:ID (I guess this takes place during xforms-rebuild.) > > But should it take place before or after xforms-recalculate? Before, > since the form author might want to do calculations based on the > (possibly new) value? Or after, since the form author might want to > calculate a new id himself? > > If it takes place as "part of" the RRR process, then it needs to be > deferred too, which might be tricky -- at least from an implementation > perspective. > > 2) Does it adhere to the readonly MIP, ie. does not change readonly > nodes? > > The form author might, for some reason, not want the nodes to be > changed, and could thus set readonly="true()" for those nodes to avoid > that. > > 3) Is it on purpose that it is stated as strict as "remain as unique > values in the instance data"? > > Should it have been more like the generate-id() function from XSLT, > which includes: > "There is no guarantee that a generated unique identifier will be > distinct from any unique IDs specified in the source document." > [http://www.w3.org/TR/xslt#misc-func] > > 4) Does the modification follow the normal XForms rules of how to > set values? > > F.x. for an element, should it only set the first text node child? > > -- > ... Allan >
Received on Wednesday, 31 May 2006 10:57:20 UTC