Norman Walsh wrote: > / Innovimax SARL <innovimax@gmail.com> was heard to say: > | The way it is written in the spec > | [[ Each p:when branch of the p:choose has a test attribute which > | **must** contain an [XPath 1.0] expression. ]] > | > | It seems it is a static error if the p:when do not contains an XPath > | If it is the case, please add a static error for reporting that > > Beyond a more general static error (not yet defined, but clearly > needing to be) for "required attribute not present" on an element? > > You don't want to check the validity of the XPath expression, do you? > Or do you? It should be an error if XProc attributes that expect XPath expressions or XSLT patterns have incorrect syntax. Personally, I think it should be a static error since it can be detected statically, but if it's made a dynamic error (presumably because that makes implementation easier?) then processors should have the option of reporting it statically. For expressions/patterns in option values, the error has to be dynamic, but processors should again be able to report them statically. Cheers, Jeni -- Jeni Tennison http://www.jenitennison.comReceived on Tuesday, 28 August 2007 18:34:22 GMT
This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:21:54 GMT