[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: Fw: [wsbpel] Issue - 157 - conf call brief recap ... and carryforward ...
So then, it is the case that part of the confusion and why we are discussing a table for 157 is that it is ambiguous in the 'to' what exactly is the location the designer wishes to copy to: In my example I want to copy 'into' the child of <a/> but <to/> only lets me point at all of <a/> and copy semantics are usually REPLACE not ADD CHILD. Alex Yiu wrote: > > Hi Rania, > (and Chris) > > To avoid confusion, I guess it would be better to answer Rania's > question with a full example. > > *[Clarification-#1]* > Say you have an element-based variable: "v1" of "foo:barElem" which > has the following content: > <foo:bar> <a/> </foo:bar> > > I expect the following <assign> will create a selectionFailure fault. > ------------------------ > <assign> > <copy> <from> concat('12','3') </from> <to>$v1*/a/text()*</to> </copy> > </assign> > ------------------------ > > On the other hand, I expect the following <assign> will NOT create a > selectionFailure fault. > ------------------------ > <assign> > <copy> <from> concat('12','3')' </from> <to>$v1*/a*</to> </copy> > </assign> > ------------------------ > > Based on the "F2F-table", I expect the <copy> will replace the empty > content of <a> with the string content of "123". That was the > ReplaceContent entry. > > *[Clarification-#2]* > > Regarding to Chris' precaution warning on 3 text nodes cases: > I want to clarify it further with 2 examples. For XPath compliant > implementation, adjacement text nodes will be merged as one text node. > See: > http://www.w3.org/TR/xpath#section-Text-Nodes > > E.g.: > <foo:bar> <a>123<![CDATA[456]]>789</a> </foo:bar> > > I expect "$v1/a/text()" will return one text node ("123456789") for > the above data. > > <foo:bar> <a>123<dummy1 /><![CDATA[456]]><dummy2 />789</a> </foo:bar> > > On the other hand, I expect "$v1/a/text()" will return 3 text nodes, > because they are not separated by <dummy1/> and <dummy2/>. > > > Thanks! > > > > Regards, > Alex Yiu > > > > Chris Keller wrote: > >>Hi Rania, >> >>It would result in a selectionFailure as there are no text nodes beneath >>element a. Also in other cases you may end up with more than one text node >>which could cause a selectionFaiilure. For example using CDATA for part of >>an elements text to avoid escaping characters, the following case may return >>3 text nodes: >> >><a> >>... >><![CDATA[...bunch of characters I don't want to escape...]]> >>... >></a> >> >>- Chris >> >>-----Original Message----- >>From: Rania Khalaf [mailto:rkhalaf@watson.ibm.com] >>Sent: Monday, June 20, 2005 9:59 AM >>To: Alex Yiu >>Cc: rkhalaf@us.ibm.com; Charlton Barreto; wsbpeltc; >>jwillis@exchange.microsoft.com >>Subject: Re: Fw: [wsbpel] Issue - 157 - conf call brief recap ... and carry >>forward ... >> >>Hi Alex, >> >>In the discussion on the call, someone mentioned that 'text' can be used >>to get to the value of an attribute or the CDATA children of an element >>(the 123 of <a>123</a>). >> >>What happens if I used it on an element <a/> ? If I am trying to copy >>'123' into '<a/>' to get <a>123</a>, does this cause selectionFailure ? >> >>thanks, >>Rania >> >> >> >>--------------------------------------------------------------------- >>To unsubscribe from this mail list, you must leave the OASIS TC that >>generates this mail. You may a link to this group and all your TCs in OASIS >>at: >>https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php >> >> >> >> >> >> >> >
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]