Marc- Just going through the archives looking for something else I stumbled across this message:
http://marc.theaimsgroup.com/?l=castor-user&m=112792379518919&w=2 I'm not sure if it will work in your specific case, but it might be helpful. Stephen On 3/21/06, Marc Respass <[EMAIL PROTECTED]> wrote: > On Mar 21, 2006, at 2:14 PM, Benoit Maisonny wrote: > > > I'm curious to know why you need to put the values in CDATA sections. > > So am I. I'm making an XML file that is sent to someone and that's > what they require. When I asked why, they ignored me. > > > The serializer escapes anyway any character that need to be escaped > > according to XML specs (if not, than the serializer is buggy). So, > > I don't see a need to use CDATA at all. > > I couldn't agree more. Besides that, the data is all simple short > values. > > > As far as I know, CDATA sections are only useful to humans, as we > > don't like to have to escape characters manually. But maybe I miss > > some weird use of CDATA. > > CDATA is sometimes useful and I've used it before but this is not a > good use of CDATA. Frankly, I suspect the guys who are receiving this > file are using String.indexOf() as their parser ;). > > However, one may desire enclosing a value in CDATA and it seems like > something useful to have. I've found a number of questions about how > to go about it and none of the answers are very easy. I was hoping > there was something that I missed in Castor but it looks like writing > my own serializer is the way. > > Marc > > > Benoit > > > > > > Marc Respass wrote: > > > >> Thanks Stephen, > >> > >> I know I have looked at an XML serializer already but maybe I > >> didn't look far enough. For the sake of the list archive, though, > >> here is what I am doing now. > >> > >> // find the element > >> // get its value into "val" > >> CDATASection cdata = document.createCDATASection(val); > >> if (cdata != null) { > >> element.removeChild(child); > >> element.appendChild(cdata); > >> } > >> > >> Still feels like a hack to me but it is working and I need to move > >> on. I'd like to have a generic solution, of course. This is a > >> really specific solution so I'm going to continue to call it a hack. > >> > >> Thanks again > >> Marc > >> > >> On Mar 21, 2006, at 1:12 PM, Stephen Bash wrote: > >> > >>> Marc- > >>> > >>> I don't know of a way to do achieve your goal in Castor, but that > >>> doesn't mean one doesn't exist. > >>> > >>> Looking through the Marshaller class though, I feel like you really > >>> want to be down at the XML serializer level. For example, when the > >>> serializer receives a call to start element "foo" with attribute > >>> "bar", just before it actually writes the output to a stream, you > >>> intercept the value of bar and replace it with <![CDATA[value of > >>> bar]]>. Anything higher up the chain I'm afraid will end up getting > >>> escaped by the serializer. Sounds like a job for AspectJ if I ever > >>> heard one :) Sorry I can't be of more help. Your post processing > >>> might be the easiest way to go. > >>> > >>> Stephen > >>> > >>> > >>> On 3/21/06, Marc Respass <[EMAIL PROTECTED]> wrote: > >>> > >>>> Hi all, > >>>> > >>>> I'm new to the list and just started using Castor. I'm impressed so > >>>> far but I've run into a problem. The party who has defined the XML > >>>> document demands that all values be wrapped in CDATA. Right now, > >>>> I am > >>>> marshalling the object with Castor and then hacking the document by > >>>> calling getElementsByTagName(), pulling the value, creating a CDATA > >>>> section, removing the element's child, and appending the new CDATA > >>>> section. It would be cool if I could get the value wrapped in CDATA > >>>> automatically during the marshalling. I tried writing a > >>>> GeneralizedFieldHandler that would wrap the value like > >>>> > >>>> value = "<![CDATA[" + value + "]]>"; > >>>> > >>>> but that is so wrong and the first "<" gets escaped. I've look > >>>> around > >>>> and tried lots of different things but I have been entirely > >>>> unsuccessful. Can anyone point me in the right direction to do > >>>> this? > >>>> Is it possible or is my hack not really a hack so I should > >>>> change my > >>>> method name from hackDocument() to fixUpDocument()? :) > >>>> > >>>> Thanks a lot > >>>> Marc > >>>> > >>>> ------------------------------------------------- > >>>> If you wish to unsubscribe from this list, please > >>>> send an empty message to the following address: > >>>> > >>>> [EMAIL PROTECTED] > >>>> ------------------------------------------------- > >>>> > >>>> > >>> > >> > >> > >> ------------------------------------------------- > >> If you wish to unsubscribe from this list, please send an empty > >> message to the following address: > >> > >> [EMAIL PROTECTED] > >> ------------------------------------------------- > > > > > > -- > > Benoit Maisonny [EMAIL PROTECTED] > > Director & Consultant http://synclude.com > > Synclude > > > > > > ------------------------------------------------- > > If you wish to unsubscribe from this list, please send an empty > > message to the following address: > > > > [EMAIL PROTECTED] > > ------------------------------------------------- > > > > > ------------------------------------------------- > If you wish to unsubscribe from this list, please > send an empty message to the following address: > > [EMAIL PROTECTED] > ------------------------------------------------- > >
