Subject: Re: node-property From: Paul Prescod <papresco@xxxxxxxxxxxxxxxx> Date: Tue, 29 Jul 1997 00:10:07 -0400 |
Richard Light wrote: > As I read it, you are fighting a losing battle. DSSSL-O seems to be > carefully constructed so that you cannot get at DTD information like > element declarations. Although the '(root' construction rule gives you > access to the sgml-document node, the commands you are then allowed to > use limit you to element-type nodes within the document instance. > > For example, the rubric for '(process-children)' quite explicitly says: > "When the current node is of class 'sgml-document', the value of the > 'document-element' property is treated as being the children of the > node." In other words, don't try getting at the prolog! > > Or am I wrong? (I would be delighted if I were!) No, you are right. That's why I was surprised that I could get at come DTD information such as entities. In a sense, entities belong to the instance, but SGML puts them syntactically in the DTD. I have since figured out that the ELEMENTS node-list should give the elements a full list of elements in the instance in tree order, not the elements in the DTD. Maybe James will comment on whether he intends to extend Jade to support a more full grove model. Such a grove would provide a single, well-documented (ISO standardized!) API for C++ programmers that would seem to have all of the power of the Jade "native API" (wouldn't it?). Paul Prescod DSSSList info and archive: http://www.mulberrytech.com/dsssl/dssslist
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
Re: node-property, Richard Light | Thread | Re: node-property, W. Eliot Kimber |
Re: WinHelp, Paul Prescod | Date | Re: node-property, W. Eliot Kimber |
Month |