[XSL-LIST Mailing List Archive Home]
[By Thread]
[By Date]
Fadi,
At 07:50 AM 10/19/2005, Mike wrote:
Sometimes one gets the impression that back in the mists of time The Committee (and it wasn't one committee it was several), or some people on it, thought that we'd no longer use entities. They come from a lexical view of markup languages that seems almost quaint now -- while how powerful it was (and is) is largely ignored.
Fadi, would you please describe more the context of your requirement (as Andrew also asked)? Do you need the entity reference because you don't (yet) know what it will refer to (that is, it's meant as a placeholder for unknown data?), or is it wanted there as a convenience to authors or someone else? Or is it some other reason entirely?
If the first reason, there may be other ways to do achieve the indirection you want that are more friendly to XSLT. If the second (or in any case), you can probably achieve the mapping to entities straightforwardly in a post-process. If the third -- please say more. As a last resort, XSLT+DOE (the optional "disable-output-escaping" feature in XSLT 1.0) might possibly be used if your architecture would tolerate it -- which might be what Mike meant by "just about possible (though not at all convenient)".
RE: [xsl] String evaluation problem
Subject: RE: [xsl] String evaluation problem From: Wendell Piez <wapiez@xxxxxxxxxxxxxxxx> Date: Wed, 19 Oct 2005 12:19:58 -0400 |
Fadi,
At 07:50 AM 10/19/2005, Mike wrote:
It's just about possible (though not at all convenient) to generate an entity reference in the result document that's expanded using the entities defined in the DTD of the result document, when the result document comes to be used. But really, XSLT and entities aren't very happy bedfellows.
Sometimes one gets the impression that back in the mists of time The Committee (and it wasn't one committee it was several), or some people on it, thought that we'd no longer use entities. They come from a lexical view of markup languages that seems almost quaint now -- while how powerful it was (and is) is largely ignored.
Fadi, would you please describe more the context of your requirement (as Andrew also asked)? Do you need the entity reference because you don't (yet) know what it will refer to (that is, it's meant as a placeholder for unknown data?), or is it wanted there as a convenience to authors or someone else? Or is it some other reason entirely?
If the first reason, there may be other ways to do achieve the indirection you want that are more friendly to XSLT. If the second (or in any case), you can probably achieve the mapping to entities straightforwardly in a post-process. If the third -- please say more. As a last resort, XSLT+DOE (the optional "disable-output-escaping" feature in XSLT 1.0) might possibly be used if your architecture would tolerate it -- which might be what Mike meant by "just about possible (though not at all convenient)".
Cheers, Wendell
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
RE: [xsl] String evaluation problem, Michael Kay | Thread | RE: [xsl] String evaluation problem, Fadi Qutaishat |
Re: [xsl] Unparse-text() string con, andrew welch | Date | RE: [xsl] String evaluation problem, Wendell Piez |
Month |
Keywords