Chronological Thread 

< Chronological >      < Thread >

---

Thanks a lot for such a complete explanation. And, btw, The solution works
for me ;-)

On Sat, Oct 26, 2002 at 11:30:22AM +0200, Joris van der Hoeven wrote:

> Block text (i.e. multiple paragraph text) and tabular text inside macros is
> a little bit tricky.

right..Actually, I was looking for the possibility to put a framed box around
pieces of text. It still fails, but I'll try to benefit for your hints.

> Since TeXmacs does not yet support "meta-information about DTD's" (it will,
> it will), the fact that a macro contains block text or tabular text should
> be detectable automatically from the macro body, both when defining *and*
> when expanding the macro. Even when the body consists of a single
> paragraph, you should make TeXmacs believe that the body is a (potentially
> multiparagraph) block with a single paragraph.

> Using scheme notation, the string
> [...]
> is again a block. If you want to *concatenate* block text with normal text,
> then you should use the "surround" primitive: this primitive surrounds
> block text on the left and the right with normal text.

So, one should use 'surround' in such places as an explicit coercion, until
the material could be automatically detected?

When I looked at Scheme exportation from TeXmacs files, 'document' and
'concat' should be understood with different meanings? For instance:

(itemize
(document
(concat (item) "bla bla 1") (1)
(concat (item) "bla bla 2"))) (2)

expresses the fact that (1) and (2) are 'block text' material, while each of
them contains 'tabular text'?

> All this will be simplified in the future, but this depends on several
> complex changes which we plan to make in the forthcoming year. In any
> case, the recognition of block text can not be fully automized, so the user
> will always need to specify this information somewhere (in the macro
> definition, application, or in a file with "meta-information about the
> DTD").

My feeling is that expecting things to be automized relies on an arbitrary
decision about *how* contents will be presented. In most cases, the rendering
will match the user expectation; at least for the already/commonly used
environments, like 'itemize' here above. However, in general, I agree with
you the user will have to specify how presentation should be done.

For that reason, my question is: does there exist a primitive which has
exactly the purpose of expressing 'that piece of text' is block/tabular/..?
Does 'surrond' exists with the very purpose of capturing this idea? I very
much like the 'with' wrappers which behaviour sounds better to me. Is
something similar possible wrt the kind of questions raised above?

Reading Scheme exports from TeXmacs documents is really interesting and other
questions arise that i'd be pleased to discuss too.

Best,
--
Philippe Audebaud.

---

• New environments and paragraphs length, Philippe Audebaud, 10/25/2002
  • Re: New environments and paragraphs length, Joris van der Hoeven, 10/26/2002
    • Re: New environments and paragraphs length, Philippe Audebaud, 10/28/2002

---

Archive powered by MHonArc 2.6.19.