[tei-council] Stand Off Markup

Tue Aug 8 13:10:04 EDT 2006

JC> As we were all actioned in the last minutes, and since I'm going
JC> away for a week's holiday, I've read 14.9, the section on
JC> stand-off markup. I didn't realise the way xinclude dealt with
JC> potentially overlapping markup as detailed in 14.9.3, that is
JC> good to know and I'm glad it is detailed there.

Thank you, James!

JC> I've got no major concerns (though I did feel that there could be
JC> more examples and that both 14.9.3 and 14.9.5 especially would
JC> not suffer from more explanation even if redundant, perhaps
JC> demonstrating the various kinds of applications people might make
JC> of it. (Not just xinclude this paragraph.)

I think you're right, but I have to admit, coming up with such
examples is a pain.

JC> By the way, the sentence in 14.9.5 after the first example
JC> mentions <text> but really means 'text' or 'text()', I think.
JC> There is no <text> in the example.

Fixed, thanks.

JC> But I think xinclude is the right idea, and the TEI XPointer
JC> Schemes detailed in 14.2.4 are wonderful.

Mind if I pass this on to the workgroup? :-)

JC> I think back to using xptr to grab ranges of verses (breaking
JC> over chapters) from the bible to import into marked-up liturgical
JC> documents, and think that if I had had a decent xinclude
JC> processor implementation and if the TEI had recommended its use,
JC> that I would have saved myself a lot of XSLT headaches.

Indeed. People 'round here have done similar things. The oXygen
editor and the xmllint commandline tool both do xinclude processing. 
At one point we even had a local copy of the Guidelines that used
XInclude rather than entity references to manage chapters, etc.

CW> You [JC] mean the sentence:
P5> ... the expression <code>range(/1/2/1.0,/1/2/11.1)</code> will
P5> select the whole poem, <gi>text</gi> <emph>and</emph>
P5> <gi>div</gi> elements <emph>and</emph> hypertext links (NB: in
P5> XPointer whitespace-only text nodes count).

The detail James was pointing out has been fixed in the Sourceforge
subversion tree, but ... now that I read this, I don't think it's
correct. The arguments to the range() scheme are supposed to be
xpointers, not paths. I'm going to bop this back to the workgroup for
review and correction.

> The general policy with the examples in this section seems to be to
> give a filename for a file that is supposed to hold the stuff in
> the grayed area (that is to say the content of the example). This
> is a very good policy, but at some places it is not implemented.
> The example in 14.9.3 has three example sections, from which the
> first one is referred to as "example1.xml" in the second one. Since
> there is no filename given on the first one, it makes it difficult
> to follow.

Oh, thank you for reminding me! This goes back aways. I felt it was
quite important to be able to associate a heading of some kind (in
this case a filename) with an example. Thus I had suggested that
<head> be permitted as a first child of <exemplum>, a suggestion that
Lou & Sebastian didn't like. Besides, <exmplum> isn't permitted in or
between paragraphs of prose as <eg> and <egXML> are, so it wouldn't
help here, anyway. Thus I accomplished this task with a kludge, and
stuck a comment in the source that it was a kludge to remind myself
to fix this. Since that time, (thanks to our wonderful class system)
<figure> can now contain <eg> or <egXML>. Thus encoding these as,
e.g., 
| 	  <figure>
| 	    <head><ident type="file">Source.xml</ident></head>
| 	    <eg><![CDATA[<content>To make a prairie it takes a clover and one bee,\n
| One clover, and a bee,\n
| And revery.\n
| The revery alone will do,\n
| If bees are few.\n
| </content>]]></eg>
| 	  </figure>
is probably the right way to go. What do you think, Lou?

CW> Apart from that, this section looks now very good for me. Good to
CW> finally see this coming to the light of P5.

I'm glad that it is finally getting some attention after 7 months,
and even happier that it looks good to Christian & James.

CW> (BTW, if you find my recent messages slightly more weird than
CW> usual, please bear in mind that the highest temperature here has
CW> been between 34"C and 38"C for 14 consecutive days -- there has
CW> to be some collateral damage)

Man, it's a bit weird to have simultaneous heat waves on opposite
sides of the planet.