[tei-council] Chapter 23 - Using the TEI - part II

[Lou's Laptop]

Brett Zamir wrote:
> Ok, here is the last installment... Enjoy! :)
Thanks for staying the course ...

>
> *
> 23.4 Implementation of an ODD System*
>
> Any plans for ideas on implementation of a TEI processor (including 
> those just acting on a subset of TEI)?
>

Lots. That's what the software SIG is all about.

> *23.4.1 Making a Unified ODD*
>
> The attribute @targeLang definition ought to have "element" and 
> "attribute" as plural. And I'm unclear as to what this definition means.
>
It's not in house style either. I've changed it.

> And I'm unclear as to why one would use @docLang.
>
Element specs may contain descriptions (<desc>s) in several languages: 
this allows you to select which one to choose if you dont want them all
> I was unclear on the meaning of this: "However, if that 
> <gi>rng:choice</gi> is itself inside a <gi>rng:zeroOrMore</gi> inside 
> a <gi>rng:group</gi>, the simplifying process *may be harder*, and the 
> resolution is *left to a later stage*."

I've revised this a bit, possibly making it more prolix in the process.

>
> *23.4.2 Generating Schemas
> *
> One paragraph begins "Finally, an application " while the next 
> paragraph begins "Finally, ODD processors "
>
I've run the two together.

> *23.4.4.2 Classes
> *
> 1) For the lines, "*Notice that the <gi>desc</gi> element is used to 
> add an <gi>a:documentation</gi> element* to the schema, which some 
> editors use to provide help during composition. The *<gi>desc</gi> 
> elements in the <gi>valList</gi> are used to create the human-readable 
> sentence *<quote>Sample values include: 1] no; 2] yes; 3] weak; 4] 
> strong</quote>", these are not actually output in the docs at 
> http://tei.oucs.ox.ac.uk/P5/Guidelines-web/en/html/USE.html
>
Formatting problem.

> 2) If the line, "An individual attribute consists of an 
> <gi>rng:attribute</gi> with a <att>name</att> attribute derived 
> according to the naming rules described above" is meant to be read 
> along with the eariler line, "the pattern name is created by appending 
> an underscore and the name of the generation sequence to the class 
> name", I don't see that the example actually uses an underscore, etc. 
> here.

No, this is a reference to the conventiuons on attribute naming; I've 
added an explicit cross reference to make it cleaer.
>
> *23.3.6 Generating Documentation*
>
> For the line, "One model of display on a web page is shown in Figure", 
> I removed "Figure" because the <ptr> that follows already gets 
> rendered with a Figure.
>
OK

> *23.4.7.1 Selection of Modules*
>
> The examples at 
> http://tei.oucs.ox.ac.uk/P5/Guidelines-web/en/html/USE.html ainclude 
> cases where & gt; is rendered as such when it shouldn't.
>
I hope these have all been fixed now...

> *23.4.7.4 Embedding Local Modifications (DTD only)
> *
> For the line, "<eg><![CDATA[<!ENTITY % macro.phraseSeq 'it | bd 
> |'>]]></eg>", isn't this only a partial listing, unlike what the 
> directions imply should take place, that the parameter entity be 
> "modified to include the generic identifiers for the new elements we 
> wish to create" (i.e., that the old items should still be present)?
>
This is just plain wrong, and correcting it will involve reinstating a 
lot of very specific technical details about DTD processing. I am 
therefore minded to fudge the issue.