[tei-council] gaiji in PCDATA, or <g> in text

Syd Bauman Syd_Bauman at Brown.edu
Sat Oct 8 22:23:31 EDT 2005


While we were in Oxford, Lou, Sebastian, and I finally hammered out a
plan for allowing <g> in places where text is allowed. Sebastian gets
the credit for the scheme, I think. 

* A new class, 'tei.gLike' which has only 1 member, <g>
* A new macro, 'macro.xtext' which is declared 
  macro.xtext = (text | tei.gLike)*
* Wherever <g> is desired, use reference macro.xtext instead of
  rng:text.

Sebastian recently implimented the system, and today I went through
various occurences and set them to either macro.xtext or rng:text.
Here is the list of the current settings, with some comments.

<altIdent>         allow <g>
<formulaNotations> doesn't matter, not used; should be nuked
<formula>          text only -- modern formulas don't contain gaijis
<charName>         text only -- must follow Unicode convention, which does not include gaijis
<glyphName>        text only -- must follow Unicode convention, which does not include gaijis
<localName>        allow <g>
<value>            allow <g>
<string>           allow <g>
<day>              allow <g>
<geog>             allow <g>
<hour>             allow <g>
<minute>           allow <g>
<month>            allow <g>
<second>           allow <g>
<week>             allow <g>
<year>             allow <g>
<schemapattern>    text only
<att>              CHANGED to tei.data.ident to match prose definition
<code>             text only
<defaultVal>       text only (should be tei.data.names?)
<eg>               text only
<egXML>            text only
<ident>            text only
<memberOf>         allow <g>
<stringVal>        text only, should be nuked
<val>              text only
<altName>          allow <g>
<collection>       allow <g> (should be macro.phraseSeq, no?)
<depth>            allow <g>
<height>           allow <g>
<institution>      allow <g> (should be macro.phraseSeq, no?)
<locus>            allow <g> (maybe should be text only?)
<origDate>         text only
<origPlace>        allow <g> (should be macro.phraseSeq, no?)
<repository>       allow <g> (should be macro.phraseSeq, no?)
<width>            allow <g>
<unicodeName>      text only -- must be a Unicode name, and they don't have gaijis

Speak up if you think any of these have been incorrectly instituted.
It would also be useful if people would express opinions on the
"should be" coments.




More information about the tei-council mailing list