When recording long quotations from another book, I want to be able to
enter more than one paragraph:
<quote><p>In a series of paintings begun in 1907, Hassam
showed a solitary woman standing by a window--the individual placed
represented a refinement of nature and culture that had been long
in the making.
<p>The key to the picture, however, is found in the sliver of the
"outside world" glimpsed through the window, where laborers are
is coming to maturity as a guardian of civilization.</p>
<bibl>Elizabeth Broun, 1994</bibl></quote>
This works fine, UNLESS I am unwise enough to put a </p> end-tag after
'making.' If I do, I get a parsing error!
This is, I assume, caused by the content model %specialPara;, which is
rather rashly constructed given SGML's interesting treatment of 'mixed
(((%m.chunk), (%component.seq)) | (%paraContent))
The first <p> is satisfied by the %m.chunk part of the model, the second
by %m.common within %component.seq, and all is well so long as there is
no 'daylight' between them (as happens when the second <p> IMPLIES the
omitted end-tag for the first <p>).
As soon as you finish the first para explicitly, you get into a situation
where the newline (loosely speaking) that follows it is treated as PCDATA
because %paraContent allows PCDATA, making it a 'mixed content' element.
However, it ain't allowed in the part of the content model you are using!
Is this a TEI bug? Or just a warning to the unwary?
[log in to unmask]