Re: Page breaks out of order for writing in multiple orientations
Martin Holmes <[email protected]>
| Newsgroups | gmane.text.tei.general |
|---|---|
| Message-ID | <[email protected]> |
HI Peter, You could use the @next and @prev attributes on pb tags as well as divs. For orientation, the Writing Modes section of the Guidelines might help: <http://www.tei-c.org/release/doc/tei-p5-doc/en/html/WD.html#WDWM> It was revised fairly recently, and provides standards-based ways of describing text sequence and orientation. You may not need it, though, if the text itself always written left-to-right, top-to-bottom, and it's only the orientation of the writing surface which changes. Cheers, Martin On 2016-10-17 09:14 AM, Peter Boot wrote: > Hello all, > > We are working on a edition of the complete writings and letters of the > painter Mondrian. One of the notebooks that he used presents an > interesting challenge to the encoder. > > The writing in the book uses all four possible orientations: > - 'normal writing', starting in the top left corner of the page, > proceeding to the next page when the page is full > - 'reversed': the book was reversed and used normally, so the writing > now begins in what is from the original perspective the bottom right > corner, and the next page of a text is now on the 'previous' page (given > the main page numbering) > - 'landscape': writing begins bottom left, lines are > vertical, proceeding to the next page when the page is full > - 'landscape reversed': writing begins top right, proceeding to previous > page when page is full > > We want to present to the user both a physical view of the notebook, > where the pages are presented in sequence, and a logical view, that > presents the reconstructed text fragments in their intended order. We'd > much prefer to construct both from a single transcription. > > In all the encoding project I was involved with up to now, it has been > the case that the physical view of the manuscript was generated from the > sequence of pb-elements. In the XML therefore, the pb-elements had to be > ordered as in the manuscript. This has never presented a problem. But if > we do that in this case, for e.g. the fragments that start at the end of > the notebook this lead would lead to unhappy consequences for div- and p > -elements. > > Suppose a div that starts on p. 100 continues on p . 99, and we keep the > pbs in sequence. Then the div that starts on p. 100 will have to be > closed on p. 99. That obviously won't work, so we'll have to split the > div and p and who knows segs and quotes and hi-elements and connect them > using next- and prev-attributes, something like this: > > <pb n="99"/> > <div xml:id="div-123" prev="#div-122"> > <p xml:id="p-789" prev="#p-788"> > <!-- text continued on p. 99 --> > </p> > </div> > <pb n="100"/> > <div xml:id="div-122" next="#div-123"> > <p xml:id="p-788" next="#p-789"> > <!-- text starts on p. 100 --> > </p> > </div> > > As that is an error-prone mess, we have no choice but to reverse the pb's: > > <pb n="100"/> > <div xml:id="div-122"> > <p xml:id="p-788"> > <!-- text starts on p. 100 --> > <pb n="99"/> > <!-- text continued on p. 99 --> > </p> > </div> > > Which is obviously better. > > The orientation of the writing on the page can be expressed using the > rotate attribute on a zone element corresponding to the pb. > > I can't think of another reasonable way of doing this, but I wanted to > ask whether there are other ideas for solving this. > > Thanks, > Peter > >