> i.e. the first half of butterfly being on page 22 and the
> second half on p. 23.
> <reg orig='butter- 23 fly>butterfly <pb n="23"></reg>
> Are there any objections to this solution?
If you do so, it will be quite complicated to automatically
reproduce the original situation from the contents of your
Why not go the other way around:
<orig reg="butterfly">butter-<pb n="23"/>fly</orig>
But why should you need to encode this? If you just write:
(using ­ for the soft hyphen character) a very simple
stylesheet will be able to produce the 'butterfly' you need. If
you don't trust the software to recognize this as one word you
might perhaps add a w-element to make the situation unambiguously