On 9/8/05, David Carlisle <davidc(_at_)nag(_dot_)co(_dot_)uk> wrote:
Is there any objective distinction between the pseudo-lines and real
lines? Maybe a test on interpunction or Capitalised words could
improve the xslt's guesses?
or of course since you know in advance what those 20 are, you could test
for their id in the stylesheet and not merge those cases (saves fixing
by hand each time you run the stylesheet)
The estiamted amount of corrections is based on manually looking
through a smaller sample so not accurate and I don't know the actual
20. However, after I have manually corrected those that I can see,
the person who is interested in the content has promised to
read-through and alert me to any which shouldn't have been merged. :-)
-James
--
James Cummings, Cummings dot James at GMail dot com