On Feb 10, 2005, at 4:17 PM, Karl Eichwalder wrote:
Lars Aronsson <lars(a)aronsson.se> writes:
I'm skeptic to your proposition that this
would indeed be a workable
solution for the problem at hand. Is there any running prototype to
prove it?
The Perseus Project uses something along these lines (IIRC, you can
retrieve complete books, chapters, or sections) - but I guess the
technology they use predates XPath/Xpointer. Or look at our beloved
"copyrighted" Adelung project
http://mdz.bib-bvb.de/digbib/lexika/adelung/text/band1/
@Generic__BookView;cs=default;ts=default
- I'm quite sure they make use of an XML storage system (Dynaweb?).
If you use idzebra you can send arbitrary XPath statements to the
database back-end and thus you can extract "quotes". Using XLink you
can even embed them properly ;)
And perseus has extensive textual apparatus, proving that even if one
can find any bit of text, on still needs the means to create the
references between it. And perseus deals only with a relatively small
corpus.
Extraction of quotes is no substitute for the means of coordinating
sources.