Another output format :
1000 ./test/005403.txt "005403.txt" 57 1 "japan" "0017".
1000 ./test/005403.txt "005403.txt" 57 1 "china" "0013".
One line for each occurence of the property.
Bill Moseley wrote:
> At 08:43 AM 04/28/01 -0700, Prosper Correa wrote:
> >The question is that Swish let the user to get some properties in the
> results (cf. -p option).
> >So, if the returned property is not the one, it would mean that such option
> >is not usefull ;-)
> >I think that returning the rearest property would be a good solution for the
> Currently, there's no way to know which property is the nearest to the word
> in your query. There's no word position stored with each property. And
> properties are not stored with the context of the meta tag it's found
> within, either.
> Again, Swish returns documents. You are trying to index sub documents
> (basically, documents within documents), it seems. Swish can do this with
> the LST type of indexing, or by splitting up your documents in some way.
> There's been three options suggested to you how to solve this: 1) spilt up
> your docs, 2) use LST (by removing the "root" tag), or 3) use the -S prog
> feature to separate your docs. Option number three means you don't need to
> alter your source documents.
> Now, how swish behaves when more than one property of the same name is
> found in a document is a topic of discussion. Should all be
> stored/returned, or just the first/last? If all then how should they be
> joined in the output so you can tell the different properties apart?
> Bill Moseley
Received on Tue May 1 17:45:36 2001