Right now we use sumy in order to summarize a page in the search results. That's because I haven't been able to get xapians search highlight feature to work from inside python.
As everything is (supposed to be) local, I'd be inclined to not store a separate copy of the text, but instead to fetch a new copy of the text from the original resource. This would be completely impractical for a real search results page, but for this one where storage space is important I think it should work fine.
Right now we use sumy in order to summarize a page in the search results. That's because I haven't been able to get xapians search highlight feature to work from inside python.
As everything is (supposed to be) local, I'd be inclined to not store a separate copy of the text, but instead to fetch a new copy of the text from the original resource. This would be completely impractical for a real search results page, but for this one where storage space is important I think it should work fine.