Storing intervals in the Lucene index

Question

I have documents with the annotated zones - say, 'title', 'body', and 'comments' (zones also may be nested). I want to search for a word 'Obama' in the 'title' zone. I can use a SpanQuery like word:'Obama' & zone:'title' matching at the same position, but it means that I need to store zone attribute for each word position in the document. Can I just store zones as interval coordinates and then perform queries only inside those intervals?

Answer 1

这似乎很混乱，但是您可以将每个单词的间隔存储为类似Dewey-Decimal的编码层次结构（请参阅我的愚蠢Lucene Tricks：层次结构 ），这将使您可以在层次结构的任何部分（所有文本，所有标题，仅标题语义等）