Why do every .term will be fully read during search ?

[xingtanzjr]

Hi all,

I am new to use Tantivy. And I encountered a case that search is relatively slow. When I debug the code I found that, every .term file were fully read during search stage, which costed large I/O.

The code to compose the query is

 let query = QueryParser::for_index(&index, vec![content_field])
                            .parse_query(&content)
                            .expect("failed to parse query");
searcher.search(&query, &TopDocs::with_limit(limit).and_offset(offset)).expect("failed to search")

Is that ok ? Do you have some insights on that ?

Thanks!