标题/内容的Lucene前缀非常慢

Question

这就是我所做的，我只有150个平均文件大小为2KB的文件。

    BooleanQuery.Builder builder = new BooleanQuery.Builder();

    // filename : prefix search
    BooleanQuery.Builder filenameBuilder = new BooleanQuery.Builder();
    for (String s : queryStr.trim().split("\\s+")) {
      PrefixQuery filenameQuery1 = new PrefixQuery(new Term(Indexer.FIELD_NAME_FILENAME, s));
      filenameBuilder.add(filenameQuery1, BooleanClause.Occur.SHOULD);
    }

    // content : prefix search
    BooleanQuery.Builder contentBuilder = new BooleanQuery.Builder();
    for (String s : queryStr.trim().split("\\s+")) {
      PrefixQuery contentQuery1 = new PrefixQuery(new Term(Indexer.FIELD_NAME_CONTENT, s));
      contentBuilder.add(contentQuery1, BooleanClause.Occur.MUST);
    }


    // Set filename and content with different weights.
    Query filenameQuery = filenameBuilder.build();
    Query contentQuery = contentBuilder.build();
    filenameQuery.setBoost(4.0f);
    contentQuery.setBoost(1.0f);

    // Merge filename and content queries : filename OR content
    builder.add(filenameQuery, BooleanClause.Occur.SHOULD);
    builder.add(contentQuery, BooleanClause.Occur.SHOULD);
    Query query = builder.build();

注意：如果我仅搜索filename，它似乎要快得多，但filename + content会超过10秒！

此外，我已尝试将PrefixQuery上的content更改为TermQuery（我的意思是进行精确搜索），但仍然无法做出任何改进。