Using BooleanQuery or write more indexes?

Question

A category tree like this:

root_1
  sub_1
  sub_2
  ... to sub_20

Every document has a sub category(like sub_2). Now, I only wrote sub_2 in lucene index:

new NumericField("category",...).setIntValue(sub_2.getID());

I want to get all root_1's documents, using BooleanQuery (merge the sub_1 to sub_20) to search or write an other category in every entry document:

new NumericField("category",...).setIntValue(sub_2.getID());
new NumericField("category",...).setIntValue(root_1.getID());//sub_2's ancestor category

Which is the better choice?

score 2 · Accepted Answer · edited May 23 '17 at 12:04

2

I would use a path enumeration/'Dewey Decimal' representation of the category hierarchy. That is, instead of just storing 'sub_2' for the second child of the first root, store instead something like '001.002'.

To find the root and all of its children, you would search on "category:001*".

To find only the children of the root, you would search on "category:001.*".

(Please also see How to store tree data in a Lucene/Solr/Elasticsearch index or a NoSQL db?.)

edited May 23 '17 at 12:04

Community

1
1

answered May 08 '12 at 12:56

Mark Leighton Fisher

5,609
2
18
29

1

I've done this before, except I didn't use IDs, I just stored the name as a path eg `category=root_1/sub_1` etc and then searched for `root_1` using `category:root_1*` – Bohemian May 08 '12 at 13:01

Using BooleanQuery or write more indexes?

1 Answers1

Linked