CODEWITHSUNDEEP
rethinkdb
RonZ
RonZ

Reputation: 467

What is a good way of getting distinct secondary index values from rethinkdb?

Lets say I have a "products" table in rethinkdb, and each product can have lots of tags. I want to retrieve all distinct tags I have in my system (without adding an additional "tags" table) Currently I'm doing this (with an addition of auto-complete filtering):

r.table('products')
    .getAll('[email protected]', { index: "email" })
    .filter(function (doc) {
      return doc.hasFields(['tags']).and(doc('tags').contains(function(tag) {
        return tag.upcase().match("^a".toUpperCase());
      }));
    })
    .pluck('tags')
    .concatMap(function (row) {
      return row('tags').filter(function(tag) {
        return tag.upcase().match("^a".toUpperCase());
      })
    })
    .distinct()

It seems that I'm not using the fact that I have a good index 'email_tags' that already probably has this information, I want to do something like this:

r.table('products_email_tags_index_table')
.between(['[email protected]', 'A'], 
         ['[email protected]', 'B'], 
         {index: 'email_tags', left_bound: 'open', right_bound: 'closed'})
.map(function(indexField) { return indexField.get(1) })

(even though it doesn't keep the casings...)

So my question is: is there any way of querying a secondary index directly?

Upvotes: 1

Views: 266

Answers (1)

mlucy
mlucy

Reputation: 5289

distinct can accept an index. Does this do what you want?

r.table('products_email_tags_index_table')
 .between(['[email protected]', 'A'], 
          ['[email protected]', 'B'], 
          {index: 'email_tags', left_bound: 'open', right_bound: 'closed'})
 .distinct({index: 'email_tags'})

Upvotes: 2

Related Questions