findAndCountAll with `GROUP BY` clause causes count to be an array of objects

What you are doing?

I’m trying to get the number of rows returned in using findAndCountAll with a GROUP BY clause however, instead of a single integer, I get an array of { count: x } objects.

const attributes = ['feed_id', [this.store.fn('COUNT', this.store.col('feed_id')), 'count']];
const where = { createdAt: { $gte: start, $lte: end } };
const group = ['feed_id'];
const order = [['count', 'DESC']];
const query = { where, attributes, order, group };

model.findAndCountAll(query).then(data => {
  console.log(data.count);
  console.log(data.rows);
});

What do you expect to happen?

I expected a single integer in count and an array of objects in row.

What is actually happening?

{
  count: [
    { count: '11' },
    { count: '20' },
    { count: '17' },
    { count: '17' },
    { count: '11' },
    ...
    { count: '12' },
    { count: '10' },
    { count: '18' } ],
  rows: [{
    dataValues: [Object],
    _previousDataValues: [Object],
    _changed: {},
    '$modelOptions': [Object],
    '$options': [Object],
    hasPrimaryKeys: true,
    __eagerlyLoadedAssociations: [],
    isNewRecord: false
  }, ...]
}

I got an array of count objects.

For me, the workaround was to stop using findAndCountAll and make an explicit model.count transaction before the findAll. I then removed the GROUP BY clause and replaced the COUNT attribute with COUNT(DISTINCT field).

SELECT COUNT(feed_id) AS "count" FROM "history" AS "history" WHERE ("history"."created_at" >= '2016-04-22 07:24:26.038 +00:00' AND "history"."created_at" <= '2016-06-21 07:24:26.040 +00:00') GROUP BY "feed_id";

SELECT COUNT(DISTINCT feed_id) AS "count" FROM "history" AS "history" WHERE ("history"."created_at" >= '2016-04-22 07:24:26.038 +00:00' AND "history"."created_at" <= '2016-06-21 07:24:26.040 +00:00')

Is this something sequelize can do automatically when it sees a GROUP BY clause?

Dialect: postgres Database version: 9.5.2 Sequelize version: 3.23.0