CODEWITHSUNDEEP
node.jsmongodbmongoosemongodb-queryaggregation-framework
denislexic
denislexic

Reputation: 11342

MongoDB/Mongoose query: How to get different sets of data based on a field's value

I'm using nodejs with mongoose library for my mongoDB interaction.

I'm building a landing page query (think Netflix page) and need to get ~15 results for each categories (think Comedy, Horror, Drama, ...).

It seems not like a great solution to send a query for each "category" and deal with the callback issues as well as a heavy request on the database for a simple page load...as illustrated below: 

const categories = ['horror','drama','comedy']
let counterCallback = 0
let allShows = []
categories.map(category =>{
    Show.find({category: category},{limit: 15}, (err, shows)=>{
       if(shows){ allShows.push(shows)}
       allShows++
       if(allShows === categories.length){ return res.json(allShows) }
    })
})

I was thinking maybe of making an $or request, but I can't limit for the number for each "or"s ...

const categories = ['horror','drama','comedy']
Show.find({
    $or:categories.map(category=> {return {category: category} })
},{limit: 10})

How can I achieve a request that limits the number of results per "category"?

Thanks

Upvotes: 1

Views: 306

Answers (2)

whoami - fakeFaceTrueSoul
whoami - fakeFaceTrueSoul

Reputation: 17915

You can take advantage of MongoDB's aggregation-framework to run complex queries, try below query :

var categories = ['horror', 'comedy', 'drama']

Show.aggregate([
    /** This $match would help to  limit docs to particular categories, as each pipeline in facet stage has to iterate to all docs in collection for thrice as 3 categories given, So at least limiting docs to needed categories. */
    { $match: { category: { $in: categories } } }, 
    {
        $facet: {
            'horror': [{ $match: { category: 'horror' } }, { $limit: 1 }],
            'comedy': [{ $match: { category: 'comedy' } }, { $limit: 1 }],
            'drama': [{ $match: { category: 'drama' } }, { $limit: 1 }]
        }
    }])

Test : MongoDB-Playground

Alternatively you can achieve that using $group as well :

Show.aggregate([
    { $match: { category: { $in: categories } } },
    { $group: { _id: '$category', shows: { $push: '$$ROOT' } } },
    { $project: { category: '$_id', _id: 0, shows: { $slice: ['$shows', 1] } } },
    /** Below two stages are optional  */
    { $group: { _id: '', data: { $push: '$$ROOT' } } },
    { $project: { _id: 0 } }
])

Test : MongoDB-Playground

Upvotes: 2

Valijon
Valijon

Reputation: 13103

You may use MongoDB Aggregation framework with $group operation.

I assume you have MongoDB v4.2. For earlier versions, instead of {$replaceWith:"$data"} use {$ReplaceRoot:{newRoot:"$data"}}

Show.aggregate([
  {
    $match: {
      category: {
        $in: [
          "horror",
          "comedy",
          "drama"
        ]
      }
    }
  },
  {
    $group: {
      _id: "$category",
      data: {
        $push: "$$ROOT"
      }
    }
  },
  {
    $addFields: {
      data: {
        $slice: [
          "$data",
          15
        ]
      }
    }
  },
  {
    $unwind: "$data"
  },
  {
    $replaceWith: "$data"
  }
]).exec((err, shows) => {
    if (err) throw err;
    console.log(shows);
})

MongoPlayground

Upvotes: 2

Related Questions