Mongo Query for nearest address search-CodePudding

`I am working on following Events Collection:

{
    title: "Birthday party",
    location: {
      address: "main street, Užupis Pizza, Užupis, Vilnius, Lithuania",
      locationDetails: {
        locality: "Vilnius",
        country: "Lithuania",
        administrative_area_level_2: "Vilnius city municipality",
        administrative_area_level_1: "Vilnius County",
       
      },
},

{
 
    title: "Wedding",
    location: {
      address: "Prie angelo, Uzupis, Vilnius, Lithuania",
      locationDetails: {
        locality: "Vilnius",
        country: "Lithuania",
        administrative_area_level_2: "Uzupis",
        administrative_area_level_1: "Vilnius",
      },
    },
  },

{
  
    title: "engagement",
    location: {
      address: "Plaza, NYC, USA",
      locationDetails: {
        locality: "New your city",
        country: "USA",
        administrative_area_level_1: "New York City",
      },
    },
  }

I am looking for an efficient query which can find the sorted documents by most relevant location (nearest location) to less relevant location (Farthest location).

Example:

Case I: If user Search event near "Uzupis pizza", or "main street Užupis" it should return

 [ {
    title: "Birthday party",
    location: {
      address: "Užupis Pizza, main street, Užupis, Vilnius, Lithuania",
      locationDetails: {
        locality: "Vilnius",
        country: "Lithuania",
        administrative_area_level_2: "Vilnius city municipality",
        administrative_area_level_1: "Vilnius County",
       
      },
    },

{
    _id: {
      $oid: "2",
    },
    title: "Wedding",
    location: {
      address: "Prie angelo, Uzupis, Vilnius, Lithuania",
      locationDetails: {
        locality: "Vilnius",
        country: "Lithuania",
        administrative_area_level_2: "Uzupis",
        administrative_area_level_1: "Vilnius",
      },
    },
  }]

Case 2: If user search "Prie angelo Vilnius" It should return both documents but change in order.

[{
    _id: {
      $oid: "2",
    },
    title: "Wedding",
    location: {
      address: "Prie angelo, Uzupis, Vilnius, Lithuania",
      locationDetails: {
        locality: "Vilnius",
        country: "Lithuania",
        administrative_area_level_2: "Uzupis",
        administrative_area_level_1: "Vilnius",
      },
    },
  },

{
 _id: {
      $oid: "1",
    },
    title: "Birthday party",
    location: {
      address: "Užupis Pizza, Užupis, Vilnius, Lithuania",
      locationDetails: {
        locality: "Vilnius",
        country: "Lithuania",
        administrative_area_level_2: "Vilnius city municipality",
        administrative_area_level_1: "Vilnius County",
       
      },
    }]

I tried solution like regex, text search but I am not sure if they are efficient enough as we are expecting 2 million event data in next couple of year. I can change in schema also or add location co-ordinates if required. Thanks.

CodePudding user response：

This might be use full.

In your case your are expecting to perform full text search.

Reference from https://www.mongodb.com/docs/manual/core/link-text-indexes/

Step 1: Create a Text index to location.address

db.event_col.createIndex({"location.address": "text"})

Step 2: Aggregate the query

//This might be use full.

db.collection.aggregate([
  {
    $match: {
      $text: {
        $search: "Prie angelo Vilnius"
      }
    }
  }
])

If you need to added geo-coordinate you can add additional pipelines to this query.

Text search for relevant data after you can go for near location search via geosearch.

I checked the above cases it's works on.