Mongodb: How to index multiple nested text fields?

Question

I have the following kind of json in my database:

{ 
    "_id" : "519817e508a16b447c00020e", "keyword" : "Just an example query", 
    "results" : 
    {
        "1" : {"base_domain" : "example1.com", "href" : "http://www.example1.com/"},
        "2" : { "base_domain" : "example2.com", "href" : "http://www.example2.com/"},
        "3" : { "base_domain" : "example3.com", "href" : "http://www.example3.com/"},
        "4" : { "base_domain" : "example4.com", "href" : "http://www.example4.com/"},
        "5" : { "base_domain" : "example5.com", "href" : "http://www.example5.com/"},
        "6" : { "base_domain" : "example6.com", "href" : "http://www.example6.com/"},
        "7" : { "base_domain" : "example7.com", "href" : "http://www.example7.com/"},
        "8" : { "base_domain" : "example8.com", "href" : "http://www.example8.com/"},
        "9" : { "base_domain" : "example9.com", "href" : "http://www.example9.com/"},
        "10" : { "base_domain" : "example10.com", "href" : "http://www.example10.com/"}
    } 
}

My goal is to have results for the following query:

> db.ranking.find({ $text: { $search: "http://www.example9.com"}})

It works when I create an index on all text fields

> db.ranking.ensureIndex({ "$**": "text" }))

But not when I create an index only over "results" field:

> db.ranking.ensureIndex( {"results" : "text"} )

Why?

John Petrone · Accepted Answer · 2014-08-15 15:08:06Z

4

The problem is that "results" is not a field, it's a sub-document. The syntax for creating an index on text fields for MongoDB requires either the notation for all fields, "$*", which you are using correctly, or a list of all text fields:

Create a text Index

You can create a text index on the field or fields whose value is a string or an array of string elements. When creating a text index on multiple fields, you can specify the individual fields or you can use wildcard specifier ($**).

http://docs.mongodb.org/manual/tutorial/create-text-index-on-multiple-fields/

In your case that would look like:

db.ranking.ensureIndex(
                           {
                             "keyword": "text",
                             "results.1.href": "text",
                             "results.1.href": "text",
                             "results.2.href": "text",
                             "results.3.href": "text",
                             "results.4.href": "text",
                             "results.5.href": "text",
                             "results.6.href": "text",
                             "results.7.href": "text",
                             "results.8.href": "text",
                             "results.9.href": "text",
                             "results.10.href": "text"
                           }
                       )

answered Aug 15, 2014 at 15:08

John Petrone

27.5k6 gold badges68 silver badges69 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

John Powell Over a year ago

Presumably this could be made a lot more efficient if results were an array.

Antoine Brunel Over a year ago

Very interesting! Indexing the first href also worked for the rest of them: db.ranking.ensureIndex({"results.1.href": "text"}) Now, is there a way to retrieve directly the key and the value without looping through all the dictionary? @JohnBarça Yes, I wish it were an array!

Antoine Brunel Over a year ago

@JohnBarça You mentioned using an array. Do you mean changing the db structure for this:

{ "_id" : "519817e508a16b447c00020e", "keyword" : "Just an example query", "results" :  [{"base_domain" : "example1.com", "href" : "http://www.example1.com/"},{"base_domain" : "example2.com", "href" : "http://www.example2.com/"},...}]

The index would be hashed or text too?

Antoine Brunel Over a year ago

@JohnBarça Any additional hints?

Antoine Brunel Over a year ago

I just asked another question since I ran into an Index key pattern too large error with code 67 because I have 100 results and not 10!

Collectives™ on Stack Overflow

Mongodb: How to index multiple nested text fields?

1 Answer 1

5 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related