SELECT Clause

The SELECT clause determines the result set.

Purpose

In a SELECT statement, the SELECT clause determines the projection (result set).

Prerequisites

For you to select data from a document or keyspace, you must have the query_select privilege on the document or keyspace. For more details about user roles, see Authorization.

Syntax

select-clause ::= SELECT [ hint-comment ] projection

Optimizer Hints

(Introduced in Couchbase Server 7.1)

In Couchbase Server 7.1 and later, you can supply hints to the optimizer within a specially-formatted hint comment. For further details, refer to Optimizer Hints.

Projection

projection ::= [ ALL | DISTINCT ]
               ( result-expr [ , result-expr ]* | ( RAW | ELEMENT | VALUE ) expr [ [ AS ] alias] ] )

ALL / DISTINCT

(Optional; default is ALL.)

SELECT ALL retrieves all of the data specified and will result in all of the specified columns, including all duplicates.

SELECT DISTINCT removes duplicate result objects from the query’s result set.

Since Couchbase Server 4.5.1, the DISTINCT clause is no longer blocking in nature since it streams the input and produces the output in-parallel, while consuming less memory.

In general, SELECT ALL results in more returned documents than SELECT DISTINCT due to DISTINCT's extra step of removing duplicates. Since DISTINCT is purely run in memory, it executes quickly, making the overhead of removing duplicates more noticeable as your recordset gets larger.

In the below table,

Q1's SELECT DISTINCT reduces the recordset to a small fraction of its original size; and while removing so many of the documents takes time, projecting the remaining small fraction is actually slightly faster than the overhead of removing duplicates.
On the other extreme, Q2's SELECT DISTINCT does not reduce the recordset at all since META().id is already unique, and thus projects the entire original recordset and shows the maximum overhead of about twice as long to execute than SELECT ALL.

N1QL Query SELECT ALL SELECT DISTINCT

N1QL Query	SELECT ALL	SELECT DISTINCT
Q1 SELECT city FROM `travel-sample`.inventory.landmark;	slightly slower	slightly faster
Q2 SELECT META().id FROM `travel-sample`.inventory.landmark;	much faster	much slower

SELECT city FROM `travel-sample`.inventory.landmark;

slightly slower

slightly faster

SELECT META().id FROM `travel-sample`.inventory.landmark;

much faster

much slower

Example 1. Query plan using the DISTINCT operator

Query

EXPLAIN SELECT DISTINCT city FROM `travel-sample`.inventory.landmark; (1)

Results

[
  {
    "plan": {
      "#operator": "Sequence",
      "~children": [
        {
          "#operator": "PrimaryScan3",
          "bucket": "travel-sample",
          "index": "def_inventory_landmark_primary",
          "index_projection": {
            "primary_key": true
          },
          "keyspace": "landmark",
          "namespace": "default",
          "scope": "inventory",
          "using": "gsi"
        },
        {
          "#operator": "Fetch",
          "bucket": "travel-sample",
          "keyspace": "landmark",
          "namespace": "default",
          "scope": "inventory"
        },
        {
          "#operator": "Parallel",
          "~child": {
            "#operator": "Sequence",
            "~children": [
              {
                "#operator": "InitialProject",
                "distinct": true,
                "result_terms": [
                  {
                    "expr": "(`landmark`.`city`)"
                  }
                ]
              },
              {
                "#operator": "Distinct" (1)
              }
            ]
          }
        },
        {
          "#operator": "Distinct" (1)
        }
      ]
    },
    "text": "SELECT DISTINCT city FROM `travel-sample`.inventory.landmark;"
  }
]

1	Lines using the `DISTINCT` operator

Result Expression

result-expr ::= ( [ path "." ] "*" | expr [ [ AS ] alias ] )

( path '.' )? '*' | expr ( 'AS'? alias )?

path ::= identifier [ '[' expr ']' ]* [ '.' path ]

identifier ('[' expr ']')* ( '.' path )?

One or more expressions that evaluate to one or more field names to be in the result set.

If no field name is specified, the input for the query is a single empty object that allows you to perform calculations with the SELECT statement, such as SELECT 10+20 AS Total; or other N1QL expression.

For details with examples, see N1QL Expressions.

RAW / ELEMENT / VALUE

(Optional; RAW and ELEMENT and VALUE are synonyms.)

SELECT RAW reduces the amount of data returned by eliminating the field attribute.

Example 2. Comparing SELECT and SELECT RAW on a basic query

Query

SELECT {"a":1, "b":2};

Query

SELECT RAW {"a":1, "b":2};

Results

[
  {
    "$1": { (1)
      "a": 1,
      "b": 2
    }
  }
]

Results

[
  { (2)
    "a": 1,
    "b": 2
  }
]

1	Added alias
2	No added alias

There are times in which this extra layer might not be desirable since it requires extra output parsing. So the RAW qualifier specifies that the expression that follows not to be qualified, as shown in the next example.

Example 3. Comparing SELECT and SELECT RAW listing 5 airport cities alphabetically

Query

SELECT city
FROM `travel-sample`.inventory.airport
ORDER BY city LIMIT 5;

Query

SELECT RAW city
FROM `travel-sample`.inventory.airport
ORDER BY city LIMIT 5;

Query

SELECT DISTINCT RAW city
FROM `travel-sample`.inventory.airport
ORDER BY city LIMIT 5;

Results

[
  {
    "city": "Abbeville"
  },
  {
    "city": "Aberdeen"
  },
  {
    "city": "Aberdeen"
  },
  {
    "city": "Aberdeen"
  },
  {
    "city": "Abilene"
  }
]

Results

[
  "Abbeville",
  "Aberdeen",
  "Aberdeen",
  "Aberdeen",
  "Abilene"
]

Results

[
  "Abbeville",
  "Aberdeen",
  "Abilene",
  "Adak Island",
  "Addison"
]

Keyspace Name

(Optional; if the keyspace is unspecified in the SELECT clause, it will be taken from the FROM clause.)

The name of the keyspace or keyspaces used, separated by a comma.

Keyspaces are equivalent to collections in Couchbase Server. A keyspace is a set of documents that may vary in structure, and is a unit of authorization and resource allocation.

When specifying the keyspace name in SELECT keyspace_name.* for all fields, the keyspace name will not appear in the result set; whereas not specifying the keyspace name in SELECT * FROM keyspace_name adds the keyspace name to the result set.

Query A

SELECT * FROM `travel-sample`.inventory.hotel;

Results

[
  {
    "hotel": { (1)
      "address": "Capstone Road, ME7 3JE",
      "alias": null,
      "checkin": null,
...
    }
  }
]

Query B

SELECT hotel.* FROM `travel-sample`.inventory.hotel;

Results

[
  { (2)
    "address": "Capstone Road, ME7 3JE",
    "alias": null,
    "checkin": null,
...
  }
]

Query C

SELECT meta().id, email, city, phone, hotel.reviews[0].ratings
FROM `travel-sample`.inventory.hotel LIMIT 5;

Results

[
  { (3)
    "city": "Medway",
    "email": null,
    "id": "hotel_10025",
    "phone": "+44 870 770 5964",
    "ratings": {
      "Cleanliness": 5,
      "Location": 4,
      "Overall": 4,
      "Rooms": 3,
      "Service": 5,
      "Value": 4
    }
  },
...
]

1	Added line with keyspace
2	No added line with keyspace
3	No added line with keyspace

Field Expression

The name of the field or fields, separated by a comma, to be in the query’s result set, such as:

Query

SELECT id, airline, stops FROM `travel-sample`.inventory.route;

To use a field within an array, use [0] after the array name, followed by a period and the field name, such as:

Query

SELECT schedule[0].day FROM `travel-sample`.inventory.route;

AS Alias

alias ::= identifier

A temporary name of a keyspace name or field name to make names more readable or unique, such as:

Query

SELECT schedule[0].day AS Weekday FROM `travel-sample`.inventory.route;

Best Practices

When possible, explicitly list all fields you want in your result set instead of the * to select all fields, since the * requires an extra trip over your network (one to get the list of field names and one to select the field names).

Examples

Example 4. Select all the fields of 1 document from the airline keyspace

Query

SELECT * FROM `travel-sample`.inventory.airline LIMIT 1;

Results

[
  {
    "airline": {
      "callsign": "MILE-AIR",
      "country": "United States",
      "iata": "Q5",
      "icao": "MLA",
      "id": 10,
      "name": "40-Mile Air",
      "type": "airline"
    }
  }
]

Example 5. Select all the fields of 1 document from the landmark keyspace

Query

SELECT * FROM `travel-sample`.inventory.landmark LIMIT 1;

Results

[
  {
    "landmark": {
      "activity": "see",
      "address": "Prince Arthur Road, ME4 4UG",
      "alt": null,
      "city": "Gillingham",
      "content": "Adult - £6.99 for an Adult ticket that allows you to come back for further visits within a year (children's and concessionary tickets also available). Museum on military engineering and the history of the British Empire. A quite extensive collection that takes about half a day to see. Of most interest to fans of British and military history or civil engineering. The outside collection of tank mounted bridges etc can be seen for free. There is also an extensive series of themed special event weekends, admission to which is included in the cost of the annual ticket.",
      "country": "United Kingdom",
      "directions": null,
      "email": null,
      "geo": {
        "accuracy": "RANGE_INTERPOLATED",
        "lat": 51.39184,
        "lon": 0.53616
      },
      "hours": "Tues - Fri 9.00am to 5.00pm, Sat - Sun 11.30am - 5.00pm",
      "id": 10019,
      "image": null,
      "name": "Royal Engineers Museum",
      "phone": "+44 1634 822839",
      "price": null,
      "state": null,
      "title": "Gillingham (Kent)",
      "tollfree": null,
      "type": "landmark",
      "url": "http://www.remuseum.org.uk"
    }
  }
]

SELECT Clause

Purpose

Prerequisites

Syntax

Optimizer Hints

Projection

ALL / DISTINCT

Result Expression

RAW / ELEMENT / VALUE

Keyspace Name

Field Expression

AS Alias

Best Practices

Examples

Related Links