8: Tips when selecting records

Since we have a place record in the database with some fields, we can use a SELECT statement to take a look at it. SELECT is followed by the fields to display and ends with FROM and the target. In this case the target is place: all place records in the database.

SELECT name, address FROM place;

Run Query

The output is an array of the address and name fields of every place record in the database.

[
  {
    address: '2025 Statement Street, Riverdale',
    name: 'Surreal Library'
  }
]

Instead of selecting all records from a table, you can add a WHERE clause to just select the ones that match a condition - or none at all, if no records match the condition.

-- Sumerian library was deleted so will return nothing
SELECT name, address FROM place WHERE name = "𒂍𒁾𒁀𒀀";

Run Query

[]

The star operator and refining what to select

Instead of typing out each field, you can use * to show all of them.

SELECT * FROM place;

Run Query

To select all fields minus a few, you can use OMIT after the star, followed by the unneeded field names.

-- Show every field except 'id'
SELECT * OMIT id FROM place;

Run Query

Using the VALUE and ONLY keywords

You can add VALUE right after the SELECT keyword if you only care about the value of a single field, and not its name.

SELECT VALUE name FROM place;

Run Query

[
  'Surreal Library'
]

The ONLY keyword works in other statements too, such as SELECT. It will always work if you give it an exact record ID, because there can only be one.

SELECT id FROM ONLY place:surreal_library;

Run Query

{
  id: place:surreal_library
}

However, if you follow it with just a table name, the database is no longer sure that only one record will be returned.

SELECT * FROM ONLY place;

Run Query

'Expected a single result output when using the ONLY keyword'

You can use the LIMIT keyword here to guarantee a maximum number of records returned. If it is followed with 1, the database will know that no more than a single record will be returned and the ONLY keyword can be used.

SELECT * FROM ONLY place LIMIT 1;

Run Query

{
  address: '2025 Statement Street, Riverdale',
  floors: 8,
  id: place:surreal_library,
  name: 'Surreal Library',
  place_type: 'library'
}

In this case, we should ensure that the single record returned is the one that we expect it to be. If we had used a random ID like place:llsqrb50cm7rkvqli2sg instead of place:surreal_library, we could use WHERE to make sure that 'Surreal Library' is the record returned.

SELECT * FROM ONLY place
WHERE name = 'Surreal Library'
LIMIT 1;

Run Query

Selecting from more than a single table name

A SELECT statement isn't just limited to a table's fields. You can create and name new fields as you see fit. They can be made from the values of existing fields, or just made up on the spot.

SELECT
    name,
    -- Use the data at 'name' for a new field
    name + '!!!' AS excited_name,
    -- Use the value 'true' for a new field
    true AS is_cool
FROM place;

Run Query

[
  {
    excited_name: 'Surreal Library!!!',
    is_cool: true,
    name: 'Surreal Library'
  }
]

And the target after FROM doesn't need to be a table name either. You can select from raw data too.

SELECT 
    name, 
    address
FROM 
    [
        { address: '10 Push to Main Street, Riverdale', name: 'Vector Park' },
        { name: 'Riverdale City Hall' }
    ];

Run Query

Because the second object has no value for the field address, it will show up as NONE.

[
  {
    address: '10 Push to Main Street, Riverdale',
    name: 'Library Park'
  },
  {
    address: NONE,
    name: 'Riverdale City Hall'
  }
]

You can even select from a combination of table name and raw data. Whatever SurrealDB sees after the FROM clause can be selected and used as output.

SELECT 
    name, 
    address
FROM 
    place, 
    { address: '10 Push to Main Street, Riverdale', name: 'Vector Park' }

Run Query

[
  {
    address: '2025 Statement Street, Riverdale',
    name: 'Surreal Library'
  },
  {
    address: '10 Push to Main Street, Riverdale',
    name: 'Vector Park'
  }
]