1.12.2022

SQL Date Range Queries

In my company work, we often use two Date or DateTime fields to express a range when the row data is valid.

CREATE TABLE test (
  id int NOT NULL PRIMARY KEY,
  title varchar(255) NULL,
  valid_from date NULL,
  valid_to date NULL,
);

Both valid_from and valid_to are nullable. NULL means an unlimited bound to one or both sides. The query to check for a valid range of a given @date is quite long due to the null checks.

SELECT * FROM test WHERE
  (valid_from IS NULL OR @date >= valid_from) AND
  (valid_to IS NULL OR @date < valid_to)

Note that this query interprets the lower bound as inclusive and the upper one as exclusive. This makes it easier to define e.g. month ranges with the same day.

Another option is the between operator which is available in most SQL dialects. The problem is that it does not work with NULL as open limit.

PostgreSQL offers range types, which greatly simplify range queries like the one above. For our example, the function daterange(date1, date2) creates a date range, either argument can be null for an unbounded range limit. With that in mind and the @> operator, the range query simplifies.

SELECT * FROM test WHERE daterange(valid_from, valid_to) @> @date

Multiple ranges can also be combined with a Multirange like datemultirange. To go even further, the range types can also be used for a column data type.

CREATE TABLE test (
  id int NOT NULL PRIMARY KEY,
  title varchar(255) NULL,
  valid daterange NOT NULL,
);

However, retrieving this object with an object-relational mapper (ORM) requires the generic type NpgsqlRange only available in the Entity Framework Core version of Npgsql.