Collection Element Ordering and Comparison

Type Order (Ascending)

• Unordered maps means we do not guarantee any order
• If this informal ordering changes and an application relies on it, the application could fail
• Ordered maps have performance advantages in 6.x

Elements with different types are ordered based on their type.

1. NIL
2. BOOLEAN
3. INTEGER
4. STRING
5. LIST
6. MAP
7. BYTES
8. DOUBLE
9. GEOJSON
10. INF

NIL

The lowest valued type. NIL is a singleton.

BOOLEAN

False < True

INTEGER

Ordered by integer value.

STRING

Order by each byte in the string.

"aa" < "b"

Strings are assumed to have UTF-8 encoding.

LIST

Order by:

1. Each element starting from index 0 [1, 2] < [1, 3]
2. Element count [1, 2] < [1, 2, 1]

caution

There is a known issue where Map/List comparisons may be incorrect for maps/lists of different lengths. If planning to utilize comparisons please update to Aerospike Database version 4.3.1 or higher.

MAP

Order by:

1. Element count
2. Each key in order stored
3. Map values if the corresponding map keys are equal

note

There is a known issue where Map/List comparisons may be incorrect for maps/lists of different lengths. If planning to utilize comparisons, update to Aerospike Database version 4.3.1 or higher.

BYTES

Order by each byte in the string.

DOUBLE

Ordered by float value.

INF

Introduced in Aerospike Database version 4.3.1.

The highest valued type. INF is a singleton.

Not a storage type. Storing INF in a bin list or map has undefined behavior.

note

Introduced in Aerospike Database version 4.3.1.

Comparison

Wildcard

The singleton WILDCARD(*) type can be passed as parameters in certain operations. WILDCARD is not a storage type.

Bin: [ [1, 1], [1, 2], [1, 3], [2, 1], [2, 2], [2, 3] ]

list_get_all_by_value([1, *]) -> [ [1, 1], [1, 2], [1, 3] ]

note

Introduced in Aerospike Database version 4.3.1.

Intervals

Intervals are inclusive-exclusive by default: start <= elements < end

Bin: [ [1, 1], [1, 2], [2, 1], [2, 2], [3, 1] ]

list_get_by_value_interval(start=[1, NIL], end=[2, NIL]) -> [ [1, 1], [1, 2] ]

INF

Using INF, we can get an inclusive-inclusive interval when using 2nd level lists.

list_get_by_value_interval(start=[1, NIL], end=[2, INF]) -> [ [1, 1], [1, 2], [2, 1], [2, 2] ]

note

Introduced in Aerospike Database version 4.3.1.