Struct IntervalCmpValue

struct IntervalCmpValue(i128);

https://github.com/postgres/postgres/blob/REL_15_2/src/backend/utils/adt/timestamp.c#L2384

Do NOT make this pub as the assumption of 1 month = 30 days and 1 day = 24 hours does not always hold in other places.

Given this equality definition in PostgreSQL, different interval values can be considered equal, forming equivalence classes. For example:

• ‘-45 days’ == ‘-1 months -15 days’ == ‘1 months -75 days’
• ‘-2147483646 months -210 days’ == ‘-2147483648 months -150 days’ == ‘-2075900865 months -2147483640 days’

To hash and memcompare them, we need to pick a representative for each equivalence class, and then map all values from the same equivalence class to the same representative. There are 3 choices (may be more): (a) an i128 of total usecs, with months and days transformed into usecs; (b) the justified interval, as defined by PostgreSQL justify_interval; (c) the alternate representative interval that maximizes abs of smaller units;

For simplicity we will assume there are only months and days and ignore usecs below.

The justified interval is more human friendly. It requires all units to have the same sign, and that 0 <= abs(usecs) < USECS_PER_DAY && 0 <= abs(days) < 30. However, it may overflow. In the 2 examples above, ‘-1 months -15 days’ is the justified interval of the first equivalence class, but there is no justified interval in the second one. It would be ‘-2147483653 months’ but this overflows i32. A lot of bits are wasted in a justified interval because days is using i32 for -29..=29 only.

The alternate representative interval aims to avoid this overflow. It still requires all units to have the same sign, but maximizes abs of smaller unit rather than limit it to 29. The alternate representative of the 2 examples above are ‘-45 days’ and ‘-2075900865 months -2147483640 days’. The alternate representative interval always exists.

For serialize, we could use any of 3, with a workaround of using (i33, i6, i38) rather than (i32, i32, i64) to avoid overflow of the justified interval. We chose the usecs: i128 option.

For deserialize, we attempt justified interval first and fallback to alternate. This could give human friendly results in common cases and still guarantee no overflow, as long as the bytes were serialized properly.

Note the alternate representative interval does not exist in PostgreSQL as they do not deserialize from IntervalCmpValue.

Tuple Fields

0: i128

Implementations

impl IntervalCmpValue

fn as_justified(&self) -> Option<Interval>

Recover the justified interval from this equivalence class, if it exists.

fn as_alternate(&self) -> Option<Interval>

Recover the alternate representative interval from this equivalence class. It always exists unless the encoding is invalid. See IntervalCmpValue for details.

Trait Implementations

impl From<Interval> for IntervalCmpValue

fn from(value: Interval) -> Self

Converts to this type from the input type.

impl Hash for IntervalCmpValue

fn hash<__H: Hasher>(&self, state: &mut __H)

Feeds this value into the given Hasher.

fn hash_slice<H>(data: &[Self], state: &mut H)

where H: Hasher, Self: Sized,

Feeds a slice of this type into the given Hasher.

impl Ord for IntervalCmpValue

fn cmp(&self, other: &IntervalCmpValue) -> Ordering

This method returns an Ordering between self and other.

fn max(self, other: Self) -> Self

where Self: Sized,

Compares and returns the maximum of two values.

fn min(self, other: Self) -> Self

where Self: Sized,

Compares and returns the minimum of two values.

fn clamp(self, min: Self, max: Self) -> Self

where Self: Sized,

Restrict a value to a certain interval.

impl PartialEq for IntervalCmpValue

fn eq(&self, other: &IntervalCmpValue) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl PartialOrd for IntervalCmpValue

fn partial_cmp(&self, other: &IntervalCmpValue) -> Option<Ordering>

This method returns an ordering between self and other values if one exists.

fn lt(&self, other: &Rhs) -> bool

Tests less than (for self and other) and is used by the < operator.

fn le(&self, other: &Rhs) -> bool

Tests less than or equal to (for self and other) and is used by the <= operator.

fn gt(&self, other: &Rhs) -> bool

Tests greater than (for self and other) and is used by the > operator.

fn ge(&self, other: &Rhs) -> bool

Tests greater than or equal to (for self and other) and is used by the >= operator.

impl Eq for IntervalCmpValue

impl StructuralPartialEq for IntervalCmpValue