Postgresql 中文操作指南
9.20. Range/Multirange Functions and Operators #
有关范围类型的概述，请参阅 Section 8.17。
See Section 8.17 for an overview of range types.
Table 9.55 显示可用于范围类型的专有运算符。 Table 9.56 显示可用于多范围类型的专有运算符。除了这些之外， Table 9.1 中所示的通常比较运算符适用于范围和多范围类型。比较运算符首先按范围下限排序，仅当这些下限相等时，它们才会比较上限。多范围运算符比较每个范围，直至遇到不相等的范围为止。这通常不会产生有用的总体排序，但会提供运算符来允许在范围内构造唯一索引。
Table 9.55 shows the specialized operators available for range types. Table 9.56 shows the specialized operators available for multirange types. In addition to those, the usual comparison operators shown in Table 9.1 are available for range and multirange types. The comparison operators order first by the range lower bounds, and only if those are equal do they compare the upper bounds. The multirange operators compare each range until one is unequal. This does not usually result in a useful overall ordering, but the operators are provided to allow unique indexes to be constructed on ranges.
Table 9.55. Range Operators
Operator Description Example(s) |
anyrange @> anyrange → boolean Does the first range contain the second? int4range(2,4) @> int4range(2,3) → t |
anyrange @> anyelement → boolean Does the range contain the element? '[2011-01-01,2011-03-01)'::tsrange @> '2011-01-10'::timestamp → t |
anyrange <@ anyrange → boolean Is the first range contained by the second? int4range(2,4) <@ int4range(1,7) → t |
anyelement <@ anyrange → boolean Is the element contained in the range? 42 <@ int4range(1,7) → f |
anyrange && anyrange → boolean Do the ranges overlap, that is, have any elements in common? int8range(3,7) && int8range(4,12) → t |
anyrange << anyrange → boolean Is the first range strictly left of the second? int8range(1,10) << int8range(100,110) → t |
anyrange >> anyrange → boolean Is the first range strictly right of the second? int8range(50,60) >> int8range(20,30) → t |
anyrange &< anyrange → boolean Does the first range not extend to the right of the second? int8range(1,20) &< int8range(18,20) → t |
anyrange &> anyrange → boolean Does the first range not extend to the left of the second? int8range(7,20) &> int8range(5,10) → t |
anyrange _- |
-_ anyrange → boolean Are the ranges adjacent? _numrange(1.1,2.2) - |
- numrange(2.2,3.3)_ → t |
anyrange + anyrange → anyrange Computes the union of the ranges. The ranges must overlap or be adjacent, so that the union is a single range (but see range_merge()). numrange(5,15) + numrange(10,20) → [5,20) |
anyrange * anyrange → anyrange Computes the intersection of the ranges. int8range(5,15) * int8range(10,20) → [10,15) |
anyrange - anyrange → anyrange Computes the difference of the ranges. The second range must not be contained in the first in such a way that the difference would not be a single range. int8range(5,15) - int8range(10,20) → [5,10) |
Table 9.56. Multirange Operators
Operator Description Example(s) |
anymultirange @> anymultirange → boolean Does the first multirange contain the second? '{[2,4)}'::int4multirange @> '{[2,3)}'::int4multirange → t |
anymultirange @> anyrange → boolean Does the multirange contain the range? '{[2,4)}'::int4multirange @> int4range(2,3) → t |
anymultirange @> anyelement → boolean Does the multirange contain the element? '{[2011-01-01,2011-03-01)}'::tsmultirange @> '2011-01-10'::timestamp → t |
anyrange @> anymultirange → boolean Does the range contain the multirange? '[2,4)'::int4range @> '{[2,3)}'::int4multirange → t |
anymultirange <@ anymultirange → boolean Is the first multirange contained by the second? '{[2,4)}'::int4multirange <@ '{[1,7)}'::int4multirange → t |
anymultirange <@ anyrange → boolean Is the multirange contained by the range? '{[2,4)}'::int4multirange <@ int4range(1,7) → t |
anyrange <@ anymultirange → boolean Is the range contained by the multirange? int4range(2,4) <@ '{[1,7)}'::int4multirange → t |
anyelement <@ anymultirange → boolean Is the element contained by the multirange? 4 <@ '{[1,7)}'::int4multirange → t |
anymultirange && anymultirange → boolean Do the multiranges overlap, that is, have any elements in common? '{[3,7)}'::int8multirange && '{[4,12)}'::int8multirange → t |
anymultirange && anyrange → boolean Does the multirange overlap the range? '{[3,7)}'::int8multirange && int8range(4,12) → t |
anyrange && anymultirange → boolean Does the range overlap the multirange? int8range(3,7) && '{[4,12)}'::int8multirange → t |
anymultirange << anymultirange → boolean Is the first multirange strictly left of the second? '{[1,10)}'::int8multirange << '{[100,110)}'::int8multirange → t |
anymultirange << anyrange → boolean Is the multirange strictly left of the range? '{[1,10)}'::int8multirange << int8range(100,110) → t |
anyrange << anymultirange → boolean Is the range strictly left of the multirange? int8range(1,10) << '{[100,110)}'::int8multirange → t |
anymultirange >> anymultirange → boolean Is the first multirange strictly right of the second? '{[50,60)}'::int8multirange >> '{[20,30)}'::int8multirange → t |
anymultirange >> anyrange → boolean Is the multirange strictly right of the range? '{[50,60)}'::int8multirange >> int8range(20,30) → t |
anyrange >> anymultirange → boolean Is the range strictly right of the multirange? int8range(50,60) >> '{[20,30)}'::int8multirange → t |
anymultirange &< anymultirange → boolean Does the first multirange not extend to the right of the second? '{[1,20)}'::int8multirange &< '{[18,20)}'::int8multirange → t |
anymultirange &< anyrange → boolean Does the multirange not extend to the right of the range? '{[1,20)}'::int8multirange &< int8range(18,20) → t |
anyrange &< anymultirange → boolean Does the range not extend to the right of the multirange? int8range(1,20) &< '{[18,20)}'::int8multirange → t |
anymultirange &> anymultirange → boolean Does the first multirange not extend to the left of the second? '{[7,20)}'::int8multirange &> '{[5,10)}'::int8multirange → t |
anymultirange &> anyrange → boolean Does the multirange not extend to the left of the range? '{[7,20)}'::int8multirange &> int8range(5,10) → t |
anyrange &> anymultirange → boolean Does the range not extend to the left of the multirange? int8range(7,20) &> '{[5,10)}'::int8multirange → t |
anymultirange _- |
-_ anymultirange → boolean Are the multiranges adjacent? _'{[1.1,2.2)}'::nummultirange - |
- '{[2.2,3.3)}'::nummultirange_ → t |
anymultirange _- |
-_ anyrange → boolean Is the multirange adjacent to the range? _'{[1.1,2.2)}'::nummultirange - |
- numrange(2.2,3.3)_ → t |
anyrange _- |
-_ anymultirange → boolean Is the range adjacent to the multirange? _numrange(1.1,2.2) - |
- '{[2.2,3.3)}'::nummultirange_ → t |
anymultirange + anymultirange → anymultirange Computes the union of the multiranges. The multiranges need not overlap or be adjacent. '{[5,10)}'::nummultirange + '{[15,20)}'::nummultirange → {[5,10), [15,20)} |
anymultirange * anymultirange → anymultirange Computes the intersection of the multiranges. '{[5,15)}'::int8multirange * '{[10,20)}'::int8multirange → {[10,15)} |
anymultirange - anymultirange → anymultirange Computes the difference of the multiranges. '{[5,20)}'::int8multirange - '{[10,15)}'::int8multirange → {[5,10), [15,20)} |
涉及到空区间或多区间时，左右/邻近运算符始终会返回 false；也就是说，空区间不会被视为在任何其他区间之前或之后。
The left-of/right-of/adjacent operators always return false when an empty range or multirange is involved; that is, an empty range is not considered to be either before or after any other range.
在其他地方，空区间和多区间会被视为加法单位：与空值联合的任何东西都是它本身。减去一个空值后的任何东西都是它本身。一个空的多区间与一个空区间具有完全相同的点。每个区间都包含空区间。每个多区间都包含任意数量的空区间。
Elsewhere empty ranges and multiranges are treated as the additive identity: anything unioned with an empty value is itself. Anything minus an empty value is itself. An empty multirange has exactly the same points as an empty range. Every range contains the empty range. Every multirange contains as many empty ranges as you like.
如果结果区间需要包含两个不重叠的子区间时，区间并集和差集运算符将失败，因为无法表示这样的区间。有单独的运算符用于并集和差集，它们采用多区间参数并返回一个多区间，即使其参数不相交，它们也不会失败。因此，如果你需要对可能不相交的区间进行并集或差集操作，你可以通过首先将区间转换为多区间来避免错误。
The range union and difference operators will fail if the resulting range would need to contain two disjoint sub-ranges, as such a range cannot be represented. There are separate operators for union and difference that take multirange parameters and return a multirange, and they do not fail even if their arguments are disjoint. So if you need a union or difference operation for ranges that may be disjoint, you can avoid errors by first casting your ranges to multiranges.
Table 9.57 显示可用于范围类型的函数。 Table 9.58 显示可用于多范围类型的函数。
Table 9.57 shows the functions available for use with range types. Table 9.58 shows the functions available for use with multirange types.
Table 9.57. Range Functions
Function Description Example(s) |
lower ( anyrange ) → anyelement Extracts the lower bound of the range (NULL if the range is empty or has no lower bound). lower(numrange(1.1,2.2)) → 1.1 |
upper ( anyrange ) → anyelement Extracts the upper bound of the range (NULL if the range is empty or has no upper bound). upper(numrange(1.1,2.2)) → 2.2 |
isempty ( anyrange ) → boolean Is the range empty? isempty(numrange(1.1,2.2)) → f |
lower_inc ( anyrange ) → boolean Is the range’s lower bound inclusive? lower_inc(numrange(1.1,2.2)) → t |
upper_inc ( anyrange ) → boolean Is the range’s upper bound inclusive? upper_inc(numrange(1.1,2.2)) → f |
lower_inf ( anyrange ) → boolean Does the range have no lower bound? (A lower bound of -Infinity returns false.) lower_inf('(,)'::daterange) → t |
upper_inf ( anyrange ) → boolean Does the range have no upper bound? (An upper bound of Infinity returns false.) upper_inf('(,)'::daterange) → t |
range_merge ( anyrange, anyrange ) → anyrange Computes the smallest range that includes both of the given ranges. range_merge('[1,2)'::int4range, '[3,4)'::int4range) → [1,4) |
Table 9.58. Multirange Functions
Function Description Example(s) |
lower ( anymultirange ) → anyelement Extracts the lower bound of the multirange (NULL if the multirange is empty has no lower bound). lower('{[1.1,2.2)}'::nummultirange) → 1.1 |
upper ( anymultirange ) → anyelement Extracts the upper bound of the multirange (NULL if the multirange is empty or has no upper bound). upper('{[1.1,2.2)}'::nummultirange) → 2.2 |
isempty ( anymultirange ) → boolean Is the multirange empty? isempty('{[1.1,2.2)}'::nummultirange) → f |
lower_inc ( anymultirange ) → boolean Is the multirange’s lower bound inclusive? lower_inc('{[1.1,2.2)}'::nummultirange) → t |
upper_inc ( anymultirange ) → boolean Is the multirange’s upper bound inclusive? upper_inc('{[1.1,2.2)}'::nummultirange) → f |
lower_inf ( anymultirange ) → boolean Does the multirange have no lower bound? (A lower bound of -Infinity returns false.) lower_inf('{(,)}'::datemultirange) → t |
upper_inf ( anymultirange ) → boolean Does the multirange have no upper bound? (An upper bound of Infinity returns false.) upper_inf('{(,)}'::datemultirange) → t |
range_merge ( anymultirange ) → anyrange Computes the smallest range that includes the entire multirange. range_merge('{[1,2), [3,4)}'::int4multirange) → [1,4) |
multirange ( anyrange ) → anymultirange Returns a multirange containing just the given range. multirange('[1,2)'::int4range) → {[1,2)} |
unnest ( anymultirange ) → setof anyrange Expands a multirange into a set of ranges. The ranges are read out in storage order (ascending). unnest('{[1,2), [3,4)}'::int4multirange) → [1,2) [3,4) |
对于空范围或多范围，lower_inc、upper_inc、lower_inf 和 upper_inf 函数都返回 false。
The lower_inc, upper_inc, lower_inf, and upper_inf functions all return false for an empty range or multirange.