than 5;

*`similar_product_ids.# = "0684824396"` – array "similar\_product\_ids"

contains string "0684824396".

*`*.color = "red"` – there is object somewhere which key "color" has value

"red".

*`*.color = "red"` – there is object somewhere which key "color" has value "red".

*`foo = *` – key "foo" exists in object.

Path selects set of JSON values to be checked using given operators. In

the simplest case path is justan key name. In general path is key names and

placeholders combined by dot signs. Path can use following placeholders:

Path selectsaset of JSON values to be checked using given operators. In

the simplest case path is justa key name. In general path is key names and

placeholders combined by dot signs. Path can usethefollowing placeholders:

*`#` – any index of array;

*`#N` – N-th index of array;

*`%` – any key of object;

*`#` – any index ofanarray;

*`#N` – N-th index ofanarray;

*`%` – any key ofanobject;

*`*` – any sequence of array indexes and object keys;

*`@#` – length of array or object,could beonly used as last component of

path;

*`$` – the whole JSON document as single value,could beonly the whole path.

*`@#` – length of array or object,mayonlybeused as the last component of

apath;

*`$` – the whole JSON document as single value,mayonly be the whole path.

Expression is true when operator is true against at least one value selected

by path.

Key names could be given either with or without double quotes. Key names

without double quotesshouldn'tcontain spaces, start with number orconcur

with jsquery keyword.

without double quotesmay notcontain spaces, start withanumber ormatch

a jsquery keyword.

The supported binary operators are:

* Check for type operators:`IS ARRAY`,`IS NUMERIC`,`IS OBJECT`,`IS STRING`

and`IS BOOLEAN`.

Expressionscould be complex. Complex expression is a set of expressions

Expressionscan be complex. Complex expression is a set of expressions

combined by logical operators (`AND`,`OR`,`NOT`) and grouped using braces.

Examples of complex expressions are given below.

Examples of complex expressions:

Prefix expressions are expressions given in the form path (subexpression).

In this case path selects JSON values to be checked using given subexpression.

Prefix expressions are expressions given in the form`path (subexpression)`.

In this case path selects JSON values to be checked usingthegiven subexpression.

Check results are aggregated in the same way as in simple expressions.

*`#(a = 1 AND b = 2)` – exists element of array which a key is 1 and b key is 2

*`%($ >= 10 AND $ <= 20)` – exists object key which values is between 10 and 20

Path alsocouldcontain following special placeholders with "every" semantics:

Pathcanalso contain the following special placeholders with "every" semantics:

*`#:` – everyindexes of array;

*`%:` – every key of object;

*`#:` – everyindex of an array;

*`%:` – every key ofanobject;

*`*:` – every sequence of array indexes and object keys.

Consider following example.

This example could be read as following: there is at least one keywhich value

is array of numerics between 0 and 1.

This example could be read as following: there is at least one keywhose value

isanarray of numerics between 0 and 1.

We can rewrite this example in the following form with extra braces.

We can rewrite this example in the following form with extra braces:

The first placeholder`%` checks that expression in braces is true for at least

one value in object. The second placeholder`#:` checks valueto be array and

all its elements satisfy expressions in braces.

The first placeholder`%` checks thattheexpression in braces is true for at least

one value intheobject. The second placeholder`#:` checksif thevalueis an array

and thatall its elements satisfy the expressions in braces.

We can rewrite this example without`#:` placeholder as follows.

We can rewrite this example withoutthe`#:` placeholder as follows:

In this example we transform assertion that every element of array satisfy some

condition to assertion that thereis noone elementwhichdoesn't satisfy the

samecondition.

In this example we transformtheassertion that every element of array satisfy some

condition toanassertion that thereare noelementswhichdon't satisfy the same

condition.

Some examples of using paths are given below.

Some examples of using paths:

*`numbers.#: IS NUMERIC` – every element of "numbers" array is numeric.

*`*:($ IS OBJECT OR $ IS BOOLEAN)` – JSON is a structure of nested objects

with booleans as leaf values.

*`#:.%:($ >= 0 AND $ <= 1)` – each element of array is object containing

*`#:.%:($ >= 0 AND $ <= 1)` – each element of array isanobject containing

only numeric values between 0 and 1.

*`documents.#:.% = *` – "documents" is array of objects containing at least

*`documents.#:.% = *` – "documents" isanarray of objects containing at least

one key.

*`%.#: ($ IS STRING)` – JSON object contains at least one array of strings.

*`#.% = true` – at least one array element isobjects which contains at least

*`#.% = true` – at least one array element isan object which contains at least

one "true" value.

Usage of path operators and braces need some explanation. When same path

operators are used multiple times they may refer different values while you can

refer same value multiple time by using braces and`$` operator. See following

examples.

The use of path operators and braces need some further explanation. When the same path

operators are used multiple times, they may refer to different values. If you want them

to always refer to the same value, you must use braces and the`$` operator. For example:

*`# < 10 AND # > 20` –exists element less than 10andexists another element

greater than 20.

*`#($ < 10 AND $ > 20)` –exists element which both less than 10 and greater

than 20 (impossible).

*`#($ >= 10 AND $ <= 20)` –exists element between 10 and 20.

*`# >= 10 AND # <= 20` –exists elementgreat or equal to 10 andexists

anotherelement less or equal to 20. Query can be satisfied by array with

no elements between 10 and 20, for instance[0,30].

*`# < 10 AND # > 20` –an element less than 10 exists, and another element

greater than 20 exists.

*`#($ < 10 AND $ > 20)` –an element which is both less than 10 and greater

than 20exists(impossible).

*`#($ >= 10 AND $ <= 20)` –an element between 10 and 20 exists.

*`# >= 10 AND # <= 20` –an elementgreater or equal to 10exists,andanother

element less or equal to 20 exists. Please note that this query also can be

satisfied by an array withno elements between 10 and 20, for instance[0,30].

Same rules apply whenyou searchinside objects andbranchy structures.

Same rules apply whensearchinginside objects andbranch structures.

Type checking operators and "every" placeholders are useful for document

schema validation. JsQuery matchig operator`@@` is immutable and can be used

points.#:(x IS NUMERIC AND y IS NUMERIC)'::jsquery));

In this example check constraint validates that in "data" jsonb column:

value of "name" key is string, value of "similar_ids" key is array of numerics,

value of "points" key is array of objects which contain numeric values in

In this examplethecheck constraint validates that inthe"data" jsonb column

thevalue ofthe"name" key isastring,thevalue ofthe"similar_ids" key is an array of numerics,

and thevalue ofthe"points" key is an array of objects which contain numeric values in

"x" and "y" keys.

See our

* jsonb\_value\_path\_ops

In each of two GIN opclasses jsonb documents are decomposed into entries. Each

entry is associated with particular value andit's path.Difference between

entry is associated withaparticular value andits path.The difference between

opclasses is in the entry representation, comparison and usage for search

optimization.

For example, jsonb document

For example,thejsonb document

would be decomposed into following entries:

* "d"."e".#.7

* "d"."e".#.false

Since JsQuery doesn't supportsearch in particular array index, we consider

Since JsQuery doesn't supportsearching in a particular array index, we consider

all array elements to be equivalent. Thus, each array element is marked with

same`#` sign inthe path.

thesame`#` sign inits path.

Major problem in the entries representation is its size. In the given example

key "a" is presented three times. In the large branchy documents with long

keyssize of naive entriesrepresentation becomesunreasonable. Both opclasses

address this issue but inaslightly differentway.

thekey "a" is presented three times. In large branchy documents with long

keyssizes of naive entries, therepresentation becomesunreasonably large.

Both opclassesaddress this issue, but in slightly differentways.

jsonb\_path\_value\_ops represents entry as pair of path hash and value.

Following pseudocode illustrates it.

Following pseudocode illustrates it:

In comparisonofentriespath hash is the higher part of entry and value is

its lower part. This determines the features of this opclass. Since path

is hashed and it is higher part of entry we need to know the full path to

the value in order to use it forsearch. However, once path is specified

When comparison entries, thepath hash is the higher part of entry and the value is

the lower part. This determines the features of this opclass. Since the path

is hashed and it's the higher part oftheentry, we need to know the full path to

a value in order to usetheit forsearching. However, once the path is specified

we can use both exact and range searches very efficiently.

jsonb\_value\_path\_ops represents entry as pair of value and bloom filter

ofpath.

jsonb\_value\_path\_ops represents entry as pair ofthevalue and a bloom filter

ofpaths:

In comparison of entries value is the higher part of entry and bloom filter of

path is its lower part. This determines the features of this opclass. Since

value is the higher part of entry we canperformonly exact value search

efficiently. Range valuesearch is possible as well but we would have to

filter all the the different paths where matching values occur. Bloom filter

over path items allows indexusage for conditions containing`%` and`*` in

thevalue is the higher part ofanentry, we can only perform exact value search

effectively. Asearchover a range of valuesis possible as well, but we have to

filter all the the different paths where matching values occur.TheBloom filter

over path items allowstheindexto be used for conditions containing`%` and`*` in

their paths.

JsQuery opclasses perform complex query optimization.Thus it's valuable for

JsQuery opclasses perform complex query optimization.It's valuable for a

developer or administrator to see the result of such optimization.

Unfortunately, opclasses aren't allowed todo any custom outputto the

EXPLAIN. That's why JsQuery providesfollowing functionswhich allows to see

how particular opclass optimizes given query.

Unfortunately, opclasses aren't allowed toput any custom outputin an

EXPLAIN. That's why JsQuery providesthese functionsto let you see

how particular opclass optimizes given query:

* gin\_debug\_query\_path\_value(jsquery) – for jsonb\_path\_value\_ops

* gin\_debug\_query\_value\_path(jsquery) – for jsonb\_value\_path\_ops

Resultof these functions is a textual representation of query tree which

leafs are GIN search entries. Following examples show different results of

query optimization by different opclasses.

The resultof these functions is a textual representation ofthequery tree

where leaves are GIN search entries. Following examples show different results of

query optimization by different opclasses:

Unfortunately, jsonb have no statistics yet. That's why JsQuery optimizer has

to do imperative decision while selecting conditions to be evaluated using

index. This decision is made byassumtion that some condition types are less

selective than others.Optimizer divides conditions into following selectivity

class (listedby descending of selectivity).

index. This decision is made byassuming that some condition types are less

selective than others.The optimizer divides conditions into following selectivity

classes (listedin descendingorderof selectivity):

1. Equality (x = c)

2. Range (c1 < x < c2)

3. Inequality (x > c)

4. Is (x is type)

5. Any (x =\*)

Optimizer evades index evaluation of less selective conditions when possible.

The optimizer avoids index evaluation of less selective conditions when possible.

For example, in the`x = 1 AND y > 0` query`x = 1` is assumed to be more

selective than`y > 0`. That's why index isn't used for evaluation of`y > 0`.

selective than`y > 0`. That's whytheindex isn't used for evaluation of`y > 0`.

With lack of statistics decisions made by optimizer can be inaccurate. That's

why JsQuery supports hints.Comments`/*-- index */`and`/*-- noindex */`

placed in the conditionsforcesoptimizer to useand not use index

correspondingly.

Withthelack of statistics, decisions made by optimizer can be inaccurate. That's

why JsQuery supports hints.The comments`/*-- index */`or`/*-- noindex */`

placed in the conditionsforce theoptimizer to useor not use an index

correspondingly:

Contribution

Please, notice, that JsQuery is still under development and while it's

stable and tested, it maycontains some bugs. Don't hesitate toraise

Please note that JsQuery is still under development. While it's

stable and tested, it maycontain some bugs. Don't hesitate tocreate

[issues at github](https://github.com/postgrespro/jsquery/issues) with your

bug reports.

If you're lacking of some functionality in JsQuery andfeeling power to

implement it then you're welcome to make pull requests.

Ifthere's some functionalityyou'd like to see added to JsQuery andyou feel

like you canimplement it, then you're welcome to make pull requests.