Document toolboxDocument toolbox

Complex data types

[ 1 About complex types ] [ 2 Set ] [ 3 Map ] [ 4 Array ] [ 5 Tuple ] [ 6 JSON ]

About complex types

In Devo, complex data types provide flexible and powerful structures for querying and manipulatingdata logs. These types include set, map, array, and tuple, each with unique properties allowing users to efficiently handle diverse data operations. This article outlines the key characteristics and operations associated with each data type.

Type

Description

Example

Type

Description

Example

Set

Unordered, unique collection of values

{1, 2, 3}

Map

Key-value pairs, unique keys

{"a": 1, "b": 2}

Array

Ordered collection, allows duplicates

["item1", "item2", "item3"]

Tuple

Ordered, immutable collection, mixed types

(1.33, "text", true)

JSON

Key-value pairs, unordered, unique keys, mixed types

{"a": "b", "c": 1, "d" : [1,2,3], "e": {"f": 1}}

These complex data types provide the flexibility for handling structured and unstructured data in Devo, enabling efficient queries and operations across vast datasets. Users can effectively manipulate and analyze data within the Devo environment by leveraging sets, maps, arrays, and tuples.

The JSON type is not considered a complex type but it is included in this documentation for similarity purposes. Currently, it is not possible to cast JSON natively from/to a complex type.


Set

A set in Devo represents an unordered collection of unique values, meaning no duplicates are allowed. The order of elements in a set is not guaranteed, making it ideal for operations where uniqueness is more important than sequence.

Example:

{1, 2, 3}

Key Characteristics:

  • Unordered: Elements do not follow a specific order.

  • Unique: No duplicates are allowed.

Related operations:

Operation

Description

Syntax

Example

Result Value

Result Type

Operation

Description

Syntax

Example

Result Value

Result Type

mkset or {}

creates a set from elements

mkset (ele_1, ... ,ele_n)
{ele_1, ... , ele_n}

mkset (1.33,2,3,4)
mkset ("One", "two", "Number 3")
mkset (domain, responseTime)
mkset (responseTime, srcPort)

{1.33, 2, 3, 4}
{One, two, Number 3}
{One, 77}
{self, 12}
{12, 37406}

set(float)
set(str)
set(str)
set(str)
set(int)

isempty

Checks if a set is empty

isempty (set)

isempty ({1, 2, 3})
isempty ({ })

false
true

bool

length

Returns the length of a set

length (set)

length ({1, 20, 3, 4, 20, 20})

4

int

has (->)

Checks the occurrence of
a specified value in a
given set

has (set, ele)
set -> ele

has ({31, 2, 77}, 77)
{31, 2, 77} -> 77

true
true

bool

in (<-)

Checks the occurrence of
a specified value in a
given set

`in` (ele, set)
ele <- set

`in` (77, {30, 2, 77})
 77 <- {30, 2, 77}

true
true

bool

has (->)

Checks whether a set
contains another set

has (set_1, set_2)
set_1 -> set_2

 has ({1, 2 , 3, 4}, {2, 1})
 {1, 2 , 3, 4} -> {2, 1}

true
true

bool

in (<-)

Checks whether a set
contains another set

`in` (set_1, set_2)
set_1 <- set_2

`in` ({1, 2 , 3, 4}, {2, 1})
 {1, 2 , 3, 4} <- {2, 1}

false
false

bool

add (+)

Adds some value to a set

add (set, ele)
add (ele, set)

add ({30, "Peter", 77}, “Ann")
 {30, "Peter", 77} + “Ann"

{30, Peter, 77, Ann}
{30, Peter, 77, Ann}

set(str)
set(str)

add (+)

Union of two sets

add (set_1, set_2)
 set_1 + set_2

add ({5, "Peter", 77}, {77, "Ann", 400})
 {5, "Peter", 77} + {77, "Ann", 400}

{5, Peter, 77, Ann, 400]
{5, Peter, 77, Ann, 400]

set(str)
set(str)

join *

Joins the strings of a given
set in a single string
(separataror can be added)
NOTE : no order granted

join (set_str)
join (set_str, sep)

join ({1, "Peter", 77, 77, "Ann"})
join ({1, "Peter", 77, 77, "Ann"} , "_x_")

1Ann77Peter
1_x_Ann_x_77_x_Peter

str

band

Intersection of two sets

band (set_1, set_2)
 set_1 & set_2

band ({30, 2, 77}, {77, 400})
band ({30, 2, 77}, {500, 400})
 {30, 2, 77} & {77, 400}
 {30, 2, 77} & {500, 400}

{77}
{}
{77}
{}

set(int)
set(int)
set(int)
set(int)

sub (-)

Deletes a value from a set

sub (set, ele)
set - ele

sub ({30, 2, 77}, 400)
 {30, 2, 77} - 400
sub ({30, "Peter", 77}, "Peter")
 {30, "Peter", 77} - "Peter"

{30, 2, 77}
{30, 2, 77}
{30, 77}
{30, 77}

•set(int)
•set(int)
•set(str)
•set(str)

sub (-)

Difference of two sets

sub (set_1, set_2)
 set_1 – set_2

sub ({30, 2, 77}, {400, 30})
 {30, 2, 77} - 400
sub ({30, "Peter", 77}, {"Peter", 30})
 {30, "Peter", 77} - {"Peter", 30}

{2, 77}
{2, 77}
{77}
{77}

set(int)
set(int)
set(str)
set(str)

Common Operations:

from siem.logtrust.web.activity
  • Create a set from values:

    select mkset(srcHost, params) as mkset_string
  • Convert an array to a set:

  • Check if a set contains a specific value:

  • Add values to a set:

  • Get distinct values in a set out of a grouped field

  • Get distinct values in a set out of a grouped field

Test them together in Data Search


Map

A map in Devo is a collection of key-value pairs where each key is unique, but values can be duplicated. Maps allow for efficient lookups by keys and are useful when working with datasets that need to associate properties or attributes to specific items.

Example:

Key Characteristics:

  • Key-Value Structure: Each pair associates a key with a value.

  • Unique Keys: Each key must be unique, though values can repeat.

Related operations:

Operation

Description

Syntax

Example

Result Value

Result Type

Operation

Description

Syntax

Example

Result Value

Result Type

mkmap or { : }

Creates a map from
keys and values

mkmap (key_1, val_1, ... , key_n, val_n)

mkmap (10, 1001, 20, 1.33)
mkmap (10, 1001, 20, "Yesterday")
mkmap ("One", 100, "Two", 200)
 { 1 : "Hello", 2 : "Bye" }

{10:1001, 20:1.33}
{10=1001, 20=Yesterday}
{One=100, Two=200}
{1=Hello, 2=Bye}

map(int, float)
map(int, str)
map(str, int)
map(int, str)

collectdistinct() (aggr)

Creates a map of distinct values after an aggregation

collectdistinct(aggr_field)

{{1, 3, 20, 4}}

map(int, int,int,int)

isempty

Checks if a map is empty

isempty (map)

isempty ({1:10, 2:20, 3:30})
isempty ({})

false
true

bool

length

Returns the length of a

map

length (map)

length ({1:"a", 2:"b", 3:"c"})

3

int

has (->)

Checks the
occurrence of a
specified key in a
given map

has (map, key)
map -> key

has ({1:"aaa", 2:"bbb", 77:"ccc"}, 77)
 {1:"aaa", 2:"bbb", 77:"ccc"} -> 77

true
true

bool

in (<-)

Checks the
occurrence of a
specified key in a
given map

`in` (key, map)
key <- map

`in` (77, {1:"aaa", 2:"bbb", 77:"ccc"})
 77 <- {1:"aaa", 2:"bbb", 77:"ccc"}

true
true

bool

add (+)

Adds some key-value
pair to a map

add (map, key, value)

add ({5:“Peter", 10:“Ann"}, 4, “xxx")

{5=Peter, 10=Ann, 4=xxx}

map(int, str)

add (+)

Concatenation of two
maps

add (map_1, map_2)
 map_1 + map_2

add ({5:"Peter", 10:"Ann"}, {4:"xxx"})
 {5:"Peter", 10:"Ann"} + {4:"xxx"}

{5=Peter, 10=Ann, 4=xxx}
{5=Peter, 10=Ann, 4=xxx}

map(int, str)
map(int, str)

at [ ]

Given a map, returns the
value associated to a
given key

at (map, key)
 map [key]

at ({"Hello":77, 4:4000}, "Hello")
 {"Hello":77, 4:4000} ["Hello"]
at ({1:"One", 2:"Two"}, 888)
 {1:"One", 2:"Two"} [888]

77
77
null
null

int
int
str
str

band

Intersection of two maps :
•Non-commutative
•Intersection by key
•Returns first map value

band (map_1, map_2)
 map_1 & map_2

band ({5:"Five", 2:"Two"}, {2:"xxx"})
 {5:"Five", 2:"Two"} & {2:"xxx"}
band ({5:"Five", 2:"Two"}, {6:“Six"})
 {5:"Five", 2:"Two"} & {6:“Six"}

{2=Two}
{2=Two}
{}
{}

map(int, str)
map(int, str)
map(int, str)
map(int, str)

band

Restrict a map to a set of
keys

band (map, set)
 map & set

band ({5:"Five", 2:"Two", 6:"Six"}, {6, 2})
 {5:"Five", 2:"Two", 6:"Six"} & {6, 2}
band ({5:"Five", 2:"Two"}, {10, 20})
 {5:"Five", 2:"Two"} & {10, 20}

{2=Two, 6=Six}
{2=Two, 6=Six}
{}
{}

map(int, str)
map(int, str)
map(int, str)
map(int, str)

sub (-)

Deletes the key-value pair
defined by a key

sub (map, key)
 map - key

sub ({5:"Five", 2:"Two", 6:"Six"}, 5)
 {5:"Five", 2:"Two", 6:"Six"} - 5}

{2:“Two", 6:“Six"}
{2:“Two", 6:“Six"}

map(int, str)
map(int, str)

sub (-)

Deletes the key-value pairs
defined by a set of keys

sub (map, set)
 map – set

sub ({5:"Five", 2:"Two", 6:"Six"}, {2, 5, 10})
 {5:"Five", 2:"Two", 6:"Six"} - {2, 5, 10}

{6=Six}
{6=Six}

map(int, str)
map(int, str)

sub (-)

Difference of two maps
(difference by key)

sub (map_1, map_2)
 map_2 – map_2

sub ({5:"Five", 2:"Two", 6:"Six"} {5:“xxx"})
 {5:"Five", 2:"Two", 6:"Six"}- {5:“xxx"}

{2:“Two", 6:“Six"}
{2:“Two", 6:“Six"}

map(int, str)
map(int, str)

keys

Returns the set of keys of a given map

keys (map)

keys ({"Hello":77, 4:4000})
keys ({99:"Monday", 4:4000})

{Hello, 4}
{99, 4}

set(str)
set(int)

values

Returns the array of values of a given map
NOTE : no order granted

values (map)

values ({"Hello":77, 4:4000})
values ({99:"Monday", 4:4000})

[77, 4000]
[4000, Monday]

array(int)
array(str)

 

Common Operations:

  • Create a map:

  • Check if a map is empty:

  • Get the length of a map (number of key-value pairs):

  • Access a value by key:

  • Append new key-value pairs:

  • Subtract pairs

  • Return the value of a give key

  • Return all keys or values from a map:

     

Test them together in Data Search


Array

An array is an ordered collection of elements where duplicates are allowed. Arrays are especially useful for performing a range of operations such as sorting, summing, and filtering. Unlike sets, arrays preserve the order of elements, making them essential for ordered data manipulation.

Example:

Key Characteristics:

  • Ordered: Elements follow a specific sequence.

  • Allows Duplicates: Identical elements can appear multiple times.

Related operations:

Operation

Description

Syntax

Example

Result Value

Result Type

Operation

Description

Syntax

Example

Result Value

Result Type

mkarray or []

Creates an array from elements

mkarray (ele_1, … ,ele_n)
[ele_1, … , ele_n]

mkarray (1.33,2,3,4) mkarray (“One", "two", "Number 3“) 

[ “One", 77 ]

mkarray (domain, responseTime)

[1.33, 2, 3, 4]
[One, two, Number 3]
[One, 77] [self, 12]

array(float) array(str) array(str) array(str)

split

Creates an array by splitting a string field

split (string, string)

split("1.2.3.4", ".")

[1, 2, 3, 4]

array

isempty

Checks if an array is empty

isempty (array)

isempty ([1, 2, 3]) isempty ([ ])

false true

bool

length

Returns the length of an array

length (array)

length ([1, 2, 3, 4, 5, 6])

6

Int

has (->)

Checks the occurrence of a specified value in a given array

•has (array, ele) •array -> ele

has ([30, 2, 77], 77)

[30, 2, 77] -> 77

true

true

bool

in (<-)

Checks the occurrence of a specified value in a given array

`in` (ele, array) ele <- array

`in` (77, [30, 2, 77]) 

77 <- [30, 2, 77]

true

true

bool

add (+)

Appends some value to the end of an array

add (array, ele)
array + ele

add ([30, "Peter", 77], "Ann")
 [30, "Peter", 77] + "Ann"

[30, Peter, 77, Ann]
[30, Peter, 77, Ann]

array(str)
array(str)

add (+)

Concatenates two arrays

add (array_1, array_2)
array_1 + array_2

add ([5, "Peter", 77], [77, "Ann", 400])
[5, "Peter", 77] + [77, "Ann", 400]

[5, Peter, 77, 77, Ann, 400]
[5, Peter, 77, 77, Ann, 400]

array(str)
array(str)

join

Joins the strings of a given
array in a single string
(separataror can be added)

join (array_str)
join (array_str, sep)

join ([1, "Peter", 77, 77, "Ann"])
join ([1, "Peter", 77, 77, "Ann"], “_x_”)

1Peter7777Ann
1_x_Peter_x_77_x_77_x_Ann

str

at or [][]

Returns the n-th element of
an array

at (array, n)
array [n]

at ([10, 20, 3.33, 40], 1)
[10, 20, 3.33, 40] [2]
at ([10, 20, 3.33, "HELLO"], 3)
at ([10, 20, 333, 40], 777)

20
3.33
HELLO
null

float
float
str
int

at or [][]

Returns a subarray between
two indices

at (array, from, to)
array [from, to]

at ([10, 20, 30, 40], 1, 3)
 ["A", "B", "C", "D"] [1, 3]

[20, 30]
["B", "C"]

array(int)
array(str)

indexof

Returns the index of the first
appearance of a value in
an array (-1 if not present)
starting at 0

indexof (array, ele)

indexof ([10, 20, 3.33, 40], 3.33)
indexof ([10, 20, 333, 40], 777)
indexof (["Hello", "my friend"],"Hello")

2
-1
0

int

dropnulls

Drops null elements of a given array

dropnulls (array)

dropnulls ([3, null, 77, null])
dropnulls ([3, null, "HELLO", null])

[3, 77]
[3, HELLO]

array(int)
array(str)

sort

Sorts the elements of a
given array

sort (array)

sort ([3, 77, 30, 1)
sort (["AAA", "aaa", "44", 5])

[1, 3, 30, 77]
[44, 5, AAA, aaa]

array(int)
array(str)

reverse

Reverses the elements of
a given array

reverse
(array)

reverse ([3, 77, 30, 1)
reverse (["AAA", "aaa", "44", 5])

[1, 30, 77, 3]
[5, 44, aaa, AAA]

array(int)
array(str)

sum

Sums the values of a
numeric array

sum (array)

sum([1,2,3,4])
sum([1,2,3.5,4.5])

10
11

Int
float

 

Common Operations:

  • Create an array:

  • Convert a set to an array:

  • Check if an array contains a value:

  • Add a value to an array:

  • Reverse an array:

  • Drop null values from an array:

  • Sum numeric arrays:

  • length of an array

     

Test them together in Data Search


Tuple

A tuple in Devo is an ordered collection of elements that can be of mixed types (e.g., strings, numbers, booleans). Unlike arrays, tuples maintain the individual types of each element and are immutable, meaning that once created, the elements cannot be changed.

Example:

Key Characteristics:

  • Ordered: Elements are stored in a specific sequence.

  • Immutable: The values in a tuple cannot be modified after creation.

  • Allows Duplicates: Identical elements can appear more than once.

  • Allows Mixed Types: Elements can be of any type (e.g., string, integer, boolean).

Related operations: 

Operation

Description

Syntax

Example

Result Value

Result Type

Operation

Description

Syntax

Example

Result Value

Result Type

mktuple

Creates a tuple with specified elements.

mktuple(elem_1, elem_2, ..., elem_n)

mktuple(username, ip4(srcHost), mm2coordinates(ip4(srcHost)), true)

(username, ip4, coordinates, true)

tuple(varied types)

Tuple literal

Creates a tuple using parentheses.

(elem_1, elem_2, ..., elem_n)

(username, srcPort, ip4(srcHost), true)

(username, srcPort, ip4, true)

tuple(varied types)

Access by index

Access tuple elements using index. Supports negative indexing.

tuple[index]

tuple[0]

tuple[-1]

First element

Last element

Matches element type

at()

Access tuple elements by index using a function.

at(tuple, index)

at(tuple, 0)

First element

Matches element type

Filter by type

Filters tuple elements based on data type or pattern.

where tuple[index] -> type

where tuple[0] -> "@"

Matches filtered criteria

bool

Filter by value

Filters tuple elements based on value comparison.

where tuple[index] in/not in (values)

where tuple[-1] not in (ip4(95.63.39.51))

Matches filtered criteria

bool

Sub-query (key lookup)

Finds the occurrence of a key in another dataset during the same period.

(from dataset group by key select key) -> key

(from siem.logtrust.web.navigation group by srcHost select srcHost) -> srcHost

Matched key occurrence

Matches key type

Sub-query (field lookup)

Returns a field from another table matching a condition.

(from dataset group by field1, field2)[key]

(from siem.logtrust.web.navigation group by userEmail, origin)[username]

origin field value

Matches field type

Sub-query (tuple matching)

Returns a tuple from another table matching specified conditions.

(from dataset group by tuple_field select fields)[key1, key2]

(from siem.logtrust.web.navigation group by userEmail, level select userEmail, count())[username, level]

(userEmail, count)

tuple(varied types)

Common Operations:

  • Create a tuple:

  • Access tuple elements by index:

  • Filter tuples by type:

  • SUB-QUERY: find the occurrence of a specific IP in another table during the same period of time

  • SUB-QUERY: return the "origin" field in another table matching by user email

  • SUB-QUERY: return the tuple (userEmail, count()) from another table matching by the tuple (email, level)

Test them together in Data Search


JSON

Then JSON format is fully supported in Devo. It is useful for representing structured data in an unordered format, enabling flexible data storage and manipulation.

  • In Data Search: you can use the "pretty-print" view to inspect the content of JSON cells and extract the values that you need. How?

Example:

Key Characteristics:

  • Unordered: The sequence of keys is not guaranteed.

  • Unique Keys: Each key must be unique within the object.

  • Mixed Types: Values can be strings, numbers, booleans, arrays, or other JSON objects.

  • Flexible: Supports nested objects and arrays for complex structures.

 

Operation

Description

Syntax

Example

Result

Result Type

Operation

Description

Syntax

Example

Result

Result Type

jqeval()

Extracts values from JSON objects using a jq filter.

jqeval(jq_filter, json_field)

jqeval(jqcompile(".email"), json)

Extracted value from JSON

Matches field type

jqcompile()

Compiles a string as a jq filter to be used in jqeval.

jqcompile(filter)

jqcompile(".email")

A compiled jq filter

jq filter

label()

Returns the data type of a value inside a JSON object.

label(json_field)

label(json.email)

"string"

string

jsonparse()

Converts a string field to a JSON data type.

jsonparse(string_field)

jsonparse(data_string)

JSON object

json

str()

Converts a JSON object or other data type to a string.

str(value)

str(jqeval(jqcompile(".email"), json))

Extracted value as a string

string

at() []

Given a map, returns the
value associated to a
given key

at (json, key)
 json [key]

json["int"]

at(json, "int")

Extract a value from a key

Matches field type

stringify()

Converts a JSON object (json data type) into string data type.

stringify(json_type)

stringify(json_type)

{“a“: 1, “b“: ”c”}

string

Common Operations:

  • Parse String to JSON

  • Extract a Property Value

  • Convert a JSON to a string

  • Determine Data Type of a JSON Value

Test them together in Data Search