Skip to content

Specify function operations in JSON

License

Notifications You must be signed in to change notification settings

cedar-team/json-operations

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

43 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

json-operations

Specify complex conditional operations in JSON. The conditional operations can be run securely against a JSON value (Python dictionary) and will return a boolean.

from json_operations import execute
operations = [">", ["key", "items"], 30]
data1 = {
    "items": 31
}
data2 = {
    "items": 29
}

execute(operations, data1) # -> True
execute(operations, data2) # -> False

Security

All operations are safe (no use of eval). It's good to enforce a length limit if you are taking input from an untrusted source.

API

execute

Run the json operations against a data dictionary. Return True/False. Can raise a JsonOperationError if operations are invalid or keys don't exist

from json_operations import execute

execute(<operations>, <data_dictionary>) -> bool

get_json_schema

Returns the JSON Schema for json operations. This is useful for validating operations before running them

from json_operations import get_json_schema

get_json_schema() -> Dict

get_keys

Returns the keys inside json operations. This is useful for getting a list of keys necessary for the json operations and validating that all keys are inside the data dictionary

from json_operations import get_keys

get_keys(<operations>) -> List[Dict] 
# [{"name": "key1", "type": "number", "index": 0}, ...]

Operators

== (Equal operator)

Check whether one value equal to another.

Syntax
["==", <operator_or_literal>, <operator_or_literal>]
Example
from json_operations import execute
operations = ["==", ["key", "items"], 30]
data1 = {
    "items": 30
}
data2 = {
    "items": 31
}

execute(operations, data1) # -> True
execute(operations, data2) # -> False

!= (Not equal operator)

Check whether one value is not equal to another.

Syntax
["!=", <operator_or_literal>, <operator_or_literal>]
Example
from json_operations import execute
operations = ["!=", ["key", "items"], 30]
data1 = {
    "items": 30
}
data2 = {
    "items": 31
}

execute(operations, data1) # -> False
execute(operations, data2) # -> True

> (Greater than operator)

Check whether one value is greater than another.

Syntax
[">", <operator_or_literal>, <operator_or_literal>]
Example
from json_operations import execute
operations = [">", ["key", "items"], 30]
data1 = {
    "items": 31
}
data2 = {
    "items": 29
}

execute(operations, data1) # -> True
execute(operations, data2) # -> False

>= (Greater than or equal operator)

Check whether one value is greater than or equal to another.

Syntax
[">=", <operator_or_literal>, <operator_or_literal>]
Example
from json_operations import execute
operations = [">=", ["key", "items"], 30]
data1 = {
    "items": 30
}
data2 = {
    "items": 29
}

execute(operations, data1) # -> True
execute(operations, data2) # -> False

< (Less than operator)

Check whether one value is less than another.

Syntax
["<", <operator_or_literal>, <operator_or_literal>]
Example
from json_operations import execute
operations = ["<", ["key", "items"], 30]
data1 = {
    "items": 31
}
data2 = {
    "items": 29
}

execute(operations, data1) # -> False
execute(operations, data2) # -> True

< (Less than or equal operator)

Check whether one value is less or equal to than another.

Syntax
["<=", <operator_or_literal>, <operator_or_literal>]
Example
from json_operations import execute
operations = ["<=", ["key", "items"], 30]
data1 = {
    "items": 31
}
data2 = {
    "items": 30
}

execute(operations, data1) # -> False
execute(operations, data2) # -> True

in (In operator)

Check whether one value is contained in another.

Syntax
["in", <operator_or_literal>, <operator_or_literal>]
Example
from json_operations import execute
operations = ["in", "my_type",  ["key", "types"]]
data1 = {
    "types": [
        "type1", "type2"
    ]
}
data2 = {
    "types": [
        "my_type", "type1"
    ]
}

execute(operations, data1) # -> False
execute(operations, data2) # -> True

!in (Not In operator)

Check whether one value is NOT contained in another.

Syntax
["!in", <operator_or_literal>, <operator_or_literal>]
Example
from json_operations import execute
operations = ["!in", "my_type",  ["key", "types"]]
data1 = {
    "types": [
        "type1", "type2"
    ]
}
data2 = {
    "types": [
        "my_type", "type1"
    ]
}

execute(operations, data1) # -> True
execute(operations, data2) # -> False

btw (Between operator)

Check whether one value is between 2 other values. Equivalent to

low <= val <= high

Both low and high value are included.

Syntax
["btw", <value_operator_or_literal>, [<low_operator_or_literal>, <high_operator_or_literal>]]
Example
from json_operations import execute
operations = ["btw", ["key", "val"],  [1, 3]]
data1 = {
    "val": 2
}
data2 = {
    "val" : 4
}

execute(operations, data1) # -> True
execute(operations, data2) # -> False

& (Intersection operator)

Check whether 2 arrays have any members in common

Syntax
["&", <value_operator_or_literal>, <value_operator_or_literal>]
Example
from json_operations import execute
operations = ["&", ["key", "val"],  [1, 2]]
data1 = {
    "val": [2,3]
}
data2 = {
    "val" : [3,4]
}

execute(operations, data1) # -> True
execute(operations, data2) # -> False

!& (Not Intersection operator)

Check whether 2 arrays have no members in common

Syntax
["!&", <value_operator_or_literal>, <value_operator_or_literal>]
Example
from json_operations import execute
operations = ["!&", ["key", "val"],  [1, 2]]
data1 = {
    "val": [2,3]
}
data2 = {
    "val" : [3,4]
}

execute(operations, data1) # -> False
execute(operations, data2) # -> True

and (And operator)

Check whether all values are True. and supports nesting other operation inside it (See complex example).

Syntax
["and", ...<operators_or_literals>]
Examples
from json_operations import execute
operations = ["and", ["key", "a"],  ["key", "b"]]
data1 = {
    "a": True,
    "b": False,
}
data2 = {
    "a": True,
    "b": True,
}

execute(operations, data1) # -> False
execute(operations, data2) # -> True

or (Or operator)

Check whether any values are True. or supports nesting other operation inside it (See complex example).

Syntax
["or", ...<operators_or_literals>]
Examples
from json_operations import execute
operations = ["or", ["key", "a"],  ["key", "b"]]
data1 = {
    "a": False,
    "b": False,
}
data2 = {
    "a": False,
    "b": True,
}

execute(operations, data1) # -> False
execute(operations, data2) # -> True

Complex Operator Example

from json_operations import execute

# (a > 1 or b < 5) and (c == 0 or d != 10) and (val in e or f >= 0)
operations = ["and", 
                ["or", [">", ["key", "a"], 1], ["<", ["key", "b"], 5]], 
                ["or", ["==", ["key", "c"], 0], ["!=", ["key", "d"], 10]],
                ["or", ["in", "val",  ["key", "e"]], [">=", ["key", "f"], 0]],
              ]
data1 = {
    "a": 0,
    "b": 5,
    "c": 0,
    "d": 9,
    "e": ["val", "another_val"],
    "f": 0,
}
data2 = {
    "a": 0,
    "b": 4,
    "c": 0,
    "d": 9,
    "e": ["val", "another_val"],
    "f": -1,
}
# (0 > 1 or 5 < 5) and (0 == 0 or 9 != 10) and (val in ["val", "another_val"] or 0 >= 0)
execute(operations, data1) # -> False

# (0 > 1 or 4 < 5) and (0 == 0 or 9 != 10) and (val in ["val", "another_val"] or -1 >= 0)
execute(operations, data2) # -> True

Differences from json-logic

https://jsonlogic.com/

  • Operations are always type safe (cannot compare different types). json-logic will automatically convert different types and compare them, which can lead to issues that are tough to find
  • Invalid operations are errors instead of ignored. json-logic tries to avoid all errors and instead continues with unexpected input. This can lead to issues
  • More compact and intuitive syntax:
["or", ["key", "a"], ["key", "b"]]

vs

{"or": [{"var": ["a"]}, {"var": ["b"]}]}
  • Only supports boolean logic. json-logic supports all types of operations (addition, subtraction, etc.), not just boolean logic