FILTER and REJECT

When you're building workflows that process arrays or objects, you often need to narrow down your data. Maybe you only want high-severity alerts, or you need to exclude test accounts from a user list. Instead of manually picking through data, Tines gives you two functions that work like opposites: FILTER and REJECT.

What FILTER does 

FILTER examines every item in an array or every key-value pair in an object and keeps only the items where your condition returns true. Everything else gets removed. Think of it like a security checkpoint: you define the criteria for what gets through, and FILTER only lets matching items pass.

FILTER basic syntax 

Both functions take two arguments: The data you want to filter (an array or object) and the criteria for filtering (either a LAMBDA function or an array of values):

FILTER(array | object, values_to_keep | LAMBDA(arg1, [arg2], expr))

FILTER example: Survey scores 

Let's say you have a surveys array of survey responses and you only want the higher-scoring ones (higher than 7):

To only keep surveys with a score greater than 7, we utilize the FILTER function:

FILTER(action_name.surveys, LAMBDA(survey, survey.score > 7))

• action_name.surveys: The upstream reference path to the target surveys array.

• survey: Represents each individual item (iteration) in the surveys array as FILTER processes it.You declare this variable, so name it something that will help you and others better understand the overall formula.

• survey.score: The reference path to the score value of the survey item (iteration).

Translation: Go through each object within the surveys array, check the score value, and keep the entire object if its score value is greater than 7.

The output: Only the surveys that matched our condition were kept:

[
  {"id": "S-001", "score": 9},
  {"id": "S-003", "score": 8}
]

What REJECT does 

REJECT examines every item in an array or every key-value pair in an object and removes the items where your condition returns true. Everything else gets kept. Think of it like a blocklist: You define what should be removed, and REJECT filters those items out, keeping everything else.

REJECT basic syntax 

Both functions take two arguments: The data you want to filter (an array or object) and the criteria for filtering (either a LAMBDA function or an array of values):

REJECT(array | object, values_to_remove | LAMBDA(arg1, [arg2], expr))

REJECT example: Survey scores 

Using the same survey score example, let's say you want to remove low-scoring surveys (score less than 5):

REJECT(action_name.surveys, LAMBDA(survey, survey.score < 5))

• action_name.surveys: The upstream reference path to the target surveys array.

• survey: Represents each individual item in the surveys array as REJECT processes it.You declare this variable, so name it something that will help you and others better understand the overall formula.

• survey.score: The reference path to the score value of the survey item (iteration).

Translation: Go through each object within the surveys array, check the score value, and remove the entire object if its score value is less than 5.

The output: The low-scoring survey was removed, and everything else was kept:

[
  {"id": "S-001", "score": 9},
  {"id": "S-003", "score": 8}
]

Complex filter with multiple conditions 

FILTER and REJECT can handle sophisticated logic using AND, OR, and nested conditions. Let's say you're processing this incidents array:

{
  "incidents": [
    {
      "id": "INC-001",
      "severity": "critical",
      "status": "open",
      "age_days": 5
    },
    {
      "id": "INC-002",
      "severity": "high",
      "status": "open",
      "age_days": 15
    },
    {
      "id": "INC-003",
      "severity": "medium",
      "status": "closed",
      "age_days": 30
    },
    {
      "id": "INC-004",
      "severity": "critical",
      "status": "open",
      "age_days": 2
    },
    {
      "id": "INC-005",
      "severity": "low",
      "status": "open",
      "age_days": 60
    }
  ]
}

We want to keep incidents that need immediate attention, so let's identify the criteria we need to check:

• Critical OR high severity AND

• Open AND

• Older than 3 days

To only keep the incidents that match these conditions, let's use this FILTER formula:

FILTER(
  action_name.incidents,
  LAMBDA(incident,
    AND(
      OR(incident.severity = "critical", incident.severity = "high"),
      incident.status = "open",
      incident.age_days > 3
    )
  )
)

Translation: Go through each object within the incidents array and check the following: Its severity is high or critical, its status is open, and its age_days is greater than 3. Only keep the objects where all conditions are met.

The output: Only the incidents that matched the multiple conditions were kept:

[
  {"id": "INC-001", "severity": "critical", "status": "open", "age_days": 5},
  {"id": "INC-002", "severity": "high", "status": "open", "age_days": 15}
]

Combine with other functions 

Both FILTER and REJECT work well in nests and chains with other functions. Let's look at this students array:

{
  "students": [
    {
      "id": "L-01",
      "first_name": "Priya",
      "last_name": "Desmond",
      "status": "active"
    },
    {
      "id": "L-02",
      "first_name": "Maya",
      "last_name": "Brennan",
      "status": "active"
    },
    {
      "id": "L-03",
      "first_name": "Jordan",
      "last_name": "Kim",
      "status": "inactive"
    },
    {
      "id": "L-04",
      "first_name": "Lucas",
      "last_name": "Ortega",
      "status": "active"
    }
  ]
}

We want to:

• Only keep active students.

• Sort them by last name.

How do we do this? By combining FILTER and SORT together, either by nesting:

SORT(FILTER(action_name.students, LAMBDA(student, student.status = "active")), "last_name")

or, by chaining:

FILTER(action_name.students, LAMBDA(student, student.status = "active")) |> SORT(%, "last_name")

We can get the results to look the way we need:

[
    {
      "id": "L-02",
      "first_name": "Maya",
      "last_name": "Brennan",
      "status": "active"
    },
    {
      "id": "L-01",
      "first_name": "Priya",
      "last_name": "Desmond",
      "status": "active"
    },
    {
      "id": "L-04",
      "first_name": "Lucas",
      "last_name": "Ortega",
      "status": "active"
    }
  ]