Filter

The Filter transform component evaluates incoming messages against a custom rule (written in an expr rule-based engine) and processes only those that match the rule.

---

Class Definition

class Filter extends Transform {
  fixed sourceName = "filter"
  code: String
}

---

Attributes

Attribute	Type	Description	Default Value
code	String	The filtering rule written in the expr engine syntax.	Required

---

I/O Types

• Input Type: map[string]any, string, []byte
• Output Type: map[string]any, string, []byte

---

Pkl Configuration Example

Basic Filter Transform

new Transforms.Filter {
  name = "example-filter"
  code = "Data.count > 5"
}

---

Description

The Filter transform component uses a custom rule defined in the expr engine to evaluate each incoming message. Only messages that satisfy the rule are passed to the next component.

---

Example from Unit Test

The following test demonstrates filtering messages where the count field in a JSON object is greater than 5:

{
	name: "test filtering string return 10",
	args: args{
		code: "Data.count > 5",
		val:  []byte("{"count":10}"),
	},
	want: []byte("{"count":10}"),
},

---

Pkl Configuration Matching with Unit Test

The equivalent Pkl configuration for the test case above:

new Transforms.Filter {
  name = "filter-count-test"
  code = "Data.count > 5"
}

---

Use Cases

1. Message Filtering
   • Filter out messages that do not meet specific criteria.
2. Data Validation
   • Process only messages with valid or desired data.

---

Notes

• The code attribute must be a valid [expr](https://github.com/expr-lang/expr] rule).
• Ensure the rule logic matches the expected structure of incoming messages.

---