Skip to main content

Filter

Category: Transform
Node id: filter

Filters an array of objects based on configurable conditions with AND/OR logic

What it does​

  • 20+ comparison operators including regex and type checks
  • AND/OR logic for combining multiple conditions
  • Detailed filter statistics: original, filtered, and removed counts
  • Nested field access for complex object filtering

Use cases​

  • Filter patient records by age range or diagnosis code
  • Select FHIR resources matching specific criteria
  • Remove invalid or incomplete records from a dataset

Runtime​

  • Performance tier: fast

Input ports​

NameTypeRequiredDescription
datajsonYesArray of objects to filter

Output ports​

NameTypeRequiredDescription
resultjsonNoFilter result: filtered_data array plus original_count, filtered_count, and removed_count
successbooleanNoWhether the filter operation completed without error

Settings​

Basic​

SettingNameTypeRequiredDefaultAllowed valuesDescription
ConditionsconditionsarrayYesArray of filter conditions; each has field, operator, and value
LogiclogicenumNoANDAND, ORHow to combine multiple conditions

Errors​

CodeDescription
INVALID_CONDITIONSConditions configuration is missing or not a valid array
INVALID_DATAInput data is not a valid array

Runtime notes​

  • The filtered items are in result.filtered_data (an array), NOT at result top level. Downstream nodes must read result.filtered_data to get the array.
  • Filter KEEPS matching items (include semantics), it does not drop them. Items that fail conditions are removed.
  • Non-object items in the input array are silently skipped — only map/object items are evaluated.
  • Operators have aliases: eq/==/equals, gt/>/ greater_than, etc. Use the long form for readability.

See Build any workflow for how to put nodes together and Ports, types and data flow for how port types connect.

Something unclear? Send documentation feedback by email.