Skip to content

Exactly One Not Null Check

Check: exactly-one-not-null-check

Purpose: Enforces that exactly one of the specified columns is non-null per row. A row fails if none or more than one of the columns are populated. Use this to enforce mutual exclusivity across optional identifiers or contact channels.

from sparkdq.checks import ExactlyOneNotNullCheckConfig
from sparkdq.core import Severity

ExactlyOneNotNullCheckConfig(
    check_id="single-contact-method",
    columns=["email", "phone", "user_id"],
    severity=Severity.CRITICAL
)

- check: exactly-one-not-null-check
  check-id: single-contact-method
  columns:
    - email
    - phone
    - user_id
  severity: critical

Typical Use Cases

  • Enforce that exactly one of several optional identifiers (e.g., email, phone, user_id) is provided per record.
  • Prevent ambiguous records where multiple mutually exclusive fields are populated simultaneously.
  • Detect rows where no identification channel is present at all.

← Row-Level Checks