Check Levels

Check levels define the severity of validation failures. They determine how failures affect the overall ValidationResult.

Available Levels

from qualink.core import Level

Level	Value	Description
`Level.ERROR`	2	Critical issues that prevent processing. Failures cause `result.success = False`.
`Level.WARNING`	1	Issues to investigate but don't block. Failures produce warnings but `success` may remain `True`.
`Level.INFO`	0	Informational observations. Row counts, profiling results, benchmarks.

How Levels Affect Results

ERROR Level

When an ERROR-level check has any failed constraints:

result.success → False
result.status → "Error"
Issues are reported with error_count incremented

check = (
    Check.builder("PK Check")
    .with_level(Level.ERROR)
    .is_complete("id")
    .is_unique("id")
    .build()
)

WARNING Level

When a WARNING-level check has any failed constraints:

result.success remains True (unless ERROR-level checks also fail)
result.status → "Warning" (if no ERROR failures)
Issues are reported with warning_count incremented

check = (
    Check.builder("Quality Check")
    .with_level(Level.WARNING)
    .has_completeness("email", Assertion.greater_than_or_equal(0.95))
    .build()
)

INFO Level

Informational checks that are tracked but don't block processing.

check = (
    Check.builder("Profiling")
    .with_level(Level.INFO)
    .has_approx_count_distinct("city", Assertion.greater_than(0))
    .build()
)

Usage Guidelines

Use Case	Recommended Level
Missing required fields	`ERROR`
Primary key violations	`ERROR`
Referential integrity failures	`ERROR`
Below-threshold data quality	`WARNING`
Unusual patterns	`WARNING`
Row count changes	`WARNING`
Row count observations	`INFO`
Profiling / statistics	`INFO`
Performance benchmarks	`INFO`

Helper Methods

level = Level.ERROR

level.as_str()           # "error"
str(level)               # "error"
level.is_at_least(Level.WARNING)  # True (ERROR >= WARNING)