Core Concepts

Detection

Use detect() to scan text for supported entities.

from fastpii import PrivacyGuard

guard = PrivacyGuard(regions=["cz"])
result = guard.detect("Email: jan.novak@example.cz, RČ: 8001011238")

detect() returns a DetectionResult with a list[Finding] in result.findings.

Each Finding contains:

• type
• value
• start
• end
• confidence
• region
• metadata

Validation

Use validate() when you already know the detector type and need to confirm whether a single value is structurally valid.

from fastpii import PrivacyGuard

guard = PrivacyGuard(regions=["cz"])
result = guard.validate("8001011238", "rodne_cislo")

validate() returns a ValidationResult with:

• is_valid
• metadata

For checksum-backed identifiers, validation uses official structural rules such as Mod 11 and weighted checksum calculations.

Anonymization

Use anonymize() to replace every detected value with the same placeholder.

guard.anonymize("Email: jan.novak@example.cz, RČ: 8001011238")

Output:

Email: [REDACTED], RČ: [REDACTED]

You can also pass a custom replacement string:

guard.anonymize("Email: jan.novak@example.cz", replacement="[PII]")

Redaction

Use redact() to replace each finding with a type-specific label.

guard.redact("Email: jan.novak@example.cz, RČ: 8001011238")

Output:

Email: [EMAIL], RČ: [RODNE_CISLO]

Masking

Use mask() to replace each finding with asterisks while preserving original length.

guard.mask("Email: jan.novak@example.cz, RČ: 8001011238")

Output:

Email: ********************, RČ: **********

Removal

Use remove() to delete detected values entirely.

guard.remove("Email: jan.novak@example.cz, RČ: 8001011238")

Output:

Email: , RČ: 

Overlap resolution

When multiple findings overlap, FastPII deduplicates them before returning the final DetectionResult.

Resolution order is based on:

1. confidence
2. type priority
3. span length

The built-in priority logic favors checksum-backed identifiers over broader personal or contact entities. In the current implementation, the priority order includes:

• checksum-backed identifiers first: rodne_cislo, ico, dic, bank_account
• then broader entities such as address, email, date_of_birth, name
• then lower-priority spans such as postal_code, vehicle_plate, phone

This reduces false positives when one detector captures a substring that belongs to a stronger match.

Pattern registry

FastPII uses a pattern registry and detector registry internally. PrivacyGuard registers built-in detectors for region cz, and custom detectors can also be registered programmatically.

This architecture is what makes regional extension possible without changing the PrivacyGuard public API.