Crate redaction 

Type-directed redaction for structured data.

This crate separates:

• Classification: what kind of sensitive data this is.
• Policy: how that data should be redacted.

The derive macro walks your data and applies the policy at the boundary when you call redact() or Redactable::redact().

Key rules:

• Use #[sensitive(Classification)] for string-like leaf values.
• Use #[sensitive] for scalars and nested Sensitive types.
• Unannotated fields pass through unchanged.
• Debug always prints "[REDACTED]" for sensitive fields; policies apply only when calling .redact().

Boxed trait objects:

• #[sensitive] supports Box<dyn Trait> by calling redact_boxed.
• Detection is conservative and only matches the simple Box<dyn Trait> syntax, not qualified paths or type aliases.

What this crate does:

• defines classification marker types and the Classification trait
• defines redaction policies and the redact entrypoint
• provides integrations behind feature flags (e.g. slog)

What it does not do:

• perform I/O or logging
• validate your policy choices

The Sensitive derive macro lives in redaction-derive and is re-exported when the derive feature is enabled.

Structs

AccountId

Classification marker for account identifiers.

BlockchainAddress

Classification marker for blockchain addresses (e.g., Ethereum, Bitcoin).

CreditCard

Classification marker for credit card numbers or PANs.

DateOfBirth

Classification marker for dates of birth.

Email

Classification marker for email addresses.

IpAddress

Classification marker for IP addresses.

KeepConfig

Configuration that keeps selected segments visible while masking the remainder.

MaskConfig

Configuration that masks selected segments while leaving the remainder unchanged.

NationalId

Classification marker for government-issued identifiers.

PhoneNumber

Classification marker for phone numbers.

Pii

Classification marker for personally identifiable information.

Secret

Classification marker for secrets such as passwords or private keys.

SessionId

Classification marker for session identifiers.

Token

Classification marker for authentication tokens and API keys.

Enums

TextRedactionPolicy

A redaction strategy for string-like values.

Constants

REDACTED_PLACEHOLDER

Default placeholder used for full redaction.

Traits

Classification

Marker trait for classification categories.

Redactable

Public entrypoint for redaction on traversable types.

RedactableBoxed

Redacts boxed trait objects that expose their own boxed redaction.

RedactionPolicy

Associates a classification type with a concrete string redaction policy.

SensitiveValue

String-like payloads that can be redacted via policies.

Functions

apply_classification

Applies a classification policy to a classifiable value.

redact

Redacts a value using classification-bound policies.

redact_boxed

Convenience helper for redacting boxed trait objects.

Derive Macros

Sensitive

Derives redaction::SensitiveType (and related impls) for structs and enums.

SensitiveError

Derives redaction::SensitiveType for types that should log without Serialize.