feat(encryption): add governed DSAR coverage

This commit is contained in:
2026-08-21 13:15:26 +02:00
parent 898d68aa1a
commit 277d5f01e4
4 changed files with 963 additions and 0 deletions
@@ -0,0 +1,527 @@
from __future__ import annotations
from collections.abc import Sequence
from dataclasses import dataclass
from datetime import datetime, timezone
from sqlalchemy import or_
from sqlalchemy.orm import Session
from govoplan_core.core.dsar import (
DsarErasureActionRef,
DsarExecutionResultRef,
DsarRecordRef,
DsarSubjectRef,
dsar_capability_name,
)
from govoplan_encryption.backend.db.models import (
ContentProtectionRecord,
EncryptionKeyOperation,
EncryptionVault,
ProtectionMigration,
RecoveryApproval,
RecoveryCeremony,
)
ENCRYPTION_DSAR_CAPABILITY = dsar_capability_name("encryption")
_MAX_RECORDS = 5_000
_CONFLICT = object()
@dataclass(frozen=True, slots=True)
class _SubjectSelectors:
account_id: str
vault_id: str | None
key_operation_id: str | None
envelope_id: str | None
migration_id: str | None
recovery_id: str | None
@property
def narrowed(self) -> bool:
return any(
(
self.vault_id,
self.key_operation_id,
self.envelope_id,
self.migration_id,
self.recovery_id,
)
)
class EncryptionDsarProvider:
provider_id = "encryption"
module_id = "encryption"
def search_subject(
self,
session: object,
*,
tenant_id: str,
subject: DsarSubjectRef,
) -> Sequence[DsarRecordRef]:
db = _session(session)
selectors = _subject_selectors(subject)
if selectors is None:
return ()
records: list[DsarRecordRef] = []
if not selectors.narrowed or selectors.vault_id:
query = db.query(EncryptionVault).filter(
EncryptionVault.tenant_id == tenant_id,
or_(
EncryptionVault.created_by == selectors.account_id,
EncryptionVault.updated_by == selectors.account_id,
),
)
if selectors.vault_id:
query = query.filter(EncryptionVault.vault_id == selectors.vault_id)
records.extend(
_vault_attribution(row, selectors.account_id)
for row in _limited(
query,
EncryptionVault.created_at,
EncryptionVault.id,
label="vault attribution",
)
)
if not selectors.narrowed or selectors.vault_id or selectors.key_operation_id:
query = db.query(EncryptionKeyOperation).filter(
EncryptionKeyOperation.tenant_id == tenant_id,
EncryptionKeyOperation.requested_by == selectors.account_id,
)
if selectors.vault_id:
query = query.filter(
EncryptionKeyOperation.vault_id == selectors.vault_id
)
if selectors.key_operation_id:
query = query.filter(
EncryptionKeyOperation.id == selectors.key_operation_id
)
records.extend(
_key_operation_attribution(row)
for row in _limited(
query,
EncryptionKeyOperation.created_at,
EncryptionKeyOperation.id,
label="key-operation attribution",
)
)
if not selectors.narrowed or selectors.vault_id or selectors.envelope_id:
query = db.query(ContentProtectionRecord).filter(
ContentProtectionRecord.tenant_id == tenant_id,
ContentProtectionRecord.registered_by == selectors.account_id,
)
if selectors.vault_id:
query = query.filter(
ContentProtectionRecord.vault_id == selectors.vault_id
)
if selectors.envelope_id:
query = query.filter(
ContentProtectionRecord.envelope_id == selectors.envelope_id
)
records.extend(
_protection_attribution(row)
for row in _limited(
query,
ContentProtectionRecord.created_at,
ContentProtectionRecord.id,
label="content-protection attribution",
)
)
if not selectors.narrowed or selectors.vault_id or selectors.migration_id:
query = db.query(ProtectionMigration).filter(
ProtectionMigration.tenant_id == tenant_id,
ProtectionMigration.requested_by == selectors.account_id,
)
if selectors.vault_id:
query = query.filter(
ProtectionMigration.target_vault_id == selectors.vault_id
)
if selectors.migration_id:
query = query.filter(ProtectionMigration.id == selectors.migration_id)
records.extend(
_migration_attribution(row)
for row in _limited(
query,
ProtectionMigration.created_at,
ProtectionMigration.id,
label="migration attribution",
)
)
if not selectors.narrowed or selectors.vault_id or selectors.recovery_id:
ceremonies = db.query(RecoveryCeremony).filter(
RecoveryCeremony.tenant_id == tenant_id,
RecoveryCeremony.requester_account_id == selectors.account_id,
)
if selectors.vault_id:
ceremonies = ceremonies.filter(
RecoveryCeremony.vault_id == selectors.vault_id
)
if selectors.recovery_id:
ceremonies = ceremonies.filter(
RecoveryCeremony.id == selectors.recovery_id
)
records.extend(
_recovery_request_attribution(row)
for row in _limited(
ceremonies,
RecoveryCeremony.created_at,
RecoveryCeremony.id,
label="recovery-request attribution",
)
)
approvals = db.query(RecoveryApproval).filter(
RecoveryApproval.tenant_id == tenant_id,
RecoveryApproval.approver_account_id == selectors.account_id,
)
if selectors.recovery_id:
approvals = approvals.filter(
RecoveryApproval.recovery_id == selectors.recovery_id
)
elif selectors.vault_id:
approvals = approvals.filter(
RecoveryApproval.recovery_id.in_(
db.query(RecoveryCeremony.id).filter(
RecoveryCeremony.tenant_id == tenant_id,
RecoveryCeremony.vault_id == selectors.vault_id,
)
)
)
records.extend(
_recovery_approval_attribution(row)
for row in _limited(
approvals,
RecoveryApproval.created_at,
RecoveryApproval.id,
label="recovery-approval attribution",
)
)
if len(records) > _MAX_RECORDS:
raise ValueError("Encryption DSAR result limit exceeded; narrow selectors.")
return tuple(
sorted(records, key=lambda item: (item.resource_type, item.resource_id))
)
def plan_erasure(
self,
session: object,
*,
tenant_id: str,
subject: DsarSubjectRef,
records: Sequence[DsarRecordRef],
) -> Sequence[DsarErasureActionRef]:
del tenant_id
_session(session)
if _subject_selectors(subject) is None:
raise ValueError("Encryption DSAR subject selectors conflict.")
actions: list[DsarErasureActionRef] = []
for record in records:
_validate_record(record)
actions.append(
DsarErasureActionRef(
action_id=(
f"encryption:retain:{record.resource_type}:{record.resource_id}"
),
provider_id=self.provider_id,
module_id=self.module_id,
kind="retain",
resource_type=record.resource_type,
resource_id=record.resource_id,
title=f"Retain {record.title}",
rationale=(
record.retention_reason
or "Cryptographic custody evidence remains immutable."
),
executable=False,
)
)
return tuple(actions)
def execute_erasure(
self,
session: object,
*,
tenant_id: str,
subject: DsarSubjectRef,
actions: Sequence[DsarErasureActionRef],
request_id: str,
) -> Sequence[DsarExecutionResultRef]:
del tenant_id
_session(session)
if _subject_selectors(subject) is None:
raise ValueError("Encryption DSAR subject selectors conflict.")
results: list[DsarExecutionResultRef] = []
for action in actions:
_validate_action(action)
if action.executable or action.kind != "retain":
raise ValueError("Encryption DSAR publishes retain actions only.")
results.append(
DsarExecutionResultRef(
action_id=action.action_id,
status="blocked",
summary=(
"Cryptographic lifecycle and custody attribution remains "
"immutable security evidence."
),
evidence={"request_id": request_id},
)
)
return tuple(results)
def _subject_selectors(subject: DsarSubjectRef) -> _SubjectSelectors | None:
references = subject.external_references
account = _coalesce(
subject.account_id,
references.get("encryption.account"),
references.get("access.account"),
)
values = {
"vault_id": _coalesce(
references.get("encryption.vault"),
references.get("encryption.vault_id"),
),
"key_operation_id": _coalesce(
references.get("encryption.key_operation"),
references.get("encryption.key_operation_id"),
),
"envelope_id": _coalesce(
references.get("encryption.envelope"),
references.get("encryption.envelope_id"),
),
"migration_id": _coalesce(
references.get("encryption.migration"),
references.get("encryption.migration_id"),
),
"recovery_id": _coalesce(
references.get("encryption.recovery"),
references.get("encryption.recovery_id"),
),
}
if account is _CONFLICT or any(value is _CONFLICT for value in values.values()):
return None
account_id = _optional_string(account)
if not account_id:
return None
return _SubjectSelectors(
account_id=account_id,
vault_id=_optional_string(values["vault_id"]),
key_operation_id=_optional_string(values["key_operation_id"]),
envelope_id=_optional_string(values["envelope_id"]),
migration_id=_optional_string(values["migration_id"]),
recovery_id=_optional_string(values["recovery_id"]),
)
def _vault_attribution(row: EncryptionVault, account_id: str) -> DsarRecordRef:
activities = []
if row.created_by == account_id:
activities.append("created_vault")
if row.updated_by == account_id:
activities.append("updated_vault")
return _record(
resource_type="vault_actor_attribution",
resource_id=row.id,
title="Encryption vault actor attribution",
data={
"vault_id": row.vault_id,
"purpose": row.purpose[:255],
"profile_kind": row.profile_kind,
"scope_type": row.scope_type,
"state": row.state,
"revision": row.revision,
"current_key_version": row.current_key_version,
"recovery_quorum": row.recovery_quorum,
"activities": activities,
"created_at": _iso(row.created_at),
"updated_at": _iso(row.updated_at),
},
observed_at=row.updated_at,
)
def _key_operation_attribution(row: EncryptionKeyOperation) -> DsarRecordRef:
return _record(
resource_type="key_operation_actor_attribution",
resource_id=row.id,
title="Encryption key-operation actor attribution",
data={
"key_operation_id": row.id,
"vault_id": row.vault_id,
"key_version": row.key_version,
"operation": row.operation,
"state": row.state,
"activity": "requested_key_operation",
"created_at": _iso(row.created_at),
"completed_at": _iso(row.completed_at),
},
observed_at=row.completed_at or row.created_at,
)
def _protection_attribution(row: ContentProtectionRecord) -> DsarRecordRef:
return _record(
resource_type="content_protection_actor_attribution",
resource_id=row.id,
title="Content-protection registration attribution",
data={
"protection_record_id": row.id,
"owner_module": row.owner_module,
"resource_type": row.resource_type,
"profile_kind": row.profile_kind,
"state": row.state,
"activity": "registered_content_protection",
"created_at": _iso(row.created_at),
"updated_at": _iso(row.updated_at),
},
observed_at=row.updated_at,
)
def _migration_attribution(row: ProtectionMigration) -> DsarRecordRef:
return _record(
resource_type="protection_migration_actor_attribution",
resource_id=row.id,
title="Protection-migration actor attribution",
data={
"migration_id": row.id,
"mode": row.mode,
"state": row.state,
"activity": "requested_protection_migration",
"created_at": _iso(row.created_at),
"completed_at": _iso(row.completed_at),
},
observed_at=row.completed_at or row.created_at,
)
def _recovery_request_attribution(row: RecoveryCeremony) -> DsarRecordRef:
return _record(
resource_type="recovery_request_actor_attribution",
resource_id=row.id,
title="Encryption recovery-request attribution",
data={
"recovery_id": row.id,
"vault_id": row.vault_id,
"state": row.state,
"quorum": row.quorum,
"revision": row.revision,
"expires_at": _iso(row.expires_at),
"activity": "requested_recovery_ceremony",
"created_at": _iso(row.created_at),
"updated_at": _iso(row.updated_at),
},
observed_at=row.updated_at,
)
def _recovery_approval_attribution(row: RecoveryApproval) -> DsarRecordRef:
return _record(
resource_type="recovery_approval_actor_attribution",
resource_id=row.id,
title="Encryption recovery-approval attribution",
data={
"recovery_approval_id": row.id,
"recovery_id": row.recovery_id,
"decision": row.decision,
"activity": "decided_recovery_approval",
"created_at": _iso(row.created_at),
},
observed_at=row.created_at,
)
def _record(
*,
resource_type: str,
resource_id: str,
title: str,
data: dict[str, object],
observed_at: datetime | None,
) -> DsarRecordRef:
return DsarRecordRef(
provider_id="encryption",
module_id="encryption",
resource_type=resource_type,
resource_id=resource_id,
category="cryptographic_custody_attribution",
title=title,
data=data,
observed_at=_aware(observed_at),
immutable_evidence=True,
retention_reason=(
"Cryptographic lifecycle and custody attribution is retained as immutable "
"security and accountability evidence."
),
)
def _limited(query, first, second, *, label: str):
rows = query.order_by(first, second).limit(_MAX_RECORDS + 1).all()
if len(rows) > _MAX_RECORDS:
raise ValueError(f"Encryption DSAR {label} limit exceeded; narrow selectors.")
return rows
def _coalesce(*values: str | None) -> str | None | object:
normalized = {str(value).strip() for value in values if str(value or "").strip()}
if len(normalized) > 1:
return _CONFLICT
return next(iter(normalized), None)
def _optional_string(value: object) -> str | None:
return value if isinstance(value, str) and value else None
def _iso(value: datetime | None) -> str | None:
aware = _aware(value)
return aware.isoformat() if aware else None
def _aware(value: datetime | None) -> datetime | None:
if value is None or value.tzinfo is not None:
return value
return value.replace(tzinfo=timezone.utc)
def _session(value: object) -> Session:
if not isinstance(value, Session):
raise TypeError("Encryption DSAR requires a SQLAlchemy Session.")
return value
_RESOURCE_TYPES = {
"vault_actor_attribution",
"key_operation_actor_attribution",
"content_protection_actor_attribution",
"protection_migration_actor_attribution",
"recovery_request_actor_attribution",
"recovery_approval_actor_attribution",
}
def _validate_record(record: DsarRecordRef) -> None:
if record.provider_id != "encryption" or record.module_id != "encryption":
raise ValueError("Encryption DSAR cannot plan a foreign provider record.")
if record.resource_type not in _RESOURCE_TYPES or not record.resource_id:
raise ValueError("Encryption DSAR record identity is invalid.")
def _validate_action(action: DsarErasureActionRef) -> None:
if action.provider_id != "encryption" or action.module_id != "encryption":
raise ValueError("Encryption DSAR cannot execute a foreign provider action.")
if not action.action_id.startswith("encryption:retain:"):
raise ValueError("Encryption DSAR action identity is invalid.")
__all__ = ["ENCRYPTION_DSAR_CAPABILITY", "EncryptionDsarProvider"]