utils/crdt-sync/crdt_sync/_e2e_cleanup.py
Krzysztof kuhy Rudnicki 3174480624
Some checks failed
crdt-sync pre-commit / pre-commit (push) Has been cancelled
crdt-sync tests / test (push) Has been cancelled
dependency freshness / dependencies on newest stable (push) Has been cancelled
file length / file-length (push) Has been cancelled
md-naming / markdown naming convention (push) Has been cancelled
palette drift / palette (push) Has been cancelled
repo-contract / repo contract (push) Has been cancelled
shared gate tests / gate unit tests (push) Has been cancelled
crdt-sync 0.11.1: verified plaintext cleanup
clear_plaintext_prefix + tool/e2e_cleanup: dry run by default, never
creates a header, deletes the old plaintext prefix only when every old
record is in the encrypted copy.

Todo: fda71ed1

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-12 10:44:44 +01:00

80 lines
2.9 KiB
Python

# Copyright (c) 2026 Krzysztof Rudnicki
"""Clear an app's old plaintext prefix once its encrypted copy is proven.
The migration never deletes the old prefix by itself: an old build may still
write there. Run this once every device runs an encrypting build. It never
creates a header (a typo'd passphrase must not lock the app for good) and
deletes nothing unless every old blob is contained in the encrypted copy.
"""
from __future__ import annotations
import hmac
from typing import TYPE_CHECKING
from crdt_sync._e2e_key import derive_key, read_header
from crdt_sync._e2e_store import E2eLayout, E2eLockedError, EncryptingStore, _blobs
from crdt_sync._log import merge_logs
from crdt_sync._store import dump_log, load_log
if TYPE_CHECKING:
from crdt_sync._e2e_key import E2eKey
from crdt_sync._remote import RemoteStore
def unverified_blobs(
store: RemoteStore, key: E2eKey, old_prefix: str, layout: E2eLayout
) -> list[str]:
"""Return old blobs (relative paths) not fully contained in the copy."""
dest = EncryptingStore(store, key, layout)
missing: list[str] = []
for rel, text in sorted(_blobs(store, old_prefix).items()):
try:
source = load_log(text)
copied = dest.get_file_text(f"{layout.prefix}/{rel}")
current = load_log(copied) if copied is not None else None
except ValueError, KeyError, TypeError:
missing.append(rel)
continue
if not source:
continue
if current is None or dump_log(merge_logs(current, source)) != dump_log(
current
):
missing.append(rel)
return missing
def _existing_key(store: RemoteStore, layout: E2eLayout, passphrase: str) -> E2eKey:
header = read_header(store, layout.e2e_path)
if header is None:
msg = f"no e2e header at {layout.e2e_path}: nothing was migrated"
raise E2eLockedError(msg)
key = derive_key(passphrase, header.salt, header.params)
if not hmac.compare_digest(key.check_for(header.salt), header.check):
msg = "wrong crdt-sync passphrase"
raise E2eLockedError(msg)
return key
def clear_plaintext_prefix(
store: RemoteStore, old_prefix: str, *, passphrase: str, apply: bool
) -> list[str]:
"""Return the old paths to delete; delete them only when ``apply``.
Raises:
E2eLockedError: No header, a wrong passphrase, or a copy that does
not yet hold every old record.
"""
layout = E2eLayout.of(old_prefix)
key = _existing_key(store, layout, passphrase)
missing = unverified_blobs(store, key, old_prefix, layout)
if missing:
msg = f"copy not verified for {', '.join(missing)}"
raise E2eLockedError(msg)
paths = [f"{old_prefix}/{rel}" for rel in sorted(_blobs(store, old_prefix))]
if apply:
for path in paths:
store.delete_file(path, message="e2e: clear plaintext prefix")
return paths