fz0x1/HPI
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

telegram_backup: order messages by date and users/chats by id for determinism

Browse source
This commit is contained in:
Dima Gerasimov 2023-03-27 02:03:48 +01:00 • committed by karlicoss
parent d2ef23fcb4
commit 74710b339a
1 changed files with 5 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
my/telegram/telegram_backup.py
View file

@ -62,6 +62,8 @@ class Message:
Chats = Dict[str, Chat]
def _message_from_row(r: sqlite3.Row, *, chats: Chats) -> Message:
ts = r['time']
# desktop export uses UTC (checked by exporting in winter time vs summer time)
# and telegram_backup timestamps seem same as in desktop export
time = datetime.fromtimestamp(ts, tz=timezone.utc)
chat = chats[r['source_id']]
sender = chats[r['sender_id']]
@ -78,12 +80,12 @@ def messages() -> Iterator[Message]:
with sqlite_connection(config.export_path, immutable=True, row_factory='row') as db:
chats: Chats = {}
for r in db.execute('SELECT * FROM chats'):
for r in db.execute('SELECT * FROM chats ORDER BY id'):
chat = Chat(id=r['id'], name=r['name'], handle=None)
assert chat.id not in chats
chats[chat.id] = chat
for r in db.execute('SELECT * FROM users'):
for r in db.execute('SELECT * FROM users ORDER BY id'):
first = r["first_name"]
last = r["last_name"]
name: Optional[str]
@ -96,8 +98,7 @@ def messages() -> Iterator[Message]:
assert chat.id not in chats
chats[chat.id] = chat
# TODO order by? not sure
for r in db.execute('SELECT * FROM messages WHERE message_type NOT IN ("service_message", "empty_message")'):
for r in db.execute('SELECT * FROM messages WHERE message_type NOT IN ("service_message", "empty_message") ORDER BY time'):
# seems like the only remaining have message_type = 'message'
yield _message_from_row(r, chats=chats)