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

whatsapp.android: expose all entities extracted from the db

Browse source
This commit is contained in:
Dima Gerasimov 2023-12-29 00:53:14 +00:00 • committed by karlicoss
parent 3ec362fce9
commit 1b187b2c1b
1 changed files with 18 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

22
my/whatsapp/android.py
View file

@ -7,7 +7,7 @@ from dataclasses import dataclass
from datetime import datetime, timezone from datetime import datetime, timezone
from pathlib import Path from pathlib import Path
import sqlite3 import sqlite3
from typing import Sequence, Iterator, Optional from typing import Union, Sequence, Iterator, Optional
from my.core import get_files, Paths, datetime_aware, Res, make_logger, make_config from my.core import get_files, Paths, datetime_aware, Res, make_logger, make_config
from my.core.common import unique_everseen from my.core.common import unique_everseen
@ -56,7 +56,10 @@ class Message:
text: Optional[str] text: Optional[str]
def _process_db(db: sqlite3.Connection): Entity = Union[Chat, Sender, Message]
def _process_db(db: sqlite3.Connection) -> Iterator[Entity]:
# TODO later, split out Chat/Sender objects separately to safe on object creation, similar to other android data sources # TODO later, split out Chat/Sender objects separately to safe on object creation, similar to other android data sources
chats = {} chats = {}
@ -73,6 +76,7 @@ def _process_db(db: sqlite3.Connection):
id=chat_id, id=chat_id,
name=subject, name=subject,
) )
yield chat
chats[chat.id] = chat chats[chat.id] = chat
senders = {} senders = {}
@ -88,6 +92,7 @@ def _process_db(db: sqlite3.Connection):
id=r['raw_string'], id=r['raw_string'],
name=None, name=None,
) )
yield s
senders[r['_id']] = s senders[r['_id']] = s
# NOTE: hmm, seems that message_view or available_message_view use lots of NULL as ... # NOTE: hmm, seems that message_view or available_message_view use lots of NULL as ...
@ -187,7 +192,7 @@ def _process_db(db: sqlite3.Connection):
yield m yield m
def _messages() -> Iterator[Res[Message]]: def _entities() -> Iterator[Res[Entity]]:
paths = inputs() paths = inputs()
total = len(paths) total = len(paths)
width = len(str(total)) width = len(str(total))
@ -200,5 +205,14 @@ def _messages() -> Iterator[Res[Message]]:
yield echain(RuntimeError(f'While processing {path}'), cause=e) yield echain(RuntimeError(f'While processing {path}'), cause=e)
def entities() -> Iterator[Res[Entity]]:
return unique_everseen(_entities)
def messages() -> Iterator[Res[Message]]: def messages() -> Iterator[Res[Message]]:
yield from unique_everseen(_messages) # TODO hmm, specify key=lambda m: m.id?
# not sure since might be useful to keep track of sender changes etc
# probably best not to, or maybe query messages/senders separately and merge later?
for e in entities():
if isinstance(e, (Exception, Message)):
yield e