mirror of
https://github.com/hwchase17/langchain
synced 2024-11-06 03:20:49 +00:00
2b14036126
Fixes #4153 If the sender of a message in a group chat isn't in your contact list, they will appear with a ~ prefix in the exported chat. This PR adds support for parsing such lines.
21 lines
741 B
Python
21 lines
741 B
Python
from pathlib import Path
|
|
|
|
from langchain.document_loaders import WhatsAppChatLoader
|
|
|
|
|
|
def test_whatsapp_chat_loader() -> None:
|
|
"""Test WhatsAppChatLoader."""
|
|
file_path = Path(__file__).parent.parent / "examples" / "whatsapp_chat.txt"
|
|
loader = WhatsAppChatLoader(str(file_path))
|
|
docs = loader.load()
|
|
|
|
assert len(docs) == 1
|
|
assert docs[0].metadata["source"] == str(file_path)
|
|
assert docs[0].page_content == (
|
|
"James on 05.05.23, 15:48:11: Hi here\n\n"
|
|
"User name on 11/8/21, 9:41:32 AM: Message 123\n\n"
|
|
"User 2 on 1/23/23, 3:19 AM: Bye!\n\n"
|
|
"User 1 on 1/23/23, 3:22_AM: And let me know if anything changes\n\n"
|
|
"~ User name 2 on 1/24/21, 12:41:03 PM: Of course!\n\n"
|
|
)
|