klbr/khoj
1
0
Fork 0
mirror of https://github.com/khoaliber/khoj.git synced 2026-03-02 13:18:18 +00:00
Code Issues Packages Projects Releases Wiki Activity

Pass deep typed chat history for more ergonomic, readable, safe code

Browse Source 
The chat dictionary is an artifact from earlier non-db chat history
storage. We've been ensuring new chat messages have valid type before
being written to DB for more than 6 months now.

Move to using the deeply typed chat history helps avoids null refs,
makes code more readable and easier to reason about.

Next Steps:
The current update entangles chat_history written to DB
with any virtual chat history message generated for intermediate
steps. The chat message type written to DB should be decoupled from
type that can be passed to AI model APIs (maybe?).

For now we've made the ChatMessage.message type looser to allow
for list[dict] type (apart from string). But later maybe a good idea
to decouple the chat_history recieved by send_message_to_model from
the chat_history saved to DB (which can then have its stricter type check)
This commit is contained in:
Debanjum
2025-06-03 15:28:06 -07:00
parent 430459a338
commit 05d4e19cb8
20 changed files with 271 additions and 248 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
tests/test_offline_chat_actors.py
View File

@@ -135,7 +135,7 @@ def test_generate_search_query_using_question_from_chat_history(loaded_model):
# Act
response = extract_questions_offline(
query,
conversation_log=generate_chat_history(message_list),
chat_history=generate_chat_history(message_list),
loaded_model=loaded_model,
use_history=True,
)
@@ -181,7 +181,7 @@ def test_generate_search_query_using_answer_from_chat_history(loaded_model):
# Act
response = extract_questions_offline(
"Is she a Doctor?",
conversation_log=generate_chat_history(message_list),
chat_history=generate_chat_history(message_list),
loaded_model=loaded_model,
use_history=True,
)
@@ -210,7 +210,7 @@ def test_generate_search_query_with_date_and_context_from_chat_history(loaded_mo
# Act
response = extract_questions_offline(
"What was the Pizza place we ate at over there?",
conversation_log=generate_chat_history(message_list),
chat_history=generate_chat_history(message_list),
loaded_model=loaded_model,
)
@@ -336,7 +336,7 @@ def test_answer_from_chat_history_and_previously_retrieved_content(loaded_model)
response_gen = converse_offline(
references=[], # Assume no context retrieved from notes for the user_query
user_query="Where was I born?",
conversation_log=generate_chat_history(message_list),
chat_history=generate_chat_history(message_list),
loaded_model=loaded_model,
)
response = "".join([response_chunk for response_chunk in response_gen])
@@ -363,7 +363,7 @@ def test_answer_from_chat_history_and_currently_retrieved_content(loaded_model):
{"compiled": "Testatron was born on 1st April 1984 in Testville."}
], # Assume context retrieved from notes for the user_query
user_query="Where was I born?",
conversation_log=generate_chat_history(message_list),
chat_history=generate_chat_history(message_list),
loaded_model=loaded_model,
)
response = "".join([response_chunk for response_chunk in response_gen])
@@ -388,7 +388,7 @@ def test_refuse_answering_unanswerable_question(loaded_model):
response_gen = converse_offline(
references=[], # Assume no context retrieved from notes for the user_query
user_query="Where was I born?",
conversation_log=generate_chat_history(message_list),
chat_history=generate_chat_history(message_list),
loaded_model=loaded_model,
)
response = "".join([response_chunk for response_chunk in response_gen])
@@ -501,7 +501,7 @@ def test_answer_general_question_not_in_chat_history_or_retrieved_content(loaded
response_gen = converse_offline(
references=[], # Assume no context retrieved from notes for the user_query
user_query="Write a haiku about unit testing in 3 lines",
conversation_log=generate_chat_history(message_list),
chat_history=generate_chat_history(message_list),
loaded_model=loaded_model,
)
response = "".join([response_chunk for response_chunk in response_gen])