sij/khoj
1
0
Fork 0
mirror of https://github.com/khoj-ai/khoj.git synced 2024-11-27 09:25:06 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions 5

Handle filter only queries. Short-circuit and return filtered results

Browse source 
- For queries with only filters in them short-circuit and return
  filtered results. No need to run semantic search, re-ranking.
- Add client test for filter only query and quote query in client tests
This commit is contained in:
Debanjum Singh Solanky 2022-09-12 16:42:41 +03:00
parent afc84de234
commit 1bfe9c4ef2
2 changed files with 29 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

5
src/search_type/text_search.py
View file

@ -112,6 +112,11 @@ def query(raw_query: str, model: TextSearchModel, rank_results=False):
if entries is None or len(entries) == 0:
return [], []
# If query only had filters it'll be empty now. So short-circuit and return results.
if query.strip() == "":
hits = [{"corpus_id": id, "score": 1.0} for id, _ in enumerate(entries)]
return hits, entries
# Encode the query using the bi-encoder
start = time.time()
question_embedding = model.bi_encoder.encode([query], convert_to_tensor=True, device=state.device)

29
tests/test_client.py
View file

@ -1,18 +1,20 @@
# Standard Modules
from io import BytesIO
from PIL import Image
from urllib.parse import quote
# External Packages
from fastapi.testclient import TestClient
# Internal Packages
from src.main import app
from src.utils.config import SearchType
from src.utils.state import model, config
from src.search_type import text_search, image_search
from src.utils.rawconfig import ContentConfig, SearchConfig
from src.processor.org_mode.org_to_jsonl import org_to_jsonl
from src.search_filter.word_filter import WordFilter
from src.search_filter.file_filter import FileFilter
# Arrange
@ -23,7 +25,7 @@ client = TestClient(app)
# ----------------------------------------------------------------------------------------------------
def test_search_with_invalid_content_type():
# Arrange
user_query = "How to call Khoj from Emacs?"
user_query = quote("How to call Khoj from Emacs?")
# Act
response = client.get(f"/search?q={user_query}&t=invalid_content_type")
@ -117,7 +119,7 @@ def test_image_search(content_config: ContentConfig, search_config: SearchConfig
def test_notes_search(content_config: ContentConfig, search_config: SearchConfig):
# Arrange
model.orgmode_search = text_search.setup(org_to_jsonl, content_config.org, search_config.asymmetric, regenerate=False)
user_query = "How to git install application?"
user_query = quote("How to git install application?")
# Act
response = client.get(f"/search?q={user_query}&n=1&t=org&r=true")
@ -129,12 +131,29 @@ def test_notes_search(content_config: ContentConfig, search_config: SearchConfig
assert "git clone" in search_result
# ----------------------------------------------------------------------------------------------------
def test_notes_search_with_only_filters(content_config: ContentConfig, search_config: SearchConfig):
# Arrange
filters = [WordFilter(), FileFilter()]
model.orgmode_search = text_search.setup(org_to_jsonl, content_config.org, search_config.asymmetric, regenerate=False, filters=filters)
user_query = quote('+"Emacs" file:"*.org"')
# Act
response = client.get(f"/search?q={user_query}&n=1&t=org")
# Assert
assert response.status_code == 200
# assert actual_data contains word "Emacs"
search_result = response.json()[0]["entry"]
assert "Emacs" in search_result
# ----------------------------------------------------------------------------------------------------
def test_notes_search_with_include_filter(content_config: ContentConfig, search_config: SearchConfig):
# Arrange
filters = [WordFilter()]
model.orgmode_search = text_search.setup(org_to_jsonl, content_config.org, search_config.asymmetric, regenerate=False, filters=filters)
user_query = 'How to git install application? +"Emacs"'
user_query = quote('How to git install application? +"Emacs"')
# Act
response = client.get(f"/search?q={user_query}&n=1&t=org")
@ -151,7 +170,7 @@ def test_notes_search_with_exclude_filter(content_config: ContentConfig, search_
# Arrange
filters = [WordFilter()]
model.orgmode_search = text_search.setup(org_to_jsonl, content_config.org, search_config.asymmetric, regenerate=False, filters=filters)
user_query = 'How to git install application? -"clone"'
user_query = quote('How to git install application? -"clone"')
# Act
response = client.get(f"/search?q={user_query}&n=1&t=org")