~95% on SimpleQA (e.g. Qwen3.6-27B on a 3090). Supports all local and cloud LLMs (llama.cpp, Ollama, Google, ...). 10+ search engines - arXiv, PubMed, your private documents. Everything Local & Encrypted.
Test Fixtures
This directory contains reusable test fixtures and mock utilities for the Local Deep Research project.
Overview
These fixtures were created based on scottvr's testing proposal (PR #361) but have been updated to address review feedback:
- Removed outdated config file references - The original MockConfigFiles class has been removed as the project no longer uses config files
- Added security validation - Includes proper URL validation for Wikipedia to address security concerns
- Improved modularity - Split fixtures into logical modules for better organization
Files
search_engine_mocks.py
Contains mock responses for various search engines:
- Wikipedia API responses
- arXiv XML responses
- PubMed search and article responses
- Semantic Scholar API responses
- Google Programmable Search Engine responses
- DuckDuckGo search responses
- Error response collections for testing error handling
Also includes:
validate_wikipedia_url()- Security function to validate Wikipedia URLs- Mock API response fixtures for successful and error cases
Usage Example
from tests.fixtures.search_engine_mocks import SearchEngineMocks, validate_wikipedia_url
def test_wikipedia_search(search_engine_mocks):
mock_response = search_engine_mocks.wikipedia_response()
# Use mock_response in your test
def test_url_security():
assert validate_wikipedia_url("https://en.wikipedia.org/wiki/Test") == True
assert validate_wikipedia_url("https://evil.com/wiki/Test") == False
Security Considerations
The validate_wikipedia_url() function addresses the security issue mentioned in PR #361 review by:
- Properly parsing URLs using
urllib.parse - Checking that the hostname ends with
.wikipedia.orgor is exactlywikipedia.org - Handling exceptions gracefully
- Rejecting malformed or malicious URLs
Integration with conftest.py
The main conftest.py file imports these fixtures to make them available throughout the test suite. The fixtures can be used directly in tests via dependency injection.