Add test infrastructure and fixtures for University Agent unit tests

[devin-ai-integration]

Description

This PR implements comprehensive unit test coverage for the University Agent components as part of Tasks 5.1, 5.4, and 5.5. It adds test infrastructure with reusable fixtures and creates 59 new unit tests covering the UniversityPlugin and GeminiChatCompletionFactory classes.

Link to Devin run: https://app.devin.ai/sessions/1bb810da59f64d8e8a22acab3c5602fc
Requested by: Shawn Azman (@ShawnAzman)

Changes

Test Infrastructure (Task 5.5)

• Created tests/conftest.py with 7 pytest fixtures:
  • mock_app_config and mock_app_config_with_api_key for configuration mocking
  • mock_requests_get and mock_requests_exception for HTTP request mocking
  • mock_google_ai_chat_completion for Google Gemini API mocking
  • university_api_responses and university_search_responses for loading test data
• Created tests/fixtures/university_api_responses.json with mock API response data (279 lines)
• Created tests/fixtures/university_search_responses.json with search scenario data

UniversityPlugin Tests (Task 5.1) - 37 tests

• Success path tests (8 tests): Verify both plugin methods (search_universities and get_universities_by_country) with various scenarios including successful responses, empty results, and result limits
• Data parsing tests (3 tests): Validate JSON-to-Pydantic model conversion, nullable field handling, and multiple domains/pages
• Search result format tests (3 tests): Verify UniversitySearchResult structure and format
• Error handling tests (9 tests): Network failures, timeouts, HTTP errors, and generic exceptions
• Input validation tests (6 tests): Empty strings, special characters, very long strings, Unicode
• Malformed response tests (8 tests): Invalid JSON, missing fields, unexpected data types, null responses

GeminiChatCompletionFactory Tests (Task 5.4) - 22 tests

• Initialization tests (4 tests): Valid API key, API key storage, missing API key, parent class initialization
• Configuration tests (1 test): Verify get_configs() returns empty list
• Valid model tests (4 tests): Parametrized tests for all 4 supported Gemini models (gemini-1.5-flash, gemini-1.5-pro, gemini-1.0-pro, gemini-2.0-flash-lite)
• Error handling tests (3 tests): Invalid model name handling for all three factory methods
• Model type tests (4 tests): Verify ModelType.GOOGLE returned for all supported models
• Structured output tests (4 tests): Verify all models support structured output
• Factory method tests (2 tests): Default and custom parameter handling for create_chat_completion()

Type of Change

• New feature (test coverage)
• Documentation (test docstrings)

Important Review Points

⚠️ Mock data accuracy: The fixture data uses "state-province" (hyphenated) which maps to state_province (underscored) in the Pydantic model. Verify this matches the actual universities.hipolabs.com API format.

⚠️ Result limit verification: Tests verify that search_universities() returns max 10 results and get_universities_by_country() returns max 20 results. Confirm these limits are implemented in the actual UniversityPlugin code (lines 59 and 105 in custom_plugins.py per task description).

⚠️ No integration tests: All HTTP requests and API calls are mocked. The plugin has never been tested against the real universities API or Google Gemini API.

⚠️ GoogleAIChatCompletion mocking: All Gemini API calls are properly mocked using unittest.mock.patch() to prevent actual API requests during testing.

Test Results

37 passed for UniversityPlugin
22 passed for GeminiChatCompletionFactory
Total: 59 new unit tests, all passing
Ruff linting: All checks passed

Additional Comments

• This PR combines multiple tasks (5.1, 5.4, 5.5) to provide complete test coverage in a single changeset
• All tests use mocks and fixtures to avoid external dependencies
• Test infrastructure is designed for reuse across future test files
• Comprehensive error handling coverage ensures robustness of both plugins

[devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR. Add '(aside)' to your comment to have me ignore it.
• Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

• Disable automatic comment and CI monitoring