## PR Summary

This PR, authored by Masterain, includes updates to the strategy router and introduces a new issue router to fetch and display GitHub issues.

## Changes by feature

<details>
  <summary>Strategy Router Updates (routers/strategy.py)</summary>

  - **Logging Improvements:** Added logging to track avatar processing and errors during strategy refresh from Miyoushe and Hoyolab.
  - **Error Handling:** Improved error logging for avatar ID retrieval failures.
  - **Redis Usage:** Modified the logic to refresh avatar strategies from Cloudflare when a TypeError occurs during Redis data loading.
</details>

<details>
  <summary>Issue Router (routers/issue.py, main.py)</summary>

  - **New Feature:** Implemented a new API endpoint to retrieve open bug issues from GitHub.
  - **Caching:** Integrated Redis caching to store and serve issue data, reducing GitHub API calls.
  - **Data Processing:** Added functions to prune unnecessary fields from the GitHub API response and calculate bug statistics based on labels.
  - **Bug Statistics:** The API now returns statistics on bug issues, including counts for "waiting_for_release", "untreated", and "hard_to_fix" issues.
  - **Filtering:** Enhanced issue filtering to focus on bug-related issues, exclude pull requests, and categorize issues based on labels.
</details>

## Key Change Highlights

- **Introduction of Issue Router:** A new router (`routers/issue.py`) is added to fetch and display open bug issues from the Snap.Hutao GitHub repository. This introduces a new dependency on the `httpx` library for making HTTP requests.
- **Strategy Refresh Logic:** The logic to refresh avatar strategies has been modified. Now, instead of calling `refresh_avatar_strategy` from `cloudflare_security_utils.mgnt` directly, it imports the function within the `try` block to resolve potential `TypeError` issues, possibly related to Redis data loading.
- **Data Structure and API Endpoint:** The `/issue/bug` endpoint now returns a structured response containing both issue details and statistics, which are calculated based on issue labels. This change affects how clients consume the API response.

## Suggested Test Methods

- **API Endpoint Tests for `/issue/bug`:**
  - Verify that the endpoint returns a `200 OK` status code.
  - Ensure that the response body contains both `details` (list of issues) and `stat` (bug statistics) keys.
  - Check that the issue details are correctly pruned to include only the required fields (`number`, `title`, `labels`, `author`, `created_at`).
  - Confirm that the bug statistics are calculated correctly based on the issue labels (e.g., "waiting_for_release", "untreated", "hard_to_fix").
  - Test the endpoint with different network conditions to ensure it handles `httpx.HTTPError` exceptions gracefully.
  - Mock the GitHub API to test different scenarios, such as empty responses or error responses.
- **Redis Caching Tests:**
  - Verify that the issue data is cached in Redis after the first request.
  - Ensure that subsequent requests are served from the cache until the cache expires.
  - Check that the cache is properly updated when new issues are added or existing issues are modified in the GitHub repository.
  - Test the cache invalidation logic when the `refresh_avatar_strategy` function is called.
- **Unit Tests for Helper Functions:**
  - Create unit tests for the `_prune_issue_fields` function to ensure it correctly filters and transforms the GitHub API response.
  - Write unit tests for the `_calc_bug_stats` function to verify that it accurately calculates bug statistics based on different label combinations.
- **Integration Tests:**
  - Test the integration between the issue router and the Redis cache to ensure that data is properly cached and retrieved.
  - Verify that the issue router correctly handles errors from the GitHub API and returns an appropriate error response to the client.