feat(spx-backend):add image filter

[thewindwillstop]

petezhuang#147

[niupilot]

🔍 PR Review: Image Filter Feature

Todo List:

• Analyze core image filter implementation
• Review configuration and environment changes
• Check new models and data structures
• Examine controller logic and API endpoints
• Review test coverage and quality
• Assess documentation and design docs
• Provide comprehensive feedback with actionable suggestions

---

📋 Overview

This PR implements a comprehensive image recommendation filtering system that prevents users from seeing duplicate image recommendations. The implementation includes:

• ✅ Well-documented design with clear architecture diagrams
• ✅ Multi-table database schema for tracking user preferences and metrics
• ✅ Sophisticated degradation strategies to handle over-filtering
• ✅ Performance optimizations with async processing and caching
• ✅ Comprehensive test coverage with unit and integration tests

🔍 Major Issues Found

🚨 High Priority Issues

1. SQL Injection Vulnerability in image_filter_service.go:353

err := s.db.WithContext(ctx).
    Model(&model.UserImageRecommendationHistory{}).
    Where("user_id = ? AND created_at > ?", userID, cutoffTime).
    Pluck("image_id", &imageIDs).Error

Issue: While this specific query uses parameterized queries correctly, the pattern could be dangerous if copied elsewhere.
Recommendation: Add query validation comments or consider using a query builder.

2. Missing Error Handling for Goroutine Failures

Location: image_filter_service.go:139-144, image_recommend.go:300-307

go func() {
    if err := s.storeFilterMetrics(context.Background(), filterCtx); err != nil {
        logger.Printf("Failed to store filter metrics: %v", err)  // Only logs, no recovery
    }
}()

Issue: Critical async operations fail silently with only logging.
Impact: Filter metrics and recommendation history may be lost without notification.

3. Race Condition Risk in Concurrent Image Generation

Location: image_recommend.go:366-419

for i := range count {
    go func(index int) {
        // Multiple goroutines accessing shared resources
        resultChan <- generateResult{...}
    }(i)
}

Issue: No synchronization mechanism for shared resource access.
Potential Impact: Data corruption under high concurrency.

4. Memory Leak Risk in Filter Service

Location: image_filter_service.go:194-196

seenImages := make(map[int64]bool, len(historyImageIDs))
for _, imageID := range historyImageIDs {
    seenImages[imageID] = true
}

Issue: For users with extensive history, this map could consume significant memory.
Recommendation: Implement pagination or use more memory-efficient data structures.

⚠️ Medium Priority Issues

5. Hard-coded Magic Numbers

Locations: Multiple files

• 0.7 similarity threshold (image_filter_service.go:278)
• 30, 15, 7, 3, 1 day windows (image_filter_service.go:217)
• 100 batch size (image_filter_service.go:392)

Recommendation: Move to configuration constants.

6. Inconsistent Error Messages

Examples:

• "filtering failed: %w" vs "failed to get user history: %w"
• Mix of sentence case and lowercase error messages

7. Potential N+1 Query Problem

Location: image_filter_service.go:464-468

err := s.db.WithContext(ctx).
    Model(&model.ImageFilterMetrics{}).
    Where("user_id = ? AND created_at > ?", userID, cutoffTime).
    Select("COUNT(*) as total_queries, AVG(filter_ratio) as avg_filter_ratio, ...").
    Scan(&stats).Error

Issue: Multiple separate queries instead of JOIN operations.

💡 Design & Architecture Issues

8. Tight Coupling Between Components

The ImageFilterService directly depends on specific GORM models and database schema. Consider using interfaces for better testability.

9. Missing Circuit Breaker Pattern

For external service calls (algorithm service), implement circuit breaker to handle service outages gracefully.

10. Configuration Validation Missing

Location: config.go:209-231
The configuration getters have fallbacks but no validation of input ranges.

✅ Positive Aspects

1. Excellent Documentation: The Chinese design document is comprehensive and well-structured
2. Performance Optimizations: Async processing, batching, and caching strategies
3. Graceful Degradation: Multi-level fallback strategies for over-filtering
4. Comprehensive Testing: Good test coverage with edge cases
5. Database Design: Well-normalized schema with proper indexes
6. Code Organization: Clean separation of concerns

🎯 Recommendations

Immediate Actions Required

1. Fix async error handling - implement proper error recovery mechanisms
2. Add input validation for all configuration values
3. Replace magic numbers with named constants
4. Add synchronization for concurrent operations

Future Improvements

1. Consider implementing Redis caching for frequently accessed filter histories
2. Add metrics monitoring for filter performance
3. Implement rate limiting for image generation requests
4. Consider database connection pooling optimization

📊 Code Quality Metrics

• Test Coverage: Good (unit + integration tests present)
• Documentation: Excellent (comprehensive design docs)
• Error Handling: Needs improvement (silent failures in async operations)
• Performance: Good (async processing, batching)
• Security: Acceptable (parameterized queries used)

👉 Want me to re-run this review? Just type /review in the comments! For more usage examples, visit CodeAgent GitHub Repository.

[gemini-code-assist]

Summary of Changes

Hello @thewindwillstop, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request significantly upgrades the AI image recommendation system by integrating a robust filtering mechanism designed to personalize user experience and avoid content repetition. It also refines the AI asset name completion tool, making it more adaptable to diverse linguistic and formatting needs. These changes are supported by comprehensive documentation, flexible configuration, and enhanced system stability, collectively aiming to deliver a more intelligent and user-friendly AI backend.

Highlights

• Image Recommendation Filtering: Introduced a new feature to prevent users from repeatedly seeing the same AI-generated image recommendations, enhancing user experience. This includes sophisticated degradation strategies to ensure a continuous flow of relevant images even when filtering aggressively.
• Bilingual Asset Name Completion: Enhanced the AI-powered game asset name completion feature with improved multilingual support (Chinese/English) and more flexible naming conventions, allowing for spaces and hyphens in suggestions.
• Comprehensive API Documentation: Added a detailed API documentation for all AI image generation, recommendation, and related services, providing clear endpoints, parameters, and response structures.
• Configuration and Metrics: Implemented new configuration options and database models for managing image filtering settings, tracking recommendation history, and monitoring filtering performance and degradation events.
• Performance and Robustness Improvements: Increased timeouts for external AI services (algorithm, SVG.IO, Recraft) to improve reliability and refactored search result processing to directly use algorithm service data, reducing potential inconsistencies.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

This pull request introduces a comprehensive image filtering feature to prevent duplicate recommendations, which is a valuable enhancement. The implementation is well-structured, including a new filtering service with degradation strategies, corresponding database models, configuration, and documentation. However, I have identified several important issues that need attention. There is a critical issue with queryID handling that could break the feedback loop, a significant performance inefficiency in the search expansion logic, and a potential regression in the themed search ranking algorithm. Additionally, the new server timeout is excessively long, and the SQL setup script is missing a crucial table definition. Please review the detailed comments for suggestions on how to address these points to ensure the feature is robust, performant, and correct.

[niupilot]

Code Review Summary

I've identified several actionable improvements for this image filter implementation. The suggestions focus on high-confidence fixes for reliability, maintainability, and performance.

Key areas addressed:
• Error handling improvements for async operations
• Replace magic numbers with named constants
• Memory efficiency optimizations
• Configuration validation enhancements

[qiniu-ci]

This PR has been deployed to the preview environment. You can explore it using the preview URL.

Warning

Please note that deployments in the preview environment are temporary and will be automatically cleaned up after a certain period. Make sure to explore it before it is removed. For any questions, contact the Go+ Builder team.

[go-wyvern]

uuid 后面空一行