Asset Cleaner

Changed

• Improved the asset usage modal so draft and revision usage can be identified directly from the displayed entry links
• Renamed user photo usage labels in the usage modal to "User profile picture" for clearer presentation

Fixed

• Split large relation scans into resumable queue batches so high-volume scans are less likely to exceed worker timeouts
• Hardened relation scan progress tracking so stale queued relation batches are less likely to overwrite later-stage scan state
• Fixed the scan progress display so the asset counter advances during long-running scans instead of remaining at 0
• Counted asset references from the users.photoId column as usage during scans and usage checks

Changed

• Split AssetUsageService into focused collaborators for entry usage resolution, relation usage resolution, and content usage scanning while preserving the existing public plugin service API
• Standardized plugin logging so operational warnings remain useful without flooding the dedicated log with temporary relation-resolution investigation traces
• Reduced scan lookup memory usage during content scans

Fixed

• Fixed relation fallback usage checks so draft-only and revision-only relation sources no longer count as usage when those inclusion options are disabled
• Fixed relation fallback handling to ignore trashed relation sources instead of treating them as generic usage
• Fixed relation source resolution across sites and owner states for relation-backed asset usage checks
• Fixed invalid field handle warnings during content usage checks by only reading HTML-capable fields from the concrete element field layout context
• Hardened content field layout discovery so problematic layouts fail safely with warnings instead of breaking scans

Added

• Added configurable scan workspace path support via the ASSET_CLEANER_SCAN_PATH environment variable or scanWorkspacePath in config/asset-cleaner.php
• Added a database-backed scan storage mode for containerized, cloud-style, and multi-worker environments where shared filesystem access is not guaranteed
• Added plugin settings for selecting the scan storage mode and configuring the file-based scan workspace path
• Added default settings for whether draft-only and revision-only asset references should count as used
• Added per-scan utility toggles for including drafts and revisions without changing the global defaults
• Added a per-scan relational fallback option that treats any row in Craft’s relations table as usage, helping preserve assets referenced by plugin-defined or otherwise unknown element types
• Added a configurable asset usage dialog on the asset edit page so draft, revision, and relational fallback behavior can be checked before reviewing usage results
• Added updated translations for the new scan and usage options

Changed

• Refactored scan persistence into storage backends so the scan coordinator can run against either file-based or database-based storage
• Batched content entry processing during scans to avoid materializing all candidate entries in memory at once
• Hardened scan workspace file writes with stricter validation, readability checks, and better failure handling
• Retained only the latest scan for restore/export workflows while allowing stale queued jobs to exit quietly if their scan has been replaced
• Made draft and revision handling explicit in scan behavior so installations can choose between canonical/live-oriented cleanup and editorial-history-aware cleanup
• Defaulted scans to the safer relational fallback mode while still allowing stricter relation resolution when needed
• Updated the utility and usage UI to expose the new scan-safety controls more clearly

Fixed

• Added path-aware diagnostics when scan metadata files are missing or unreadable, including resolved storage context details to help debug container and shared-storage setups
• Improved support for container-based installs by allowing scan state to be stored in the database instead of relying on shared local storage

Fixed

• Replaced the cache-backed asset scan flow with a file-backed staged scan pipeline stored under @storage/asset-cleaner/scans/<scanId>
• Fixed large scans failing when the cached content index expired before the queue finished processing
• Removed the need to pass the full asset ID list through each queued batch job

Changed

• Split background scans into dedicated setup, relations, content, and finalize queue stages
• Snapshotted assets into chunk files and resolved usage from relations and content in one pass per scan
• Updated scan progress reporting to reflect the current scan stage
• Restored legacy CSV path formatting for unused asset exports

Added

• Plugin icon with broom and dust particles for control panel and plugin store
• Custom Logger helper class for dedicated error logging to storage/logs/asset-cleaner_YYYY-MM-DD.log
• Comprehensive error handling with try/catch blocks in all controller actions
• Translation support for 20 languages: Arabic, Bulgarian, Czech, German, English, Spanish, French, Hungarian, Italian, Japanese, Korean, Dutch, Polish, Portuguese, Romanian, Russian, Slovak, Turkish, Ukrainian, and Chinese

Changed

• Updated utility icon to use custom SVG icon

Added

• Initial release
• Scan assets across multiple volumes to identify unused assets
• Per-volume results display with separate tables
• Bulk actions: Download CSV, Download ZIP, Move to Trash, Delete Permanently
• ZIP download with folder structure preservation option
• Memory-efficient ZIP creation for large files using chunked processing
• Smart filename generation with snake_case sanitization and timestamps
• Progress bar for scanning operations
• "View Usage" button on asset edit pages showing where assets are used
• Soft delete (trash) and permanent delete options with confirmation dialogs
• Asset usage detection in entry relations and content fields
• Support for Craft CMS 5.0+