Czkawka

Czkawka provides several different interfaces depending on your platform and needs:

• Krokiet: The new generation GUI frontend. It is simple, multiplatform (Linux, macOS, Windows), and uses the Slint framework. This is the recommended version for new users.
• Czkawka (GUI): The older GTK 4-based frontend. Note: Version 12.0 is the last released version; no new binaries will be provided.
• Cedinia: A touch-friendly Android GUI frontend built with Slint.
• Czkawka CLI: A command-line interface for easy automation.
• Czkawka Core: The underlying Rust library used by all frontends and available for external integration.

Cedinia provides a mobile interface for Czkawka Core scanning tools.

Supported Tools

It supports the same scanning tools as Krokiet, including:

• Duplicates
• Similar images and videos
• Empty files and folders
• Large files
• Invalid extensions
• Duplicate music
• And more

Missing Features

The following features are currently unavailable because they require FFmpeg:

• Similar video detection based on frames
• Detection of corrupted video files
• Video optimization

Known Issues

• Keyboard Support: Multiple issues exist due to bugs in the Slint framework.
• Display: Only portrait mode is supported with specific aspect ratios, which may cause layout issues on tablets or certain devices.

The Krokiet interface consists of six main areas:

1. Left panel: Tool selector containing 14 tools and a Settings gear.
2. Results area: Displays scan results as grouped rows.
3. Action bar: Contains Scan/Stop buttons, selection helpers, and action buttons (delete, move, etc.). Includes toggles for the Directory panel and Error log.
4. Status bar: Shows current scan state or result summary (e.g., "Found N items in Xms").
5. Directory panel: Toggled via the folder icon. Shows Included Paths (to be scanned) and Excluded Paths (to be ignored).
6. Right pane: Optional panel for image previews or tool sub-settings (e.g., for Similar Images).

The Czkawka GTK interface is organized into several functional areas:

• Image Preview (1): Displays images for duplicate or similar file comparison. It cannot be resized and can be disabled.
• Main Notebook (2): Used to switch between different scanning tools (e.g., Duplicate Files, Similar Images).
• Main Results Window (3): Displays scan results; used for choosing, deleting, and configuring files.
• Bottom Panel (4): Contains action buttons (like selection tools) and visibility toggles for GUI elements.
• Text Panel (5): Displays system messages, warnings, and errors. This panel can be hidden.
• Directory Panel (6): Used to manage included/excluded directories, allowed file extensions, and file size limits.
• Settings & About (7): Buttons to access application settings and information.

The duplicate finder uses a two-stage pipeline to optimize performance. Because of this, you will see two distinct cache files:

1. Prehash Cache: Stores fast, partial hashes computed from small chunks at the start and end of files. This is used to quickly eliminate non-duplicates. It is computationally cheap to build.
2. Full Hash Cache: Stores cryptographically strong hashes of the entire file content. This is only computed for files that passed the prehash stage and serves as the authoritative signal for deduplication.

Managing the cache:

• To fix corruption: Delete the files in the cache directory (e.g., ~/.cache/czkawka/). The next scan will rebuild them.
• To edit paths manually: Enable "Save also as JSON" in the GUI settings. You can then edit the .json file and delete the .bin file to force the app to load your edited version.

Developers can reuse Czkawka's functionality by using the czkawka_core crate. This core library is written in Rust and powers all official frontends.

External projects can interact with Czkawka in several ways:

1. Direct Dependency: Depend on the czkawka_core Rust crate.
2. Python Bindings: Use the official Python bindings available on PyPI.
3. CLI Wrapper: Wrap czkawka_cli to perform scans and retrieve results in JSON format.
4. Tauri: Use existing implementations like czkawka-tauri for web-tech based GUIs.

Czkawka identifies duplicates using a multi-stage hashing pipeline (size + prehash + full hash) rather than a byte-by-byte comparison. This is highly efficient and reliable.

Hashing Pipeline

1. Grouping: Files are first grouped by exact size.
2. Hashing: Only files with identical sizes are compared using hashes.
3. Algorithms:
   • Blake3 (Default): Fast and cryptographically strong. Recommended for deduplication.
   • CRC32: Fast checksum, not cryptographic.
   • XXH3 (xxHash): Very fast, not cryptographic.

Content vs. Name

• Hash method: Files with different names but identical content (same size and same hash) will be reported as duplicates.
• Name method: If you only want to find duplicates with the same filename, use the "Name" check method instead.

Scanning Symlinks

By default, Czkawka does not follow symbolic links. A symlink is treated as its own filesystem entry, and the scan will not traverse into the target. To scan the contents of a symlinked directory, add the target path directly to the included directories.

The "Symlink" Action

When you perform the "Symlink" action on selected (duplicate) files:

1. Czkawka replaces the selected files with symlinks.
2. These symlinks point to the first unselected file in the group.
3. The original (unselected) file remains intact.
4. The selected files are replaced in-place.

Note: On Windows, creating symlinks requires elevated privileges or Developer Mode to be enabled.