- Introduced `IAiExtractionService` and its implementation `AiExtractionService` for processing images and extracting text.
- Created `AiResultItem` model to hold results from AI extraction.
- Added `ImageProcessingCoordinator` to manage image processing tasks and provide progress updates.
- Implemented view models for AI settings, path settings, processing state, race upload settings, and visual settings to support UI binding.
- Updated `Program.cs` to register new services and dependencies.
- Modified project file to skip MinVer execution during local builds.
- Added ImageCreatorImageSharp class for image creation, handling EXIF orientation, resizing, and saving images.
- Replaced GDI+ dependencies with ImageSharp for cross-platform compatibility.
- Introduced methods for drawing text and logos on images, including handling transparency and positioning.
- Created a test plan for validating ImageCreatorImageSharp functionality, focusing on image resizing, text positioning, logo features, and EXIF orientation.
- Added documentation for the test plan outlining goals, project structure, and implementation notes.
- Implement GeneralTabView for source and destination path settings, options, processing parameters, and image library selection.
- Create LogoTabView for logo selection, preview, and positioning options.
- Add PhotoTabView for photo dimensions and JPEG quality settings.
- Introduce RaceUploadTabView for race setup and processed photo upload functionality, including API integration.
- Develop TextTabView for horizontal and vertical text settings, font options, and race time configuration.
- Implement ThumbnailsTabView for thumbnail creation options and settings.
- Implemented IRaceUploadCommunicationClient with methods for saving races, creating race points, indexing race points, retrieving race details, and uploading files to the receiver.
- Added ReceiveFilePath option to CatalogCommunicationOptions for file transfer configuration.
- Enhanced CatalogCommunicationServiceCollectionExtensions to validate ReceiveFilePath.
- Developed RaceUploadCommunicationClient to handle race-related API interactions, including saving race data and uploading processed images.
- Updated API documentation to reflect new race upload and file transfer endpoints.
- Modified Avalonia UI to support race creation and processed image uploads, including new input fields and buttons.
- Introduced RaceSaveRequest and ReceiveFileUploadRequest models for structured data handling.
Refactored image creation APIs to use byte[] for logo data instead of System.Drawing.Image, enabling cross-platform support. Wrapped all GDI+/Windows-specific code in #if WINDOWS and updated project files to conditionally include Windows-only dependencies. Defaulted to ImageSharp on non-Windows, and updated UI and settings to reflect platform capabilities. Application now builds and runs on Linux/macOS with Avalonia and ImageSharp, while retaining full Windows functionality.
Introduce Avalonia as a new cross-platform UI option, including new XAML and code-behind files for the application and main window. Update Program.cs to support a --avalonia launch argument and add corresponding launch profile. Integrate Avalonia NuGet packages and ensure DataModel supports UI-thread invocation for all frontends. All business logic and state are shared via DI, enabling consistent behavior across WinForms, WPF, and Avalonia.
Refactored Program.cs to accept command-line arguments and select between WinForms (default) and WPF UI based on the presence of --wpf. Removed old commented-out UI selection code. Updated launchSettings.json to provide separate profiles for WinForms and WPF launches. If WPF is requested but unavailable, the app falls back to WinForms.
- Moved "Create Thumbnails" checkbox to the Thumbnails section.
- Relocated "Add Time" and "Add Race Time" controls to a new "Tempo Gara" section in the text options area.
- Enabled multiline input and wrapping for the Vertical Text box.
- Replaced text color selection with MahApps ColorPicker for better UX.
- Moved "Select Transparent Color" button to the logo options section.
- Made minor UI layout improvements for clarity.
Integrated MahApps.Metro.IconPacks for Material icons in tab headers and buttons, enhancing visual clarity and user experience. Refactored resource dictionary to define theme brushes at the top level for easier theme switching. All button content now combines icons and text for a more modern interface. No functional changes; UI/UX improvements only.
Updated text rendering logic to center each line horizontally
based on its estimated width, ensuring multi-line text is
visually centered. Previously, all lines were drawn at a fixed
horizontal position. Both shadow and main text now use the
calculated centered position.
- Use vertical text variant (TestoFirmaV) for rotated images to ensure correct orientation of overlays.
- Adjust font sizing logic for multi-line text to better fit within image bounds, accounting for line count and spacing.
- Set rotation flags (FotoRuotaADestra, FotoRuotaASinistra) before drawing text, enabling correct text variant selection.
- Ensure vertical text field is initialized from settings for consistent behavior.
Ensures EXIF orientation is cleared when cloning and after rotation,
preventing incorrect display in EXIF-aware viewers. Failures are
logged as debug messages without interrupting processing.
- Enhance ImageCreatorAlternate to support multi-line text rendering with correct sizing, spacing, and shadow for each line.
- Only draw logos on non-thumbnail images to match GDI behavior.
- Add a status bar to MainWindow showing the app version at runtime.
- Upgrade MinVer to 7.0.0 and adjust versioning to avoid WPF/XAML assembly identity issues.
- Refactor XAML layout to accommodate the new status bar.
Enhance private NuGet source setup in build_windows and publish_release jobs to support fallback authentication using NUGET_USERNAME/NUGET_PASSWORD when CI_JOB_TOKEN is unavailable. Update log messages for clarity. Add diagnostic steps in build_windows to verify NuGet feed and package visibility using configured credentials. No changes to core build or publish logic.
Switched private NuGet source authentication in build_windows and publish_release jobs to use CI_JOB_TOKEN instead of NUGET_USERNAME/NUGET_PASSWORD. Updated related log messages and ensured PowerShell variable usage is correct for environment variables. This improves compatibility with GitLab CI best practices.
Renamed the private NuGet source from 'GitLab' to 'Nuget-GitLab-AIFotoONLUS' in CI scripts for clarity and consistency. Added a repository-level NuGet.Config with packageSourceMapping to ensure only specific private packages are restored from the GitLab feed, preventing 401 errors when restoring public packages from nuget.org.
- Add .gitlab-ci.yml for Windows CI/CD: build, publish, and GitLab Release with artifact upload
- Replace GitVersion.MsBuild with MinVer for versioning; update all versioning logic in .csproj
- Remove GitVersion CLI fallback for Visual Studio builds
- Update post-publish renaming to use MinVer's Version property
- Replace multiple thumbnail text checkboxes with a single ComboBox selector in MainWindow.xaml
- Change default UI startup: prefer WPF MainWindow, fallback to WinForms MainForm
Replaced multiple mutually-exclusive boolean properties for thumbnail text options with a single enum (`ThumbnailOption`) in the data model. Updated WinForms UI to use a ComboBox instead of radio buttons for selecting thumbnail mode. Adjusted designer, mapping profile, settings DTO, and settings service for enum support and backward compatibility. Simplified thumbnail generation logic and improved maintainability by ensuring only one mode can be selected at a time.
Implement mutually exclusive logic for thumbnail display modes in DataModel and MainForm. Add ThumbnailMode property for authoritative state. Replace radio button bindings with event handlers. Synchronize UI with model changes. Explicitly map thumbnail flags to PicSettings. Force WinForms startup, disabling WPF branch.
Integrates optional AI/OCR (AIFotoONLUS.Core) support to extract numbers from images after processing. Adds new "AI" tab in the UI for enabling extraction, selecting models folder, specifying CSV output, and previewing results. Results can be exported to CSV. Uses reflection for AI library invocation, with fallback simulation if unavailable. Persists new AI settings. Updates related NuGet packages and adds theme resources.
Introduced MainWindow.xaml and code-behind for a modern WPF interface with tabbed settings. Enabled WPF in project file. Refactored startup to launch WPF or WinForms UI based on availability. Registered WPF MainWindow in DI. Improved DataModel cancellation to be synchronous and log exceptions. Minor logging and comment updates. App now supports both WPF and WinForms frontends.
