[PowerOCR][WIP] migrate PowerOCR to WinUI 3

[moooyo]

Summary of the Pull Request

Migrates Text Extractor (PowerOCR) from WPF to WinUI 3 while preserving its Runner integration, settings contract, activation behavior, telemetry, and UI Automation identifiers.

• Replaces the WPF overlay with native WinUI 3 WindowEx overlays, one per display, with mixed-DPI and negative-origin coordinate handling.
• Extracts OCR, bitmap preparation, text formatting, table analysis, and selection geometry into the UI-independent PowerOCR.Core library.
• Migrates localization from .resx to .resw/x:Uid, and updates process lifetime, Named Event dispatch, clipboard handling, and deterministic resource cleanup.
• Moves managed Text Extractor outputs into WinUI3Apps and updates the native launcher, signing manifest, installer resources, and installation verification list.
• Adds Core unit tests and updates the existing PowerOCR UI automation/checklist for the WinUI controls.

Part of the WinUI 3 migration effort tracked in #46463.

PR Checklist

• Closes: Migrate Text Extractor to WinUI3 #49656
• Communication: I've discussed this with core contributors already. If the work hasn't been agreed, this work might be rejected
• Tests: Added/updated and all pass — Core tests pass, but interactive UI tests and several visual/multi-monitor checks remain environment-blocked as detailed below
• Localization: All end-user-facing strings can be localized
• Dev docs: Added/updated
• New binaries: Added on the required places
  • JSON for signing for PowerToys.PowerOCR.Core.dll and the WinUI3Apps paths
  • WXS for installer — handled by the existing WinUI3Apps component generator; removed the obsolete PowerOCR satellite-resource component from installer/PowerToysSetupVNext/Resources.wxs
  • YML for CI pipeline — N/A; the new test project is included in PowerToys.slnx and uses existing test discovery
  • YML for signed pipeline — N/A; no per-module pipeline entry is required
• Documentation updated: If checked, please file a pull request on our docs repo and link it here: #xxx — N/A; no end-user behavior change

Detailed Description of the Pull Request / Additional comments

Core and project structure

• Adds src/modules/PowerOCR/PowerOCR.Core/ and PowerOCR.Core.UnitTests/.
• Removes production UseWPF, System.Windows, WPF imaging types, MEF construction, and the WPF-only Common.UI dependency.
• Converts src/modules/PowerOCR/PowerOCR/PowerOCR.csproj to an unpackaged, self-contained WinUI 3 application with XAML under PowerOCRXAML/.

Overlay and OCR flow

• Captures every display before showing overlays, then creates one borderless topmost overlay per DisplayArea.
• Uses cached physical-pixel screenshots for clicked-word, region, single-line, and table OCR modes.
• Converts WinUI DIP pointer coordinates through each window's rasterization scale and supports displays with negative virtual-desktop origins.
• Replaces the WPF CombinedGeometry selection clip with four mask rectangles plus a selection Border.
• Synchronizes language and formatting state across all overlay windows and preserves Escape, S, T, and D1-D9 keyboard behavior.

Lifetime and compatibility

• Preserves the existing process name, mutex, Runner PID argument, GPO behavior, Named Events, standalone keyboard hook, settings JSON, telemetry events, and existing AutomationIds.
• Adds cancellation-aware Named Event handling and deterministic disposal for windows, screenshots, WinRT image sources, OCR crops, pointer capture, and cursor clipping.
• Keeps PowerOCRModuleInterface.dll in the install root and launches WinUI3Apps\PowerToys.PowerOCR.exe.

Validation Steps Performed

• Built PowerOCR.Core.UnitTests for x64 Debug and ran all tests with vstest.console.exe: 23/23 passed.
• Built PowerOCR-UITests for x64 Debug successfully.
• Built PowerOCR for x64 Release and ARM64 Release successfully.
• Built PowerOCRModuleInterface for x64 Release and ARM64 Release successfully.
• Ran XAML Styler against the branch: 3 PowerOCR XAML files processed with no changes.
• Confirmed the required x64/ARM64 outputs, including PowerToys.PowerOCR.Core.dll and PowerToys.PowerOCR.pri, are generated in the expected locations.
• Ran build+sideload checklist validation against the local WinUI 3 bits: 15 PASS, 0 FAIL, 16 BLOCKED.
  • Blocked items require an attached interactive desktop, WinAppDriver v1.2.1, multiple displays/mixed DPI, real pointer drag/input, or visual theme inspection.
  • No product defects were observed in the checks that could be executed.
• The legacy UI test runtime remains blocked because WinAppDriver is unavailable and the current RDP input desktop is detached; the tests compile successfully.