Support Voice Memos file drops for meeting transcription(#219)

[Yaomeng1749]

Fixes #219

Summary

• Accept file promise drops from Voice Memos in Meeting Transcription.
• Allow selecting and queuing multiple local files for sequential transcription.
• Add focused coverage for supported URL filtering, drop/pasteboard type registration, and item-provider representation detection.

Tests

• swiftlint --strict --config .swiftlint.yml
• xcodebuild test -quiet -project Fluid.xcodeproj -scheme Fluid -destination 'platform=macOS' -derivedDataPath /Users/yaomeng/Library/Developer/Xcode/DerivedData/Fluid-afdlxoaqyckzxcdgcnblcegocvnz -clonedSourcePackagesDirPath /Users/yaomeng/Library/Developer/Xcode/DerivedData/Fluid-afdlxoaqyckzxcdgcnblcegocvnz/SourcePackages -disableAutomaticPackageResolution -only-testing:FluidDictationIntegrationTests/DictationE2ETests/testMeetingTranscriptionFileImport_filtersAndDedupesSupportedURLs -only-testing:FluidDictationIntegrationTests/DictationE2ETests/testMeetingTranscriptionFileImport_registersFileURLAndPromiseDropTypes -only-testing:FluidDictationIntegrationTests/DictationE2ETests/testMeetingTranscriptionFileImport_detectsAudioItemProviderRepresentations CODE_SIGNING_ALLOWED=NO

Local result bundle summary: 3 selected tests passed, 0 failed on macOS (My Mac). Xcode printed a CoreSimulator version warning before running the macOS tests.

[Yaomeng1749]

Rebased on current main and verified locally.

This fixes the import path that Voice Memos uses when dragging recordings into Meeting Transcription: those drops arrive as promised files, not just plain file URLs. The view now accepts both normal file URLs and file promises, supports multiple selected files, and queues the selected files for sequential transcription.

Checks run:

• swiftlint --strict --config .swiftlint.yml
• targeted xcodebuild test for the three new MeetingTranscriptionFileImport regression tests: 3 passed, 0 failed on macOS/My Mac.