Add HEIC/HEIF EXIF metadata support for PowerRename

[ThanhNguyxn]

Summary of the Pull Request

This PR enables PowerRename to extract EXIF metadata from HEIC/HEIF images (Apple's photo format) for use in file renaming patterns.

References and Relevant Issues

Fixes #43758

Detailed Description of the Pull Request / Additional comments

Root Cause Analysis

HEIC/HEIF files store EXIF metadata under a different path structure than JPEG/TIFF files:

• JPEG/TIFF: /app1/ifd/exif/{ushort=...} and /app1/ifd/gps/{ushort=...}
• HEIC/HEIF: /ifd/exif/{ushort=...} and /ifd/gps/{ushort=...}

The existing implementation only queried the standard JPEG/TIFF paths, so HEIC metadata was not found.

Changes Made

1. Helpers.cpp: Added .heic and .heif to the supportedExtensions set in isMetadataUsed() function
2. WICMetadataExtractor.cpp:
   • Added 23 HEIC-specific metadata query paths (18 EXIF + 5 GPS)
   • Added 4 fallback methods: ReadDateTimeWithFallback, ReadStringWithFallback, ReadIntegerWithFallback, ReadDoubleWithFallback
   • Updated ExtractAllEXIFFields() to use fallback methods (try standard path first, then HEIC path)
   • Updated ExtractGPSData() to use fallback paths for GPS data
3. WICMetadataExtractor.h: Added declarations for the 4 fallback methods

Testing

• Tested with HEIC images from iPhone containing EXIF data
• Verified that standard JPEG/TIFF metadata extraction still works correctly

Validation Steps Performed

• Built and verified no compilation errors
• Tested metadata extraction from HEIC files
• Tested that JPEG/TIFF extraction still works (no regression)

[ThanhNguyxn]

@microsoft-github-policy-service agree

[Copilot]

Pull request overview

This PR adds support for extracting EXIF metadata from HEIC/HEIF image files (Apple's modern photo format) to PowerRename. HEIC files store EXIF data under different WIC metadata paths (/ifd/...) compared to JPEG/TIFF files (/app1/ifd/...), which prevented metadata extraction in the original implementation.

Key Changes

• Added .heic and .heif extensions to the supported file types for metadata extraction
• Implemented fallback mechanism that tries standard JPEG/TIFF paths first, then HEIC-specific paths
• Added 23 HEIC-specific metadata path constants for EXIF and GPS data

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File	Description
src/modules/powerrename/lib/Helpers.cpp	Added .heic and .heif to the list of supported image extensions for metadata extraction
src/modules/powerrename/lib/WICMetadataExtractor.h	Added declarations for four fallback helper methods (ReadDateTimeWithFallback, ReadStringWithFallback, ReadIntegerWithFallback, ReadDoubleWithFallback)
src/modules/powerrename/lib/WICMetadataExtractor.cpp	Added 23 HEIC-specific metadata path constants, implemented four fallback helper methods, and updated ExtractAllEXIFFields and ExtractGPSData to use fallback paths for HEIC support

[Copilot]

Missing test coverage for HEIC/HEIF metadata extraction. The existing WICMetadataExtractorTests.cpp has comprehensive tests for JPEG files, but no tests for the new HEIC/HEIF support. Consider adding test cases with sample HEIC files to verify:

1. EXIF metadata extraction from HEIC files
2. GPS data extraction from HEIC files
3. Fallback behavior works correctly when HEIC paths are used

This is especially important given that the HEIC metadata paths differ from JPEG/TIFF and could have edge cases.

[Copilot]

[nitpick] The ExtractGPSData method uses a repetitive if-check pattern for each GPS field fallback. Consider refactoring this to use helper methods similar to ReadDoubleWithFallback for consistency and reduced code duplication. For example, you could create a ReadMetadataWithFallback helper that takes primary and fallback paths.

[daverayment]

Hi.

I have a few concerns with this approach. Most importantly, there are no unit tests here. Please add some specific unit tests to cover each of the formats you aim to support. (There are existing tests for the JPEG format in the WICMetadataExtractorTests.cpp file. It may be possible to parameterise these, but that would need investigating.)

Secondly, while your code may work for the specific HEIC test case, it relies on the duplication of every metadata path and hard-coding specific fallback logic. This violates DRY and isn't scalable, creating a maintenance burden. For example, if we wanted to add support for PNG or WebP in the future, we'd have to declare more constants and more fallbacks.

A more flexible solution would be to separate the EXIF tag IDs (the {ushort=nnn} strings) from the paths. The tags are identical across the file formats, and the only difference is that the JPEG path differs from the TIFF/HEIC one, e.g. /app1/ifd/exif/ versus /ifd/exif.

I'd recommend restarting with a more extensible approach. For example:

• Remove the duplicate constants
• Declare each of the EXIF tags once
• Create a vector of the valid root paths
• Refactor the reading of metadata to search through the root paths combined with the tag IDs until a match is found
• etc.

This will reduce the repetition, the need for fallback logic and allow us to extend the approach to other formats in the future.

[ThanhNguyxn]

@daverayment Thank you for the detailed feedback! I've completely refactored the implementation based on your suggestions.

Changes Made

1. Extensible Path System (DRY Principle)

• Separated EXIF tag IDs from root paths: Each tag ID (e.g., {ushort=36867} for DateTimeOriginal) is now declared once
• Created root path vectors: IFD_ROOT_PATHS, EXIF_ROOT_PATHS, GPS_ROOT_PATHS that can be easily extended
• Added BuildMetadataPaths() helper: Combines root paths with tag IDs to build full query paths

// Root paths for EXIF sub-IFD metadata
const std::vector<std::wstring> EXIF_ROOT_PATHS = {
    L"/app1/ifd/exif",  // JPEG (APP1 marker)
    L"/ifd/exif"        // HEIC/HEIF/TIFF
};

// Tag IDs (identical across formats)
const std::wstring TAG_DATE_TAKEN = L"{ushort=36867}";

2. Multi-Path Reading Methods

• Added ReadDateTimeFromPaths(), ReadStringFromPaths(), ReadIntegerFromPaths(), ReadDoubleFromPaths(), ReadMetadataFromPaths()
• These methods iterate through all paths until metadata is found
• Removed the repetitive WithFallback methods

3. Future Extensibility

To add support for new formats (PNG, WebP, etc.), you simply:

1. Add new root paths to the vectors
2. No changes needed to extraction logic

4. Unit Tests for HEIC/HEIF

• Added ExtractHEICMetadataTests test class
• Tests gracefully skip when HEIC test files aren't available
• Tests handle missing HEIF Image Extensions gracefully

Note: I didn't include HEIC test files because:

1. Need to find/create properly licensed test images
2. HEIC testing requires Microsoft's HEIF Image Extensions to be installed

Would you like me to:

1. Create a simple HEIC test image with known metadata?
2. Add any additional test cases?

Thank you again for the guidance on making this more maintainable!