get_image_collection() returns empty Vec when reading xlsx with embedded images

[wolfiesch]

Description

When reading an .xlsx file that contains embedded images (inserted via Excel), worksheet.get_image_collection() returns an empty vector. The images are present in the file and visible when opened in Excel.

Note: The write path works correctly — images added via umya's API are properly embedded and visible in Excel. This issue only affects the read path.

Steps to Reproduce

1. Create an .xlsx file in Excel with one or more embedded images (Insert > Pictures)
2. Read the file with umya-spreadsheet
3. Call worksheet.get_image_collection() or iterate media objects

Expected Behavior

Returns a vector containing the embedded image objects with their position, dimensions, and image data.

Actual Behavior

Returns an empty vector ([]).

Possibly Related

• Issue has_pivot_table() and get_pivot_table() not pulling pivot table #278 (pivot table read returning empty) — may share a similar root cause in the relationship/parts parsing logic.

Environment

• umya-spreadsheet 2.2.x
• Tested with .xlsx files generated by Microsoft Excel (macOS)
• Discovered via ExcelBench fidelity testing

[MathNya]

@wolfiesch
Thank you for your report.
We are currently working on addressing the pivot table-related issue.
Resolving this may also resolve your issue.
We will look into it.