[switch-nz]

Found by comparing the volume tree for a GPT disk image against FTK Imager's. Three differences, each fixed here.

What changes for examiners

• HFS+ volume name was never shown. The tree titled an HFS+ partition only "Partition N"; NTFS, exFAT, FAT and ext volumes show their label. The name is now read from the catalog's root folder record, which sorts first in the catalog, so it is read from the first leaf node without indexing the whole catalog (a large volume's catalog is not walked just to name it).
• GPT partition names were only in the details panel. The tree now shows the partition table's own name for the partition beside size and filesystem ("GPT name “…”"), apart from the filesystem's label in the title.
• The GPT's own tables counted as unpartitioned space. The first gap started at sector 1, so it included the GPT header and entry array, and the backup header and array at the end of the disk produced a trailing "Unpartitioned (trailing)" entry even on a disk with no spare room. Gaps are now bounded by the GPT's usable LBA range; gaps between partitions and spare room inside the usable range are still reported (hidden data can sit there).

Verification

• New tests/imagebuild_gpt.py and tests/test_volume_gpt.py (8 tests: names, HFS+ label, label without indexing the catalog, damaged catalog gives no label rather than an error, gap boundaries, trailing gap, no gaps on a full disk). Five fail and one errors on the old code.
• Checked in a browser on a synthetic GPT disk: title shows the HFS+ name, detail line shows the GPT name, no table gaps.
• Full suite 1306 tests OK (14 skipped).
• Not changed, but noted: Strata lists the small alignment gaps between partitions that FTK Imager does not; that is deliberate. APFS volume names are still not shown.

🤖 Generated with Claude Code

https://claude.ai/code/session_017DPjyZZHvaivtvJ3N4hZhz

---

Generated by Claude Code