Fix VisuCoreWordType spelling for 8-bit unsigned 2dseq (_8BIT_UNSGN_INT)

[gdevenyi]

Problem

config/properties_2dseq_core.json selects the numpy dtype for 8-bit unsigned 2dseq data by matching VisuCoreWordType == "_8BIT_USGN_INT". That spelling is wrong — ParaVision spells the enum _8BIT_UNSGN_INT (with the N):

• RECO_WORDTYPE in prog/include/recotyp.h: { _32BIT_SGN_INT, _16BIT_SGN_INT, _8BIT_UNSGN_INT, _32BIT_FLOAT }
• real visu_pars files store ##$VisuCoreWordType=_8BIT_UNSGN_INT

Because the stored value never matches _8BIT_USGN_INT, no numpy_dtype branch fires, the property is never set, and Schema.__init__ raises MissingProperty("numpy_dtype") — an 8-bit-unsigned 2dseq cannot be loaded. The 2dseq numpy_dtype selection has no ACQ_sw_version gate, so the failure is version-independent and depends only on VisuCoreWordType.

Fix

Correct both occurrences (little- and big-endian branches) to _8BIT_UNSGN_INT. The signed 16/32-bit and float branches were already correct.

Public dataset that reproduces it

The MRIReco.jl Bruker test data contains a reconstruction stored as 8-bit unsigned:

• Download (public, no login): https://media.tuhh.de/ibi/mrireco/MRIRecoTestData.tar.gz  (sha256 c5a421aab7f3ea3fb20c469704db33c17b659861d9d77dbd29e353db2d5c8fc7)
• Inside the archive: BrukerFile/2D_FLASH/pdata/7/ — its visu_pars has ##$VisuCoreWordType=_8BIT_UNSGN_INT and ##$VisuCoreByteOrder=littleEndian.

from brukerapi.dataset import Dataset
Dataset("<extracted>/BrukerFile/2D_FLASH/pdata/7/2dseq")
# before this change: MissingProperty: numpy_dtype
# after:  loads as a uint8 array

🤖 Generated with Claude Code

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.