Nexus file generates wrong omega for multipin when alternating direction

See /dls/i03/data/2026/mx23694-160/auto/TestInsulin/20250428-mp1601/

It seems nexus files contain omega 360 -> 720 "even" rotation sweeps instead of 360 -> 0, this manifest in xia2-dials failing for every other sample in multipin.

```
[yqf46943@ws502 20250428-mp1601]$ h5ls -d 20250428-mp1601_4.nxs/entry/sample/sample_omega
omega Dataset {3600}
Data:
(0) 360, 360.1, 360.2, 360.3, 360.4, 360.5, 360.6, 360.7, 360.8, 360.9, 361, 361.1, 361.2, 361.3, 361.4, 361.5, 361.6, 361.7, 361.8, 361.9, 362, 362.1, 362.2, 362.3, 362.4, 362.5, 362.6, 362.7, 362.8, 362.9, 363,
(31) 363.1, 363.2, 363.3, 363.4, 363.5, 363.6, 363.7, 363.8, 363.9, 364, 364.1, 364.2, 364.3, 364.4, 364.5, 364.6, 364.7, 364.8, 364.9, 365, 365.1, 365.2, 365.3, 365.4, 365.5, 365.6, 365.7, 365.8, 365.9, 366,

```

Suspect this is due to scanspec being generated incorrectly for the even sweeps as it doesn't appear to account for rotation direction parameter. It appears that rotation likely works correctly because it does not use the scanspec to program the zebra