[seb5g]

Stacked on #34 (base is its branch; retarget to 5.0.x once #34 is merged).

Before: daq_1Dviewer_ShamrockSCMOS and daq_1Dviewer_ShamrockSCMOSLegacy import daq_move_plugins.daq_move_Shamrock, which #32 removed, so they cannot be imported on 5.0.x.

After:

• ShamrockSCMOS (composite) uses DAQ_Move_ShamrockPll for the spectrometer, as ShamrockCCDPll does: SetNumberPixelsSR, SetPixelWidthSR and GetCalibrationSR become set_number_pixels, set_pixel_width and get_calibration() (metres, converted to nm), and the wavelength is read with get_wavelength(). The number of pixels is set again before each calibration read, because the ROI may have changed.
• ShamrockSCMOSLegacy (double inheritance) is left as it is and imports the actuator from pymodaq_plugins_andor.legacy (Keep the replaced plugins in a legacy folder #34).
• Both declare supported_platforms = ('win32',): they load the Andor SDK3 (and, for the legacy one, the Shamrock SDK) with ctypes.windll. With the declaration, PyMoDAQ >= 5.3.5 refuses to initialize them elsewhere and the plugin tests skip them instead of failing.

Tested: no hardware. The file compiles and check_plugin was run locally on Linux. Not verified: the SDK3 camera and the pylablib Shamrock in the same process, and the calibration values and flip of the wavelength axis against a real spectrometer. The CI of 5.0.x is red for reasons I cannot read in the logs.

🤖 Generated with Claude Code

https://claude.ai/code/session_01Tdiso1u1ZhikSutj2Qs5Cg

[seb5g]

tests / tests is red here as on the base (5.0.x after #32, and #34). The only information I can read is "Process completed with exit code 1" on the Tests step: the job logs are not readable from my session, so I cannot tell yet whether the SCMOS imports were part of the failure or whether another plugin of the package fails first. I will not guess a fix; the failing log lines are needed to continue.

---

Generated by Claude Code