[seb5g]

Summary

The BeamShaping extension builds its Shaper from the Shaper control module (a DAQ_Move), using the contract of pymodaq_plugins_utils.hardware.shaper (pymodaq_plugins_utils 5.3.4, PyMoDAQ/pymodaq_plugins_utils#16): two read-only settings groups, shaper_geometry (pixels, pixel sizes) and shaper_properties (name, modulator type, has_internal_calibration, use_internal_calibration).

• extensions/spatial_beam_shaping.py: the shaper property returns Shaper.from_module(actuator) once the plugin is initialized, else the default shaper. use_internal_calibration is read live from the actuator. update_from_shaper() re-reads it and rebuilds the fields (input and target sizes, pixel sizes, masks); it is called in do_things_after_experiment_set, when the actuator emits init_signal (the plugin may be initialized after the experiment is set) and when the configuration changes.
• utilities/shaper.py: Shaper and ModulatorType come from pymodaq_plugins_utils (re-exported). The per-SLM configuration is removed: the [SLM] section of the config template, slm/slm.py (unused) and the config lookups in utilities/sizing.py, the loaders and the mock plugin. A single [default_shaper] section (1920 x 1080 pixels of 8 µm, phase, no internal calibration) is the fallback for all shapers, editable by the user; get_shaper() returns the shaper last told by the actuator (set_shaper), so sizing and the algorithms follow it. An old [SLM] section left in a user's config file is ignored.
• DAQ_Move_MockShaper fills both groups with the default shaper.
• New dependency: pymodaq_plugins_utils>=5.3.4.

Tests

tests/test_shaper_interface.py (not initialized / initialized actuator, user choice of the internal calibration, no actuator / no groups, default shaper from the configuration, sizes follow the shaper, mock) and test_custom_apps.py (extension rebuilds its fields for the shaper told by the actuator). The init_signal connection to a real DAQ_Move is not tested. Not tested with a real SLM.

🤖 Generated with Claude Code

https://claude.ai/code/session_01CAhZwmbyCSpSgfTyicUxuS

[seb5g]

the idea is to remove entirely the need of the shaper configuration. Only a default fallback (for all shapers) should be done

[seb5g]

in fact the best woul dbe to have the settings from the real instrument if any, else fallback but in that case one should be able to update them => I guess this is possible by tweaking the preferences then?