FastCS eiger testing 17/12/25

- `grid_detect_then_xray_centre` not called by `pin_centre_then_xray_centre`, meaning fastcs eiger is never prepared. I would guess this also means old eiger isnt staged? But that seems to work fine, need to ask @DominicOram - relates to https://github.com/DiamondLightSource/mx-bluesky/issues/1522
- Need to set up `beamline_specific_parameters` before Preparing eiger. Update: this doesn't work as the path provider is different. Need to `update()` the i03 path provider.
- Need to pass correct eiger into `kickoff_and_complete_gridscan`
- `mx_bluesky.common.utils.exceptions.ISPyBDepositionNotMadeError: No ISPyB IDs received by the end of self.triggering_plan='do_fgs'` - probably due to wrong key in `doc["data"]` in zocalo callback. Fix with `filename = doc["data"]["fastcs_eiger-odin-id"]`
- Getting `StopIteration` in `zocalo_callback`, think fixed with above
- Seems to hang at end of collection, then got `timed out waiting for zocalo results`. Caused by no data files being written, only master files. Maybe something to do with path provider? Currently we are waiting on zocalo results before doing eiger complete, we should swap this around so we get a sensible error message if there's no data.