write positions.csv log during acquisition

[ieivanov]

pycromanager loses information on the position labels as discussed in micro-manager/pycro-manager#575. To keep track of the position labels we are currently writing a positions.csv log file as described in #29. The log file is currently written manually, after the acquisition has finished, using write_pos_csv.py where I copy-paste the positions dictionary from the acquisition log file. Instead, the positions.csv file should be generated during acquisition, writing one row for each new position during the first timepoint (only).

This is a temporary fix until micro-manager/pycro-manager#575 is resolved, the position label is the only unique metadata contained in this file.

An example positions.csv file can be found in rawdata/mantis/2023_04_20 HEK RAC1 PCNA/timelapse_2

[mattersoflight]

👍🏼, it goes without saying that these positions should be saved also with the reconstructed data. I see that there is already a PR to transfer this metadata from NDTIFF to ome-zarr (czbiohub-sf/iohub#118).