While upgrading looper and ensuring its compatibility with the new version of peppy, I encountered several issues. One of these issues pertained to the use of the __reduce__ method in `peppy.Sample for object serialization. I successfully addressed this problem in the Samples class (unfortunately without inherited Project - Sample is independent).
Now, a question arises: Given that we have implemented the__reduce__ method in the Sample class, should we also consider adding serialization capabilities to the Project class? However, the current challenge is that the __init__ method of the Project class currently only accepts paths to physical files. To enable serialization, we would need to introduce an additional argument in the __init__ method to accommodate a dictionary. This approach was discussed a few months ago, and decided not to add this argument, but instead make separate method from_dict
While upgrading looper and ensuring its compatibility with the new version of peppy, I encountered several issues. One of these issues pertained to the use of the
__reduce__
method in `peppy.Sample for object serialization. I successfully addressed this problem in the Samples class (unfortunately without inherited Project - Sample is independent).Now, a question arises: Given that we have implemented the
__reduce__
method in theSample
class, should we also consider adding serialization capabilities to theProject
class? However, the current challenge is that the__init__
method of theProject
class currently only accepts paths to physical files. To enable serialization, we would need to introduce an additional argument in the__init__
method to accommodate a dictionary. This approach was discussed a few months ago, and decided not to add this argument, but instead make separate methodfrom_dict