pyiceberg/avro/reader.py (2 lines):
	- line 336: # TODO: Implement struct-reuse
	- line 356: # TODO: Implement struct-reuse


pyiceberg/table/puffin.py (1 line):
	- line 56: # TODO: Optimize this


pyiceberg/io/pyarrow.py (1 line):
	- line 2694: # TODO: As a next step, we could also play around with yielding instead of materializing the full list


pyiceberg/table/update/validate.py (1 line):
	- line 59: # TODO: Maybe do the IO in a separate thread at some point, and collect at the bottom (we can easily merge the sets