Verify or calculate offset from FileLength

[mkitti]

Currently the code reads a "FileLength" attribute from the header, but this is currently unused.

https://github.com/janelia-cosem/fibsem-tools/blob/f4bedbfc4ff81ec1b83282908ba6702baf98c734/src/fibsem_tools/io/fibsem.py#L587

One use of this would be to verify or calculate the OFFSET location rather than assuming that it is at 1024. That is OFFSET == FileLength - np.prod(shape)*np.dtype(dtype).itemsize

[d-v-b]

As long as the offset is in fact a constant, I think we should continue to treat it that way in the .dat reader. As for doing a validation, unfortunately it's possible for data to be incomplete (due to some problem during acquisition), and then the file length couldn't be used to infer the offset. Rather, in this situation the offset is needed for validating the file length.

[mkitti]

Perhaps we should actually record the offset somewhere then. It may be needed in case the metadata needs to expand.

[mkitti]

Is the FileLength verified anywhere in this repository?