don't add hashes to file names

[donovanr]

trading clarity for efficiency with an option to not mangle the names of the files would make using this a lot more friendly.

the cardio fish data set is an example, where mysterious hash like strings confused pretty much everyone, and will definitely be odd to end-user biologists.

[evamaxfield]

How do you propose I should counteract this:

# Just using val.name could result in files that shouldn't be grouped being grouped
# Example column:
# SourceReadpath
# a/0.tiff
# a/1.tiff
# b/0.tiff
# b/1.tiff
# Even though there are four files, this would result in both a/0.tiff and b/0.tiff, and,
# a/1.tiff and b/1.tiff being grouped together. To solve this we can prepend a the first couple
# of characters from a hash of the fully resolved path to the logical key.