Solve error for big .fli files

[rharkes]

Big files gave an integer overflow on line 191.

[sbesson]

Thanks @rharkes. Including the changes for the daily CI builds. We do not have any representative .fli file that shows such an integer overflow. Would you have one you would be able to share with us so that we could add it to our non-regression tests?

[rharkes]

@sbesson Yes I have one, but it is 2.3 GB in size, since the problem does not occur below 2GB.

[melissalinkert]

@rharkes: alternate instructions for sending large files have been sent to you via private email.

[melissalinkert]

Thank you for uploading sample files, @rharkes.

@openmicroscopy/formats, files copied to inbox/gh-3303/ for testing.

[dgault]

In this particular case the Integer overflow is being caused by the large number of timepoints as opposed to the bytes per plane. As such the change makes sense and other instances of plane size do not need to be updated.

I tested using the large sample file provided.

Without the PR opening the image and reading one of the later timepoints results in:
java.lang.IllegalArgumentException: Negative position

With the PR included I was able to open and read all of the planes within the image without issue.

This is good for inclusion in Bio-Formats 6.0.0