Removed empty bytes at channel subchunk ends for issue #31

[ggirelli]

I tried to address issue #31.

This edit checks if the chunk data size does not match the expected, based on image axes size. It then verifies that all unwanted bytes are zero-bytes and removes them if that is the case. It also supports (unseen) cases in which more than one unwanted zero-byte is present per step.

Scenarios:

• If the unwanted byte distribution is not consistent (i.e., not a multiple of column length -i.e., height-), an assert stops execution.
• If unwanted non-zero bytes are present, then a warning is triggered and the parsing crashes as usual down the line.
• If all unwanted bytes are zero-bytes and properly distributed, they are removed and a warning is issued.

[rbnvrw]

Thanks @ggirelli for taking a look at this, I really appreciate it! 👍
Did you try this also for the example files in issue #31? And does it work?
I've left some small comments to make the warnings hopefully more user friendly and to avoid some code duplication.
On top of that, because it is possibly a dangerous change to the parser, we really need to add some unit tests to check the proper functioning of this part of the code. After that, I'd be really happy to merge it and I hope it will solve these long-standing problems!

[ggirelli]

Hej @rbnvrw, thanks for the code revision! Yes, I tested it on the files provided on issue #31 and it works fine. I'll have a better look at your revs and prepare the unit tests next week 😄

[rbnvrw]

That sounds fantastic! Thank you.

[ggirelli]

Hi @rbnvrw, I tried to address all your code review points, thanks again for the comments. I added a simple unit test for this. As I did not have a smaller sample image with this issue and creating an artificial one would take a bit longer than I am willing to put on this, the code download the sample stitched image from #31 (2019 MB!) the first time one runs the tests, and then keeps using the same file from then onward. So the first time the test is run it takes some time (1-2 min) depending on your connection speed (and requires internet connection!). If you have any ideas on how to improve this, please let me know! 😄

[rbnvrw]

Thank you very much @ggirelli for your efforts! 🥇

I think this looks good. Might decide to migrate the file to something like git LFS but I'll have to look into that.