1/4 TIFF processing simplification by 'shifting' base of random access reader

[don-vip]

This is the first step towards solving #278

Port of this commit: https://github.com/drewnoakes/metadata-extractor-dotnet/commit/2281ceaad7dff2b8d6ccebadc992c758c6b964c2

[don-vip]

(not ready yet, I got test failures to fix)

[drewnoakes]

Looking good so far. The best way to validate correctness is to clone https://github.com/drewnoakes/metadata-extractor-images and run your updated code against all those files to check for regressions. It's not a process I've documented, sorry, so I can do that for you if you like, but doing it yourself would give you a tighter feedback loop.

[don-vip]

@drewnoakes thank you! I'll do it :) And probably create a PR to improve the documentation as well :)

[drewnoakes]

Here's an overview of the process.

• There is a .NET program here: https://github.com/drewnoakes/metadata-extractor-images/tree/master/src/dotnet
• It assumes you have sibling repositories for metadata-extractor, metadata-extractor-dotnet and metadata-extractor-images
• That .NET solution references the source from metadata-extractor-dotnet (here), so it'll build the .NET code as part of its build. You must currently manually build the Java code separately however.
• Running that .NET application will update all the metadata/*.txt files, and compute updated diffs.

[don-vip]

thank you! now the PR is ready for review :)