Update ImageInput docs to not use deprecated APIs

[jessey-git]

Description

Update the ImageInput documentation to not use the deprecated read_image APIs.

Based on the discussion in #3902

Tests

Checklist:

• I have read the contribution guidelines.
• If this is more extensive than a small change to existing code, I
  have previously submitted a Contributor License Agreement
  (individual, and if there is any way my
  employers might think my programming belongs to them, then also
  corporate).
• I have updated the documentation, if applicable.
• I have ensured that the change is tested somewhere in the testsuite
  (adding new test cases if necessary).
• My code follows the prevailing code style of this project.

[jessey-git]

I tried to keep each sample relatively similar to each other and changed a few uses of non-standard dynamic arrays to just use vector.

I've never used DCO before. Do I just edit the above message with the special text or do I need to do something else before I use the special text?

[lgritz]

The easiest way to correct this is:

git commit --amend -s
git push --force origin deprecateddocs

The -s (which you could have used on any commit) appends the "Signed-off-by" line to your commit message.

[lgritz]

https://github.com/OpenImageIO/oiio/blob/master/CONTRIBUTING.md

Read the section about DCO signoff. There's also a tool you can use (per repo) that will automatically sign all your commits, if you are like me and will forget otherwise.

[lgritz]

This in particular: https://github.com/coderanger/dco

[lgritz]

Just as an FYI, if I were writing this for real, I think the best idiom would be:

auto pixels = std::unique_ptr<unsigned char[]>(new unsigned char [spec.width * spec.height * spec.nchannels]);

That would be freed upon scope exit (like std::vector), but it's much more efficient -- new[] doesn't initialize the memory, but std::vector will fill it with zeroes unnecessarily!

[jessey-git]

How about I change them to the following. Seem ok?
auto pixels = std::make_unique<unsigned char[]>(spec.width * spec.height * spec.nchannels);

[lgritz]

Will that really work without the new?

[jessey-git]

Yes, make_unique does work like that, however it looks like the semantics of make_unique include value initialization. After trying it out on compiler explorer the make_unique calls will include some form of memset basically (rep stosq)... c++ why are you the way you are... https://godbolt.org/z/9PEYKbY6f

I'll replace all of them with your suggestion then.