Add Block, Paragraph, Word, Symbol models for Full Text Annotation

[HighLiuk]

Models are based on the Google Cloud Vision API's docs

[faithoflifedev]

@HighLiuk , thanks for helping with the package.

In running some checks after merging your PR, I'm seeing some Warnings (unrelated to your changes), so I've submitted a PR appsup-dart/x509#25 to the https://pub.dev/packages/x509 package to resolve that. I'm hoping that the x509 package will get updated and then I can published these changes. If x509 isn't updated soon, then I'll publish regardless, since the issue is just warnings.

[HighLiuk]

Thanks @faithoflifedev for the clarification. I have one more (unrelated) improvement if you are interested into it.
Check it out at my dev branch.
I may make a PR for that if you want to.
The purpose is to let GoogleVision to be initialized by passing the google-credentials.json file by only knowing its contents. Otherwise you have to make workarounds like saving a temporary file with that just to read the credentials.

P.S. not sure that JwtFromStringGeneratoris a proper name for that.