Fix a TypeError on complex number format with small enough number

[ma-gnesium]

TypeError occurs with numbers that convert to scientific notation (ie small enough ones) when using number formats that require the complex algorithm of NumberFormatter.

This is:

- [x] a bugfix
- [ ] a new feature
- [ ] refactoring
- [ ] additional unit tests

Checklist:

• Changes are covered by unit tests
  • Changes are covered by existing unit tests
  • New unit tests have been added
• Code style is respected
• Commit message explains why the change is made (see https://github.com/erlang/otp/wiki/Writing-good-commit-messages)
• CHANGELOG.md contains a short summary of the change
• Documentation is updated as necessary

Why this change is needed?

This fixes #3128

[oleibman]

Your change works well for your use case. It does not seem to work well for large numbers, e.g. 1E20 (which are also not handled well by existing code). Might it be possible to to solve that problem as well, and add test cases for that and intermediate numbers (e.g. 1E5)?

[oleibman]

I've experimented a little with large numbers. number_format seems to return some flaky results. I think it would be fairly straightforward with the BCMath extension, but that is not a requirement at the moment.

[oleibman]

A function which might be usable can be found at https://stackoverflow.com/questions/66135441/converting-float-to-string-without-exponential-notation
You probably want to do something about trailing zeros.

[ma-gnesium]

Thanks you for the feedback @oleibman.
Reconsidering the original bug report, I'm closing this PR because the effective fix from my point of vue is to use a correct format syntax, ie using comma as a thousands separator.