☂️ Formatter stability improvements

[ematipico]

Description

The formatter is in a good shape, but it doesn't have enough stability to be used. In order to reach a point where we can have a good integration with LSP, we still need to improve it and add some more features.

Top priority

• 📎Create an API to format a range #2009

  Essentially we want to be able to give a portion of text to the formatter and return the formatted result (either formatted or keep as is because it's a broken piece of code). Ideally, we could use the same APIs we have now.

• Create a workflow where we format some big code bases and report its results into the CI. #2030

  This is something that we already have with our parser. We could do the same with our formatter. We could use the babel repository for JavaScript and the the old rome repository for typescript. Using the old rome branch can be useful to see if what we format doesn't change

• Expand testing suite with different options: different indent style and indent size. #2044

• Expand testing suite by adding Windows and make sure it doesn't crash #2042

• Better error handling for formatter, by using Diagnostic

• Better handling of trailing and leading comments #2006

Nice to have

• Automatic code generation for formatter #1997
  At the moment, every time we create new grammar, we have to manually add those nodes to our formatter suite. It would be great if we can automate that part by creating the new stuff automatically
• Reach 100% implementation coverage
  • Template literals @ematipico feat(rome_formatter): template literals and static initializers  #2021
  • import syntax @ematipico feat(rome_formatter): format import syntax #1992
  • export syntax @ematipico feat(rome_formatter): format export syntax  #2008
  • etc.
• Create helpers for Formatter #2002