Consider using Jest's docblock parser for pragmas

[nickmccurdy]

Clear and concise description of the problem

We're maintaining our own regex based docblock pragma parsing logic, which is probably slower and less reliable than it could be as it search entire files instead of just comments at the top.

Suggested solution

We could use jest-docblock's parse function instead of our custom logic. It's a relatively small package with one dependency. They already have good support with many test cases, so we'd have less maintenance burden for maintaining compatibility with Jest (for example #1962 would be easier to fix).

Alternative

Refactor our parser so it's more performant and less reliant on regexes. I don't think we need a full CFG/BNF parser as long as we can focus on comments at the top of the file and parse pragmas. This would also reduce bundle size, in exchange for increasing maintenance effort and decreasing compatibility.

Additional context

jest-docblock doesn't support plain comments (//), so we'd technically be losing support for that unless we added a patch or workaround. However, we'd have to starting dealing with edge cases around groups of comments if we weren't using docblocks exclusively.

Validations

• Follow our Code of Conduct
• Read the Contributing Guidelines.
• Read the docs.
• Check that there isn't already an issue that request the same feature to avoid creating a duplicate.