Should favour names in DU members (and maybe pattern matching?)

There are two ways to write things, named or unnamed.

Unnamed (types):

type Data =
    | TwoParts of string * string
    | OnePart of string

Named (types):

type Data =
    | TwoParts of part1: string * part2: string
    | OnePart of part1: string

We could have a rule that warns about the former and prefers the latter. This would be enough to make a v1 of a rule.

For a 2nd version of the rule, we could also examine match blocks:

Unnamed (match blocks):

let examineData x =
    match data with
    | OnePart p1 -> p1
    | TwoPart (p1, p2) -> p1 + p2

Named (match blocks):

let examineData x =
    match data with
    | OnePartData(part1=p1) -> p1
    | TwoPartData(part1=p1; part2=p2) -> p1 + p2

The issue does not name files, tests, or entry points. Start by locating how existing FSharpLint rules detect and report union-case declarations, then define completion as warnings for unnamed DU fields; pattern matching is described as a possible second version.