seq[int].T is int should compile (and be true) according to spec; also fails with type seq2[T] = T

[timotheecour]

import typetraits

when defined(case_ok):
  # ok
  block:
    type seq2[T] = (T,int)
    echo seq2[int].T
    doAssert seq2[int].T is int
  block:
    type seq2[T] = (T,)
    echo seq2[int].T
  block:
    type seq2[T] = int
    echo seq2[int].T
  block:
    type seq2[T] = object
      discard
    echo seq2[int].T

when defined(case2):
  type seq2[T] = T
  # Error: undeclared field: 'T'
  echo seq2[int].T

when defined(case_seq):
  # Error: undeclared field: 'T'
  echo seq[int].T

this breaks generic code and user defined concepts isSeq that check whether some type is a seq[T] for some T

• likewise for other magic types in system.nim

NOTE: this should be feasible according to (my interpretation of) docs:

Procs written with the implicitly generic style will often need to refer to the type parameters of the matched generic type. They can be easily accessed using the dot syntax:
type Matrix[T, Rows, Columns] = object
proc [](m: Matrix, row, col: int): Matrix.T =

in any case it's useful

[timotheecour]

@zah actually I kind of dislike this feature MyGeneric.T : it's convenient but can cause hijacking issues in case T is in scope and result doesn't match user expectations. Could a macro achieve same goal (and fix this issue as well as #8459), maybe using same tricks as I used in #8531 for undistinct macro?
eg: extractGeneric(MyGeneric, T)

[zah]

Can you clarify what these hijacking issues are?

[timotheecour]

sent out a PR (which in fact provides better workaround for this issue) and moved discussion there: #8554