fix REPL summary for types

[vtjnash]

Some improved help formatting:

help?> Base.GenericIOBuffer
  No documentation found.

  Summary
  ≡≡≡≡≡≡≡≡≡

  mutable struct Base.GenericIOBuffer{T<:AbstractVector{UInt8}}

  Fields
  ≡≡≡≡≡≡≡≡

  data     :: T<:AbstractVector{UInt8}
  readable :: Bool
  writable :: Bool
  seekable :: Bool
  append   :: Bool
  size     :: Int64
  maxsize  :: Int64
  ptr      :: Int64
  mark     :: Int64

  Supertype Hierarchy
  ≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡

  Base.GenericIOBuffer{T<:AbstractVector{UInt8}} <: IO <: Any

help?> Core.Intrinsics
  No documentation found.

  No docstring found for module Core.Intrinsics.

  Exported names:
  ≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡≡

  abs_float, add_float, add_float_fast, add_int, add_ptr, and_int, arraylen, ashr_int, bitcast, bswap_int, ceil_llvm, cglobal, checked_sadd_int, checked_sdiv_int, checked_smul_int, checked_srem_int,
  checked_ssub_int, checked_uadd_int, checked_udiv_int, checked_umul_int, checked_urem_int, checked_usub_int, copysign_float, ctlz_int, ctpop_int, cttz_int, div_float, div_float_fast, eq_float,
  eq_float_fast, eq_int, flipsign_int, floor_llvm, fma_float, fpext, fpiseq, fpislt, fptosi, fptoui, fptrunc, le_float, le_float_fast, llvmcall, lshr_int, lt_float, lt_float_fast, mul_float,
  mul_float_fast, mul_int, muladd_float, ne_float, ne_float_fast, ne_int, neg_float, neg_float_fast, neg_int, not_int, or_int, pointerref, pointerset, rem_float, rem_float_fast, rint_llvm, sdiv_int,
  sext_int, shl_int, sitofp, sle_int, slt_int, sqrt_llvm, sqrt_llvm_fast, srem_int, sub_float, sub_float_fast, sub_int, sub_ptr, trunc_int, trunc_llvm, udiv_int, uitofp, ule_int, ult_int, urem_int,
  xor_int, zext_int

Replaces #21126
Fixes #21029