changelog-delay: move the delay/divert logic inside the (inner) revlog Instead of hacking throught the vfs/opener, we implement the delay/divert logic inside the `_InnerRevlog` and `randomaccessfile` object. This will allow to an alternative implementation of the `_InnerRevlog` that does not need to use Python details. As a result, the new implementation can use the transaction less agressively and avoid some extra output since no data had been written yet. That seems like a good side effect.

  $ hg init repo
  $ cd repo
  $ echo This is file a1 > a
  $ hg add a
  $ hg commit -m "commit #0"
  $ ls -A
  .hg
  a
  $ echo This is file b1 > b
  $ hg add b
  $ hg commit -m "commit #1"
  $ hg co 0
  0 files updated, 0 files merged, 1 files removed, 0 files unresolved

B should disappear

  $ ls -A
  .hg
  a