[PATCH] chgserver: remove Python 2 file descriptor logic

[Raphaël Gomès]

Queued, thanks!

On 3/4/22 03:12, Yuya Nishihara wrote:
> # HG changeset patch
> # User Yuya Nishihara <yuya at tcha.org>
> # Date 1646357326 -32400
> #      Fri Mar 04 10:28:46 2022 +0900
> # Node ID 3a9729bead90ef7caaf4d25138680c83a888be78
> # Parent  46b3ecfb16e2781ede9752d972dc22f0e1dfea87
> chgserver: remove Python 2 file descriptor logic
>
> Follows up 0bb28b7736bc "chgserver: remove Python 2 support code."
>
> On Python 2, we had to close newfp prior to restoring the original file
> description since "delete newfp" would otherwise close the file descriptor
> shared with the long-lived fp:
>
>    in attachio():
>      newfp = os.fdopen(fp.fileno(), mode, bufsize)
>    in _restoreio():
>      newfp.close()  # temporarily close newfp.fileno() (= fp.fileno())
>      os.dup2(fd, fp.fileno())  # reopen fp.fileno() with original fd
>
> On the other hand, we shouldn't call newfp.close() on Python 3 since
> any function calls are proxied to the underlying file object by
> procutil.LineBufferedWrapper.
>
> diff --git a/mercurial/chgserver.py b/mercurial/chgserver.py
> --- a/mercurial/chgserver.py
> +++ b/mercurial/chgserver.py
> @@ -438,14 +438,8 @@ class chgcmdserver(commandserver.server)
>           nullfd = os.open(os.devnull, os.O_WRONLY)
>           ui = self.ui
>           for (ch, fp, fd), (cn, fn, mode) in zip(self._oldios, _iochannels):
> -            newfp = getattr(ui, fn)
> -            # On Python 3, newfp is just a wrapper around fp even if newfp is
> -            # not fp, so deleting newfp is safe.
> -            if newfp is not fp:
> -                newfp.close()
> -            # restore original fd: fp is open again
>               try:
> -                if newfp is fp and 'w' in mode:
> +                if 'w' in mode:
>                       # Discard buffered data which couldn't be flushed because
>                       # of EPIPE. The data should belong to the current session
>                       # and should never persist.
>
> _______________________________________________
> Mercurial-devel mailing list
> Mercurial-devel at mercurial-scm.org
> https://www.mercurial-scm.org/mailman/listinfo/mercurial-devel