On Wed, Dec 9, 2020 at 4:17 PM Yuya Nishihara <y...@tcha.org> wrote:
>
> On Tue, 08 Dec 2020 19:12:20 +0530, Pulkit Goyal wrote:
> > # HG changeset patch
> > # User Pulkit Goyal <7895pul...@gmail.com>
> > # Date 1606897517 -19800
> > #      Wed Dec 02 13:55:17 2020 +0530
> > # Node ID 13496bf926221a034c890135954fdc72c895f6d7
> > # Parent  8b20d469a0c819c2ade8635c25e9fcf0af553796
> > # EXP-Topic chg-test
> > procutils: don't try to get `.buffer` if sys.stdin is None
>
> > diff -r 8b20d469a0c8 -r 13496bf92622 mercurial/utils/procutil.py
> > --- a/mercurial/utils/procutil.py     Fri Apr 03 20:30:36 2020 +0530
> > +++ b/mercurial/utils/procutil.py     Wed Dec 02 13:55:17 2020 +0530
> > @@ -124,7 +124,9 @@
> >      # Python 3 implements its own I/O streams.
> >      # TODO: .buffer might not exist if std streams were replaced; we'll 
> > need
> >      # a silly wrapper to make a bytes stream backed by a unicode one.
> > -    stdin = sys.stdin.buffer
> > +
> > +    # sys.stdin can be None
> > +    stdin = sys.stdin.buffer if sys.stdin else os.fdopen(0, 'rb')
>
> Oops, os.fdopen() fails if fd=0 isn't available, so we'll need other
> workaround. Sorry.
>
>   % python3 -c 'import os; os.fdopen(0, "rb")' 0<&-
>   Traceback (most recent call last):
>     File "<string>", line 1, in <module>
>     File "/usr/lib/python3.9/os.py", line 1023, in fdopen
>       return io.open(fd, *args, **kwargs)
>   OSError: [Errno 9] Bad file descriptor
>
> We can simulate Py2 behavior by opening /dev/null and closing the underlying
> fd, but that might be a bit hacky.
>
>   f = open(os.devnull, 'rb')
>   os.close(f.fileno())
>
> Any thoughts?
>
> For stdout/stderr, we do need something. Adding null check to procutil
> and dispatch isn't enough.

Seems like we can introduce a wrapper around sys.stdin/out/err. I
guess procutil.stdin/out/err is that wrapper only.
I am not sure what's correct because my understanding in this area is limited.

(My V1 patches got pushed, sorry about that. Actually I first sent
them on phabricator and realized later that I should have a look from
you. I forgot to update phabricator about conversation here and they
are pushed. I will update this series to be based on those patches.)
>
>   % python3 hg version  1<&-
>   ...
>   Traceback (most recent call last):
>     File "mercurial/dispatch.py", line 434, in _runcatchfunc
>       return _dispatch(req)
>     File "mercurial/dispatch.py", line 1039, in _dispatch
>       extensions.loadall(lui)
>     File "mercurial/extensions.py", line 310, in loadall
>       ui.warn(
>     File "mercurial/ui.py", line 1766, in warn
>       self._writemsg(self._fmsgerr, type=b'warning', *msg, **opts)
>     File "mercurial/ui.py", line 1241, in _writemsg
>       _writemsgwith(self._write, dest, *args, **opts)
>     File "mercurial/ui.py", line 2381, in _writemsgwith
>       write(dest, *args, **opts)
>     File "mercurial/ui.py", line 1184, in _write
>       self._writenobuf(dest, *args, **opts)
>     File "mercurial/ui.py", line 1196, in _writenobuf
>       self._fout.flush()
>   AttributeError: 'NoneType' object has no attribute 'flush'
>
>   During handling of the above exception, another exception occurred:
>
>   Traceback (most recent call last):
>     File "mercurial/dispatch.py", line 453, in _callcatch
>       return scmutil.callcatch(ui, func)
>     File "mercurial/scmutil.py", line 155, in callcatch
>       return func()
>     File "mercurial/dispatch.py", line 436, in _runcatchfunc
>       ui.flush()
>     File "mercurial/ui.py", line 1253, in flush
>       self._fout.flush()
>   AttributeError: 'NoneType' object has no attribute 'flush'
>
>   During handling of the above exception, another exception occurred:
>
>   Traceback (most recent call last):
>     File "mercurial/dispatch.py", line 268, in dispatch
>       ret = _runcatch(req) or 0
>     File "mercurial/dispatch.py", line 444, in _runcatch
>       return _callcatch(ui, _runcatchfunc)
>     File "mercurial/dispatch.py", line 494, in _callcatch
>       if not handlecommandexception(ui):
>     File "mercurial/dispatch.py", line 1346, in handlecommandexception
>       ui.warn(warning)
>     File "mercurial/ui.py", line 1766, in warn
>       self._writemsg(self._fmsgerr, type=b'warning', *msg, **opts)
>     File "mercurial/ui.py", line 1241, in _writemsg
>       _writemsgwith(self._write, dest, *args, **opts)
>     File "mercurial/ui.py", line 2381, in _writemsgwith
>       write(dest, *args, **opts)
>     File "mercurial/ui.py", line 1184, in _write
>       self._writenobuf(dest, *args, **opts)
>     File "mercurial/ui.py", line 1196, in _writenobuf
>       self._fout.flush()
>   AttributeError: 'NoneType' object has no attribute 'flush'
>
>   During handling of the above exception, another exception occurred:
>
>   Traceback (most recent call last):
>     File "hg", line 59, in <module>
>       dispatch.run()
>     File "mercurial/dispatch.py", line 115, in run
>       status = dispatch(req)
>     File "mercurial/dispatch.py", line 291, in dispatch
>       req.ui.flush()  # record blocked times
>     File "mercurial/ui.py", line 1253, in flush
>       self._fout.flush()
>   AttributeError: 'NoneType' object has no attribute 'flush'
_______________________________________________
Mercurial-devel mailing list
Mercurial-devel@mercurial-scm.org
https://www.mercurial-scm.org/mailman/listinfo/mercurial-devel

Reply via email to