Your message dated Sun, 3 Jan 2021 00:53:59 +0100
with message-id <20210102235359.GA10796@nuc>
and subject line Re: Bug#930773: dialog: args provided using --file behave
differently when provided directly in CLI
has caused the Debian Bug report #930773,
regarding dialog: args provided using --file behave differently when provided
directly in CLI
to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact [email protected]
immediately.)
--
930773: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=930773
Debian Bug Tracking System
Contact [email protected] with problems
--- Begin Message ---
Package: dialog
Version: 1.3-20160828-2
Severity: normal
Dear Maintainer,
There are cases where providing arguments to dialog using `--file'
results in failure but providing them directly in CLI works. It seems to
me that there is a string escaping problem.
I bumped into this because I use python-dialog, which invokes dialog
with arguments passed using the `--file' argument. I started seeing backslashes
in titles when colorizing them, and even in some cases dialog failed to
display widgets.
The easiest way to reproduce it is the following:
1) Invoke dialog passing args to CLI:
$ dialog "--msgbox" "\"test\"" "0" "0"
* A message box is displayed with the string `"test"` in it. *
2) Invoke dialog using the --file argument:
$ cat args
"--msgbox" "\"test\"" "0" "0"
$ dialog --file args
Error: Expected at least 3 tokens for --msgbox, have 1.
Use --help to list options.
It seems that version `1.3-20190211-1' from Buster works as expected in
this matter, so perhaps it should be backported to Stretch.
-- System Information:
Debian Release: 9.9
APT prefers stable
APT policy: (990, 'stable'), (500, 'stable-updates')
Architecture: amd64 (x86_64)
Kernel: Linux 4.19.5-arr-00009-g4694f1517d3f (SMP w/1 CPU core)
Locale: LANG=C, LC_CTYPE=C (charmap=ANSI_X3.4-1968), LANGUAGE=C
(charmap=ANSI_X3.4-1968)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)
Versions of packages dialog depends on:
ii debianutils 4.8.1.1
ii libc6 2.24-11+deb9u4
ii libncursesw5 6.0+20161126-1+deb9u2
ii libtinfo5 6.0+20161126-1+deb9u2
dialog recommends no packages.
dialog suggests no packages.
-- no debconf information
--- End Message ---
--- Begin Message ---
Version: 1.3-20190211-1
[ Sorry for the late reply ]
On Fri, Mar 20, 2020 at 05:40:54PM -0400, Thomas Dickey wrote:
> On Thu, Jun 20, 2019 at 01:02:03PM +0300, Alex Tsitsimpis wrote:
> > Package: dialog
> > Version: 1.3-20160828-2
> > Severity: normal
> >
> > Dear Maintainer,
> >
> > There are cases where providing arguments to dialog using `--file'
> > results in failure but providing them directly in CLI works. It seems to
> > me that there is a string escaping problem.
> >
> > I bumped into this because I use python-dialog, which invokes dialog
> > with arguments passed using the `--file' argument. I started seeing
> > backslashes
> > in titles when colorizing them, and even in some cases dialog failed to
> > display widgets.
> >
> > The easiest way to reproduce it is the following:
> >
> > 1) Invoke dialog passing args to CLI:
> > $ dialog "--msgbox" "\"test\"" "0" "0"
> > * A message box is displayed with the string `"test"` in it. *
> >
> > 2) Invoke dialog using the --file argument:
> > $ cat args
> > "--msgbox" "\"test\"" "0" "0"
> > $ dialog --file args
> > Error: Expected at least 3 tokens for --msgbox, have 1.
> > Use --help to list options.
>
> This is probably due to
>
> 2018/06/21
> + modify dlg_string_to_argv() to change the quoting behavior to be
> more consistent with shell behavior (patch by Denilson Sa Maia).
>
> > It seems that version `1.3-20190211-1' from Buster works as expected in
> > this matter, so perhaps it should be backported to Stretch.
>
> Santiago's free to disagree, but reading the guidelines for bug severity
> and backports, it seems that the level of effort for a backport makes it
> unlikely to do except for really urgent bugs:
>
> https://www.debian.org/Bugs/Developer#severities
>
> https://backports.debian.org/Contribute/
>
> (and the bug reporting guidelines don't seem to be geared toward requesting
> a backport...).
I have the same feeling, this issue was not important enough to make a backport.
However, this is not the same as saying that a backport would have
been unthinkable. It would have been just a matter of finding somebody
willing to make the backport, which could be any Debian maintainer,
not necessarily me.
(I hope that you found a way to workaround this bug, either by making
a backport for private use, or by just upgrading to Debian 10).
Thanks a lot.
--- End Message ---
