coding system and gdb

coding system and gdb

[Rusi]

Some students of mine are working on a hand-compiled python and debugging with gdb in emacs.
Unicode characters dont seem to be flowing properly between (inferior) python and
emacs; seems to work better when python is started outside emacs.

Is there some process-related-coding-system that is to be set for gdb?

How/When/Where?

Re: coding system and gdb

[Eli Zaretskii]

> Date: Wed, 8 Jul 2015 09:23:39 -0700 (PDT)
> From: Rusi <rustompmody@gmail.com>
> 
> Some students of mine are working on a hand-compiled python and debugging with gdb in emacs.
> Unicode characters dont seem to be flowing properly between (inferior) python and
> emacs; seems to work better when python is started outside emacs.
> 
> Is there some process-related-coding-system that is to be set for gdb?
> 
> How/When/Where?

Please provide some minimal details about the GDB commands you tried,
and their results.  GDB is generally agnostic to encoding issues, it's
all just data for it, so I'm having hard time guessing what exactly
didn't work for your students.

In general, process-coding-system-alist is the feature you are
supposed to use to set up encoding for communicating with processes,
in case the locale's defaults aren't TRT.

Re: coding system and gdb

[Rusi]

On Wednesday, July 8, 2015 at 10:27:02 PM UTC+5:30, Eli Zaretskii wrote:
> > Date: Wed, 8 Jul 2015 09:23:39 -0700 (PDT)
> > From: Rusi 
> > 
> > Some students of mine are working on a hand-compiled python and debugging with gdb in emacs.
> > Unicode characters dont seem to be flowing properly between (inferior) python and
> > emacs; seems to work better when python is started outside emacs.
> > 
> > Is there some process-related-coding-system that is to be set for gdb?
> > 
> > How/When/Where?
> 
> Please provide some minimal details about the GDB commands you tried,
> and their results.  GDB is generally agnostic to encoding issues, it's
> all just data for it, so I'm having hard time guessing what exactly
> didn't work for your students.
> 
> In general, process-coding-system-alist is the feature you are
> supposed to use to set up encoding for communicating with processes,
> in case the locale's defaults aren't TRT.

I'll find out more details and repost.
For now Ive this mail from them:

--------
I tried a='ब' in emacs gdb but it is interpreting it as. a= '  ' and it is not 
printing anything when we do print(a) but it is printing correctly in terminal 
with our compiled python.

Re: coding system and gdb

[Eli Zaretskii]

> Date: Wed, 8 Jul 2015 10:09:55 -0700 (PDT)
> From: Rusi <rustompmody@gmail.com>
> 
> I tried a='ब' in emacs gdb but it is interpreting it as. a= '  ' and it is not 
> printing anything when we do print(a) but it is printing correctly in terminal 
> with our compiled python.

That leaves out too many important details.  Like was a='ब' typed at
the program's prompt or at GDB's prompt, and what exactly does "in
terminal" mean.

Re: coding system and gdb

[tomas]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Wed, Jul 08, 2015 at 09:23:39AM -0700, Rusi wrote:
> Some students of mine are working on a hand-compiled python and debugging with gdb in emacs.
> Unicode characters dont seem to be flowing properly between (inferior) python and
> emacs; seems to work better when python is started outside emacs.
> 
> Is there some process-related-coding-system that is to be set for gdb?

Locale from the environment? I.e. the value of LANG?

- -- t
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAlWdYjUACgkQBcgs9XrR2kbPKwCeMxDGE6BZRpZWP5kj/7bMBuMQ
8m8AnAttMaTlrFla6ffEh0rFp9vOK7Op
=67WS
-----END PGP SIGNATURE-----

Re: coding system and gdb

[Eli Zaretskii]

> Date: Wed, 8 Jul 2015 19:47:33 +0200
> From: <tomas@tuxteam.de>
> Cc: help-gnu-emacs@gnu.org
> 
> Locale from the environment? I.e. the value of LANG?

Not only LANG, but the other LC_* variables as well.  See the node
"Language Environments" in the Emacs manual.

Re: coding system and gdb

[Rusi]

On Wednesday, July 8, 2015 at 11:17:42 PM UTC+5:30, tomas wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> On Wed, Jul 08, 2015 at 09:23:39AM -0700, Rusi wrote:
> > Some students of mine are working on a hand-compiled python and debugging with gdb in emacs.
> > Unicode characters dont seem to be flowing properly between (inferior) python and
> > emacs; seems to work better when python is started outside emacs.
> > 
> > Is there some process-related-coding-system that is to be set for gdb?
> 
> Locale from the environment? I.e. the value of LANG?

As you seem to have guessed it was not one of the UTF-8 variants.
[I could have sworn I had set that right the last time I met them...Seem to be losing my memory...
]

And what exactly is the difference between LANG and LANGUAGE?
I am just assuming all the LC_* can be left unset (and inferred?)

Its all a bit arcane to me!

Re: coding system and gdb

[Eli Zaretskii]

> Date: Thu, 9 Jul 2015 10:11:04 -0700 (PDT)
> From: Rusi <rustompmody@gmail.com>
> 
> And what exactly is the difference between LANG and LANGUAGE?
> I am just assuming all the LC_* can be left unset (and inferred?)

Please read the node in the manual which I pointed to, it's all
described and explained there.

Re: coding system and gdb

[Rusi]

On Thursday, July 9, 2015 at 11:09:44 PM UTC+5:30, Eli Zaretskii wrote:
> > Date: Thu, 9 Jul 2015 10:11:04 -0700 (PDT)
> > From: Rusi 
> > 
> > And what exactly is the difference between LANG and LANGUAGE?
> > I am just assuming all the LC_* can be left unset (and inferred?)
> 
> Please read the node in the manual which I pointed to, it's all
> described and explained there.

Seems to be working now -- thanks.

[After an interval of 2 days in which said student crashed his laptop by
mis-editing /etc/default/locale file.
A garbled locale makes the machine unbootable evidently!!]