"help echo" forgets about "\u"

2018-12-07 Thread Egmont Koblinger
Hi, With 4.4 as well as the 20181205 snapshot: the output of "help echo" doesn't mention the escape characters "\u" and "\U". (They are mentioned in the manual page.) cheers, egmont

Re: Feature request: PROMPT_COMMANDS array variable

2018-02-03 Thread Egmont Koblinger
Hi, I'd like to second this feature request. As Daniel mentioned, the biggest problem with the current single variable approach that it's hardly usable due to lack of coordination between distributions as well as apps that wish to modify PROMPT_COMMAND. Let's look at it at more details. There is

Re: SIGSEGV: rl_redisplay: Long Directory Name with EM Dash character in Graphical Terminal

2018-02-03 Thread Egmont Koblinger
Hi, I can reproduce this problem with 4.4.12 on Ubuntu Artful (in gnome-terminal, xterm, urxvt - so it doesn't seem to matter). I cannot reproduce with devel bash. (Sorry if this email doesn't show up nicely in the thread index, I have no clue how to properly reply to an email just by seeing it i

Re: Bracketed paste mode breaks cooked mode's tab + backspace

2018-01-30 Thread Egmont Koblinger
Hi Chet, > It seems like adding \r to the end of the sequence to turn off bracketed > paste mode should do the trick. Yup, I confirm that adding \r to BRACK_PASTE_FINI fixes the problem, this looks the simplest approach. cheers, egmont

Re: Bracketed paste mode breaks cooked mode's tab + backspace

2018-01-30 Thread Egmont Koblinger
Hi Phi, > This is not directly related to bash, it depend on the line discipline > on the tty, I use ^H as erase and got no problems with . For me the same bug occurs after "stty erase ^H" and pressing ^H, too. I would be really surprised if the kernel's logic on backspacing over tabs (which is

Bracketed paste mode breaks cooked mode's tab + backspace

2018-01-29 Thread Egmont Koblinger
Configuration Information [Automatically generated, do not change]: Machine: x86_64 OS: linux-gnu Compiler: gcc Compilation CFLAGS: -DPROGRAM='bash' -DCONF_HOSTTYPE='x86_64' -DCONF_OSTYPE='linux-gnu' -DCONF_MACHTYPE='x86_64-pc-linux-gnu' -DCONF_VENDOR='pc' -DLOCALEDIR='/usr/share/locale' -DPACKAGE

Re: Colored prompt + cmdline ending at right margin => no explicit newline

2017-11-18 Thread Egmont Koblinger
A bit of technical clarification: Terminal emulators seem to default to lines ending in an explicit newline. It's the implicit overflow (printing a letter that wraps to the next line) which makes the previous line soft-wrapped, and a \e[K (clear to EOL) and perhaps a few other similar sequences (e

Colored prompt + cmdline ending at right margin => no explicit newline

2017-11-18 Thread Egmont Koblinger
Configuration Information [Automatically generated, do not change]: Machine: x86_64 OS: linux-gnu Compiler: gcc Compilation CFLAGS: -DPROGRAM='bash' -DCONF_HOSTTYPE='x86_64' -DCONF_OSTYPE='linux-gnu' -DCONF_MACHTYPE='x86_64-pc-linux-gnu' -DCONF_VENDOR='pc' -DLOCALEDIR='/usr/share/locale' -DPACKAGE

Re: Segfault: Lone surrogate followed by locale change

2017-11-10 Thread Egmont Koblinger
On Fri, Nov 10, 2017 at 2:19 PM, Eduardo A. Bustamante López wrote: > I think the fix looks something like this: > + localconv = (iconv_t)-1; /* initialize */ I can confirm that the crash is gone with this fix. Thanks a lot, Eduardo! e.

Segfault: Lone surrogate followed by locale change

2017-11-10 Thread Egmont Koblinger
Configuration Information [Automatically generated, do not change]: Machine: x86_64 OS: linux-gnu Compiler: gcc Compilation CFLAGS: -DPROGRAM='bash' -DCONF_HOSTTYPE='x86_64' -DCONF_OSTYPE='linux-gnu' -DCONF_MACHTYPE='x86_64-pc-linux-gnu' -DCONF_VENDOR='pc' -DLOCALEDIR='/usr/share/locale' -DPACKAGE

Re: Builtin echo interrupted by WINCH

2017-09-24 Thread Egmont Koblinger
24, 2017 at 7:48 PM, Chet Ramey wrote: > On 9/23/17 4:55 PM, Egmont Koblinger wrote: > >> Is there anything bash could do to prevent incorrect behavior of its >> built-in commands when a SIGWINCH is encountered? And of course I >> don't mean the "echo" comma

Builtin echo interrupted by WINCH

2017-09-23 Thread Egmont Koblinger
Configuration Information [Automatically generated, do not change]: Machine: x86_64 OS: linux-gnu Compiler: gcc Compilation CFLAGS: -DPROGRAM='bash' -DCONF_HOSTTYPE='x86_64' -DCONF_OSTYPE='linux-gnu' -DCONF_MACHTYPE='x86_64-pc-linux-gnu' -DCONF_VENDOR='pc' -DLOCALEDIR='/usr/share/locale' -DPACKAGE

Re: Finding the last foreground command in the current session

2015-07-12 Thread Egmont Koblinger
Hi guys, Friendly ping... is there anything we could do? Chet, I couldn't find anything related to this in 4.4-alpha's changes. Is there any chance to get this feature to 4.4-final? thx, egmont On Fri, May 22, 2015 at 11:15 AM, Debarshi Ray wrote: > Hey, > > Any further comments / questions /

Re: Finding the last foreground command in the current session

2015-03-01 Thread Egmont Koblinger
Hi Chet & everyone, The goal is to notify the user if a long-running command (e.g. a compilation) finishes, and the notification should mention the given command. $BASH_COMMAND seems to contain the command that's about to be executed, whereas what we'd need is the command that just terminated. $B

winch trap delayed until keypress

2014-05-20 Thread Egmont Koblinger
Hi, Execute this in an interactive bash and then resize the window: trap 'stty size' winch In bash-4.2, the trap was executed immediately upon resize. In bash-4.3, it is delayed until the next keypress. Is this an intentional, expected change? There seems to be no relevant entry in NEWS. I us

Re: multiline ^C regression

2014-03-27 Thread Egmont Koblinger
Awesome, thanks for the super fast fix :) e. On Thu, Mar 27, 2014 at 7:26 PM, Chet Ramey wrote: > On 3/27/14 3:40 AM, Egmont Koblinger wrote: > > Hi, > > > > Type a multiline command, and press ^C. > > > > bash-4.2: The new prompt appeared below the complet

Re: multiline ^C regression

2014-03-27 Thread Egmont Koblinger
Sorry, I wasn't clear enough. By "multiline" I mean a simple command that is so long (e.g. takes so many parameters) that it cannot fit into a single line. (You reach the right margin of your terminal and keep on typing.) On Thu, Mar 27, 2014 at 8:50 AM, Chris Down wrote: >

multiline ^C regression

2014-03-27 Thread Egmont Koblinger
Hi, Type a multiline command, and press ^C. bash-4.2: The new prompt appeared below the complete multiline command. bash-4.3: The new prompt appears right under the previous prompt, overwriting parts of the aborted command line and leaving garbage after the new prompt. I believe this is a reg

Two ideas similar to ignorespace

2014-02-11 Thread Egmont Koblinger
Hi, I've been using HISTCONTROL=ignorespace for a long time, and I often begin a command with space if I don't want that command to be stored in history (typically "dangerous" commands, e.g. I almost always type "rm something" instead of "rm something"). I have two feature request ideas for you t

Re: [PATCH] display bug in multi-line command when a utf-8 is shifted to next line

2013-02-17 Thread Egmont Koblinger
On Sun, Feb 17, 2013 at 10:43 PM, Chet Ramey wrote: > On 2/17/13 8:37 AM, Egmont Koblinger wrote: > > Hi Chet, > > > > Friendly ping - could you please look at the bug description below and > > review the attached patch? This is a really trivial bugfix (just 1 > &

Re: [PATCH] display bug in multi-line command when a utf-8 is shifted to next line

2013-02-17 Thread Egmont Koblinger
Hi Chet, Friendly ping - could you please look at the bug description below and review the attached patch? This is a really trivial bugfix (just 1 character), should be obvious to verify. Thanks a lot, egmont On Tue, Jan 29, 2013 at 2:06 PM, Egmont Koblinger wrote: > Hi, > >

Re: readline display bug with UTF-8 and char insertion

2013-02-17 Thread Egmont Koblinger
t me know if there's anything more I could do to get this one fixed. thanks, egmont On Mon, Jan 21, 2013 at 9:16 PM, Egmont Koblinger wrote: > Hi, > > On a fully UTF-8 environment, in certain easily reproducible cases, the > command line becomes messed up (the cursor ends up in the

[PATCH] display bug in multi-line command when a utf-8 is shifted to next line

2013-01-29 Thread Egmont Koblinger
Hi, In UTF-8 environment, with bash-4.2.42, type a command that takes up at least 3 lines, and contains an accented (UTF-8) letter in the first line. Move the cursor to the beginning, and insert more characters. Just when the UTF-8 accented letter is pushed down to the second line, the last line

readline display bug with UTF-8 and char insertion

2013-01-21 Thread Egmont Koblinger
Hi, On a fully UTF-8 environment, in certain easily reproducible cases, the command line becomes messed up (the cursor ends up in the wrong column, and from then on it's tons of cumulative mistakes). I'm using bash-4.2.42, but the bug also affects other readline-based applications. To reproduce:

Re: Wrong prompt and incorrect parsing if completion is used

2010-08-15 Thread Egmont Koblinger
Hi Chet, Thanks for fixing it! Sorry I didn't provide the output you requested, I was on vacation for a week, and now it seems you no longer need it :) egmont On Aug 15, 2010 6:22 PM, "Chet Ramey" wrote: > On 8/9/10 12:02 PM, Egmont Koblinger wrote: > >> Sta

Wrong prompt and incorrect parsing if completion is used

2010-08-09 Thread Egmont Koblinger
Hello, I've found a weird behavior. In some circumstances, if I use TAB completion during typing a command line, the final command is parsed differently than if I typed it all along. I'm on Ubuntu Lucid. The bug described below occurs only if I use the bash-completion package (that is, I source

sh, pwd -P and $PWD

2006-11-13 Thread Egmont Koblinger
Hi, According to the bash manpage and to "help pwd", "pwd -P" should only print something but not change the state of bash. However, if bash is launched as "sh", the "pwd -P" command alters $PWD to the resolved canonical full path name, and if the prompt contains \w, it is also updated to the can

Re: regression with color prompt + utf8

2006-02-06 Thread Egmont Koblinger
On Sun, Feb 05, 2006 at 03:49:38PM -0500, Chet Ramey wrote: > If I'm reading this right, it's not a regression, per se. It's always > possible to fool the redisplay code by violating the assumptions it > starts with, e.g., that the cursor always begins at column 0. > [...] Hi, Thanks for your a

regression with color prompt + utf8

2006-02-04 Thread Egmont Koblinger
Configuration Information [Automatically generated, do not change]: Machine: i586 OS: linux Compiler: i586-uhu-linux-gcc Compilation CFLAGS: -DPROGRAM='bash' -DCONF_HOSTTYPE='i586' -DCONF_OSTYPE='linux' -DCONF_MACHTYPE='i586-uhu-linux' -DCONF_VENDOR='uhu' -DLOCALEDIR='/usr/share/locale' -DPACKAG