Re: [PATCH v2] read: non-raw-mode fixes

[Chet Ramey]

On 7/26/23 7:53 PM, Grisha Levit wrote:
> On Wed, Jul 26, 2023 at 2:23 PM Grisha Levit <grishalevit@gmail.com> wrote:
> > If the last character read was an (unescaped) backslash, store it as
> > such instead of as a CTLESC.  Avoids:
> >
> > $ printf '\\' | { read; echo "${REPLY@Q}"; }
> > bash: DEBUG warning: dequote_string: string with bare CTLESC
> > $'\001'
>
> Sorry that was wrong, the CTLESC should just be removed.  Much simpler
> patch below.

Thanks for the report and patch. I simplified it a little more by making
saw_escape into a real counter, and it will be in the next devel branch push.

Chet

> ---
> diff --git a/builtins/read.def b/builtins/read.def
> index 5b2621fe..ce5bcc38 100644
> --- a/builtins/read.def
> +++ b/builtins/read.def
> @@ -751,11 +751,11 @@ read_builtin (WORD_LIST *list)
>         if (pass_next)
>          {
>            pass_next = 0;
> -         if (c == '\n')
> +         if (c == '\n' || c == '\0')
>              {
>                if (skip_ctlesc == 0 && i > 0)
>                  i--;            /* back up over the CTLESC */
> -             if (interactive && input_is_tty && raw == 0)
> +             if (interactive && input_is_tty && raw == 0 && c == '\n')
>                  print_ps2 = 1;
>              }
>            else
> @@ -825,6 +825,11 @@ add_char:
>         if (nchars > 0 && nr >= nchars)
>          break;
>       }
> +
> +  if (pass_next && skip_ctlesc == 0)
> +    i--;               /* back up over the CTLESC */
> +  if (skip_ctlnul && saw_escape && i == 1 && input_string[0] == CTLNUL)
> +    saw_escape = 0;    /* Avoid dequoting bare CTLNUL */
>     input_string[i] = '\0';
>     check_read_timeout ();

--
``The lyf so short, the craft so long to lerne.'' - Chaucer
                 ``Ars longa, vita brevis'' - Hippocrates
Chet Ramey, UTech, CWRU    chet@case.edu    http://tiswww.cwru.edu/~chet/