Hey again
On Thu, 2026-08-06 at 11:22 -0400, Chet Ramey wrote:
> This time, the LLM is correct. There are two stages of prompt
> expansion:
> first the expansion of backslash escapes, then word expansions.
I had just one new case of that "prompt corruption", but AFAICS all my
escape sequences produced from expansions are properly surrounded by
\001 and \002.
\[\e[0m\]\[\e[32m\]\u\[\e[0m\]\[\e[2;37m\]@\[\e[0m\]\[\e[32m\]\h\[\e[0m\]\[\e[2;37m\]:\[\e[0m\]\[\e[1;34m\]\w\[\e[0m\]$(
[ "$?" -eq 0 ] && { [ -n "$(jobs -s; jobs -r)" ] && printf '%s{%s%s\j%s%s}%s'
'\001\e[2;37m\002' '\001\e[0m\002' '\001\e[36m\002' '\001\e[0m\002'
'\001\e[2;37m\002' '\001\e[0m\002'; true; } || { [ -n "$(jobs -s; jobs -r)" ]
&& printf '%s{%s%s\j%s%s}%s' '\001\e[2;37m\002' '\001\e[0m\002'
'\001\e[36m\002' '\001\e[0m\002' '\001\e[2;37m\002' '\001\e[0m\002'; false; }
)$( [ "$?" -eq 0 ] && printf '%s' '\001\e[2;37m\002' || printf '%s'
'\001\e[31m\002' )\$\[\e[0m\]\[\e]0;\u@\h:\w\a\]
The only thing that I could think of is that I have that \j (number of
jobs) produced from a command substitution... could that cause any
troubles?
Maybe it was just some other glitch though.
Thanks,
Chris.