28 Haziran 2020 Pazar tarihinde felix <fe...@f-hauri.ch> yazdı: > Machine: x86_64 > OS: linux-gnu > Compiler: gcc > Compilation CFLAGS: -g -O2 -fdebug-prefix-map=/build/bash-2bxm7h/bash-5.0=. > -fstack-protector-strong -Wformat -Werror=format-security -Wall > -Wno-parentheses -Wno-format-security > uname output: Linux medium 4.19.0-9-amd64 #1 SMP Debian 4.19.118-2 > (2020-04-29) x86_64 GNU/Linux > Machine Type: x86_64-pc-linux-gnu > > Bash Versions: 3.2.57(1)-release, 5.0.3(1)-release, 5.1.0(1)-alpha > > In order to reduce forks and make some tasks a lot quicker, I run > forked filters as background tasks, with dedicated I/O fd. > > For sample, to convert human datetime to UNIX SECONDS, instead of running > _out=$(date -d "$_string" +%s) > many time in same script, I run something like: > > _fifo=$(mktemp -u /tmp/fifo-XXXXXXXX) > mkfifo $_fifo > exec 9> >(exec stdbuf -o0 date -f - +%s >$_fifo 2>&1) > exec 8<$_fifo > rm $_fifo > > Then to convert human datetime to UNIX SECONDS: > > echo >&9 $_string > read -t 1 -u 8 _out > > become a lot quicker! > > But I recently discovered another way for this: > > exec 8<> <(:) > exec 9> >(exec stdbuf -o0 date -f - +%s >&8 2>&8) > > usable in same way: > > echo >&9 $_string > read -t 1 -u 8 _out > > maybe a little more quicker...
Seems like a bug to me. The FIFO created for <(:) shouldn't be writable. > > This was tested using https://f-hauri.ch/vrac/date1fork-demo.sh under > latest version of bash: 5.1.0(1), debian stable version: 5.0.3(1) and > old 3.2: 3.2.57(1). > > There is maybe something to document or even create a new feature > about open2 and open3... > > (I will have to rewrite https://f-hauri.ch/vrac/shell_connector.sh.txt ;) > > -- > Félix Hauri - <fe...@f-hauri.ch> - http://www.f-hauri.ch > > -- Oğuz
