I have file with hex symbols like below
cat demo
\x22count\x22
\x22count\x22
\x22count\x22
I need conversion like:
echo $(cat demo)
"count" "count" "count"
How to get unicode symbols in pipeline with newline symbol? Something like:
cat demo | ???
"count"
"count"
"count"
You can use printf's %b conversion specification. This will print the output you want:
printf '%b\n' "$(<demo)"
Note: %b causes printf to expand other backslash escape sequences as well (e.g., \n, \t etc.)
cat /tmp/ssv_count | xargs -I{} printf '%b\n' {} . My command returns x22countx22 three times
xargs is error-prone and rarely useful in a bash script. You shouldn't use it unless you have to. But if you really insist, this should work for GNU xargs : cat /tmp/ssv_count | xargs -d '\n' -I{} printf '%b\n' {} . (-d option is not specified in the POSIX standard, but is implemented by GNU systems).cat demo | xargs -d '\n' -I{} printf '%b\n' {} | grep count > resultprintf '%b\n' "$(<demo)" | grep count > result if you want to grep after the %b conversion, or, printf '%b\n' "$(grep count demo)" > result if you want to grep before the %b conversion.You could use printf to convert the hexadecimal data. Depending on the size of your input, you could read the lines into an array then use IFS to delimit the output:
join() {
local IFS="$1"
shift
echo "$*"
}
arr=( $(while read -r line; do printf "$line "; done < demo) )
join $' ' "${arr[@]}"
"count" "count" "count"
join $'\n' "${arr[@]}"
"count"
"count"
"count"
printf can be used for this purpose, but the way it is used in this answer is incorrect. printf's format string shouldn't come from user input (unless the input is supposed to be a format string): It won't produce correct output if the input contains format specifications (eg, %s, %d, etc.).