Converting a string into Unsigned int 8 array-CodePudding

I am new to bash and I am trying to convert a swift obfuscation into a bash script.

Basically, I want to convert a string into an Unsigned-Int 8 array (UTF-8).

For example,

"hey" = [104, 101, 121] (UTF-8 UINT8 value)
"example" = [101, 120, 97, 109, 112, 108, 101] (UTF-8 UINT8 value)

Does anyone know if this is possible?

CodePudding user response：

The following shell script converts input in the for of hey into the string [104, 101, 121].

# Print hey
printf "%s" hey |
# convert to hex one per line
xxd -p -c 1 |
# convert to decimal one per line
xargs -I{} printf "%d\n" 0x{} |
# Join lines with comma
paste -sd, |
# Add spaces after comma
sed 's/,/, /g' |
# Add [ ]
{ echo -n '['; tr -d '\n'; echo ']'; }
# echo "[$(cat)]"

The script is not aware of input encoding - the script only translates bytes representation. The input string has to be already in the desired encoding. Use iconv to convert between encodings.

CodePudding user response：

Using pure bash, no external programs:

#!/usr/bin/env bash                                                                                                                                                                                                                              

to_codepoints() {
    local LC_CTYPE=C IFS=, n
    local -a cps
    # Iterate over each byte of the argument and append its numeric value to an array                                                                                                                                                            
    for (( n = 0; n < ${#1}; n   )); do
        cps =( $(printf "%d" "'${1:n:1}") )
    done
    printf "[%s]\n" "${cps[*]}"
}

to_codepoints hey
to_codepoints example
to_codepoints $'\u00C4ccent'

outputs

[104,101,121]
[101,120,97,109,112,108,101]
[195,132,99,99,101,110,116]