[SOLVED] Bash convert word with "echo" to binary and back?

blumenwesen · 03-20-2022, 11:46 AM

Hello, is it possible to convert a word to binary and back to utf only with the command "echo" in bash, as in the following example?

Is it also possible to convert a word to hex and back to utf only with the "echo" command?

bash:

Code:

a=$(echo "test" | tr -d "\n" | xxd -p)
echo $a
echo $a | xxd -r -p

Thank you

NevemTeve · 03-20-2022, 12:34 PM

It's not 'only with `echo`', you're using `xxd` and `tr` as well. Please describe your problem more precisely. For example, what would be word 'AARDVARK' in binary?

blumenwesen · 03-20-2022, 01:41 PM

I have try this but have no input from word to integer.
Is there no way without xxd / bc / od like in the secound example?

Code:

z=({0..1}{0..1}{0..1}{0..1}{0..1}{0..1}{0..1}{0..1})
echo ${z[65]}

st1="74657374"
for ((z=0; z<${#st1}; z+=2)); do a+=$(echo -ne $(echo ${st1:$z:2}) " "); done; st2=$a; unset a;
echo $st2

astrogeek · 03-20-2022, 01:52 PM

As already asked, please describe in simple sentences, rather than with code, the problem you are actually trying to solve. Your descriptions so far seem ambiguous and inconsistent. For example, define exactly what you mean by "word" as used in your question.

Please review the Site FAQ for guidance in asking well formed questions. Especially visit the link from that page, How to Ask Questions the Smart Way for discussion of things to consider when asking others for help.

blumenwesen · 03-20-2022, 02:21 PM

conversion:
word="helloworld"

to:
results="01101000 01100101 01101100 01101100 01101111 01110111 01101111 01110010 01101100 01100100"

code example:

Code:

st1="helloworld"
EXECUTE="???" CODE TO EXECUTE
for ((z=0; z<${#st1}; z+=2))
    do 
        a+=$(echo -ne "$EXECUTE")
    done
st2=$a; unset a
echo $st2

pan64 · 03-20-2022, 03:53 PM

although it is not impossible it is not really supported by [pure] bash. Would be better to use an external converter (like od, xxd or similar)

NevemTeve · 03-21-2022, 12:15 AM

(Guess this is still the 'Shadowrun decker' project.)

dugan · 03-21-2022, 12:07 PM

You can do this in Python pretty easily.

dugan · 03-21-2022, 06:14 PM

The Python:

Code:

>>> "".join(f"{ord('h'):08b} " for c in "helloworld").strip()
'01101000 01101000 01101000 01101000 01101000 01101000 01101000 01101000 01101000 01101000'

GazL · 03-21-2022, 09:04 PM

That was an interesting challenge. Pure bash solution:

Code:

#!/bin/bash

declare -A nibbles=(
  [0]='0000'  [1]='0001'  [2]='0010'  [3]='0011'
  [4]='0100'  [5]='0101'  [6]='0110'  [7]='0111'
  [8]='1000'  [9]='1001'  [a]='1010'  [b]='1011'
  [c]='1100'  [d]='1101'  [e]='1110'  [f]='1111'
)

for (( i=0 ; i < ${#1} ; i++ ))
do
  byte=$( printf "%x" "'${1:i:1}" )
  for (( j=0 ; j < ${#byte} ; j++ ))
  do
	n=${byte:j:1}
	printf "%s" ${nibbles[$n]}
  done
  printf " "
done
printf "\n"

Code:

$ /tmp/binbash.sh "A€B"
01000001 0010000010101100 01000010 
$

'€' is unicode 20ac, so it gets more bits.

pan64 · 03-22-2022, 12:47 AM

Quote:

Originally Posted by GazL

That was an interesting challenge. Pure bash solution:

Yes, that was what I was thinking about (or at least something similar).

https://unix.stackexchange.com/quest...nversion-tools

And what about the other direction?

Code:

#!/bin/bash
for c; do printf "\\$(printf "%o" "$((2#$c))")"; done

does not work on multibyte chars (like €)

GazL · 03-22-2022, 06:07 AM

Quote:

Originally Posted by pan64

And what about the other direction?

Code:

#!/bin/bash
for c; do printf "\\$(printf "%o" "$((2#$c))")"; done

does not work on multibyte chars (like €)

The only way I could think of is to use a eval and $'\u____'

bash functions for both directions:

Code:

str_to_binary()
{
  # Converts a locale specific charset string, passed as arg1, into a string
  # of unicode code-points using an ascii representation of base 2.

  local arg="$1"
  local i j n wchar

  declare -A nibbles=(
	[0]='0000'  [1]='0001'  [2]='0010'  [3]='0011'
	[4]='0100'  [5]='0101'  [6]='0110'  [7]='0111'
	[8]='1000'  [9]='1001'  [a]='1010'  [b]='1011'
	[c]='1100'  [d]='1101'  [e]='1110'  [f]='1111'
  )

  for (( i=0 ; i < ${#arg} ; i++ ))
  do
	printf -v wchar '%x' "'${arg:i:1}"
	for (( j=0 ; j < ${#wchar} ; j++ ))
	do
	  n=${wchar:j:1}
	  printf '%s' ${nibbles[$n]}
	done
	printf " "
  done
  printf "\n"
}

binary_to_str()
{
  # Converts a string of unicode code-points in an ascii representation,
  # of base 2 back into a locale specific charset string.

  local arg=$1
  local byte ucode

  for byte in $arg
  do
	if ucode=$(( 2#$byte )) ; then
	  printf -v ucode '\\u%x' "$ucode"
	  eval printf \'%s\' \$\'$ucode\'
	else
	  >&2 printf "error: invalid binary value: %s\n" "$byte"
	  return 1
	fi
  done
  return 0
}

Of course, this is an insane way to represent characters, so really, just for entertainments sake.

blumenwesen · 03-22-2022, 08:58 AM

thank you solved my problem.