Replace specific string using sed

Question

I have a code.txt file that contains morse code for example

.- .-.

I have a function called decode inside a bash file called morse as this:

decode (){ 
   sed -i 's/ \.-/A/g' $1
   sed -i 's/ \.-./R/g' $1
   cat $1
}

When I type in terminal $bash morse decode code.txt

I receive:

AA.

The output I want is :

AR

How can it see separate that the string .- is A and the .-. is R?

Run a single sed like this: sed -i 's/ \.-\./R/g; s/ \.-/A/g' "$1" — anubhava
– anubhava, Commented Feb 6, 2020 at 15:00
If all you are trying to do is output the translation to stdout, there is no need to edit the file or use cat. You could just run the combined sed anubhava suggested without the -i option. Then the output will go to stdout without the use of cat. Then you can use your function like this: decode code.txt > decoded.txt — Jason
– Jason, Commented Feb 6, 2020 at 15:06

Matias Barrios · Accepted Answer · 2020-02-06 16:26:26Z

1

If your intention is to encode and decode Morse messages with any tool then something like this will do :

#!/usr/local/bin/python3
import re

alphabet = { 'A':'.-', 'B':'-...', 'C':'-.-.', 'D':'-..', 'E':'.', 'F':'..-.', 'G':'--.', 'H':'....', 'I':'..', 'J':'.---', 'K':'-.-', 'L':'.-..', 'M':'--', 'N':'-.', 'O':'---', 'P':'.--.', 'Q':'--.-', 'R':'.-.', 'S':'...', 'T':'-', 'U':'..-', 'V':'...-', 'W':'.--', 'X':'-..-', 'Y':'-.--', 'Z':'--..', '1':'.----', '2':'..---', '3':'...--', '4':'....-', '5':'.....', '6':'-....', '7':'--...', '8':'---..', '9':'----.', '0':'-----', ', ':'--..--', '.':'.-.-.-', '?':'..--..', '/':'-..-.', '-':'-....-', '(':'-.--.', ')':'-.--.-',' ':'  '} 

def encode(message): 
    return "".join([ ( alphabet[letter.upper()] + ' ' ) if letter != ' ' else '  ' for letter in message])

def decode(message):
    return "".join([ list(alphabet.keys())[list(alphabet.values()).index(item if item != '|' else '  ')] for item in re.sub(r' {2,}', ' | ',message).split(' ')])

print(encode('THIS IS FINE'))
print(decode('- .... .. ...   .. ...   ..-. .. -. .'))

Hope it helps too.

answered Feb 6, 2020 at 16:26

Matias Barrios

5,0864 gold badges31 silver badges71 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

P.Bes Over a year ago

That a good effort. I am searching for something like that but in a bash file with unix not in python..I would appreciate if you can help me there.

Matias Barrios Over a year ago

@P.Bes Well you are in the right track if thats your goal. sed looks like a good start. Let me see how much of that Python script I can translate to Bash. Can you use awk at least? or you need pure bash?

P.Bes Over a year ago

pure bash I think

Ivan · Accepted Answer · 2020-02-07 07:20:34Z

Wow interesting idea! Based on @MatiasBarrios alphabet i made this.

#!/bin/bash

string=$1

declare -A morse=(
    [A]='.-'    [B]='-...'  [C]='-.-.'  [D]='-..'   [E]='.'
    [F]='..-.'  [G]='--.'   [H]='....'  [I]='..'    [J]='.---'
    [K]='-.-'   [L]='.-..'  [M]='--'    [N]='-.'    [O]='---'
    [P]='.--.'  [Q]='--.-'  [R]='.-.'   [S]='...'   [T]='-'
    [U]='..-'   [V]='...-'  [W]='.--'   [X]='-..-'  [Y]='-.--'
    [Z]='--..'

    [1]='.----'  [2]='..---'  [3]='...--'  [4]='....-'  [5]='.....'
    [6]='-....'  [7]='--...'  [8]='---..'  [9]='----.'  [0]='-----'
    [(]='-.--.'  [)]='-.--.-' [/]='-..-.'  [-]='-....-' [+]='.-.-.'
    [.]='.-.-.-' [,]='--..--' [?]='..--..' [!]='-.-.--' [ ]='  '
)

morse () {
    while [[ "$string" ]]; do
        symbol="${string::1}"
        printf -- "${morse["${symbol^}"]} "
        string="${string:1}"
    done
}

demorse () {
    declare -A demorse
    for item in "${!morse[@]}"; { demorse["${morse["$item"]}"]="$item"; }
    while [[ $@ ]]; do
        printf -- "${demorse["$1"],}"
        shift
    done
}

case $string in
    demorse) shift; demorse "$@";;
    *      )          morse     ;;
esac

Usage

$ ./morse 'hello world!'
.... . .-.. .-.. ---    .-- --- .-. .-.. -.. -.-.--

Demorse also worsk but, spaces have to be printed like this ' '

$ ./morse demorse .... . .-.. .-.. --- '  ' .-- --- .-. .-.. -.. -.-.--
hello world!

Wiktor Stribiżew · Accepted Answer · 2020-02-07 12:49:16Z

You need to run s/ \.-\./R/g replacement first. Note the second . must be escaped to only match a dot.

Hence, use

sed 's/ \.-\./R/g;s/ \.-/A/g' file

See the online demo

Or, another way:

sed -e 's/ \.-\./R/g' -e 's/ \.-/A/g' file

Replace the file with "$1" in your code.

UPDATE

Here is the translation of encoding / decoding Python function posted by Matias below:

#!/bin/bash

### Encoding:

declare -A MORSE=( [A]='.-' [B]='-...' [C]='-.-.' [D]='-..' [E]='.' [F]='..-.' [G]='--.' [H]='....' [I]='..' [J]='.---' [K]='-.-' [L]='.-..' [M]='--' [N]='-.' [O]='---' [P]='.--.' [Q]='--.-' [R]='.-.' [S]='...' [T]='-' [U]='..-' [V]='...-' [W]='.--' [X]='-..-' [Y]='-.--' [Z]='--..' [1]='.----' [2]='..---' [3]='...--' [4]='....-' [5]='.....' [6]='-....' [7]='--...' [8]='---..' [9]='----.' [0]='-----' [',']='--..--' ['.']='.-.-.-' [';']='-.-.-.' [':']='---...' ['?']='..--..' ['!']='-.-.--' ['/']='-..-.' ['-']='-....-' ['+']='.-.-.' ['(']='-.--.' [')']='-.--.-' ['_']='..--.-' ['"']='.-..-.' ["'"]='.----.' ['$']='...-..-' ['@']='.--.-.' ['&']='.-...' ['  ']=' '  )

function encode {
  res=''
  s="$1"
  for (( i=0; i<${#s}; i++ )); do
    letter="${s:$i:1}"
    if [[ "$letter" == ' ' ]]; then
      res="${res}  "
    else
      res="${res}${MORSE[${letter^^}]} ";
    fi
  done
  printf "%s" "$res"
}

echo "$(encode "THIS IS FINE")"

### Now, decoding

declare -A MORSEDEC=( ['-.--.-']=')' ['..--..']='?' ['--..--']=', ' ['-....-']='-' ['.-.-.-']='.' ['...--']='3' ['-.--.']='(' ['---..']='8' ['-..-.']='/' ['....-']='4' ['-....']='6' ['----.']='9' ['.----']='1' ['..---']='2' ['.....']='5' ['--...']='7' ['-----']='0' ['-...']='B' ['-..-']='X' ['-.-.']='C' ['--..']='Z' ['--.-']='Q' ['.-..']='L' ['-.--']='Y' ['..-.']='F' ['.--.']='P' ['.---']='J' ['...-']='V' ['....']='H' ['-..']='D' ['---']='O' ['..-']='U' ['...']='S' ['.--']='W' ['-.-']='K' ['.-.']='R' ['--.']='G' ['-.']='N' ['..']='I' ['--']='M' ['.-']='A' ['  ']=' ' ['.']='E' ['-']='T' )

function decode {
  res=''
  tmp="$(sed 's/ \{2,\}/ | /g' <<< "$1")";
  for word in $tmp; do
    if [[ "$word" == '|' ]]; then
      res="${res}${MORSEDEC['  ']}";
    else
      res="${res}${MORSEDEC[$word]}";
    fi
  done
  printf "%s" "$res"
}
echo "$(decode "- .... .. ...   .. ...   ..-. .. -. .")"

See Bash demo online.

Yes but that was an example..What can I do in the case of having all letters of alphabet..For example also "L" is .-.. this means that I must do the same for L
@P.Bes That is the same idea: the longer patterns should be handled before the shorter.

Gem Taylor · Accepted Answer · 2020-02-06 17:29:36Z

0

The easy answer in RE engines that support look-ahead and look-behind would be to treat the spaces as look-ahead and look-behind triggers, but sed does not support this.

Another option that avoids needing to order the letters is to inject extra symbols to help you mark each letter. Say we inject = round each space, then we can replace delimited sequences in any order, and finally get rid of the delimiters:

echo  .- .-.|sed -e 's/^\(.*\)$/=\1=/;s/ /= =/g' -e 's/=\.-\.=/=R=/g;s/=\.-=/=A=/g' -e 's/= =//g;s/^=//;s/=$//'

If you have rules that need to preserve multiple spaces, then that can be accommodated.

answered Feb 6, 2020 at 17:29

Gem Taylor

5,7351 gold badge11 silver badges34 bronze badges

Collectives™ on Stack Overflow

Replace specific string using sed

4 Answers 4

3 Comments

Comments

2 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

3 Comments

Comments

2 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related