String Alteration Using Sed

Question

Once i execute | xxd -bi, I get the output as

00000000: 01101000 01100101 01101100 01101100 01101111 00001010 hello.

Now I am trying to remove initial 0000000, which I did using sed and I am trying to remove "hello" (variable text) which is stored in $covertthis using sed but I can't achieve so, as shown below in the example please suggest me the changes.

I have tried using cut aswell but can't alter both 0000000 and last variable word.

echo "What you want to input numbers or string?"
read input

if [[ "$input" == "number" ]] || [[ "$input" == "Number" ]] || [[ "$input" == "NUMBER" ]] ;then
        echo "Number selected 1"
elif [[ "$input" == "String" ]] | [[ "$input" == "STRING" ]] || [[ "$input" == "string" ]] ;then
        echo "String selected"
        echo "Please give me the string to be XOR'ed"
        read convertthis
        echo  $convertthis | xxd -bi > bin-store
        $(sed -i -e 's/00000000://g' bin-store)
        $(sed -i -e 's/($convertthis).//g' bin-store)
else
        echo "Please re-run the script, input is wrong"
fi

Answer 1

You need only one sed for that. Separate commands using ;

# read convert_this
echo "00000000: 01101000 01100101 01101100 01101100 01101111 00001010  hello." | 
sed -E "s/^[^[:blank:]]*[[:blank:]]+//;s/[[:blank:]]+$convert_this\.$//" 

should do it.

Output

01101000 01100101 01101100 01101100 01101111 00001010

Explanation

• The -E option with sed enables the use of Extended regular expressions.
• s/^[^[:blank:]]*[[:blank:]]+//; The ^ in the beginning looks for the start of the line. [] in sed is meant for ranges. If the range begins with a ^ that means you're negating a range. The * looks is meant for zero or more and + is meant for one or more. The [:blank:] is a character class matching any blank characters like whitespaces tabs and so. In short we are looking for any non-blank characters in the beginning followed by one or more spaces. Then we substitute it with nothing effectively deleting it.
• The second substitute replaces the string stored in $convert_this and any full-stop that follows with nothing.

All good :-)

---

^{Sidenote: You need to use double quotes to wrap the sed commands so that your bash variables are expanded.}

Answer 2

Following awk may also help you in same.

awk -v val="hello." '{sub(/^0+: +/,"");sub(val,"")} 1'   Input_file

Or if you have a shell variable which we want to pass to awk command then following may help you too:

convertthis="hello."
awk -v val="$convertthis" '{sub(/^0+: +/,"");sub(val,"")} 1'   Input_file

Output will be as follows.

01101000 01100101 01101100 01101100 01101111 00001010

Answer 3

Use

sed -i -e "s/$convertthis.$//" bin-store 

instead of

$(sed -i -e 's/($convertthis).//g' bin-store)

to use the content of your convertthis variable instead of its literal name

Answer 4

Try this command:

sed "s/\(^[0-9a-z]\{8\}:[[:space:]]\)\([^a-z]*\)\([a-z]*\)/\2/"

The regex splits the input into three group matches; the groups are delineated using escaped brackets \( \). The groups are as follows:

1. The first group is alphanumeric and 8 characters, prepended by a colon and a space. We have ^ to match the beginning then and 8 alphanumeric characters matched by [0-9a-z]\{8\}: in this expression the square braces enclose a character class (alphanumeric). The space can be matched by [[:space:]].

2. The second group contains no alphabetic characters which we match as [^a-z]* where the ^ inverts the match in square braces.

3. The third group matches the alphabetic characters at the end.

We can then replace the string by any of the three groups: group 2 (\2) is the one that we want.

This works just as well for multi-line input as sed matches line by line.

---

Side note : for POSIX compliance to match spaces [[:space:]] should be used not \s (see this question).