Bash-Script (separating string)

Question

I have an exercise in which I have to get a string from the user read str and to print the number of times which the last character of this string is being repeated.

For example, the string is: ab2aa2b122 and the output will be : 4 because 2 is being repeated 4 times in the string.

My problem is to divide the string into single characters. I want that every character will be separated from the other character by space " " or by a \n in order to use the tr command and uniq command.

I have to use uniq and sort commands, in order to use these I have to split the characters, which means that every character must be separated by space or by \n.

Answer 1

A 100% pure bash solution:

read -r -p "Enter your string: " a
l=${a:(-1)}
b=${a//[^$l]/}
n=${#b}
echo "Last character is \`$l' and is repeated $n times in the string \`$a'"

No loop, no external process, nothing!

As a one-liner (to show it's really short, and probably the shortest possible!):

echo "ab2aa2b122"|{ read -r a;b=${a//[^${a: -1}]/};echo ${#b};}

Edit

Explanations of the first version (the second one is exactly the same, but suited for a code-golf contest).

• First line line reads user input and stores that input in variable a.
• l=${a:(-1)} will store the last character of a in l. More generally, ${a:i:j} expands to the substring of a starting at index i, of length j. When i=-1, it expands to the last character.
• Parameter expansion is used in the next line: ${a//P/R} will expand to a withall the P's replaced by R. Here, R is empty, and P is [^$l], so this matches all characters distinct from l. Bottom line: ${a//[^$l]/} expands to a where all the characters different from the last are deleted. So we're left with just 2222 in your case.
• ${#b} expands to the size of b (here, its length). So we're done!

Answer 2

Not sure that you can use sed in your exercise, but with sed it can be done like this:

bash-3.1$ PPP=`echo ab2aa2b122 | sed 's/\(.\)/\1 /g'`
bash-3.1$ echo $PPP
a b 2 a a 2 b 1 2 2

Answer 3

awk one-liner:

kent$  echo "ab2aa2b122"|awk  'BEGIN{FS=""}{for(i=1;i<=NF;i++)a[$i]++;print "last char "$NF" : "a[$NF]" times";}'
last char 2 : 4 times

if you just want to get the number "4":

kent$  echo "ab2aa2b122"|awk  'BEGIN{FS=""}{for(i=1;i<=NF;i++)a[$i]++;print a[$NF]}'
4

Answer 4

From the OP:

My problem is to divide the string into single characters. I want that every character will be seperated from the other character by space (" ") or by a \n (next line)-in order to use the tr command and uniq command.

From a comment left as a comment to my other answer:

Iv'e learnt basic commands such as cat,cut,tr,sort,uniq,for and while loops,if, cp,wc,input-output (>|),abit of globe style and extended regular expressions (egrep..) but I'm allowed to use egrep and Extended reg expressions, and globe style in the current excercise. Thanks!

To "divide" your string using grep:

• Using a terrible echo-pipe and subshell:

  echo "$string" | grep -o '.'
  

• Using a here string:

  grep -o '.' <<< "$string"
  

Hope this helps!

Answer 5

Only using wc, cut and tr:

#/bin/bash

input="ab2aa2b122"

length=$(wc -c <<< $input)

last=$(cut -c $((length-1)) <<< $input)

tr -dc $last <<< $input | wc -c

wc -c counts the number of characters in the string.

cut -c $len-1 gets the last character in the string

tr -dc $last removes all the characters that aren't the last (i.e 2) leaving the string 2222 and then use wc -c to count the length again.

./script.sh 
4