Get unique elements from Array

Shell Programming and Scripting
1

I have an array code and output is below:

    
echo $1

while read -r fline; do
    echo "%%%%%%$fline%%%%%"
    fmy_array+=("$fline")
    done <<< "$1"

Output:

CR30903 YU0007 SRIL CR30903 Yogesh SRIL
%%%%%%CR30903    YU0007     SRIL%%%%%
%%%%%%CR30903    Yogesh SRIL%%%%%

Desired output:

TKT: 30903

ACTOR:
YU0007
Yogesh

DEPT: SRIL

As you can see TKT & DEPT gets single unique value printed while ACTOR has two values because they are not unique.

I know how to get unique element from array using sort -u but here we are dealing with a sub-element of an array.

Can you please suggest how can achieve the desired output ?

2

Hi
To get started, simply print the value of this parameter

echo "$1"

--- Post updated at 17:15 ---

Something like this?

cat file
30903	 YU0007	SRIL
30903	 Yogesh	SRIL
awk '{ B[$2] } END { print "TKT:", $1 RS "ACTOR:"; for (i in B) print i; print "DEPT:", $3 }' file

--- Post updated at 17:20 ---

or maybe

if [[ "${my_array[@]}" =~ $fline ]]; then
    continue
fi
3

Unfortunately I moved away from my system. Can we do without that information?

4

here's another

for fline in $1; do
if ! [[ "${my_array[@]}" =~ "$fline" ]]; then
        my_array+=("$fline")
fi
done
echo ${my_array[@]}
5

This won't work for dynamic data. We need to eliminate duplicate entries only if it exists. You are nowhere checking for unique or duplicate entries in your logic.

6

Assuming the input file has got many TKT records in sequence, an associative array is perfect:

awk '
NF>=3 {
  A[$1]=(A[$1] ORS $2)
  D[$1]=$3
}
END {
  for (t in A) {
    print "TKT:", t
    print  "ACTOR:", A[t]
    print "DEPT:", D[t]
    print ""
  }
}' file

Also doable in bash 4 (using its associative arrays).

7 
awk '!T[$0]++' RS='[[:space:]]+' <<<"$1"

or

grep -o '\S*' <<<"$1" | sort -u

Further, if you want, you can arrange this into an array, here is a sample

my_array=($(grep -o '\S*' <<<"$1" | sort -u))

And please pay attention to post number #4

1 Like
8

There is no file involved here.

I tried your solution but i get syntax error

    while read -r fline; do
      fmy_array+=("$fline")
echo $fline | awk 'NF>=3 {A[$1]=(A[$1] ORS $2) D[$1]=$3 } END {   for (t in A) {    print "TKT:", t    print  "ACTOR:", A[t]    print "DEPT:", D[t]    print ""  }}'
    done <<< "$1"

error output:
awk: cmd. line:1: NF>=3 {A[$1]=(A[$1] ORS $2) D[$1]=$3 } END {   for (t in A) {    print "TKT:", t    print  "ACTOR:", A[t]    print "DEPT:", D[t]    print ""  }}
awk: cmd. line:1:                                  ^ syntax error

--- Post updated at 12:52 PM ---

This solution did not help. Although it took care of the duplicate entries there is no way to determine the three field of each array element into "TKT", "ACTOR" & "DEPT"

here is the output I got:

--- Post updated at 01:06 PM ---

There is no files involved here.

Also, I have updated the Original post with the output of $1.

--- Post updated at 01:43 PM ---

grep -o '\S*' <<<"$1" | sort -u

<- This works but there is no order so i dont know which element is what as requested in the desired output. Can you please help assign this to get displayed as below:

Current output:

9 
egrep -o '(\S+\s+){2}\S+' <<<"$1" |
  awk '
        {A[$1]; B[$3]; C[$2]}
  END   { print "TKT:" cat(A) "\n\nDEPT:" cat(B) "\n\nACTOR:" cat(C)}
  function cat(T)
        { s=""; for (i in T)
                  s = s RS i
                  return s
        }'
1 Like