Ytdyklly

current input is from the command

grep -i "final_model" /dir1/dir2/dir3/user/dir4/assemble.preprocessing

current output is:

final_model=</dir1/dir2/dir3/user/dir4/test_F00ME001.inp>

required Output-1 is:

test_F00ME001.inp

required Output-2 is:

F00ME001

Assuming that

final_model=</dir1/dir2/dir3/user/dir4/test_F00ME001.inp>

is a line of text that your grep command is producing,

grep ... | sed 's#.*/##; s/>$//' | tee output1 |

           sed 's/^[^_]*_//; s/..*//' >output2

This would save the two modified lines into output1 and output2

Testing:

$ printf '%sn' 'final_model=</dir1/dir2/dir3/user/dir4/test_F00ME001.inp>' |

  sed 's#.*/##; s/>$//' | tee output1 |

  sed 's/^[^_]*_//; s/..*//' >output2

$ cat output1

test_F00ME001.inp

$ cat output2

F00ME001

The first sed call operates on the original data and first removes everything up to and including the last / character. It then removes the > at the end.

The tee saves this into output1 and also passes it on to the next stage of the pipeline.

The second sed call operates on the modified data and first removes everything up to and including the first _ character. It then removes everything from the first dot onwards and redirects the result into output2.

Tried with below commands to achieve same

final_model=</dir1/dir2/dir3/user/dir4/test_F00ME001.inp>



o1=`echo $final_model| awk -F "/" '{print $NF}'|sed "s/[^[a-zA-Z_0-9.]//g"`



echo $o1



o2=`echo $o1|awk -F [_.] '{print $2}'`



echo $o2



F00ME001

$ final_model="</dir1/dir2/dir3/user/dir4/test_F00ME001.inp>"

$ o1=$(echo "$final_model" | sed -e 's/<(.*)>/1/')   #remove <>

$ o1=$(basename "$o1")   #get basename (test_F00ME001.inp)

$ echo "$o1"

test_F00ME001.inp

$ o2=$(echo "$o1" | sed -e 's/test_(.*).inp/1/')   #get text between test_ and .inp

$ echo "$o2"

F00ME001

Try more use google next time, you can find both commands answered somewhere.

sed command source: https://stackoverflow.com/a/13242517/7351855

replace the grep -i final_model with perl, and you capture the data you need:

$ perl -nE '/^final_model=.*/(test_([^.]+)[^>]+)/ && say "$1n$2"' assemble.preprocessing

test_F00ME001.inp

F00ME001

The first set of capturing parentheses grabs the stuff after the last slash: (test_([^.]+)[^>]+)

The second set grabs the stuff between the underscore and the dot.

Then you can capture the 2 lines into variables with this construct:

{ read output1; read output2; } < <(

    perl -nE '/^final_model=.*/(test_([^.]+)[^>]+)/ && say "$1n$2"' assemble.preprocessing 

)

We are redirecting a Process Substitution into the grouping of 2 read commands.

$ echo "$output1"

test_F00ME001.inp

$ echo "$output2"

F00ME001