CODEWITHSUNDEEP
bash
James Starlight
James Starlight

Reputation: 523

bash: assign variable on the fly according to the pattern in file name

My bash script operates with different files located within the same directory and then execute some program for them. Each files match the following pattern where the ${receptor} keyword is always located after the second "_" of the file name. This keyword defines the group of the file.

something_something_${receptor}_repX.pdb

According to the ${receptor}, all files are devided in three groups. So I create a special function which set the name of the prefix and execute all operations in flow only for the files contained this index:

set_receptor () {
#receptor='5r82dim'
#receptor='2zu2dim'
receptor='7dr8dim'
}


set_receptor

for pdb in "${docking_pdb}"/*_${receptor}_*.${lig_ext}; do
pdb_name=$(basename "$pdb" .${lig_ext})
pdb_name="${pdb_name/_rep1}"
echo "Converting ${output}: the system ${pdb_name}
done

This means that each time when I need to switch between different groups I need to uncomment $receptor manually in the set_receptor function. How could I determine $receptor automatically in the workflow in order that my script could be executed only 1 time for any number of groups and determine each group automatically in the begining of FOR loop execution?

Upvotes: 0

Views: 40

Answers (1)

choroba
choroba

Reputation: 242038

If the keyword is always after the second underscore, you can use parameter expansion to extract it:

#!/bin/bash
for pdb in "$docking_pdb"/*_*_*."$lig_ext" ; do
    receptor=${pdb##*/}       # Remove the path.
    receptor=${receptor#*_}   # Up to the first underscore.
    receptor=${receptor#*_}   # Up to the second underscore.
    receptor=${receptor%%_*}  # Remove everything after the 1st underscore.
    ...

done

Upvotes: 1

Related Questions