passing awk variable to bash script

Question

I am writing a bash/awk script to process hundreds of files under one directory. They all have name suffix of "localprefs". The purpose is to extract two values from each file (they are quoted by ""). I also want to use the same file name, but without the name suffix.

Here is what I did so far:

#!/bin/bash

for file in *  # Traverse all the files in current directory.
   read -r name < <(awk ' $name=substr(FILENAME,1,length(FILENAME)-10a) END {print name}' $file) #get the file name without suffix and pass to bash. PROBLEM TO SOLVE!
   echo $name # verify if passing works.
   do
   awk  'BEGIN { FS = "\""} {print $2}' $file #this one works fine to extract two values I want.
   done

exit 0

I could use

awk '{print substr(FILENAME,1,length(FILENAME)-10)}' to extract the file name without suffix, but I am stuck on how to pass that to bash as a variable which I will use as output file name (I read through all the posts on this subject here, but maybe I am dumb none of them works for me).

If anyone can shed a light on this, especially the line starts with "read", you are really appreciated.

Many thanks.

konsolebox · Accepted Answer · 2013-09-01 17:07:39Z

2

Try this one:

#!/bin/bash

dir="/path/to/directory"

for file in "$dir"/*localprefs; do
    name=${file%localprefs}  ## Or if it has a .: name=${file%.localprefs}
    name=${name##*/}  ## To exclude the dir part.
    echo "$name"
    awk 'BEGIN { FS = "\""} {print $2}' "$file"  ## I think you could also use cut: cut -f 2 -d '"' "$file"
done

exit 0

answered Sep 1, 2013 at 17:07

konsolebox

76.3k13 gold badges110 silver badges114 bronze badges

Sign up to request clarification or add additional context in comments.

6 Comments

user46292 Over a year ago

it works like charm. Thank you so much! I will use yours as my template to beatify the output.

user46292 Over a year ago

#!/bin/bash dir="/path/to/dir" diro="/path/to/dirout" for file in "$dir"/*localprefs; do name=${file%localprefs} ## Or if it has a .: name=${file%.localprefs} name=${name##*/}; ## To exclude the dir part. touch "$diro"/"$name" awk 'BEGIN { FS = "\""} {print $2}' "$file" >>"$diro"/"$name" done exit 0

user46292 Over a year ago

How can I write the for loop to get the name first and create the file with touch, then run the awk part to put two lines into the file? See above. It is very close but it has two duplicates in the output file

konsolebox Over a year ago

@user46292 Your script seems the right way to do it. What is the error? And does awk produce two lines always? Also would we assume that $diro exists? Or you have to create it?

user46292 Over a year ago

Per your suggestion, it now produces only two lines. Thanks again.

|

stark · Accepted Answer · 2013-09-01 17:20:50Z

1

To just take sbase name, you don't even need awk:

for file in * ; do
   name="${file%.*}"
   etc
done

edited Sep 1, 2013 at 17:20

answered Sep 1, 2013 at 17:06

stark

13.3k3 gold badges39 silver badges56 bronze badges

3 Comments

user46292 Over a year ago

Here is what I got with your suggestion:line 12: syntax error near unexpected token FILE=awk '{print substr(FILENAME,1,length(FILENAME)-10)}'' line 12: FILE=awk '{print substr(FILENAME,1,length(FILENAME)-10)}''

Kevin Over a year ago

No, no, no. Use $()

stark Over a year ago

Edited - it was force of habit to use backticks

Collectives™ on Stack Overflow

passing awk variable to bash script

2 Answers 2

6 Comments

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

6 Comments

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related