Shell Programming and Scripting

I have directory path in which there are several sub directories. In all these sub dir there will be one env.cnf file. I want to copy this env.cnf file from
each sub dir's and place them in destination path by creating same filename as sub dir_env.cnf.

After copying env.cnf files from source to destination dir, I have to run few unix commands(grep) on each files and print the grep command o/p in result files in output dir.

I have seen that we can use the below command to do the recursive file copy part but don't know how to process grep command on each files

Code:

 
find /src_path/to/ -iname '*.cnf' -exec cp {} /new/directory/ \;

Ex:
path : /tmp/config/ has 3 sub dir's

Code:

SubA ---- has env.cnf
SubB ---- has env.cnf
SubC ---- has env.cnf
.
.
.

after running the script, it should place the env.cnf file in input dir in /tmp/Config/input as name:

Code:

SubA_env.cnf
SubB_env.cnf
SubC_env.cnf
.
.
.

now run the same script again on each files present in the input dir to perform grep command and put the output in output dir in /tmp/Config/output/

Code:

SubA_env.txt
SubB_env.txt
SubC_env.txt

ManyThanks,

One solution is to pipe the find output to a read loop and process pathname within loop:

Code:

find /src_path/to/ -iname '*.cnf' -print | while read src
do
   dest=$(dirname "$src" | sed 's:.*/::')
   dest="${dest}_$(basename "$src")"
   log=$(basename "$dest" .cnf).txt
   cp "$src"  "/tmp/Config/output/$dest"
   grep SEARCH_STRING "/tmp/Config/output/$dest" > "/tmp/Config/output/$log"
done

Thanks Chubler. your solution did work as expected but am trying change the ouput file name as *_result.txt. I tried to update this line like this but didn't work

Code:

log=$(basename "$dest"_result .cnf).txt

Could you please tell me what am doing is correct in above stmt and if possible can you tell me how to acrchive this generated file (input/output) to someother path. so that if we run this script again it should genreated the new set of input/output file then it should generate new compare result file by comparing pervious result and new result file.

like :
1. generate input/result file (already done by above script)

2. If we run the above script again, it should check if there is any files present in output folder if yes, then move this genereated files to someother location with datetime stamp in the filename(archive) . if no files found then just generate files.

3. from step 2, if new set of results files generated (if files archived), then should compare archived result file to newly genereated result file.
ie, something like this

Code:

sdiff -s old_result_file new_result_file

please do help me on this.

ManyThanks,

Use:

Code:

log=$(basename "$dest" .cnf)_result.txt

for 2 you could do this:

Code:

[ -d /tmp/Config/output ] && mv /tmp/Config/output /tmp/Config/output_$(date +%Y%M%d%H%M)
mkdir -p /tmp/Config/output

Thanks Chubler. it worked as i was expecting. Can you please advise me on how to achive the point 3.

from step 2, if new set of results files generated (if files archived), then should compare archived result file to newly genereated result file.