[SOLVED] bash find

1s440 · 06-10-2022, 01:35 PM

Hi all,

I wanted to find the "temp.txt" files recursively in directories and subdirecotires and then display the output of temp.txt as well.

#!/bin/bash
filename='temp.txt'
n=1
while read line; do
# reading each line
echo "$line"
n=$((n+1))
done < $filename

I am not sure if i can just add find command to the above script.

pan64 · 06-10-2022, 02:05 PM

Quote:

Originally Posted by 1s440

Hi all,

I wanted to find the "temp.txt" files recursively in directories and subdirecotires and then display the output of temp.txt as well.
#!/bin/bash
filename='temp.txt'
n=1
while read line; do
# reading each line
echo "$line"
n=$((n+1))
done < $filename

please use code tags: https://www.linuxquestions.org/quest...do=bbcode#code
Why don't you use: cat $filename ?

Quote:

Originally Posted by 1s440

I am not sure if i can just add find command to the above script.

Why not?

ondoho · 06-10-2022, 02:22 PM

Looks a lot like homework to me.

fwiw, you can use find to simply find all files named temp.txt (option: -name), and execute a command like cat on each (option: -exec).
Please read:

Code:

man find

Please use CODE tags for full command output and code (see my signature).

teckk · 06-10-2022, 02:24 PM

I wanted to find the "temp.txt" files recursively in directories

Code:

find . -name temp.txt

1s440 · 06-10-2022, 11:26 PM

[QUOTE=pan64;6359997]please use code tags: https://www.linuxquestions.org/quest...do=bbcode#code
Why don't you use: cat $filename ?

You mean

Code:

#!/bin/bash
filename='temp.txt’
find . -name $filename
cat $filename

1s440 · 06-11-2022, 12:29 AM

Quote:

Originally Posted by ondoho

Looks a lot like homework to me.

fwiw, you can use find to simply find all files named temp.txt (option: -name), and execute a command like cat on each (option: -exec).
Please read:

Code:

man find

Please use CODE tags for full command output and code (see my signature).

It’s not a homework. I think this group is to help people when people get stuck at some point..

1s440 · 06-11-2022, 12:31 AM

Quote:

Originally Posted by teckk

I wanted to find the "temp.txt" files recursively in directories

Code:

find . -name temp.txt

If I need to find files across the all the servers then I think bash script won’t help isn’t it ?

NevemTeve · 06-11-2022, 12:52 AM

Now that smalltalk is over, can you ask your actual question?

rnturn · 06-11-2022, 01:09 AM

Quote:

Originally Posted by 1s440

Hi all,

I wanted to find the "temp.txt" files recursively in directories and subdirecotires and then display the output of temp.txt as well.

#!/bin/bash
filename='temp.txt'
n=1
while read line; do
# reading each line
echo "$line"
n=$((n+1))
done < $filename

I am not sure if i can just add find command to the above script.

Your script appears to read a single file line by line, counting each line (but never using that result for anything). I'm not sure how you could insert a recursive find into that script either.

Code:

find . -type f -name temp.txt -exec cat {} \;

will display all the files named 'temp.txt' in and beneath the current directory.

If I was correct in assuming that you were trying to count all the lines in the files you find, to get that, you could use:

Code:

wc -l $( find . -type f -name temp.txt ) | grep total | awk '{print $1}'

Instead of using 'find' twice, use it once to capture the list of files in an array:

Code:

declare -a FILES
FILES=( $( find . -type f -name text.txt ) )

While it's not strictly necessary to declare the array, it's not a bad idea to document what that variable is used for (when you come back to make changes to your script next year).

Once you have an array of filepaths, you can use it in other commands:

Code:

for F in ${FILES[@]}
do
    cat ${F}
done
echo "$( wc -l ${FILES[@]} | grep total ) lines."

Putting it all together (and simplifying it a bit) and you'd get:

Code:

declare -a FILES
FILES=( $( find . -type f -name text.txt ) )
cat ${FILES[@]}
echo "$( wc -l ${FILES[@]} | grep total ) lines."

Suggested leisure reading: Advanced Bash-Scripting Guide -- An in-depth exploration of the art of shell scripting

Have a lot of fun.

rnturn · 06-11-2022, 01:56 AM

Quote:

Originally Posted by 1s440

If I need to find files across the all the servers then I think bash script won’t help isn’t it ?

Hmm... that information might have been nice to know up front.

You could accomplish what I think you want to do using a utility like Ansible (or something similar) assuming you have an inventory of "all the servers". But it doesn't make much sense to propose solutions when the problem hasn't been fully presented.

At this point, however, I think the discussion would benefit greatly from a clear description of the problem you are trying to solve. What other information can you provide? It appears that locating text files may only be a small part of what you are attempting to do.

NevemTeve · 06-11-2022, 07:43 AM

Code:

for i in host1 host2; do
    echo "=== $i ==="
    ssh "$i" 'find / -type f -name temp.txt' </dev/null
done

1s440 · 06-11-2022, 10:05 AM

Quote:

Originally Posted by rnturn

Your script appears to read a single file line by line, counting each line (but never using that result for anything). I'm not sure how you could insert a recursive find into that script either.

Code:

find . -type f -name temp.txt -exec cat {} \;

will display all the files named 'temp.txt' in and beneath the current directory.

If I was correct in assuming that you were trying to count all the lines in the files you find, to get that, you could use:

Code:

wc -l $( find . -type f -name temp.txt ) | grep total | awk '{print $1}'

Instead of using 'find' twice, use it once to capture the list of files in an array:

Code:

declare -a FILES
FILES=( $( find . -type f -name text.txt ) )

While it's not strictly necessary to declare the array, it's not a bad idea to document what that variable is used for (when you come back to make changes to your script next year).

Once you have an array of filepaths, you can use it in other commands:

Code:

for F in ${FILES[@]}
do
    cat ${F}
done
echo "$( wc -l ${FILES[@]} | grep total ) lines."

Putting it all together (and simplifying it a bit) and you'd get:

Code:

declare -a FILES
FILES=( $( find . -type f -name text.txt ) )
cat ${FILES[@]}
echo "$( wc -l ${FILES[@]} | grep total ) lines."

Suggested leisure reading: Advanced Bash-Scripting Guide -- An in-depth exploration of the art of shell scripting

Have a lot of fun.

Thank you so much.

1s440 · 06-11-2022, 10:17 AM

Quote:

Originally Posted by rnturn

Your script appears to read a single file line by line, counting each line (but never using that result for anything). I'm not sure how you could insert a recursive find into that script either.

Code:

find . -type f -name temp.txt -exec cat {} \;

will display all the files named 'temp.txt' in and beneath the current directory.

If I was correct in assuming that you were trying to count all the lines in the files you find, to get that, you could use:

Code:

wc -l $( find . -type f -name temp.txt ) | grep total | awk '{print $1}'

Instead of using 'find' twice, use it once to capture the list of files in an array:

Code:

declare -a FILES
FILES=( $( find . -type f -name text.txt ) )

While it's not strictly necessary to declare the array, it's not a bad idea to document what that variable is used for (when you come back to make changes to your script next year).

Once you have an array of filepaths, you can use it in other commands:

Code:

for F in ${FILES[@]}
do
    cat ${F}
done
echo "$( wc -l ${FILES[@]} | grep total ) lines."

Putting it all together (and simplifying it a bit) and you'd get:

Code:

declare -a FILES
FILES=( $( find . -type f -name text.txt ) )
cat ${FILES[@]}
echo "$( wc -l ${FILES[@]} | grep total ) lines."

Suggested leisure reading: Advanced Bash-Scripting Guide -- An in-depth exploration of the art of shell scripting

Have a lot of fun.

Quote:

Originally Posted by rnturn

Hmm... that information might have been nice to know up front.

You could accomplish what I think you want to do using a utility like Ansible (or something similar) assuming you have an inventory of "all the servers". But it doesn't make much sense to propose solutions when the problem hasn't been fully presented.

At this point, however, I think the discussion would benefit greatly from a clear description of the problem you are trying to solve. What other information can you provide? It appears that locating text files may only be a small part of what you are attempting to do.

Yes I also thought about using ansible to fetch for the text files but I am new to environment and find little complicated to understand it.
First login : user@server:
Second login: ssh server2
Third login: they use alias to connect to the server.. ( here the temp.txt files ) are stored.

Code:

user@server:~$ su - s1
ssh server
[s1@ts1/server ~]$ switchuser
switchuser@server ----------->

this is how we connect and from swichuser@server.. there are temp.txt files under /app/tmp/

in the sameway: we have many users say s2, s3, s4...

for ex: to find second temp.txt on s2, but this is manual..

Code:

user@server:~$ su - s2
ssh server
[s2@ts1/server ~]$ switchuser
switchuser@server ----------->

So I am totally confused.. ��

NevemTeve · 06-11-2022, 12:46 PM

Code:

for i in user1@host1 user2@host2; do
    echo "=== $i ==="
    ssh "$i" 'find / -type f -name temp.txt' </dev/null
done

1s440 · 06-11-2022, 04:24 PM

Quote:

Originally Posted by NevemTeve

Code:

for i in user1@host1 user2@host2; do
    echo "=== $i ==="
    ssh "$i" 'find / -type f -name temp.txt' </dev/null
done

Thank you.. but this would do for multiple users isn’t it ?