Shell script to copy the files - shell

Question

I worked very little with scripts and I don't know..
I need to create a script (in Ubuntu) that copies only files where a certain user modified more than 20 lines at a given time.
I know that to copy a file elsewhere I use this code:
$ ls dir1/
dir2/
$ cp -r dir1/ dir1.copy
$ ls dir1.copy
dir2/
And to count lines: wc -l file1
But how could I check if a user has modified more than 20 lines in a file (eg a simple txt, for example today)?
Thank you in advance !

Answer 1

In the first place, if by "modifying lines in a file" you mean "adding lines to a file", then you can do something about it. If you are literally talking about modifying lines in files, there is nothing you can do to track that activity without setting up some version control first.
So, assuming we are talking about files in which your users will be adding lines, a workaround for that may consist of setting up some scheduled tasks to check the line numbers of those files "at a given time" (as you said) and compare that value to a previous result, and then if there are more than 20 additional lines than from the last value, copy the files elsewhere.
First things first, counting the lines of your files is something you have already mentioned and it is right: I will propose using wc -l too.
Once here you will need two things: one place (tipically a file) to periodically save the number of lines of your files at a given time and one trigger that would start copying the files in case there have been more than 20 lines added.
So for example, in this case you can set up a cron job, like this (i.e. to run every hour):
0 */1 * * * cat ${FILE} |wc -l > /tmp/${FILE}_counter
That one will check the number of lines of a given file and send the output to a temporary file that we will be using soon. In case you have multiple files you can easily script that and make a loop, like this:
#!/bin/bash
for FILE in file1 file2 file3; do
cat ${FILE} |wc -l > /tmp/${FILE}_counter
done
Don't forget to add the path to the script in the cron job if you do that this way. After that, you will have something like this in your /tmp directory:
/tmp/file1_counter
/tmp/file2_counter
/tmp/file3_counter
...
At this point you only need a trigger, which can be another script, to compare the current number of lines of a file at a given time and start copying it elsewhere in case there are more than 20 additional lines than in the previous check. Consider this:
#!/bin/bash
LAST_VALUE=$(cat /tmp/${FILE}_counter)
CURRENT_VALUE=$(cat ${FILE} |wc -l)
if [ ${CURRENT_VALUE} -gt $(expr ${LAST_VALUE} + 20) ]; then
# Your cp stuff here
fi
Of course you can add a loop here too in case of handling multiple files:
#!/bin/bash
LAST_VALUE=$(cat /tmp/${FILE}_counter)
CURRENT_VALUE=$(cat ${FILE} |wc -l)
for FILE in file1 file2 file3; do
if [ ${CURRENT_VALUE} -gt $(expr ${LAST_VALUE} + 20) ]; then
# Your cp stuff here
fi
done
Then you only have to add this last script to a cron job too, and you should be done.
Hope you find this useful.

Answer 2

You can use diff to compare 2 files. With the -u0 option, it will show you the added/deleted/modified lines, prefixed with "+" or "-". You can then count lines starting with "+" or "-" with grep and it's -c option.
So for the number of lines added or modified, which begin with "+" :
diff -u0 $file_before $file_after | grep -c '^+'
and this will count the deleted or modified lines, which start with "-" :
diff -u0 $file_before $file_after | grep -c '^-'
Note that there are 2 header lines in this format, which also start with "+" and "-", so you may want to take that in account.