Shell Programming and Scripting

I need to know if file1 is a subset of file2 i.e all the contents of file1 are present in file2 or not.

Here is how i would do it.

Read line by line file1 and grep every line in file2 in a for loop. any failing grep would means that it is not a subset.

Is there a quicker or easier way to check that ?

Code:

uname -a
SunOS mymac 5.11 11.2 sun4u sparc SUNW,SPARC-Enterprise

You can also use the -vf flags of grep something like this:-

Code:

grep -vf file2 file1

This will display lines from file1 that are not in file2. No output infers that file1 lines are all found in file2, however there is no consideration on the order of the lines, if there are duplicate lines etc.


Would this do what you need?

I haven't got a SunOS server available to test on, so you might need to adjust your code to perhaps use fgrep and drop the -f flag.



Robin

Works but how can i ignore newlines / whitespace lines ? Becoz the files shows as different just becoz if has a few new / blank lines.

rbatte's approach is a good one.

Consider using tr or grep -v to remove blank line in both files.
By blank I assume '/n' only on a "blank line"

Code:

grep -v '^$' file1> tmpfile1

Use tmpfile1 instead of file1

But.
If you have files with thousands of lines, the run time for the approach will be large. And if you have to do this for a large number of files you could think of this project as a full-time hobby for the next few weeks.

Check back after you try it. If it works well for your dataset, wonderful. If not, well, we may be able to help you set up some kind of parallelism to reduce run-times by a large factor.