comm(1)

comm – select or reject lines common to two files

comm takes two files as input. It prints three columns: Lines only in file 1; Lines only in file 2; lines in both files. This is effectively set logic (picture a Venn diagram). Here’s some examples:

$ cat > colors1.txt <<-EOF
blue
green
red
yellow
EOF
$ cat > colors2.txt <<-EOF
green
orange
yellow
EOF
# Print items only in colors1.txt
$ comm -23 colors1.txt colors2.txt
blue
red
# Print items only in colors2.txt
$ comm -13 colors1.txt colors2.txt
orange
# Print items in both files
$ comm -12 colors1.txt colors2.txt
green
yellow
# Print all colums at once
$ comm colors1.txt colors2.txt
blue
    green
  orange
red
    yellow

The last example has the strangest output due to how comm prints multiple columns. The -1, -2, and -3 options suppress the relevant columns for easier consumption.

comm requires that files are sorted and do not contain duplicate items. Thus, this style invocation is common: comm -23 <(sort -u file1.txt) <(sort -u file2.txt). The <() writes the output of the command (sort -u, -u means unique items only) to temporary file and returns the temporary file. This is a great workaround for comm because input files may not meet requirements.

The -i uses case insensitive comparison.

Recap

This guide is shorter than the others because there’s not too much to show. Here are some use cases to help crystalize it it:

• Set subtraction and intersection (and’ing)
• Given two files containing list of files; use comm to determine which files to keep or delete

Options are easy as well:

• -1 Suppress the first column (items only in file 1)
• -2 suppress the second column (items only in file 2)
• -3 suppress the third column (items on both files)
• -i case insensitive comparison