You have an original directory. You copy the files in it to another directory, edit some of them, and add a few others. Later, you want to know the differences between the two directories. If your system's has a -r (recursive) option, you can use that. System V has dircmp. The output of dircmp is formatted with ; you get 66-line-long pages with headings:
dircmp a bSep 16 09:26 1991 a only and b only Page 1 ./foo.tmp ./defs.h ... Sep 16 09:26 1991 Comparison of a b Page 1 directory . same ./Makefile directory ./data same ./data/test1 same ./data/test2 ... different ./pqp.help same ./pqs.help ...
a only and b only listing, files only in the first directory
are in the first column and files only in the second directory are in the
Comparison of a b listing compares files that are in both
The comparison is recursive - if there are any subdirectories,
dircmp checks those, too.
The dircmp -s option stops the "identical file" messages. Use -d to run diff on files that are different; dircmp prints a new page for each diff it runs:
dircmp -d -s a bSep 16 09:35 1991 a only and b only Page 1 ... Sep 16 09:35 1991 Comparison of a b Page 1 ... Sep 16 09:35 1991 diff of ./pqp.help in a and b Page 1 3c3,4 < -# "Only this printer"... 'pqp -3' would print on #3. -- > -# "Only this printer"... 'pqp -3' would print only on #3; > other printer queues will be held. ...
The designers assumed you'd want to send the output to a printer. I usually read it on my screen with the pager and its -s option, which squeezes out the multiple blank lines. If you don't have less or more -s, try piping the output through or .