CORRECTION Re: [ILUG] comparing files

[Brian Foster]

  | Date: Mon, 02 Feb 2004 20:12:10 +0100
  | From: Brian Foster <blf at utvinternet.ie>
  |[ ... ]
  |  be aware, however, that both uniq(1) and `comm'
  |  require the input to be sorted.
  |  ( I must confess I have never understood why the
  |   input must be sorted.  `uniq' could still deal
  |   with _adjacent_ duplicate lines (e.g., N and N+1),
  |   and `comm' could compare line N with line N.
  |   why the insistence on sorting? )   [ ... ]

 sorry for replying to my own post.
 the above is (almost) all gibberish.
 uniq(1) does not require its input to be sorted,
 unless you want to find/remove all duplicates.
 comm(1) requires sorted input because making
 a tri-state decision (file1-only, file2-only,
 or both-files) on the basis of a line-by-line
 (bi-valued) comparison is computationally infeasible.

 now I just hope the work-related e-mail I sent
 late last night is more coherent than the above
 nonsense!  ;-\ 
cheers!
	-blf-

p.s. my clear(er?) memory now in the morning is it was
    older, any maybe only non-GNU, versions of join(1)
    which inexplicitly insisted on sorted input.
    however, I note the GNU join man(1) page does not
    mention sorting, so perhaps that nonsense has been
    removed.  or I am just imagining it?

--
«How many surrealists does it take to    |  Brian Foster      Montpellier,
 change a lightbulb?  Three.  One calms  |  blf at utvinternet.ie      France
 the warthog, and two fill the bathtub   |    Stop E$$o (ExxonMobile)!
 with brightly-colored machine tools.»   |        http://www.stopesso.com