Expressing AND, OR, and NOT in a Single Pattern (Perl Cookbook, 2nd Edition)

6.18.3. Discussion

When in a normal program you want to know whether something doesn't match, use one of:

    if (!($string =~ /pattern/)) { something( ) }   # ugly
    if (  $string !~ /pattern/)  { something( ) }   # preferred
unless (  $string =~ /pattern/)  { something( ) }   # sometimes clearer

To see whether both patterns match, use:

if ($string =~ /pat1/ && $string =~ /pat2/ ) { something( ) }

To see whether either of two patterns matches:

if ($string =~ /pat1/ || $string =~ /pat2/ ) { something( ) }

Instead of trying to do it all within a single pattern, it's often more efficient and clearer to use Perl's normal Boolean connectives to combine regular expressions. However, imagine a trivially short minigrep program that reads its single pattern as an argument, as shown in Example 6-9.

Example 6-9. minigrep

  #!/usr/bin/perl
  # minigrep - trivial grep
  $pat = shift;
  while (<>) {
      print if /$pat/o;
  }

To tell minigrep that some pattern must not match, or that it has to match both subpatterns in any order, you're at an impasse. The program isn't built to accept multiple patterns. How can you do it using one pattern? This need comes up in programs reading patterns from configuration files.

The OR case is pretty easy, since the | metacharacter provides for alternation. The AND and NOT cases, however, are more complex.

For AND, you have to distinguish between overlapping and non-overlapping needs. If, for example, you want to see whether a string matches both "bell" and "lab" and allow overlapping, the word "labelled" should be matched. But if you don't want to count overlaps, it shouldn't be matched. The overlapping case uses a lookahead assertion:

"labelled" =~ /^(?=.*bell)lab/s

Remember: in a normal program, you don't have to go through these contortions. Simply say:

$string =~ /bell/ && $string =~ /lab/

To unravel this, we'll spell it out using /x and comments. Here's the long version:

if ($murray_hill =~ m{
             ^              # start of string
            (?=             # zero-width lookahead
                .*          # any amount of intervening stuff
                bell        # the desired bell string
            )               # rewind, since we were only looking
            lab             # and the lab part
         }sx )              # /s means . can match newline
{
    print "Looks like Bell Labs might be in Murray Hill!\n";
}

We didn't use .*? to end early, because minimal matching is more expensive than maximal matching. It's more efficient to use .* over .*?, given random input where the occurrence of matches at the front or the end of the string is completely unpredictable. Of course, sometimes choosing between .* and .*? may depend on correctness rather than efficiency, but not here.

To handle the non-overlapping case, you need two parts separated by an OR. The first branch is THIS followed by THAT; the second is the other way around:

"labelled" =~ /(?:^.*bell.*lab)|(?:^.*lab.*bell)/

or in long form:

$brand = "labelled";
if ($brand =~ m{
        (?:                 # non-capturing grouper
              bell          # look for a bell
              .*?           # followed by any amount of anything
              lab           # look for a lab
          )                 # end grouper
    |                       # otherwise, try the other direction
        (?:                 # non-capturing grouper
              lab           # look for a lab
              .*?           # followed by any amount of anything
              bell          # followed by a bell
          )                 # end grouper
    }sx )                   # /s means . can match newline
{
    print "Our brand has bell and lab separate.\n";
}

Neither of those patterns matches the test data of "labelled", since there "bell" and "lab" do overlap.

These patterns aren't necessarily efficient. $murray_hill =~ /bell/ && $murray_hill =~ /lab/ scans the string at most twice, but the pattern-matching engine's only option is to try to find a "lab" for each occurrence of "bell" with (?=^.*?bell)(?=^.*?lab), leading to quadratic worst-case running times.

If you followed those examples, the NOT case should be a breeze. The general form looks like this:

$map =~ /^(?:(?!waldo).)*$/s

Spelled out in long form, this yields:

if ($map =~ m{
        ^                   # start of string
        (?:                 # clustering grouper
            (?!             # look ahead negation
                waldo       # is he ahead of us now?
            )               # is so, the negation failed
            .               # any character (cuzza /s)
        ) *                 # repeat that grouping 0 or more
        $                   # through the end of the string
    }sx )                   # /s means . can match newline
{
    print "There's no waldo here!\n";
}

How would you combine AND, OR, and NOT? It's not a pretty picture, and in a regular program, you'd almost never do this. But you have little choice when you're reading from a config file or pulling in arguments from the command line, because you specify only one pattern. You just have to combine what we've learned so far. Carefully.

Let's say you wanted to run the Unix w program and find out whether user tchrist were logged on anywhere but a terminal whose name began with ttyp; that is, tchrist must match, but ttyp must not.

Here's sample input from w:

 7:15am  up 206 days, 13:30,  4 users,  load average: 1.04, 1.07, 1.04
USER     TTY      FROM              LOGIN@  IDLE   JCPU   PCPU  WHAT
tchrist  tty1                       5:16pm 36days 24:43   0.03s  xinit
tchrist  tty2                       5:19pm  6days  0.43s  0.43s  -tcsh
tchrist  ttyp0    chthon            7:58am  3days 23.44s  0.44s  -tcsh
gnat     ttyS4    coprolith         2:01pm 13:36m  0.30s  0.30s  -tcsh

Here's how to do that using the minigrep program previously outlined or with the tcgrep program from the end of this chapter:

% w | minigrep '(?!.*ttyp)tchrist'

Decoding that pattern:

m{
    (?!                     # zero-width look-ahead assertion
        .*                  # any amount of anything (faster than .*?)
        ttyp                # the string you don't want to find
    )                       # end look-ahead negation; rewind to start
    tchrist                 # now try to find Tom
}x

Of course, this example is contrived: any sane person would call the standard grep program twice, once with a -v option to select only non-matches.

% w | grep tchrist | grep -v ttyp

The point is that Boolean conjunctions and negations can be coded up in one single pattern. You should comment this kind of thing, though, having pity on those who come after you—before they do.

One last thing: how would you embed that /s in a pattern passed to a program from the command line? The same way as you would a /i modifier: by using (?i) in the pattern. The /s and /m modifiers can be painlessly included in a pattern as well, using (?s) or (?m). These can even cluster, as in (?smi). That would make these two reasonably interchangeable:

% grep -i 'pattern' files
% minigrep '(?i)pattern' files

When you turn on a modifier that way, it remains on for the entire pattern. An alternative notation restricts the scope of the modifier. Use a clustering parenthesis set, (?:...), and place the modifiers between the question mark and the colon. Printing out a qr// quoted regex demonstrates how to do this:

% perl -le 'print qr/pattern/i'
(?i-xsm:pattern)

Modifiers placed before a minus are enabled for just that pattern; those placed after the minus are disabled for that pattern.

Example 6-9. minigrep

6.18. Expressing AND, OR, and NOT in a Single Pattern

6.18.1. Problem

6.18.2. Solution

6.18.3. Discussion

6.18.4. See Also