regex search

[zuma53]

This is probably a really silly question, but I am stumped! Let's say I have these strings:

$A:  abc.def.ghi
$B:  abc

To ask "does $A start with $B" is trivial:

$A =~ /^${B}/;

But what is the complement of this? That is, "is $B a prefix of $A?"

$B =~ /?$A?/;

Here, I am taking the fact that $B is shorter than $A, but this logic fails when $B is longer than $A. I can probably do some clunky len/substring preprocessing, but I am certain there's a more elegant regex way (isn't there always?)

Thanks in advance.

J

[AnomalousMonk]

I, too, am confused by the whole question of the 'complimentary' (or complementary, or converse) question. Maybe the question can be usefully (and symmetrically) reframed as "Are there any characters in common at the beginnings of both strings?". This could be answered by index as well as by a regex.

>perl -wMstrict -le
"my @set = ('abc.def.ghi', 'abc', '...');
 ;;
 for my $p (@set) {
   for my $q (@set) {
     print qq{'$p', '$q' common at start? },
           $q =~ m{ \A \Q$p\E }xms || $p =~ m{ \A \Q$q\E }xms
           ? 'yes' : 'no';
     }
   }
 ;;
 print '';
 for my $p (@set) {
   for my $q (@set) {
     print qq{'$p', '$q' common at start? },
           0 == index($p, $q) || 0 == index($q, $p) ? 'yes' : 'no';
     }
   }
"
'abc.def.ghi', 'abc.def.ghi' common at start? yes
'abc.def.ghi', 'abc' common at start? yes
'abc.def.ghi', '...' common at start? no
'abc', 'abc.def.ghi' common at start? yes
'abc', 'abc' common at start? yes
'abc', '...' common at start? no
'...', 'abc.def.ghi' common at start? no
'...', 'abc' common at start? no
'...', '...' common at start? yes

'abc.def.ghi', 'abc.def.ghi' common at start? yes
'abc.def.ghi', 'abc' common at start? yes
'abc.def.ghi', '...' common at start? no
'abc', 'abc.def.ghi' common at start? yes
'abc', 'abc' common at start? yes
'abc', '...' common at start? no
'...', 'abc.def.ghi' common at start? no
'...', 'abc' common at start? no
'...', '...' common at start? yes

[aaron_baugher]

To ask "does $A start with $B" is trivial:

$A =~ /^${B}/;

But what is the complement of this? That is, "is $B a prefix of $A?"

In my understanding of English, those don't complement each other; they mean the same thing. A "prefix" is a group of letters at the beginning of a word, so if $B is a prefix of $A, then $A starts with $B. Your line of code above tests for that situation, whichever you call it.

[cheekuperl]

With "compliment" you probably mean searching for $A at the beginning of $B- which essentially fails since $A is longer than $B!

[zuma53]

The compliment (that I am looking for) would be that $B is a prefix of $A.

J

[muba]

But your first regex is checking that.

[muba]

Given that AnomalousMonk's response above doesn't answer your question, I think I get you now. What you're trying to achieve, actually, is this: you have a situation, in which you have these set of strings:

my $superstring = "123abc.def.ghi";
my $substring = "abc.def.ghi";
my $prefix = "123";

And what you want to do is to see whether $substring appears in $superstring, but only if it's preceded by $prefix?

Like this?

#### (pfff where is my head at. This won't work of course) $superstrin
+g =~ m/$B$A/
$superstring =~ m/$prefix$superstring/;

If that's not it, then you'll have to explain what you mean with 'prefix' in this context.

Edit:Caught a stupid brainfail in the regex code. Fixed.