Re: Capturing regex from map

[thezip]

Try:

my @online_services = map { /^online.*svc:(.*?):/; $1 } `svcs`;

[LanX]

> try my @online_services = map { /^online.*svc:(.*?):/; $1 } `svcs`;

not equivalent:

  DB<217> map { /(1\d*)/;$1 } 5..15
 => (undef, undef, undef, undef, undef, 10, 11, 12, 13, 14, 15)

  DB<218> map { /(1\d*)/ } 5..15
 => (10, 11, 12, 13, 14, 15)

update

I think you meant:

  DB<225> map { @matches=/(1\d*)/;@matches} 5..15
 => (10, 11, 12, 13, 14, 15)

[thezip]

That's what I get for not testing! </sheepish grin>

[Anonymous Monk]

I'd say map { /^online.*svc:(.*?):/; $1 || () }

just to be clear on what the expected transformation is. I feel that a regexp-match inside a map is fairly abstruse.

[Anonymous Monk]

Some other variants:

map { /^online.*svc:(.*?):/; $1 // () }
map { /^online.*svc:(.*?):/ and $1 or () }
map { /^online.*svc:(.*?):/ ? $1 : () }

The latter two are easy to extend if there are multiple captures. (The middle one is my favourite.)

[brianski]

This is definitely clearer, I'd recommend it.

[BrowserUk]

I think you're wrong about clearer.

But bedsides that, it does not do (and cannot me made to do), the same thing:

@a = map{ m[(..)(.)]g } 'the quick brown fox', 'jumps over the lazy do
+g';;
print @a;;
th e  q u ic k  b r ow n  f o ju m ps   ov e r  t he   la z y  d

@a = map{ m[(..)(.)]g; $1 || () } 'the quick brown fox', 'jumps over t
+he lazy dog';;
print @a;;
th ju

Perl is context sensitive. If you do not understand what that is, or try ignore it, you will never make best use of Perl.

On the other hand, if you take a few hours or days to understand context sensitivity, the OPs code is clear, concise, and very readable.

And very powerful. It requires many extra, contorted (and ultimately redundant) lines of code to achieve the same thing without using the effects of list context. It is a tool, that once you've realised its powerful simplicity, you will not want to do without.