in reply to Why aren't perl regular expressions really regular expressions?
They've answered the what, here's the why:
The 'regular languages' that could be defined in terms of those original 'regular expressions' were so limited as to be practically useless beyond the kind of pointless exams questions so beloved of academia.
They had to made irregular, to become useful. And despite that the purists hate the irregularity, they've not come up with anything better.
A bit like the Perl language itself; much of its power derives directly from its irregularity.
Rant:
The same irregularities, that if the 'new breed' of Perl programmers and architects get their way, will disappear under cloak of mundane uniformity that is PBP, PerlCritic, Mo((o|u)(se)), Modern Perl et. al.
The problem with -- or rather for -- Perl; is that it allows you to write (Java|Javascript|C|Clos|Lisp|...) in Perl.
As such, there is a (new) breed of "Perl programmer" whom have never bothered to learn 'Perl'.
And C in Perl is slower; Java in Perl, (even) more verbose; JS in Perl slower and more memory hungry; -- so they migrate back to X; which they use anyway when writing Perl.
Perl -- the real Perl; unadulterated by "post-modernisms", nor "old-foggyisms" -- is hard to master; but repays you time over with concise; fast; productive solutions to many (most) of your daily programming needs.
But every time you reject a perlism in favour of an XXX language dogmatism; you diminish Perl & its pragmatic approach to making the difficult: easy; and the impossible: possible.
And you diminish your own power and productivity in the process.
Every time you bow to the tacky, tack-on (f)utility of the JUnit wannabe: Test::*; you diminish your power; along with that of Perl.
Every time you fall for the "post-modern" introversion of the Clos wannabe: Mo((o|u)(se)), you adopt the 19th Century practice of driving your code behind a man walking with a red-flag.
|
|---|