Re: ^\s not equal \S?

[Abigail-II]

$ perl -Dr -ce '/[\S*\W*]/'
Compiling REx `[\S*\W*]'
size 13 Got 108 bytes for offset annotations.
first at 1
   1: ANYOF[\0-\377!utf8::IsSpacePerl !utf8::IsWord](13)
  13: END(0)
stclass `ANYOF[\0-\377!utf8::IsSpacePerl !utf8::IsWord]' minlen 1 
Offsets: [13]
        1[8] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 0[0] 9[
+0] 
Omitting $` $& $' support.

EXECUTING...

-e syntax OK
Freeing REx: `"[\\S*\\W*]"'

Now, pay attention to the 'ANYOF' part. It includes all the ASCII and LATIN-1 characters (it also includes lots of Unicode characters, but that's not important right now).

Abigail

[davido]

Now that's a neat trick. After getting up off the floor following fainting at the very sight of it, I'll have to read perlrun again to understand how you did it. It's high time I dig into perldebug too, I see.

Thanks for the motivation / lesson. :)

[Abigail-II]

I find '-Dr' far more useful than 'YAPE::Regex::Explain'. The latter just parrots back what it was given, but then in English. '-Dr' shows how perl compiles it. As shown in the thread 'YAPE::Regex::Explain' doesn't notice the overlap between \S and \W, not even that '*' is mentioned twice. '-Dr' shows what's really going on, although the output is sometimes hard to grog.

And '-Dr' really shines at runtime, showing how Perl actually matches a regexp.

Abigail