[Tutor] non-greedy matching

[A.T.Hofkamp]

spir wrote:
> Hello,
> 
> imagine you need to match such a pattern:
> 
> pat : (@@ [charset]* @@) | [charset]*
> ... where [charset] has to include '@'
> 
> My questions are:
> * Is there any other way than using a non-greedy form of [charset]* ?

Something like this?
(in pseudo-RE syntax)

"(@@" ( [^@]* "@" [^@] | [^@]* "@" "@"+ [^@)] )* [^@]* "@" "@"+ ")"

to understand, break the above down in pieces:

"(@@" ( A | B )* C

with A: [^@]* "@" [^@]
         # lots of chars, then a @, and a non-@

      B: [^@]* "@" "@"+ [^@)]
         # lots of chars then at least two @, and a non-closing bracket
         # (the non-@ at the end is for forcing all @ to be matched in "@"+)

      C: [^@]* "@" "@"+ ")"
         # lots of chars, then at least two @, and finally a closing bracket


> * How, actually, is non-greedy character string matching performed?

That's what I'd like to know too.
(and while we are at it, what about the \b )?


Sincerely,
Albert