Re^2: Refactoring challenge.

This seems like a rather problematic approach, in both before and after variants:

  my ($pos, $first ) = ( pos $string, $1 ) if $string =~ m/([\[\{\]\},
+])/g;
  return unless defined $pos;
[download]

which I'd rather write as:

  $string =~ m/([\Q[]{},\E])/ or return;
  my($pos, $first) = ($+[0], $1);
[download]

With the code as is I'm not sure what you expect $pos and $first to be when the match fails, but if my reformulation is appropriate, it becomes clear that neither will ever be set to undef - that should either allow the remaining code to be simplified some, or point to a problem.

Hugo

Comment on Re^2: Refactoring challenge. Select or Download Code

Replies are listed 'Best First'.
Re^3: Refactoring challenge. by BrowserUk (Patriarch) on Mar 07, 2005 at 16:43 UTC
That is a much better way to write it, except that even using `$+[0]` rather than pos, the /g is still required? I do not pretend to understand that, but it is so. Which also explains why it broke when I tried to use `$-[0]` originally. I would never have thought that @- and @+ where dependant upon /g? Is that a bug? I mentioned I was embarassed by the code. My only excuse is that it evolved into the state where it worked and I've been concerned with more fundemental restructuring. In-situ, with the /g, it makes no difference to the functioning of the code--which means your version is now in the real code. I'm not sure what you expect $pos and $first to be when the match fails... Nothing! If the match fails, there is nothing (or nothing further) that can be done in the call and the code just returns to the caller. If the match fails, I do not have any break point upon which to decide to output anything, so all I can do is wait until more of the string accumulates. Ie. the next call, or the one after that. Examine what is said, not who speaks. Silence betokens consent. Love the truth but pardon error.	[reply] [d/l] [select]
Re^4: Refactoring challenge. by Roy Johnson (Monsignor) on Mar 07, 2005 at 20:33 UTC
I would never have thought that @- and @+ where dependant upon /g? Is that a bug? You need the /g because you are doing iterative matching in a while loop. Caution: Contents may have been coded under pressure.	[reply]