Re^2: Break perl foreach loop

The STDERR did not provide any output. Only see something if I 'print' to document. See example output above where it matches after total. It does print based on the match but if I put Last WID before print, it does not print. Only prints if followed by Last WID.

My sense is that since I don't completely understand the top of the code (I did not write it), from what I see it's putting everything into $x and then performing the matching and print operations after. So ending the foreach comes too late. I think I'd have to delete everything at the end of the array after I first match total.

Maybe?

Comment on Re^2: Break perl foreach loop

Replies are listed 'Best First'.
Re^3: Break perl foreach loop by ELISHEVA (Prior) on Mar 03, 2009 at 19:50 UTC
Sorry, I was adding thoughts to my post while you were responding. Yes, I think the strategy is to start by commenting out stuff until you no longer get the symptoms and then adding it back. I think your hunch about something not quite matching up may be right. I don't think it is the problem here but I remember once driving myself crazy over a bug that was caused by a single missing `;` - for some reason the code compiled without syntax errors but line 2 was executing before line 1 and I couldn't figure out why... until I realized that Perl thought line 2 was a parameter to line 1 and so was evaluating it first. Sometimes Perl makes sense of things we rather it wouldn't. Best, beth	[reply] [d/l]
Re^3: Break perl foreach loop by Marshall (Canon) on Mar 03, 2009 at 19:56 UTC
You could try putting a $\|=1; statement at top of program. This turns force flushing on. Each print will get flushed to output as it happens. Maybe the loop really is finishing, but some stuff is stuck in print buffer that you don't see until later? I can't see anything wrong with syntax. Of course you don't need the WID label (just like not need for the next's), but you probably already know that. Update:I like Beth's suggestion more. Something is strange here that Perl is doing its best with, but not what was intended.	[reply]
Re^4: Break perl foreach loop by chromatic (Archbishop) on Mar 03, 2009 at 20:09 UTC
You could try putting a `$\|=1;` statement at top of program. This turns force flushing on. Each print will get flushed to output as it happens. That won't affect `STDERR` at all, for two reasons. First, unless `STDERR` is the currently-selected filehandle, disabling buffering won't affect `STDERR` at all. Second, `STDERR` is likely already unbuffered if it's attached to a tty.	[reply] [d/l] [select]
Re^5: Break perl foreach loop by Marshall (Canon) on Mar 04, 2009 at 17:08 UTC
Completely correct! STDOUT is by default buffered. There are a lot of ways that this can "fool you". For example I have one .pl program that runs for many hours, but is not very verbose. $>long_prog.pl >somefile. You watch the size of somefile and it's not getting bigger! But yes, it does get "bigger" it just takes awhile for that to show up in the file size results. It can be that the buffer is like 4KB - Mileage varies. If you have a debug or error statement that prints to STDERR, that can fool you when watching the terminal...the time sequence may be off ...a "normal print" to STDOUT may show up on the screen after a print to STDERR even if the "normal print" happened first. I think we are in agreement here. I don't see any issue.	[reply]