RE: How do I remove blank lines from text files?

Update: as davorg correctly points out, I had a strange and (mostly) unnecessary construct. This is my updated version.
As an explanation: I had the <IN> in a while loop, to take advantage of the diamond operator's auto-assign-to-$_ feature. This code just makes the assignment explicit.

Thanks, davorg!

This will write a new file with blank lines (where there are two or more newlines together) removed:

local $/ = undef;
open IN, "<junk" or die "Couldn't open junk: $!";
open OUT, ">fixed" or die "Couldn't open fixed: $!";

($_ = <IN>) =~ s/\n{2,}/\n/g;
print OUT;

close IN; close OUT;
[download]

Variations on this theme (like whitespace, as noted elsewhere in this thread) would be similar...

Russ

Comment on RE: How do I remove blank lines from text files? Download Code

Replies are listed 'Best First'.
RE: RE: How do I remove blank lines from text files? by perlmonkey (Hermit) on Jul 11, 2000 at 08:29 UTC
Just to nitpick a tiny bit, I want to warn about your modification to $/. By doing things the way to did, you are setting yourself up for crazy-ass bugs. You properly localize the variable, but the scope of the localization is way too broad. I suggest getting in the habit of anonymous blocks every time you localize a global variable. Like This: `open OUT, ">fixed" or die "Couldn't open fixed: $!"; open IN, "<junk" or die "Couldn't open junk: $!"; { local $/ = undef; ($_ = <IN>) =~ s/\n{2,}/\n/g; print OUT; } close IN; close OUT;` [download] This way you properly limit the scope of local $/ to just the lines you need it. And wont accidentilly screw yourself up later.	[reply] [d/l]
RE:(3) remove blank lines? (localizing block for $/) by Russ (Deacon) on Jul 11, 2000 at 08:37 UTC
Good point. I should have included that in my snippet. Russ Brainbench 'Most Valuable Professional' for Perl	[reply]
RE: RE: How do I remove blank lines from text files? by davorg (Chancellor) on Jul 10, 2000 at 23:32 UTC
As you've undefined `$/` and are therefore acting in 'slurp' mode, you'll read the whole file in with your first call to `<IN>`. Doesn't that make your `while` loop unecessary? -- <http://www.dave.org.uk> European Perl Conference - Sept 22/24 2000, ICA, London <http://www.yapc.org/Europe/>	[reply]