Re: Re: Extract lines and output...

Hi Fletch,

I liked your approach. I'm not sure you need the sub in it
though; you're already doing variable interpolation when you
open( OUT, ">$outfile" ) or die ... so you could lean it up a bit:
open( OUT, ">${_}_distilled" ) or die ...

Here's what your code inspired me to do

#!/usr/bin/perl

for(glob '*')
{
  /^$0$/ and next;  # don't process the script
  open(I,"<$_") or die("$_: $!");
  open(O,">$_.$$") or die("$_: $!");

  print O $_,"\n";
  { $. > 4 ? last : print O while(<I>); }

  close(I);
}

exit;
[download]

While writing and subsequently edting above program I discov-
ered the redundancy of local() in certain contexts. I originally wrote
{ local $_ = undef; $. > 4 ? last : print O while(<I>) }
RRedundancy bad, so local went away.

Update: changed >$outfile_distilled
to ${_}_distilled which actually makes sense
in the context of the original program.

blyman

Comment on Re: Re: Extract lines and output... Select or Download Code

Replies are listed 'Best First'.
Re: Re: Re: Extract lines and output... by tachyon (Chancellor) on Oct 10, 2001 at 08:17 UTC
If you like brief you might like this: `#!/usr/bin/perl -w while(<*>) { next if /^$0$/; open I, $_ or die "$_: $!"; open O,">$_.bak" or die "$_.bak: $!"; print O $_, "\n", (<I>)[0..3]; close I, close O; }` [download] cheers tachyon s&&rsenoyhcatreve&&&s&n.+t&"$'$`$\"$\&"&ee&&y&srve&&d&&print	[reply] [d/l]
Re: Re: Re: Extract lines and output... by broquaint (Abbot) on Oct 10, 2001 at 13:42 UTC
Shorter still - `open(LOG, ">logfile.txt"); while($FL = <*.extension>) { open(FL) and print LOG "$_\n",(<FL>)[0..3] and close(FL) } close(LOG);` [download] But we could golf this 'til the cows come home :o) HTH broquaint	[reply] [d/l]
Re: Re: Re: Re: Extract lines and output... by belden (Friar) on Oct 10, 2001 at 08:58 UTC
I like it really brief. The only reason I have `close(I)` is to reset $. You don't ever use $. so $. doesn't need to be reset - therefore the close calls aren't needed. This works just as well for me... `#!/usr/bin/perl -w while(<>) { next if /^$0$/; open I, $_ or die "$_: $!"; open O,">$_.bak" or die "$_.bak: $!"; print O $_, "\n", (<I>)[0..3]; #close I, close O; }` [download] A few things I picked up from your code: You used .bak instead of .$$, which makes your solution portable beyond nix systems. I like your `<*>` more than my `glob`. Your treatment of the filehandle as though it were an array is also new to me. blyman	[reply] [d/l] [select]
Re: Re: Re: Re: Re: Extract lines and output... by tachyon (Chancellor) on Oct 10, 2001 at 11:10 UTC
As you say you don't have to close filehandles in many cases. It is just tidy to do so. If you want to do it the more politiacally correct way you use a lexically scoped variable as FH. This is destroyed when you go out of scope and is well scoped. `while(<*>) { /^$0$/ && next; open my $IN, $_ or die "$_: $!";; open my $OUT,">$_.bak" or die "$_.bak: $!"; print $OUT $_, $/, (<$IN>)[0..3]; }` [download] cheers tachyon s&&rsenoyhcatreve&&&s&n.+t&"$'$`$\"$\&"&ee&&y&srve&&d&&print	[reply] [d/l]