simple file and array operation problem

[panuvin]

Hello,

I'm having a lot of trouble getting a program working. I'm trying to open a file, seperate it into an array of words.
I've opened the file and split it into an array of
words. Then I need to take out the punctuation and test to see if any of the words begin with '[' and end in ']'. I'm not sure how to do this with regular expressions. This is what I have so far:

########Code Begin
open($fh,$param) or die "Cannot open file";
$file=<$fh>;
chomp $file;
@words=split(/ /,$file);
$size=@words;
########Code End

I'm then trying to cycle through the array and pull
the punctuation or test for a beginning/ending
character of '[' and ']'

It's not working...and I can't figure out why

########Code Begin
for($ct=0;$ct<$size;$ct++)
{
$theword=@words[$ct];
$_ = $theword;
if (/.,;!?$/)
{
print "test";
}
#######Code End

What am I doing wrong?
Using the regular expression on the $_ variable, it
should notice these punctuation marks at the end of
the word. Also, I'm wondering why I can't access the
first element of a scalar.....like the first
letter....or for this program, the first and last
being brackets.

Please help,

Very Confused,

Justin

 

[KevinADC]

your question is not very clear, it might help to post some sample lines of text you are working with. If your files are one liners or you only want to read the first line of a file this will work OK:

$file=<$fh>;

because that only reads the first line of the file.

your loop is poorly constructed:

for($ct=0;$ct<$size;$ct++)
{
  $theword=@words[$ct];
  $_ = $theword;
  if (/.,;!?$/)
  {
  print "test";
  }
}

should maybe be more like this:

foreach(@words) {
  if (/[.,;!?]$/) {#use a character class [.,;!?]
      print "Has punctuation: .,;!? at the end\n";
  }
  if (/^\[/ || /\]$/) { 
      print "Has a [] at the beginning or end\n";  
}

 

[KevinADC]

correcting a typo in the last code block above:

foreach(@words) {
  if (/[.,;!?]$/) {#use a character class [.,;!?]
      print "Has punctuation: .,;!? at the end\n";
  }
  if (/^\[/ || /\]$/) {
      print "Has a [] at the beginning or end\n";  
  }# <-- ending bracket was ommited
}

 

[mikevh]

I don't know if this helps you, but Perl has a built-in character class [unct:] that matches anything except letters, digits, control characters, or space characters. s/[[unct:]]//g will remove all "punctuation" chars from a string. Note, however, that this will remove [ and ] also, so if it's crucial to keep these, maybe this is not the way to go. Just thought I'd mention it in case ...

 

[KevinADC]

Good point, I tend to forget POSIX supported syntax. But using tr/// should still be faster and a better habit to get into for just trading/removing characters when no further pattern matching is required.