SubSearch - "Line starts with"

Saved has asked for the wisdom of the Perl Monks concerning the following question:

I Search for "Blank_Line_Separated" Records with the following script. I would now like to Sub-Search the Matched Records for records containing lines which start with "<WhatEver>" I have been trying foreach loops but not getting the results I expected.

#!/usr/bin/perl
#<"Search.pl"> Split File
$LOGIN=$ENV{"LOGNAME"};
system("mkdir /tmp/${LOGIN} 2>/dev/null");
$STATUS="/tmp/${LOGIN}/STATUS";
undef $/;

if ( $#ARGV == 1 ) {
  $OPTION=$ARGV[0];
  $SEARCH=$ARGV[1];
  $spid = open(SFILE, ">>$STATUS") or die "STATUS File Not Found: $!\n
+";
  @finfo=split(/\n\n/, <STDIN>);
  $tRecsInfo = @finfo;
}
elsif ( $#ARGV == 2 ) {
  $OPTION=$ARGV[0];
  $SEARCH=$ARGV[1];
  $FLNAME=$ARGV[2];
  $fpid = open(XFILE, $FLNAME) or die "Search File Not Found: $!\n";
  $spid = open(SFILE, ">$STATUS") or die "Status File Not Found: $!\n"
+;
  @finfo=split(/\n\n/, <XFILE>);
  $tRecsInfo = @finfo;
}
else
{
print ("Usage: $0 <with|not> <SearchString> [<File2Search>STDIN if mis
+sing]\n");
exit;
}

@MATCHING = ();
if ( $OPTION eq "with" )
{
    @MATCHING = grep /$SEARCH/i, @finfo;
}
else
{
    @MATCHING = grep !/$SEARCH/i, @finfo;
}
$tMATCHING = @MATCHING;
print SFILE "\n$OPTION, $SEARCH, $FLNAME";
print SFILE "\nI read ",  $tRecsInfo, " RECORDS,";
print SFILE " I Selected ",  $tMATCHING, " RECORDS\n";
print join "\n\n", @MATCHING;
print "\n";
close(SFILE);
$spid = open(SFILE, "<$STATUS" ) or die "The Status File Not Found: $!
+\n";
print STDERR <SFILE>;
close(SFILE);
[download]

All help appreciated muchly

Comment on SubSearch - "Line starts with" Download Code

Replies are listed 'Best First'.
Re: SubSearch - "Line starts with" by roboticus (Chancellor) on Mar 31, 2010 at 14:26 UTC
Saved: You really ought to make your question a bit more complete and concise (see How To Ask Questions The Smart Way and I know what I mean. Why don't you?). And you really should `use strict;` and `use warnings;` in your code (though I don't know if they'll tell you anything useful in this case ... it's just a good habit to get into). Anyway, my suggestion is to set the record delimiter to a blank line so each read will get one record. Then you can structure your code a bit more like: `my ($cnt_rex, $cnt_match) = (0, 0); while (<XFILE>) { ++$cnt_rex; next unless /$SEARCH/i; ++$cnt_match; #.... do your processing here .... }` [download] ...roboticus	[reply] [d/l] [select]
Re: SubSearch - "Line starts with" by jethro (Monsignor) on Mar 31, 2010 at 15:11 UTC
if the string $SEARCH has any regex special characters in it (like . or * or ()) they will match something unexpected (expect if that is what you want). You can use the function quotemeta() to escape any potential special characters in a string: `$SEARCH= quotemeta($SEARCH);` [download] To anchor a regex to the start of a string you have to start it with ^: `@MATCHING = grep /^$SEARCH/i, @finfo;` [download] But to anchor a regex to the start of a line(!) in a string you have to add a m flag to the regex so that perl knows it is a multiline string (and interprets ^ differently): `@MATCHING = grep /^$SEARCH/im, @finfo;` [download] PS: Dividing a file on empty lines is not what I would call "searching". That did confuse me a bit when I tried to make sense of your question	[reply] [d/l] [select]