如何grep在Perl中捕获文件的多行模式

Question

I have a file that looks something like this:

Random words go here
/attribute1
/attribute2
/attribute3="all*the*things*I'm*interested*in*are*inside*here**
and*it*goes*into*the*next*line.*blah*blah*blah*foo*foo*foo*foo*
bar*bar*bar*bar*random*words*go*here*until*the*end*of*the*sente
nce.*I*think*we*have*enough*words"

I want to grep the file for the line \\attribute3= then I want to save the string found inside the quotation marks to a separate variable.

Here's what I have so far:

#!/bin/perl
use warnings; use strict;
my $file = "data.txt";
open(my $fh, '<', $file) or die $!;
while (my $line = <$fh>) {
    if ($line =~ /\/attribute3=/g){
        print $line . "\n";
    }
}

That's printing out /attribute3="all*the*things*I'm*interested*in*are*inside*here** but

I want all*the*things*I'm*interested*in*are*inside*here**and*it*goes*into*the*next*line.*blah*blah*blah*foo*foo*foo*foo*bar*bar*bar*bar*random*words*go*here*until*the*end*of*the*sentence.*I*think*we*have*enough*words .

So what I did next is:

#!/bin/perl
use warnings; use strict;
my $file = "data.txt";
open(my $fh, '<', $file) or die $!;
my $part_I_want;
while (my $line = <$fh>) {
    if ($line =~ /\/attribute3=/g){
        $line =~ /^/\attribute3=\"(.*?)/;   # capture everything after the quotation mark
        $part_I_want .= $1;   # the capture group; save the stuff on line 1
        # keep adding to the string until we reach the closing quotation marks
        next (unless $line =~ /\"/){
             $part_I_want .= $_;    
        }
    }
}

The code above doesn't work. How do I grep capture a multiline pattern between two characters (in this case it's quotation marks)?

Answer 1

my $str = do { local($/); <DATA> };
$str =~ /attribute3="([^"]*)"/;
$str = $1;
$str =~ s/\n/ /g;

__DATA__
Random words go here
/attribute1
/attribute2
/attribute3="all*the*things*I'm*interested*in*are*inside*here**
and*it*goes*into*the*next*line.*blah*blah*blah*foo*foo*foo*foo*
bar*bar*bar*bar*random*words*go*here*until*the*end*of*the*sente
nce.*I*think*we*have*enough*words"

Answer 2

将整个文件读入一个变量，然后使用/attribute3=\\"([^\\"]*)\\"/ms

Answer 3

From the command line:

perl -n0e '/\/attribute3="(.*)"/s && print $1' foo.txt 

This is basically what you had, but the 0 flag is the equivalent of undef $/ within the code. From the man page:

-0[octal/hexadecimal]

specifies the input record separator ($/) as an octal or hexadecimal number. If there are no digits, the null character is the separator.