如何使用Perl正则表达式解析lshw的输出？

Question

我正在尝试用这段代码将lshw输出解析成一个哈希值，到目前为止还行得通。

  use strict;
  use warnings;

  my (%lshw,$key,$value);
  while (<>){
  s/#.*//;                # no comments
  s/^\s+//;               # no leading whites
  s/\s+$//;               # no trailing whites
  next unless length;     # anything left?
  if (/(?<key>.*?):\s+(?<value>.*)/x){
    $lshw{$+{key}} = $+{value};
  }
}

# remove white spaces in hash keys
for $key (keys %lshw){
  $value = delete $lshw{$key};
  for ($key){
    s/\s+//g;
   }
  $lshw{$key} = $value;
  }

my $logname   = $lshw{'logicalname'};
print "Logical name\t $logname\n";

但当我遇到这样的配置时，我会感到困惑：

clock: 33Mhz 
width: 32 bits 
capacity: 1Gbit/s 
configuration:autonegotiation=on broadcast=yes driver=igb driverversion=5.3.0-k duplex=full firmware=1.63, 0x800009fa ip=[REMOVED] latency=0 link=yes multicast=yes port=twisted pair speed=1Gbit/s`

我尝试了一种hoh方法，但没有找到如何拆分键/值的解决方案，因为它包含像port=twisted pair这样的多字值。关键始终是一个单词。

谁能给我一个提示如何解决这个问题？

(感谢simbabque的严格/警告提示)

Answer 1

您需要的是捕获等号后面的所有字符，这些字符后面没有模式somekeyname=。

#!/usr/bin/env perl

use strict;
use warnings;

my $s = q{configuration: autonegotiation=on broadcast=yes driver=igb driverversion=5.3.0-k duplex=full firmware=1.63, 0x800009fa ip=[REMOVED] latency=0 link=yes multicast=yes port=twisted pair speed=1Gbit/s};

my ($key, $rest) = split /:\s*/, $s, 2;

my %params = ($rest =~ / (\w+) = ((?:. (?! \w+ = ))+) /gx);

use YAML::XS;
print Dump \%params;

输出：

---
autonegotiation: on
broadcast: yes
driver: igb
driverversion: 5.3.0-k
duplex: full
firmware: 1.63, 0x800009fa
ip: '[REMOVED]'
latency: '0'
link: yes
multicast: yes
port: twisted pair
speed: 1Gbit/s

此外，还可以改进您的初始循环：

 while (<>) {
     next if /^#/; # skip comments
     /\S/ or next; # skip blank lines
     s/^\s+//;
     s/\s+\z//;
     # ...
}

Answer 2

您只需按如下方式split配置字符串

use strict;
use warnings 'all';
use feature 'say';

my $s = 'configuration: autonegotiation=on broadcast=yes driver=igb driverversion=5.3.0-k duplex=full firmware=1.63, 0x800009fa ip=[REMOVED] latency=0 link=yes multicast=yes port=twisted pair speed=1Gbit/s';

say for split /\s+(?=[^\s=]+=)/, $s;

输出

configuration:
autonegotiation=on
broadcast=yes
driver=igb
driverversion=5.3.0-k
duplex=full
firmware=1.63, 0x800009fa
ip=[REMOVED]
latency=0
link=yes
multicast=yes
port=twisted pair
speed=1Gbit/s

现在，您就有了一个键及其值的列表，该列表由键名称正确划分。这应该是易于处理的

Answer 3

Borodin的方法就是它。

如果您想用regexp解析它，这将会起作用，并将键与它的值分开。

#!/usr/bin/env perl

use warnings FATAL => 'all';
use strict;
my $s = 'configuration: autonegotiation=on broadcast=yes driver=igb driverversion=5.3.0-k duplex=full firmware=1.63, 0x800009fa ip=[REMOVED] latency=0 link=yes multicast=yes port=twisted pair speed=1Gbit/s';

while ($s =~ m/(?<key>[A-Za-z0-9]+)=(?<value>([\/\[\]A-Za-z0-9., -]+)(?= [a-z]+)|([\/\[\]A-Za-z0-9., -]+))/g) {
    print "$+{key} >>  $+{value}\n";
    $s =~ s/$+{key}//;
}

输出

autonegotiation >>  on
broadcast >>  yes
driver >>  igb
driverversion >>  5.3.0-k
duplex >>  full
firmware >>  1.63, 0x800009fa
ip >>  [REMOVED]
latency >>  0
link >>  yes
multicast >>  yes
port >>  twisted pair
speed >>  1Gbit/s

Pros

• 键/值分离

Cons

正则表达式中的

• 开销很大的正向先行

重构提案

消除正则表达式[\/\[\]A-Za-z0-9., -]

• get中的复杂字符组
• 替换循环

中找到的模式