如何使用Perl从文件中选择行

Question

I am trying to check each line in a file for a string and print that line if it has that string.

Example of the array:

---
BIP1288I: Message flow 'flow1' on execution group 'EG1' is running.

Additional thread instances: '0'
Deployed: '11/12/13 1:54 AM' in Bar file '/MSD/deploy/ENV/EG1/flow1.bar'
Last edited: '5/24/13 4:38 PM'.
Long description: ''
User-defined property names:
Keywords:

---
BIP1288I: Message flow 'flow2' on execution group 'EG1' is running.

Additional thread instances: '0'
Deployed: '11/12/13 1:54 AM' in Bar file '/MSD/deploy/ENV/EG1/flow2.bar'
Last edited: '5/24/13 4:38 PM'.
Long description: ''
Keywords:

----
BIP1288I: Message flow 'flow3' on execution group 'EG1' is running.

Additional thread instances: '0'
Deployed: '11/12/13 1:54 AM' in Bar file '/MSD/deploy/ENV/EG1/flow3.bar'
Last edited: '5/24/13 4:38 PM'.
Long description: ''
User-defined property names:
  'DBSchema' = 'SIDBT01'
  'LogLevel' = 'ERROR'
Keywords:

I am checking each line of this file to check for BIP1288I and flow name 'flow1'. I want my output to contain just this

Additional thread instances: '0'
Deployed: '11/12/13 1:54 AM' in Bar file '/MSD/deploy/ENV/EG1/flow1.bar'
Last edited: '5/24/13 4:38 PM'.

This is my Perl code

foreach $line (@flows) {
  next if /^(\s)*$/;
  if ($line =~ "BIP1288I" && $line =~ "flow1") {
    $msg = "Flow found\n$line\n";
print "$msg";
  }}

Answer 1

It is unclear to me if you actually need to use perl , but this is quite simple with awk :

awk '/BIP1288I/ && /flow1/ { split($0, a, "\n"); print a[4],a[5],a[6]}' \
    RS=--- OFS=\\n input-file

You can use a2p to generate the similar script in perl . ( a2p is the awk to perl translator. perl was largely modelled on awk . Never forget your roots.) Or:

perl -wnE 'BEGIN{ $/="---"; $,="\n"}; 
    @a=split("\n"); say $a[3],$a[4],$a[5] if (m/BIP1288I/ && m/flow1/)' input-file

This is not a tight test, since it will act if the strings appear anywhere in the record, but it should be sufficient. You might need to match BIP1288I.*flow1 or restrict the match to the first line, depending on your requirements.

Answer 2

It's pretty unclear to me what you mean, but maybe grep can help you:

grep "pattern1" your_file | grep "pattern2"

will print lines containing BOTH pattern1 and pattern2

egrep "pattern1|pattern2" your_file

will print lines containing either pattern.

grep -v "pattern" your_file

will print lines not matching pattern.

So maybe you can use somethinhg like:

egrep "^Additional|flow1|^Last" your_file

Answer 3

I assume that the array @flows contains all the lines of your posted data.

use warnings; #always use these modules!
use strict; #always use these modules!

my $flag=0; #a flag to track if we found BIP1288I & flow1 in order to print the lines that follow
for my $line (@flows) {
    if ($flag==1) {
        if ($line=~/^(?:Additional thread|Deployed)/) {
            print $line; #we print the line only if flag is 1 and the line start with the words "Additional thread" or "Deployed" or "Last edited"
        }
        elsif ($line=~/^Last edited/) {
            print $line; #we print the line only if flag is 1 and the line start with the words "Additional thread" or "Deployed" or "Last edited" 
            $flag=0; #we reset the flag
        }
    }
    if ($line=~/BIP1288I: Message flow 'flow1'/) { #this the flow we are interested in
        $flag=1; #set the flag to 1 so we can start printing the lines that will follow
    }
}

This will print:

Additional thread instances: '0'
Deployed: '11/12/13 1:54 AM' in Bar file '/MSD/deploy/ENV/EG1/flow1.bar'
Last edited: '5/24/13 4:38 PM'.

Answer 4

Perhaps the following will be helpful:

use strict;
use warnings;

local $/ = '---';

while (<>) {
    print "$1$2$3\n"
      if /BIP1288I:.+'flow1'.+\n(Addi[^\n]+\n)(Depl[^\n]+\n)(Last[^\n]+\n)/s;
}

Usage: perl script.pl inFile [>outFile]

The last, optional parameter directs output to a file.

Output on your dataset:

Additional thread instances: '0'
Deployed: '11/12/13 1:54 AM' in Bar file '/MSD/deploy/ENV/EG1/flow1.bar'
Last edited: '5/24/13 4:38 PM'

Assuming that "---" separates records, you can set Perl's record separator ( $/ ) to "---" so you file's read in chunks of those records. Then you can match for "BIP1288I" and flow name "flow1" and capture the other info you want.

Hope this helps!