How to get correct regex

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

I have below text file. I just want to get "NO."(1 to 21 or more), "device" and "name" to another file. The "NO." sometimes have different numbers. When it does the "name" also different. What regex can i use? any suggestion?

..more text...
..more text...
..more text...
Information Summary for Head #1, Site1:
  No.            P       Device          Percent     Name
 -----------    ---     ------------    -------     -----------------
     1           F        14             0.06      OPEN
     2           F        216            0.97      CONTACT
     4           F        326            1.47      LEFT
     5           F        232            1.05      MID
     6           F        54             0.24      RIGHT
     7           F        6              0.03      AT
     8           F        6              0.03      AT_GHT
     11          F        184            0.83      5DBLT
     12          F        151            0.68      2P5D
     21          P        21009          94.64     Bin P

..more text...
..more text...
..more text...
[download]

Comment on How to get correct regex Download Code

Replies are listed 'Best First'.
Re: How to get correct regex by toolic (Bishop) on Jan 13, 2015 at 13:18 UTC
split `use warnings; use strict; while (<DATA>) { chomp; my @cols = split /\s+/, $_, 6; print "@cols[1,3,5]\n"; } __DATA__ 1 F 14 0.06 OPEN 2 F 216 0.97 CONTACT 4 F 326 1.47 LEFT 5 F 232 1.05 MID 6 F 54 0.24 RIGHT 7 F 6 0.03 AT 8 F 6 0.03 AT_GHT 11 F 184 0.83 5DBLT 12 F 151 0.68 2P5D 21 P 21009 94.64 Bin P` [download] Outputs: `1 14 OPEN 2 216 CONTACT 4 326 LEFT 5 232 MID 6 54 RIGHT 7 6 AT 8 6 AT_GHT 11 184 5DBLT 12 151 2P5D 21 21009 Bin P` [download] See also unpack if fixed width applies.	[reply] [d/l] [select]
Re^2: How to get correct regex by Anonymous Monk on Jan 13, 2015 at 13:46 UTC
Thanks for the updated parameters of split, which corrects my version.	[reply]
Re^3: How to get correct regex by BillKSmith (Monsignor) on Jan 14, 2015 at 14:17 UTC
Sorry for the late reply. I like your idea of a regex. use warnings; use strict; while (<DATA>) { print "$1 $2 $3\n" if /^\s(\d+)\s+\S+\s+(\d+)\s+\S+\s+(\w+\|\w+\s\ +w+)\s$/; } __DATA__ ..more text... ..more text... ..more text... Information Summary for Head #1, Site1: No. P Device Percent Name ----------- --- ------------ ------- ----------------- 1 F 14 0.06 OPEN 2 F 216 0.97 CONTACT 4 F 326 1.47 LEFT 5 F 232 1.05 MID 6 F 54 0.24 RIGHT 7 F 6 0.03 AT 8 F 6 0.03 AT_GHT 11 F 184 0.83 5DBLT 12 F 151 0.68 2P5D 21 P 21009 94.64 Bin P ..more text... ..more text... ..more text... [download] Bill	[reply] [d/l]
Re^4: How to get correct regex by AtlasFlame (Novice) on Jan 15, 2015 at 11:32 UTC
Re^2: How to get correct regex by AtlasFlame (Novice) on Jan 13, 2015 at 13:29 UTC
I forgot to mention. There are some data above and below the text that i post. Direct split does not work.	[reply]
Re^3: How to get correct regex by AnomalousMonk (Archbishop) on Jan 13, 2015 at 18:45 UTC
Direct split does not work. You need to write some code to step through your data until you get to the data you wish to process, then process data (e.g., with split) until the end of the data that is of interest to you. Then stop. Give a man a fish: `<%-(-(-(-<`	[reply] [d/l]
Re: How to get correct regex by Anonymous Monk on Jan 13, 2015 at 13:16 UTC
Seems like a job for ... Text::CSV(_XS)!	[reply]
Re^2: How to get correct regex by Anonymous Monk on Jan 13, 2015 at 13:20 UTC
... or if OP does insist on a regex, split can use one ... `( $no, $p , $dev , $percent , $name ) = split /\s+/ , $line , 5` [download]	[reply] [d/l]