Matching Across Lines

mkb125 has asked for the wisdom of the Perl Monks concerning the following question:

Ok, so I searched the archives but I still don't get it. I have the following text in a file that I want to match:

Total Datafile Size MB
----------------------
73602.1406

Total Database Size MB
----------------------
48107.7656

Total Free Space MB
-------------------
25462.8125

The code I tried to match the following:

Total Datafile Size MB
----------------------
73602.1406

is (and many variations at which point I gave up):

local $/;  # slurp the entire file
local *FH;
open (FH, $file_name) or die "Cannot open";
my $text = <FH>

if ($text =~ /[^Total Datafile Size MB\s*][^-{22}]/s)
{
   print "Matched\n";
}
[download]

My ultimate goal is to match this:

Total Free Space MB
-------------------
25462.8125

and then grab the number 25462.8125.

Any ideas?

many thanks

mohammed

Edited by Chady -- Added code tags.

Comment on Matching Across Lines Download Code

Replies are listed 'Best First'.
Re: Matching Across Lines by duff (Parson) on Feb 24, 2004 at 22:12 UTC
You need to use the /m modifier so that `^` can match the start of lines within your string. Otherwise it just matches the start of your string and not lines within it. Also, you might want to use the /x modifier so that you can comment your RE a little bit. Here's a snippet for you to modify as is your wont: `#!/usr/bin/perl -l $text = do { local $/; <DATA> }; print $1 if $text =~ / ^Total\s+Datafile\s+Size\s+MB\s* # header ^-+\s* # separator ^([\d.]+) # the number I want /mix; __DATA__ Total Datafile Size MB ---------------------- 73602.1406 Total Database Size MB ---------------------- 48107.7656 Total Free Space MB ------------------- 25462.8125` [download] Update: Some other notes. I'm not sure why you're using `local *FH;` but if you have a modern perl, you can do something like this instead: `open(my $fh, $file_name) or die; my $text = do { local $/; <$fh> }; # slurp close $fh;` [download] duff	[reply] [d/l] [select]
Re: Re: Matching Across Lines by mkb125 (Initiate) on Feb 25, 2004 at 18:43 UTC
matija and duff Many thanks for your replies. I got it working using duff's solution. The local FH stuff I think I picked out from Programming Perl or some where on the net. Not sure. Anyway, I changed it as described in the solution. One thing I'm still confused about are the m and s modifiers. Still don't understand: /m matches ^$ next to embedded \n -- not sure what this means /s . match newline and ignores deprecated $ -- same here, not sure what this means Anyway, thanks both for replying and getting my script working!!!	[reply]
Re: Re: Re: Matching Across Lines by mkb125 (Initiate) on Feb 25, 2004 at 19:37 UTC
OK, ok, ok...I understand now.... I re-read http://www.infocopter.com/perl_corner/perlre.htm and http://lists.w3.org/Archives/Public/public-qt-comments/2003Jul/0091.html I also got both solutions working. Here is majita's solution which I just figured out: if ($text =~ /^Total Datafile Size MB(-+)\s+(\d.+)/s) { print $2; } Again, thanks both for helping me out. mohammed	[reply]
Re: Matching Across Lines by mkb125 (Initiate) on Feb 24, 2004 at 21:38 UTC
Ok, wait I re-read the archs again. I changed my regexp to `if ($text =~ /[^Total Datafile Size MB\s\n(.)^-{22}]/s)` Seems to be working (I think) Now all I need is to match for the number on the 3rd line and grab it. mohammed	[reply] [d/l]
Re: Re: Matching Across Lines by matija (Priest) on Feb 24, 2004 at 21:48 UTC
To do that, the regexp should be `if ($text =~ /^Total Datafile Size MB\s\n(.)^-{22}\s\n(\d+\.\d+)/s)` [download] The second number will now be in $2. While you're at it, you should also change the `.` in the first paren with something like `\d+\.\d+`. I realize both work right now, but if you get some funny stuff in your file, the stricter version will make sure you don't pick up a non-number.	[reply] [d/l] [select]