Regex with Newline

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

Dear Monks, I have a long list in the following style:

__DATA__
Some Company; Some Street 52
12345 City1
Another Company; Another Street 63
23456 City2
Yet Another Company; New Street 11
34546 City3
[download]

I need it in one line per company (i.e. the newline between the house number and the zip code should disappear). This code does not work:

use strict;
use warnings;


while(<DATA>)    {
    
    $_=~s/([\w; ]+\d+)\n(\d{5})/$1$2/gms;
    print;
}
[download]

I got something with the following workaround:

use strict;
use warnings;


while(<DATA>)    {
$_=~s/(\d{5})/;$1;/g;
    $_=~s/\n//gms;
    $_=~s/(\d{5}; \w+)/$1\n/g;
    
    print;
}
[download]

but this seems ugly. Where is the catch? Many thanks! VE

Comment on Regex with Newline Select or Download Code

Replies are listed 'Best First'.
Re: Regex with Newline by JavaFan (Canon) on Aug 26, 2011 at 16:02 UTC
`while (<DATA>) { s/\n/ / if /;/; print; }` [download]	[reply] [d/l]
Re^2: Regex with Newline by ricDeez (Scribe) on Aug 27, 2011 at 04:21 UTC
so simple and elegant... I like it, a lot!!!	[reply]
Re: Regex with Newline by toolic (Bishop) on Aug 26, 2011 at 14:31 UTC
Consider this approach: `use warnings; use strict; my $prev; while(<DATA>) { if (/;/) { chomp; $prev = $_; } else { s/(\d{5})/$1;/; print "$prev; $_"; } } __DATA__ Some Company; Some Street 52 12345 City1 Another Company; Another Street 63 23456 City2 Yet Another Company; New Street 11 34546 City3` [download]	[reply] [d/l]
Re: Regex with Newline by RichardK (Parson) on Aug 26, 2011 at 14:37 UTC
As ';' is the only way to identify a company name, why not something like this? `while (<DATA>) { chomp; print "\n" if /;/; print $_; } print "\n";` [download]	[reply] [d/l]
Re^2: Regex with Newline by Marshall (Canon) on Aug 26, 2011 at 16:04 UTC
I like the simplicity of this.	[reply]
Re: Regex with Newline by hardburn (Abbot) on Aug 26, 2011 at 14:24 UTC
The `while(<DATA>)` only reads one line at a time. In the second example, the first regex replace will match the first line, while the second will match all lines (note that you'd be better off using `chomp()` for that purpose). In the first example, there is no single line that will match that regex, so the replacement never happens. "There is no shame in being self-taught, only in not trying to learn in the first place." -- Atrus, Myst: The Book of D'ni.	[reply] [d/l] [select]
Re: Regex with Newline by Anonymous Monk on Aug 26, 2011 at 14:43 UTC
Thank you very much! Now it works with the following code: `use strict; use warnings; while(<DATA>) { chomp; $_=~s/(\d{5} \w+)/ \1\n/g; print; } __DATA__ Some Company; Some Street 52 12345 City1 Another Company; Another Street 63 23456 City2 Yet Another Company; New Street 11 34546 City3` [download] Is this a reasonable code or I just have luck now? Many thanks! VE	[reply] [d/l]
Re^2: Regex with Newline by Marshall (Canon) on Aug 26, 2011 at 15:51 UTC
I think that is a bit "too clever". I was surprised but, this \1 idea did work although it generates a warning message. I would argue that easy to understand code is better. And it often executes faster! `#!/usr/bin/perl -w use strict; my $previous_line; while (<DATA>) { if (/;/) { chomp; $previous_line = $_; } else { print "$previous_line; $_"; } } __DATA__ Some Company; Some Street 52 12345 City1 Another Company; Another Street 63 23456 City2 Yet Another Company; New Street 11 34546 City3` [download]	[reply] [d/l]