Workaround for my CSV file

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

Hello Monks
As subject tells , i needed some workaround for the issue iam facing in creation of CSV file.
iam not using TEXT::CSV_XS module due to issues and other factors in our environments, currently iam creating CSV file manually as below.

#!/usr/bin/perl
use strict;
use warnings;

open my $fh,">","tmp.csv" or die "Unable to open $! \n";
#read the file line by line and delimt with comma
while(<DATA>){
        my @array = split /~/;
        my $str = join(",",@array);     # Split seperated by comma
        print $fh "$str";
}
close $fh;

__DATA__
col1~col2~col3~col4~col5
data11~data12~data13~data14~data15
data21~data22~data23~data24~data25
data31~data32~data33~data34~data35
data,data41~data42~data43~data44~data45
data51,data52,data,junk,specialchar,sometingdata53~data54~data55
[download]

The data will be seperated by ~ sysmbol, iam going to split it and join it by comma(which is equivalent to replacing ~ with comma).
i have issues when there is a comma in one of the column data as shown in above sample.
i needed your advice on how do i overcome this for timebeing, we are considering to upgrade to TEXT:CSV_XS module sooner.
iam thinking of enclosing each value of array into double quotes to preserve original comma before writing to CSV file, how can i achieve that?
if at all i have enclose the data in double quotes, is it worth to enclose all values into double quotes or only value having special characters to be enclosed?
in either case i need monks advise on how can i proceed further till i implement CPAN module.

Comment on Workaround for my CSV file Download Code

Replies are listed 'Best First'.
Re: Workaround for my CSV file by GrandFather (Saint) on Dec 05, 2008 at 07:34 UTC
Assuming that you really can't use a module (but see Yes, even you can use CPAN), a simple regex substitution does the trick: `use strict; use warnings; open my $fh, ">", "tmp.csv" or die "Unable to open $! \n"; #read the file line by line and delimit with commas print join ",", map {s/(.,.)/"$1"/; $_} split /~/ while <DATA>; close $fh; __DATA__ col1~col2~col3~col4~col5 data11~data12~data13~data14~data15 data21~data22~data23~data24~data25 data31~data32~data33~data34~data35 data,data41~data42~data43~data44~data45 data51,data52,data,junk,specialchar,sometingdata53~data54~data55` [download] Prints: `col1,col2,col3,col4,col5 data11,data12,data13,data14,data15 data21,data22,data23,data24,data25 data31,data32,data33,data34,data35 "data,data41",data42,data43,data44,data45 "data51,data52,data,junk,specialchar,sometingdata53",data54,data55` [download] Perl's payment curve coincides with its learning curve.	[reply] [d/l] [select]
Re^2: Workaround for my CSV file by gone2015 (Deacon) on Dec 05, 2008 at 11:44 UTC
...and, if you want to allow '`"`' in the fields you can extend that to: `use strict ; use warnings ; print join ",", map { if (m/[,"]/) { s/(\A\|"\|\n\|(?<!\n)\Z)/"$1/g } ; $ +_ } split /~/ while <DATA>; __DATA__ col1~col2~col3~col4~col5 data11~data12~data13~data14~da,data15 data21~"data22"~d"ata"23~data24~da"ta"25 data31~data32~data33~data34~"data35" data,data41~data42~data43~data44~data45 data51,data52,data,junk,specialchar,sometingdata53~data54~data55` [download] which gives: col1,col2,col3,col4,col5 data11,data12,data13,data14,"da,data15" data21,"""data22""","d""ata""23",data24,"da""ta""25" data31,data32,data33,data34,"""data35""" "data,data41",data42,data43,data44,data45 "data51,data52,data,junk,specialchar,sometingdata53",data54,data55 (the last item on each line has the line ending attached to it... so have to box a little clever to get the trailing '`"`' right in all cases. GrandFather used `(.*)`, which won't match a line ending, unless you tell it to.) Or you can just stick '`"`' around every item: `print join '",', map { s/(\A\|"\|\n)/"$1/g ; $_ } split /~/ while <DATA> +;` [download] which is slightly more straightforward: "col1","col2","col3","col4","col5" "data11","data12","data13","data14","da,data15" "data21","""data22""","d""ata""23","data24","da""ta""25" "data31","data32","data33","data34","""data35""" "data,data41","data42","data43","data44","data45" "data51,data52,data,junk,specialchar,sometingdata53","data54","data55" One assumes that you don't expect your '`~`' item separators to appear in any item, not even within '`"`' or any other escaping mechanism... The deeper you go into this kind of thing, the more you find how useful stuff on CPAN is !	[reply] [d/l] [select]
Re^2: Workaround for my CSV file by harishnuti (Beadle) on Dec 05, 2008 at 08:00 UTC
Thank you Mr Grandfather, that was a good regular expression and solves current issues.. and indeed good learning for me. appreciate your time	[reply]
Re: Workaround for my CSV file by Anonymous Monk on Dec 05, 2008 at 07:22 UTC
Text::CSV_PP	[reply]
A reply falls below the community's threshold of quality. You may see it by logging in.
Re: Workaround for my CSV file by moritz (Cardinal) on Dec 05, 2008 at 07:39 UTC
If you can't use a module, you can at least look at how the standard modules solve your problem, and mimic that in your own code.	[reply]
Re: Workaround for my CSV file by brsaravan (Scribe) on Dec 05, 2008 at 08:37 UTC
Here is the simple regex logic without spliting the input `#! /usr/bin/perl use strict; use warnings; open my $fh,">","tmp.csv" or die "Unable to open $! \n"; while(<DATA>){ $_ =~ s/(\S+,)+([^~]*)/"$1$2"/g if $_ =~ /,/; $_ =~ s/~/,/g; print $fh "$_"; } close $fh; __DATA__ col1~col2~col3~col4~col5 data11~data12~data13~data14~data15 data21~data22~data23~data24~data25 data31~data32~data33~data34~data35 data,data41~data42~data43~data44~data45 data51,data52,data,junk,specialchar,sometingdata53~data54~data55` [download]	[reply] [d/l]