reg exp to parse a csv

[SM777]

I have a csv file which I want to strip out the " quotes. The problem is that some of the fields have , in the middle of them e.g. 3 fields:

"intel","fast, expensive","$45"

when I replace s/"//g I end up with 4 fields:

intel,fast, expensive,$45

I guess I need to strip out the first " on the first char of a row, then any s/",// then s/,"// then the physical last " on the row.

Is this the right way? how do I detect the first and last "
Another way I was thinking of is to change all the , which are not surrounded by " with a ; so I then get:

"intel";"fast, expensive";"$45"

and then I can just do a s/"//g to get

intel;fast, expensive;$45


Cheers.

 

[toolkit]

SM777

Something like the following should work. I'm sure some regexp guru can provide an even more compact solution:

while(<>) {

    chomp( @fields = split /&quot;,&quot;/ );

    $fields[0]        =~ s/&quot;//;
    $fields[$#fields] =~ s/&quot;//;

    print join( &quot;;&quot;, @fields ), &quot;\n&quot;;
}

Cheers, NEIL s-)

 

[nattugan]

If you are encountering something like the one below it might help you..

$temp='"intel","fast,expensive","$45"';
$_=$temp;
$pattern='","';
$replacewith=";";

s/$pattern/$replacewith/g;
s/"//g;
print($_."/n"

Good luck
ganesh :-D