Hi,
 I have a script to rewrite a data file which is comma
delimited (sub openfile). Whenver there is a line return in a field, that
field
is wrapped with the double quote character.

I can't grab that field to write into the new data file properly.
In the pasted"sample.txt" file (bottom), line 2 and 3 are actually one
record.

I've attempted a work around (sub strip_wrap), trying to strip out any line
returns for
any field. This does not remove the line return.

Any suggestions

Below is the sub to open the Comma Delimited file:

sub openfile {
#work with each field in uploaded file

close DATA;

$Count2 = 0;

open (LIST, ">>$BaseDir/1_$foo");


if ($CField[5] ne "") {
$AdFile[$Count2] = $Ad;
## serial number $wum to renumber records
$wum = $Count2;
 $Count2++;
 &strip_wrap;

print LIST
qq!$wum|$CField[0]|$CField[1]|$CField[2]|$CField[3]|$CField[4]|$CField[5]|$C
Field[6]|$CField[7]|$CField[8]|$CField[9]|$CField[10]|$CField[11]|$CField[12
]|$CField[13]|$CField[14]|$CField[15]|$CField[16]|$CField[17]|$CField[18]|$C
Field[19]|$CField[20]|$CField[21]|$CField[22]|$CField[23]|$CField[24]|$CFiel
d[25]|0\n!;

sub strip_wrap {

## need device to remove return from within a comma delimited field
$CField[0] =~ s/%0D%0A/ /go;
$CField[0] =~ s/\n/ /go;
$CField[0] =~ s/\r/ /go;

## ... and so on through all the fields

--

sample.txt snip <<
Line 1: First Name,Last Name,Middle Name,Name,Nickname,E-mail Address,Home
Street,Home City,Home Postal Code,Home State,Home Country,Home Phone,Home
Fax,Mobile Phone,Personal Web Page,Business Street,Business City,Business
Postal Code,Business State,Business Country,Business Web Page,Business
Phone,Business Fax,Pager,Company,Job Title,Department,Office Location,Notes

Willow
Line 3: Ave",city,zip,state,Canada,phone,fax,cell,url,Bus
address,bcity,bzip,bprov,bcountry,,phone,fax,pager,ASM,btitle,bdept,office
num,test
Line 4: Adrian,Dan,,Adrian Dan,,,,,,,,(416) 245-1927,,,,,,,,,,,,,,,,,
Line 5: Adrian,Heaps,,Adrian

Line 6: Alan,Burt,,Alan Burt,,,,,,,,(416) 266-4682,,,,,,,,,,,,,,,,,

Chris Bell

Tel: (416) 698-5763

"Interesting Code" !

Maybe not your full answer but don't forget to check for \t in your sub as
well.

I think your problem is caused by the fact that you haven't told perl anything
about the "text text [newline] text" being different from just normal [newline]
so it thinks every newline it finds is the end of a record.   I'd suggest "slurp
mode". This means reading the entire file into a single scalar variable rather
than as a series of lines.  You can then substitute the newlines inside double
quoted strings and leave the real newlines alone.

I've tested this with a very small file and it appeared to work.

Hope this helps

Graham Wood

===================================================================
# open your file
open(INPUT,"input.csv") || die "Can't open input.csv $!\n";

# $/ is the Input record separator (see perldoc perlvar for details). If you
undef it
# perl doesn't split an input file into separate lines

undef $/;

# ... which means you can slurp the whole file into a single scalar variable,
newlines
#  and all
$input=<INPUT>;

close INPUT;

# You can then replace any newlines in between " characters with
# spaces (or anything else you want to) and remove the " characters
# This may need repeating for \r if \n doesn't catch everything.
#  Using [^"] in the pattern to match after the \n means the first " you hit
after the \n
# will terminate the pattern match.

$input=~ s/"(.*)\n([^"]*)"/$1 $2/g;

# then you can split $input at the remaining newlines and loop through
# them
$wum=1;
foreach(split /\n/,$input){
    #
    # print record number and fields in $_ delimited by "|" then a newline
    #
    print join("|", $wum++ , split(/,/,$_)    ),    "\n";

Other folks answered your question, but I have an unrelated
suggestion.  Instead of this:

Consider writing

It will be easier to read and understand, and you won't have to change

I hope this is helpful.

[snip stuff... want to turn comma seperated into pipe seperated]

# This assumes that if a field has either quotes and a newline in it,
# any quotes are doubled... that is, if the original was something
# like qq[This has a " quote and a \n newline], then the version
# in the file will be qq["This has a "" quote and a \n newline"].

my $quotepart = qq["[^"]*(?:""[^"]*)*"];
my  $csvregex = qq[(?:(?!")[^,\n]*|$quotepart)];
sub openfile {
   open( my($csv), "<", "$BaseDir/sample.txt")
      or die "Couldn't open $BaseDir/sample.txt: $!";
   open( my($psv), ">>", "$BaseDir/1_$foo")
      or die "Couldn't open $BaseDir/1_$foo: $!";
   my ($buf, $record, $line) = ("", 1, 1);
   while( sysread( $csv, $buf, 8192, length $buf ) ) {
      while( s/^((?:$csvregex,)*$csvregex\n)//, my $x = $1 ) {

         /^"/ ?
            # if quoted, remove quotes if unneeded.
            (s/^"([^"\n|]*)"$/$1/s && s/""/"/g) :
            # if not quoted, add quotes if needed.
            (/\|/ and (s/"/""/g, $_ = qq["$_"]))


         ++$records; $lines += $x =~ tr/\n//;
      }
      next unless $buf =~ /^$quotepart([^"\n,])/;
      die( "Malformed input, line $line record $record : q[$buf]\n" .
           "Unexpected char q[$1] at position $-[1].\n" );
   }
   warn "Incomplete last record q[$buf]\n" if length $buf;
   close $csv;
   close $psv or die "Couldn't close $BaseDir/1_$foo: $!";

[untested]

--
print reverse( ",rekcah", " lreP", " rehtona", " tsuJ" )."\n";