sql server - Removing "blank" or bad rows in a csv file using Perl. -

i working on project exports .csv files old application (a dos base application). export works, truncates column (it seems @ random) , inserts next line. it's last half of word or of sort. know, if there way use perl remove lines. use perl script remove comma delimiters , replace them pipe (|) delimiter. found using pipe makes easier insert sql use. i'm very, new perl code works replace comma's pipe. there way use , remove "bad" rows. have tried importing sql bulk insert, receive errors such "bulk load data conversion error (truncation) row 49, column 17." have looked @ export , these errors occurring there full column 17 rows, next line |cial) or |3. hope made clear. thank help.

use strict; use warnings;  use text::csv_xs;  $in_file = "deal_log.csv"; $out_file = "newdeal.csv";  open $fh, '<', $in_file or die "$in_file: $!"; open $out_fh, '>', $out_file or die "$out_file: $!";  $in_csv = text::csv_xs->new; $out_csv = text::csv_xs->new( { sep_char => '|', eol => "\n" } );  while( $row = $in_csv->getline( $fh ) ) {      $out_csv->print( $out_fh, $row ); } 

do not output line immediatelly. instead, read in next line , check number of columns. if there 17, output previous line, otherwise append line previous remembered line.

#!/usr/bin/perl use warnings; use strict;  use text::csv_xs;  $in_file  = 'deal_log.csv'; $out_file = 'newdeal.csv'; $size     = 17;  open $fh,     '<', $in_file  or die "$in_file: $!"; open $out_fh, '>', $out_file or die "$out_file: $!";  $in_csv  = 'text::csv_xs'->new; $out_csv = 'text::csv_xs'->new( { sep_char => '|',                                      eol      => "\n",                                    } );  $previous = []; while( $row = $in_csv->getline($fh) ) {     if (@$row == $size) {         $out_csv->print($out_fh, $previous) if @$previous;         $previous = $row;     } else {         $previous->[-1] .= $row->[0];     } } $out_csv->print($out_fh, $previous);  # not forget print last remembered line.