Perl解析在线银行对账单井

Question

I am parsing out the Wells Fargo online banking statement. The card number is truncated and the number starting with a P or an S is a unique transaction identifier. I changed out the unique identifier and last four digits of card number. I consider myself safe from any privacy concerns. I parse, or filter each line through a long ugly regex - which I am not sure you are supposed to do. There might be a better way, I don't know.

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

#my $filename = 'wellsfargo_balanceStatement.txt';
#open(my $fh, '<:encoding(UTF-8)', $filename)
#  or die "Could not open file '$filename' $!";

#while (my $row = <$fh>)  {
while (my $line = <DATA>) {
  chomp $row;
  if ($row =~ /(\d{2}\/\d{2}\/\d{2})\s(PURCHASE).*\d{2}\/\d{2}(.*)\w\d{10}\d+.*\$(\d+\.\d+)/) {
      my $date=$1;
      my $purchs=$2;
      my $pur_plce=$3;
      my $pur_amt=$4;
      print "$date $purchs $pur_plce $pur_amt\n";
      }
  sleep .5 ;
}



__DATA__

  09/18/17 PURCHASE AUTHORIZED ON 09/17 CVS/PHARM 06062--200 W Manhattan NY P00000000032583371 CARD 4184   $4.87
  09/18/17 PURCHASE AUTHORIZED ON 09/16 JUBILEE MARKET NEW YORK NY S467259862756690 CARD 4184   $8.78
  09/18/17 PURCHASE AUTHORIZED ON 09/16 LOWE'S #3292 NEW YORK NY P00307259724475616 CARD 6029   $23.39
  09/18/17 PURCHASE AUTHORIZED ON 09/16 NYSC WEST END NEW YORK NY S587259513187673 CARD 4184   $39.00
  09/18/17 PURCHASE AUTHORIZED ON 09/16 JUBILEE MARKET NEW YORK NY S587259468801533 CARD 4184   $21.73
  09/18/17 PURCHASE AUTHORIZED ON 09/15 7-ELEVEN NEW YORK NY P00000000840668487 CARD 4184   $12.75
  09/18/17 PURCHASE AUTHORIZED ON 09/15 DUNKIN #351200 NEW YORK NY S307258635156794 CARD 4184   $2.82
  09/18/17 PURCHASE AUTHORIZED ON 09/15 DUNKIN #351200 NEW YORK NY S587258634843803 CARD 4184   $2.82

The transaction identifier beginning with the consonant S or P which is followed by an number anywhere from 10 to 17 digits. I thought I was pretty crafty with the regex \\w\\d{10}\\d+ however when the first consonant of the unique identifier is a 'P' it prints out 5 characters, and when it is an 'S' it prints out 3. Frankly, I don't even want the unique identifier, and do not know how it is getting in there.

09/18/17 PURCHASE  CVS/PHARM 06062--200 W Manhattan NY P00000 4.87
09/18/17 PURCHASE  JUBILEE MARKET NEW YORK NY S467 8.78
09/18/17 PURCHASE  LOWE'S #3292 NEW YORK NY P00307 23.39
09/18/17 PURCHASE  NYSC WEST END NEW YORK NY S587 39.00
09/18/17 PURCHASE  JUBILEE MARKET NEW YORK NY S587 21.73
09/18/17 PURCHASE  7-ELEVEN NEW YORK NY P00000 12.75
09/18/17 PURCHASE  DUNKIN #351200 NEW YORK NY S307 2.82
09/18/17 PURCHASE  DUNKIN #351200 NEW YORK NY S587 2.82

Eventually I am going to add comma delimit the file and enter it into excel. So I can create bar graphs, circle charts and whatever. This is what I want

09/18/17 PURCHASE  CVS/PHARM 06062--200 W Manhattan NY 4.87
09/18/17 PURCHASE  JUBILEE MARKET NEW YORK NY 8.78
09/18/17 PURCHASE  LOWE'S #3292 NEW YORK NY 23.39
09/18/17 PURCHASE  NYSC WEST END NEW YORK NY 39.00
09/18/17 PURCHASE  JUBILEE MARKET NEW YORK NY 21.73
09/18/17 PURCHASE  7-ELEVEN NEW YORK NY 12.75
09/18/17 PURCHASE  DUNKIN #351200 NEW YORK NY 2.82
09/18/17 PURCHASE  DUNKIN #351200 NEW YORK NY 2.82

Answer 1

I'm a big fan of using .* in regexps as little as possible, because it makes it much easier to read and understand. Also, you can use other delimiters than / , which will allow you to use / in your regexp without escaping it.

I'd suggest something like this:

if ($row =~ m!(\d\d/\d\d/\d\d) (PURCHASE) \w+ ON \d\d/\d\d (.*) \w+ CARD \d+\s+\$(\d+\.\d+)$!) {

I've also taken the liberty to change \\d{2} to \\d\\d - I find that much easier to read, because my brain doesn't have to go into "hey, there has to be a certain number of \\d here, the { is important" mode.