in reply to Regular Expression to Parse Data from a PDF
G'day Kevin,
Firstly, I'm not a user of CAM::PDF; in fact, I didn't even have it installed. I suspect the getPageText() method is not the best choice for this: as you noted, you can't split lines easily and dollar amounts have an embedded space — I can't advise of a better choice; perhaps another monk can.
I would strongly recommend that you do not write lengthy regexes the way you did in the last example in your code: they are incredibly difficult to read; even more difficult to maintain; and extremely error-prone. See my code below for a much better way to do this. Also, take a look at Regexp::Debugger: I find it very helpful and, in fact, used it to check some of the fiddlier parts of the regex in the code below.
I've ignored the PDF download part of the code. You didn't ask about that: I'm assuming you've got that working satisfactorily. I just downloaded the two PDFs you referenced and accessed them from a local disk.
Some notes on how I've dealt with lack of information:
- You said you wanted to "capture all of the columns except comments". You said nothing about header information or the various totals, so I've simply ignored them.
- You also said nothing about output: I've captured the columnar data; I'll leave you to decide what you want to do with it.
- The two example PDFs you linked only had one page each, so looping through all pages seems somewhat superfluous; however, I've left that for loop almost exactly as you wrote it.
- It's also unclear whether you want to capture data by page, document, or some other grouping: again, I'll leave you to decide.
Here's the code:
#!/usr/bin/env perl use strict; use warnings; use constant { AMOUNT => 3, ADDL_RATE_PER => 4, DISCOUNT_PRICE => 6, }; use CAM::PDF; use Data::Dumper; my $jacket_id = $ARGV[0]; my $pdf_file = "pm_11113472_$jacket_id.pdf"; my $pdf = CAM::PDF::->new($pdf_file) or die $CAM::PDF::errstr; my $re = qr{(?x: \A \s*? ((?:A|)) # Awd \s+ (\d+-\d+) # Contractor Code \s+ ([^\$]+?) # Name \s+ (\$\s[0-9,.]+) # Amount \s+ (\$\s[0-9,.]+\s[A-Z]) # Add'l Rate/PER \s+ ([0-9.]+\s+\d+) # Discount % Days \s+ (\$\s[0-9,.]+) # Discount Price \s+ ([\D]+?) # Bidders Name \s+ (\S+) # Date Received \s+ (\(\d+\)\s\d+-\d+) # Phone Number )}; for my $page_num (1 .. $pdf->numPages) { my $text = $pdf->getPageText($page_num); my @lines; my $wanted_line = 0; for my $line (split /$jacket_id/, $text) { next unless $wanted_line++; my @fields = $line =~ $re; $fields[AMOUNT] =~ y/ //d; $fields[ADDL_RATE_PER] =~ s/ //; $fields[DISCOUNT_PRICE] =~ y/ //d; push @lines, [ $jacket_id, @fields ]; } print Dumper(\@lines); }
Here's the first part of the output using your first example PDF:
$ ./pm_11113472_pdf_parse.pl 746810 $VAR1 = [ [ '746810', 'A', '140-89226', 'UNION HOERMANN PRESS', '$844.00', '$15.00 C', '1 20', '$835.56', 'Randy Sigman', '01/22/2020', '(563) 582-3631' ], [ '746810', '', '190-38407', 'GRAPHIC VISIONS', '$869.00', '$140.00 M', '0.5 20', '$864.66', 'Howard Roskosky', '01/22/2020', '(301) 987-5586' ],
Open the spoiler to see full output for both example PDFs.
— Ken
|
---|
Replies are listed 'Best First'. | |
---|---|
Re^2: Regular Expression to Parse Data from a PDF
by kevyt (Scribe) on Feb 27, 2020 at 16:35 UTC | |
Re^2: Regular Expression to Parse Data from a PDF
by kevyt (Scribe) on Feb 28, 2020 at 02:25 UTC | |
by kcott (Archbishop) on Feb 28, 2020 at 06:28 UTC | |
by kevyt (Scribe) on Feb 28, 2020 at 06:52 UTC |