Read/Change files in a Directory

LF has asked for the wisdom of the Perl Monks concerning the following question:

Hi,

I am simply trying to read all the (XML) files from within a directory and make a few text changes to each:

The changes are meant to change all occurrances of XML lines like:

<dc:subject xml:lang="en-US">18th century 17th century</dc:subject>

To:

<dc:subject xml:lang="en-US">17th century</dc:subject>
<dc:subject xml:lang="en-US">18th century</dc:subject>

And then to output the files to a new directory.

However, it seems like there are some errors in my code, esp. for opening/closing the files. Can anyone tell me what is wrong? This is my code:

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

my $record_element = 'dcterms:available';
my $filename_element = 'dc:identifier';

my ($file, $dir) = @ARGV;
defined($file) && defined($dir) || usage();
chop($dir) if $dir =~ m#/$#;
#$/ = "</$record_element>";
mkdir $dir;

open(INPUT, "<$ARGV[0]") || die "Failed to open $ARGV[0]\n";
  $file = "INPUT";

my $record_count = 0;
my $header1 = '<?xml version="1.0" encoding="UTF-8"?>';
my $header2 = '<qualifieddc xmlns="http://example.org/appqualifieddc/"
+';  
my $header3 = 'xmlns:dc="http://purl.org/dc/elements/1.1/"';
my $header4 = 'xmlns:dcterms="http://purl.org/dc/terms/"';
my $header5 = 'xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"';
my $header6 = 'xsi:schemaLocation="http://example.org/appqualifieddc/'
+;
my $header7 = 'http://dublincore.org/schemas/xmls/qdc/2003/04/02/appqu
+alifieddc.xsd"';
my $closing = '</qualifieddc>';

$/ ='terms/">';
<INPUT>;

$/ = "</$record_element>";

while(<INPUT>){
  next if m/^\s*$/;    #Disregard 0+ trailing spaces
  s/^\s*//;            #Collapse 0+ leading spaces
  $record_count++;
  if (m#<$filename_element xsi:type="dcterms:URI">(\w+) (\w+) (\w+)_(\
+w+)</$filename_element>#s)
      {
    msg("Just matched $1 $2 $3_$4");
    my $outfile = "$dir/$4.xml";
    open (OUTPUT, '>', $outfile) or die "Can't create/open $outfile fo
+r writing";
        print OUTPUT $header1, "\n", $header2, "\n", $header3, "\n", $
+header4, "\n", $header5, "\n", $header6, "\n", $header7, "\n";      
    print OUTPUT $_, "\n";
     print OUTPUT $closing;

    close (OUTPUT) or die "Couldn't close $outfile after writing recor
+d";
      } 
  else {
    warn("Couldn't find filename in record $record_count, skipping..."
+);
       }
      }

close (INPUT) or die "Couldn't close $file";
close (OUTFILE);

$record_count--;
msg("Processed $record_count records from $file");

sub msg {
    print @_, "\n";
}

sub usage {
    msg("Usage: $0 <file> <directory>");
    exit(1);
}
[download]

Comment on Read/Change files in a Directory Download Code

Replies are listed 'Best First'.
Re: Read/Change files in a Directory by osunderdog (Deacon) on Nov 09, 2004 at 17:51 UTC
Lacking a few details, but here is how I would approach this: use strict; use XML::Simple; my $xs = XML::Simple->new(KeepRoot=>1); my $data; { local $/; $data = <DATA>; } my $inStruct = $xs->XMLin($data); my $outStruct; foreach my $tag (@{$inStruct->{Root}->{'dc:subject'}}) { my @matches = $tag->{content} =~ /(\d+th century)/g; foreach my $match (@matches) { my %item = %$tag; $item{content} = $match; push(@{$outStruct->{Root}->{'dc:subject'}}, \%item); } } my $resultXML = $xs->XMLout($outStruct); print $resultXML; __DATA__ <Root> <dc:subject xml:lang="en-US">18th century 17th century</dc:subject> <dc:subject xml:lang="en-US">18th century 14th century</dc:subject> <dc:subject xml:lang="en-US">18th century 12th century</dc:subject> <dc:subject xml:lang="en-US">18th century 11th century</dc:subject> <dc:subject xml:lang="en-US">18th century 10th century</dc:subject> <dc:subject xml:lang="en-US">18th century 12th century</dc:subject> <dc:subject xml:lang="en-US">18th century 12th century 12th century 12 +th century 12th century</dc:subject> </Root> [download] *"Look, Shiny Things!"* is not a better business strategy than compatibility and reuse. OSUnderdog	[reply] [d/l]
Re: Read/Change files in a Directory by zigdon (Deacon) on Nov 09, 2004 at 17:25 UTC
I must be terribly confused - but from what I can decode in your program, it does nothing like what you described as your goal? Perhaps you can post a short example file, or explain more what you're trying to do? Since I can't understand what you're trying to do, I'll just give you the one tip - whenever you find yourself using variables like $header1...$header7, you probably will be better off using an array @headers. Or, if this is really your whole program, all the headers could be put in a single scalar $header. -- zigdon	[reply]
Re: Read/Change files in a Directory by bgreenlee (Friar) on Nov 09, 2004 at 17:21 UTC
It's hard to troubleshoot your code without more information about what errors you're getting, but if you're really just making a simple change like that, and this is a one-off thing, this sounds like a good candidate for a little perl-one-liner: `perl -pi -0777 -e "s{(<dc:subject xml:lang=\"en-US\">)(\w+) (\w+) (\w+ +) (\w+)(</dc:subject>)}{$1$2 $3$6\n$1$4 $5$6}gs" *.xml` [download] Experiment with this on a backup copy of the directory you're working with, as it will change the files in place. -b	[reply] [d/l]