Re: problem in XML parsing

[hipowls]

There are lots of modules on CPAN that parse XML. You really don't want to do it yourself except as as a learning exercise. And then you will use a module from CPAN;)

When dealing with data structures it is very helpful to print the data you are working with. I've use YAML others use Data::Dumper, either is fine.

Update: I've changed the code below in light of ikegami's comments.

This is one approach using XML::Simple

#!/net/perl/5.10.0/bin/perl

use strict;
use warnings;

use YAML;

use XML::Simple;

my $teams = XMLin(
    do { local $/; <DATA> },
    ForceArray => [qw( football_team Player )],
);

print Dump $teams;

foreach my $team ( @{ $teams->{football_team} } ) {
    print "$team->{Name}\n";
    foreach my $player ( @{ $team->{Players}{Player} } ) {
        print "  $player->{ForeName} $player->{LastName}\n";
    }
}
__DATA__

<football_teams>
  <football_team>
    <Name>ONE</Name>
    <DateCreated>
      <Year>1990</Year>
      <Month>12</Month>
      <Day>04</Day>
    </DateCreated>
    <Players>
      <Player>
        <LastName>Wonguso</LastName>
        <ForeName>Jimm</ForeName>
        <Initials>JW</Initials>
      </Player>
      <Player>
        <LastName>Fearless</LastName>
        <ForeName>A</ForeName>
        <Initials>AF</Initials>
      </Player>
    </Players>
    <Hometown>New York</Hometown>
  </football_team>
  <football_team>
    <Name>TWO</Name>
    <DateCreated>
      <Year>1978</Year>
      <Month>9</Month>
      <Day>23</Day>
    </DateCreated>
    <Players>
      <Player>
        <LastName>Pedro</LastName>
        <ForeName>Therry</ForeName>
        <Initials>TP</Initials>
      </Player>
      <Player>
        <LastName>Haywardis</LastName>
        <ForeName>Richie</ForeName>
        <Initials>RH</Initials>
      </Player>
      <Player>
        <LastName>Eswa</LastName>
        <ForeName>Jeyan</ForeName>
        <Initials>JE</Initials>
      </Player>
      <Player>
        <LastName>Leongus</LastName>
        <ForeName>John</ForeName>
        <Initials>JL</Initials>
      </Player>
      <Player>
        <LastName>Bentson</LastName>
        <ForeName>Billie</ForeName>
        <Initials>BB</Initials>
      </Player>
    </Players>
    <Hometown>California</Hometown>
  </football_team>
</football_teams>

which produces

---
football_team:
  - DateCreated:
      Day: 04
      Month: 12
      Year: 1990
    Hometown: New York
    Name: ONE
    Players:
      Player:
        - ForeName: Jimm
          Initials: JW
          LastName: Wonguso
        - ForeName: A
          Initials: AF
          LastName: Fearless
  - DateCreated:
      Day: 23
      Month: 9
      Year: 1978
    Hometown: California
    Name: TWO
    Players:
      Player:
        - ForeName: Therry
          Initials: TP
          LastName: Pedro
        - ForeName: Richie
          Initials: RH
          LastName: Haywardis
        - ForeName: Jeyan
          Initials: JE
          LastName: Eswa
        - ForeName: John
          Initials: JL
          LastName: Leongus
        - ForeName: Billie
          Initials: BB
          LastName: Bentson
ONE
  Jimm Wonguso
  A Fearless
TWO
  Therry Pedro
  Richie Haywardis
  Jeyan Eswa
  John Leongus
  Billie Bentson

[ikegami]

I don't like XML::Simple. Remove a team from the sample data and you get

Not an ARRAY reference at 675580.pl line 16.

Remove a player from ONE and you'll get

ONE
Not an ARRAY reference at 675580.pl line 18.

The call to XMLin for this schema should be *at least*

my $teams = XMLin(
    $xml_text,
    ForceArray => [qw( football_team Player )],
);

And then there's ForceContent.

XML::Simple is deceptively complex to use because it returns data in an unpredictable format, forcing you to use ref all over the place or to specify arguments to tell it what it already knows but chooses to destroy. All that because XML::Simple assumes attributes and elements are interchangeable. (Are they ever?)

[hipowls]

Thanks for pointing out the potential problem with my code, I've updated the original post.

I agree that XML::Simple isn't simple but at some point you have to provide a mapping between XML and perl data structures and nothing is truly simple. For reading XML without attributes it is good enough, though you do need to be aware of the traps (but that's what the man page and perlmonks are for;).

For "real" XML processing I'd look elsewhere but this wasn't that question and then I'd be far more likely to seek advice than offer it.