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

I have the below complex yet small XML snippet saved in a file, with multiple entries of ENVDETAILS.

<OUTMOST>
        <MYENVDETAILS>
                <ENVDETAILS id="abc" >
                        <DMGRHOST>a.b.c.d</DMGRHOST>
                        <CELL>XYZ</CELL>
                        <NODE nodeid="1" >
                                <NODEHOST>v.x.y.z</NODEHOST>
                                <IHSNODE>v_ihsnode</IHSNODE>
                                <IHSJVM ihsjvmtype="log_ihs" >
                                        <IHSJVMNAME>abc_logihs1</IHSJV
+MNAME>
                                </IHSJVM>
                                <IHSJVM ihsjvmtype="appihs" >
                                        <IHSJVMNAME>abc_appihs1</IHSJV
+MNAME>
                                </IHSJVM>
                                <JVM jvmtype="nodeagent" >
                                        <JVMNAME>nodeagent</JVMNAME>
                                </JVM>
                                <JVM jvmtype="api" >
                                        <JVMNAME>abc_api1.1</JVMNAME>
                                        <JVMNAME>abc_api1.2</JVMNAME>
                                </JVM>
                        </NODE>
                </ENVDETAILS>
        </MYENVDETAILS>
</OUTMOST>

[download]

Few things we must know:

a). nodeid is not unique.

b). ENVDETAILS id is unique.

I wanted to achieve/perform the following things:

1). Display all the ENVDETAILS ids: [output should have all ids: abc]

2). Display all the jvms corresponding to ENVDETAILS ids & after checking the jvm type: [input: (id=abc)(jvmtype=api) output should have all jvms of type 'api' from ENV id 'abc' i.e.: abc_api1.1, abc_api1.2 ]

3). Display the NODEHOST of JVMNAME is entered as input: [input: jvm=abc_logihs1output should be NODEHOST of jvm 'abc_logihs1' i.e.: v.x.y.z]

4). Display all the jvmtype present on the NODEHOST entered as a input: [input: NODEHOST=v.x.y.z output should be: log_ihs, appihs, nodeagent, api]

5). Display the CELL of NODEHOST is entered as a input: [input: NODEHOST=v.x.y.z output should be: XYZ ]

6). Display the nodeid when NODEHOST is entered as a input: [input: NODEHOST=v.x.y.z output should be: 1 ]

7). Display the DMGR when JVMNAME is entered as input: [input: jvm=abc_api1.1output should be : a.b.c.d ]

8). Display the ENVDETAILS when JVMNAME is entered as input: [input: jvm=abc_api1.1output should be : abc ]

Please help. I am using XML::Smart to achieve few things, however I am learning perl.

I have only written code for 1st point.


use XML::Smart;
use strict;
use warnings;

$ENVFILE="/appl/TOPOLOGY.xml";
my $ENVxml = XML::Smart->new($ENVFILE) || die ("Could not find details
+ file.");
my $env = $ARGV[0];

if ( $ARGV[0] eq "fetch" ) {
        my @envname=$ENVxml->{OUTMOST}->{MYENVDETAILS}->{ENVDETAILS}('
+[@]','id');
foreach my $i (@envname) {
                print "$i\n";
}
}

[download]

Replies are listed 'Best First'.
Re: parsing & retrieving from an XML file
by marto (Cardinal) on Nov 06, 2017 at 17:39 UTC

    I'd use Path::Tiny to read the file into a variable ($xml), then Mojo::DOM to process it, e.g.

    # no XML declaration, in example provided
# so no auto detect fancyness, force xml mode:

my $dom = Mojo::DOM->new->xml(1)->parse( $xml );

for my $e ( $dom->find('ENVDETAILS')->each ){
  say $e->{id};
}

    [download]

    Familiarise yourself with Mojo::DOM, use it to filter the data based upon your other requirements, post your attempts and how they failed. See also this example using selectors to get just the data required.

[reply]
[d/l]
[select]
      Ty very much for the reply, I will try & post the results. Thanks :)
[reply]
Re: parsing & retrieving from an XML file
by Discipulus (Canon) on Nov 06, 2017 at 19:50 UTC
    Hello sbafna and welcome to the monastery and to the wonderful world of Perl!

    I abandoned the use of XML::Smart long time ago. iirc the smart part ends in it's name..

    In a recent thread you can review all serious tools to parse XML using Perl. My homenode has a bunch of links about the matter.

    I normally use XML::Twig that has also a website with tutarials and docs.

    L*

    There are no rules, there are no thumbs..
    Reinvent the wheel, then learn The Wheel; may be one day you reinvent one of THE WHEELS.
[reply]
[d/l]
[select]
      Ty very much for the warm welcome!! :) Thanks for sharing, I will try them & post the results.
[reply]
Re: parsing & retrieving from an XML file
by 1nickt (Canon) on Nov 06, 2017 at 17:22 UTC

    Hi, welcome, so does the code accomplish the 1st point successfully?

    Regarding the remaining points, what is the question?

    Regarding your choice of XML-handling module: I'm not familiar with XML::Smart but I see its last release was in 2013 and it has some unaddressed issues; this would lead me to be cautious and possibly consider a different module.


    The way forward always starts with a minimal test.
[reply]
      Ty very much for the reply. Yes, first point does the needful. However, I know there is an easy way out there..just wanted to know. I have got few wonderful replies. I will try them & post the results.
[reply]
Re: parsing & retrieving from an XML file
by tangent (Parson) on Nov 06, 2017 at 20:57 UTC
    Here is a way to extract all the necessary info from your XML using XML::LibXML and simple xpath queries. You will probably want to save the results to a data structure (an array of hashes should work) rather than printing them out like I do, but printing them out first allows you to see if it is working correctly.

    See XML::LibXML::Node and XML::LibXML::Element for more details.

    use XML::LibXML;

my $doc = XML::LibXML->load_xml(location => '/path/to/file.xml');

my @envdetails = $doc->findnodes('//ENVDETAILS');

for my $envdetail (@envdetails) {
    my $id = $envdetail->getAttribute('id');
    print "id: $id\n";
    my $dmgrhost = $envdetail->findvalue('DMGRHOST');
    print "DMGRHOST: $dmgrhost\n";
    my @nodes = $envdetail->findnodes('NODE');
    for my $node (@nodes) {
        my $nodeid = $node->getAttribute('nodeid');
        print "nodeid: $nodeid\n";
        my $nodehost = $node->findvalue('NODEHOST');
        print "NODEHOST: $nodehost\n";
        my @jvms = $node->findnodes('JVM');
        for my $jvm (@jvms) {
            my $jvmtype = $jvm->getAttribute('jvmtype');
            if ($jvmtype eq 'api') {
                my @jvmnames = $jvm->findnodes('JVMNAME');
                for my $jvmname (@jvmnames) {
                    my $name = $jvmname->textContent;
                    print "JVMNAME: $name\n";
                }
            }
        }
    }
}

# OUTPUT

id: abc
DMGRHOST: a.b.c.d
nodeid: 1
NODEHOST: v.x.y.z
JVMNAME: abc_api1.1
JVMNAME: abc_api1.2

    [download]
[reply]
[d/l]
      Thanks a lot for giving it a try!! Appreciate it. I will try & post the results. What I understood is that I need to create my own data structure to save & retrieve it.
[reply]
Re: parsing & retrieving from an XML file
by sbafna (Novice) on Nov 08, 2017 at 09:03 UTC

    I have written the code using XML::Smart, as perl in my box is not an updated one. Also, I have tweaked the XML a little bit.

    XML:

    <OUTMOST>
        <MYENVDETAILS>
                <ENVDETAILS id="abc" >
                        <DMGRHOST>a.b.c.d</DMGRHOST>
                        <CELL>XYZ</CELL>
                        <NODEDETAIL nodeid="1" NODEHOST="v.x.y.z" >
                                <IHSNODE>v_ihsnode</IHSNODE>
                                <IHSJVM ihsjvmtype="log_ihs" >
                                        <IHSJVMNAME>abc_logihs1</IHSJV
+MNAME>
                                </IHSJVM>
                                <JVM jvmtype="api" >
                                        <JVMNAME>abc_api1.1</JVMNAME>
                                        <JVMNAME>abc_api1.2</JVMNAME>
                                </JVM>
                        </NODEDETAIL>
                </ENVDETAILS>
        </MYENVDETAILS>
</OUTMOST>

    [download]

    Execution

    ----- myaccount  v.x.y.z /appl/ -----
$ ./my_topologyParser.pl fetch
abc
----- myaccount  v.x.y.z /appl/ -----
$ ./my_topologyParser.pl nodesall
v.x.y.z
----- myaccount  v.x.y.z /appl/ -----
$ ./my_topologyParser.pl abc nodewrtseg
v.x.y.z
----- myaccount  v.x.y.z /appl/ -----
$ ./my_topologyParser.pl abc jvmtype api
abc_api1.1
abc_api1.2
----- myaccount  v.x.y.z /appl/ -----
$ ./my_topologyParser.pl abc jvmwrtnode v.x.y.z api
abc_api1.1
abc_api1.2
----- myaccount  v.x.y.z /appl/ -----
$ ./my_topologyParser.pl abc nodewrtjvm abc_api1.2
v.x.y.z
----- myaccount  v.x.y.z /appl/ -----

    [download]

    Actual Code

    #!/appl/perl/bin/perl

use XML::Smart;
use strict;
use warnings;

$ENVFILE="/appl/test_MYENV_Top.xml";

my $ENVxml = XML::Smart->new($ENVFILE) || die ("Could not find file.")
+;

my $env = $ARGV[0];

if ( $ARGV[0] eq "fetch" ) {
        my @envname=$ENVxml->{OUTMOST}->{MYENVDETAILS}->{ENVDETAILS}('
+[@]','id');
        foreach my $i (@envname) {
                print "$i\n";
}
}
elsif ($ARGV[0] eq "nodesall"){
        my @envname=$ENVxml->{OUTMOST}->{MYENVDETAILS}->{ENVDETAILS}('
+[@]','id');
        foreach my $i (@envname) {
                my @nodename=$ENVxml->{OUTMOST}->{MYENVDETAILS}->{ENVD
+ETAILS}->('id','eq',"$i")->{NODEDETAIL}('[@]','NODEHOST');
                foreach my $c (@nodename) {
                        print "$c\n";
                }
        }
}
elsif ($ARGV[1] eq "nodewrtseg"){
        my @nodename=$ENVxml->{OUTMOST}->{MYENVDETAILS}->{ENVDETAILS}-
+>('id','eq',"$env")->{NODEDETAIL}('[@]','NODEHOST');
        foreach my $i (@nodename) {
                print "$i\n";
        }
}
elsif ($ARGV[1] eq "jvmtype") {
        my $typeCheck=$ARGV[2];
        my @jvmlist;
        my @nodename=$ENVxml->{OUTMOST}->{MYENVDETAILS}->{ENVDETAILS}-
+>('id','eq',"$env")->{NODEDETAIL}('[@]','NODEHOST');
        if ($#nodename < 0){
                exit 1;
        }
        foreach my $i (@nodename) {
                my @names=@{$ENVxml->{OUTMOST}->{MYENVDETAILS}->{ENVDE
+TAILS}->('id','eq',"$env")->{NODEDETAIL}->('NODEHOST','eq',"$i")->{JV
+M}->('jvmtype','eq',"$typeCheck")->{JVMNAME}};
                if ($#names > 0){
                        push(@jvmlist,@names);
                }
        }
        @jvmlist = grep /\S/, @jvmlist;
        if ($#jvmlist < 0){
                print " No $typeCheck type jvm found on $env . Incorre
+ct jvm type passed.\n";
        exit 1;
        }
        foreach my $c (@jvmlist) {
                print "$c\n";
        }
}
elsif ($ARGV[1] eq "jvmwrtnode") {
        my $nodeCheck=$ARGV[2];
        my $typeCheck=$ARGV[3];
        if (defined $nodeCheck && defined $typeCheck){
                my @names=@{$ENVxml->{OUTMOST}->{MYENVDETAILS}->{ENVDE
+TAILS}->('id','eq',"$env")->{NODEDETAIL}->('NODEHOST','eq',"$nodeChec
+k")->{JVM}->('jvmtype','eq',"$typeCheck")->{JVMNAME}};
                if ($#names < 0){
                        print " No $typeCheck type jvm found on $nodeC
+heck node. Incorrect type/nodename passed.\n";
                        exit 1;
                }
                @names = grep /\S/, @names;
                foreach my $c (@names) {
                        print "$c\n";
                }
        }
        else
        {
                print " No Node type or nodename passed. Incorrect no.
+ of arguments passed.\n";
                exit 1;
        }
}
elsif ($ARGV[1] eq "nodewrtjvm") {
        my $jvmCheck=$ARGV[2];
        if (defined $jvmCheck){
                my @nodename=$ENVxml->{OUTMOST}->{MYENVDETAILS}->{ENVD
+ETAILS}->('id','eq',"$env")->{NODEDETAIL}('[@]','NODEHOST');
                if ($#nodename < 0){
                        exit 1;
                }
               OUTER: foreach my $i (@nodename) {
                        my @jvmtypes=$ENVxml->{OUTMOST}->{MYENVDETAILS
+}->{ENVDETAILS}->('id','eq',"$env")->{NODEDETAIL}('NODEHOST','eq',"$i
+")->{JVM}('[@]','jvmtype');
                        foreach my $k (@jvmtypes) {
                                my @names = @{$ENVxml->{OUTMOST}->{MYE
+NVDETAILS}->{ENVDETAILS}->('id','eq',"$env")->{NODEDETAIL}->('NODEHOS
+T','eq',"$i")->{JVM}('jvmtype','eq',"$k")->{JVMNAME}};
                                if ($#names >= 0){
                                        foreach (@names){
                                         if ( $_ eq $jvmCheck ){
                                                print "$i\n";
                                                last OUTER;
                                                }
                                        }
                                }
                        }
                }
        }
}
else{
        print "\nThis parser is argument specific.";
        print "\n1). $0 fetch : To list out environments";
        print "\n2). $0 nodesall : To list all nodes of all environmen
+ts.";
        print "\n3). $0 <segID> nodewrtseg : to list all nodes of spec
+ific environment.";
        print "\n4). $0 <segID> jvmtype <typeofjvm> : to list all node
+s having specified jvmtypes of the specific environment.";
        print "\n5). $0 <segID> jvmwrtnode <nodename> <typeofjvm> : to
+ list all jvms present on specific node of specific type of the envir
+onment.";
        print "\n6). $0 <segID> nodewrtjvm <jvmname> : to list the nod
+e of specific jvm of specified environment.\n";
}

    [download]

    I know I could have used getopts, however, I was in a hurry :)

[reply]
[d/l]
[select]

Back to Seekers of Perl Wisdom