view release on metacpan or  search on metacpan

lib/XML/TreePP/XMLPath.pm  view on Meta::CPAN

Second, parse the elements from step 1 that have key/val pairs, such that
each single key/val is contained by the [ and ] characters

    my $el = charlexsplit (
        string        => q( path[@key='val'][@key2='val2'] ),
        boundry_start => '[',
        boundry_stop  => ']',
        tokens        => [qw( ' ' " " )],
        boundry_begin => 0,
        boundry_end   => 0
        );
    print Dumper( $el );

Output:

    ["\@key='val'", "\@key2='val2'"]

Step 3:

Third, parse the elements from step 2 that is a single key/val, the single
key/val is delimited by the = character

    my $el = charlexsplit (
        string        => q{ @key='val' },
        boundry_start => '=',
        boundry_stop  => '=',
        tokens        => [qw( ' ' " " )],
        boundry_begin => 1,
        boundry_end   => 1
        );
    print Dumper( $el );

Output:

    ["\@key", "'val'"]

Note that in each example the C<tokens> represent a group of escaped characters
which, when analyzed, will be collected as part of an element, but will not be
allowed to match any starting or stopping boundry.

So if you have a start token without a stop token, you will get undesired
results. This example demonstrate this data error.

    my $el = charlexsplit   (
        string        => q{ path[@key='val'][@key2=val2'] },
        boundry_start => '[',
        boundry_stop  => ']',
        tokens        => [qw( ' ' " " )],
        boundry_begin => 0,
        boundry_end   => 0
        );
    print Dumper( $el );

Undesired output:

    ["\@key='val'"]

In this example of bad data being parsed, the C<boundry_stop> character C<]> was
never matched for the C<key2=val2> element.

And there is no error message. The charlexsplit method throws away the second
element silently due to the token start and stop mismatch.

=head2 Method: parseXMLPath

    use XML::TreePP::XMLPath qw(parseXMLPath);
    use Data::Dumper;
    
    my $parsedPath = parseXMLPath(
                                  q{abcdefg/xyz/path[@key1='val1'][key2='val2']/last}
                                  );
    print Dumper ( $parsedPath );

Output:

    [
      ["abcdefg", undef],
      ["xyz", undef],
      ["path", [["-key1", "val1"], ["key2", "val2"]]],
      ["last", undef],
    ]

=head2 Method: filterXMLDoc

Filtering an XML Document, using an XMLPath, to find a node within the
document.

    #!/usr/bin/perl
    use XML::TreePP;
    use XML::TreePP::XMLPath qw(filterXMLDoc);
    use Data::Dumper;
    #
    # The XML document data
    my $xmldata=<<XMLEND;
        <level1>
            <level2>
                <level3 attr1="val1" attr2="val2">
                    <attr3>val3</attr3>
                    <attr4/>
                    <attrX>one</attrX>
                    <attrX>two</attrX>
                    <attrX>three</attrX>
                </level3>
                <level3 attr1="valOne"/>
            </level2>
        </level1>
    XMLEND
    #
    # Parse the XML document.
    my $tpp = new XML::TreePP;
    my $xmldoc = $tpp->parse($xmldata);
    print "Output Test #1\n";
    print Dumper( $xmldoc );
    #
    # Retrieve the sub tree of the XML document at path "level1/level2"
    my $xmlSubTree = filterXMLDoc($xmldoc, 'level1/level2');
    print "Output Test #2\n";
    print Dumper( $xmlSubTree );
    #
    # Retrieve the sub tree of the XML document at path "level1/level2/level3[@attr1='val1']"
    my $xmlSubTree = filterXMLDoc($xmldoc, 'level1/level2/level3[@attr1="val1"]');