If I understand correctly you want to extract the text and its enclosing tag if it has one, and just the text if it doesn't. This might help:

use XML::LibXML;

my $xml = q|
<seg>
    <foo mid="0" mtype="seg">
        <g id="1">Need to export this text</g>
    </foo>
    <foo mid="1" mtype="seg">
        Need to export this text also
    </foo>
</seg>|;

my $doc = XML::LibXML->load_xml(string => $xml);
my @foos = $doc->findnodes('//foo');

for my $foo (@foos) {
    my $mid = $foo->getAttribute('mid');
    print "mid: $mid ";
    my @childnodes = $foo->childNodes();
    if (@childnodes) {
        for my $node (@childnodes) {
            print $node->toString, "\n";
        }
    }
    else {
        print $foo->textContent, "\n";
    }
}
[download]

Output:

mid: 0 
<g id="1">Need to export this text</g>

mid: 1 
Need to export this text also
[download]

You may need to trim leading and trailing whitespace if your XML contains it.