Regex: keep everything in <profession></profession> tags

Question

I have a large XML, looking like this:

<gender>M</gender>
<last-name>*</last-name>
<profession>2165dda2-dc59-41af-acb5-06d8914c4841</profession>
<first-name>*</first-name>
<mail-confirmation>1</mail-confirmation>
<fax-confirmation>1</fax-confirmation>

I only want to keep the tags. I found a way to search IN the tag, like this:

<profession[^>]*>([^<]*?)</profession>

but how do I search everything outside of it? I tried to just flip it, like:

</profession[^>]*>([^<]*?)<profession>

or

</profession>([^<]*?)<profession[^>]*>

but that won't work.

Answer 1

What about

# Perl
$xml =~ s/^<profession>.*<\/profession>$/<profession><\/profession>/m;

Just make sure to use the multiline modifier.

Answer 2

Don't use regular expressions to parse XML. Use an XML parser:

#!/usr/bin/perl

use strict; use warnings;

use XML::LibXML::Reader;

my $reader = XML::LibXML::Reader->new(IO => \*DATA)
    or die "Cannot read from \\*DATA\n";

while ( $reader->read ) {
    print $reader->readInnerXml if $reader->localName eq 'profession';
}

$reader->finish;
print $reader->document->toString(1);

__DATA__
<person>
<gender>M</gender>
<last-name>*</last-name>
<profession>2165dda2-dc59-41af-acb5-06d8914c4841</profession>
<first-name>*</first-name>
<mail-confirmation>1</mail-confirmation>
<fax-confirmation>1</fax-confirmation>
</person>

Output:

C:\Temp> t
2165dda2-dc59-41af-acb5-06d8914c4841

See XML::LibXML::Reader.

Answer 3

Strictly you can't parse XML with a regex.

Quick and dirty solution with sed is to grep the lines with profession then replace "profession" and "/profession" with "" (markup is stripping the < > )