Question

For example, from the following string

<?xml version="1.0"?><root><point><message>hello world 1</message></point><point><data><message>hello world 2</message></data></point></root>

if I want to extract message, the result should be

hello world 1
hello world 2

Is there an easy way to do this?

All I can think of is to first find out the position of and and then generate substrings in a loop. Is there a better way?

Answer 1

Your data is not XML, so I guess you’ll have to use a regular expression for that:

perl -n -E'say $1 while m{<message>(.*?)</message>}g' your_file_here.xml 

If your file was proper XML, then XML::Twig would work nicely. You could even use the xml_grep tool that comes with it to do just what you want.

update: with valid XML you can then do

xml_grep --text_only message mes.xml 

or

xml_grep2 --text_only '//message' mes.xml # xml_grep2 is in App::xml_grep2

or

perl -MXML::Twig -E'XML::Twig->new( twig_handlers => 
                                      { message => sub { say $_->text; }, })
                             ->parsefile( "mes.xml")'