Re: Passing a list as a subroutine's return value ( list of HTML::Element objects )

Replies are listed 'Best First'.
Re^2: Passing a list as a subroutine's return value ( list of HTML::Element objects ) by ikegami (Patriarch) on Sep 27, 2023 at 13:53 UTC
Not quite that simple, cause you now have a memory leak. The entire tree is never freed.	[reply]
Re^2: Passing a list as a subroutine's return value ( list of HTML::Element objects ) by choroba (Cardinal) on Sep 27, 2023 at 09:16 UTC
Is it enough? Declaring $tmp outside the unescape_entitites sub makes the output even larger and identical to the expected one. So it seems you need to have the HTML::TreeBuilder::XPath object around to be able to access the nodes it matched. Update: it seems the difference is the _parent contents which just contains the XPath internals. Disregard this node. `map{substr$_->[0],$_->[1]\|\|0,1}[\\|\|{},3],[[]],[ref qr-1,-,-1],[{}],[sub{}^ARGV,3]`	[reply] [d/l]
Re^3: Passing a list as a subroutine's return value ( list of HTML::Element objects ) by tobyink (Canon) on Sep 27, 2023 at 15:37 UTC
Yeah, the `_parent` is a weak reference, so becomes undef after `$tmp` goes out of scope. Hire me at Toby Ink Ltd or Join my OnlyFans	[reply] [d/l] [select]
Re^2: Passing a list as a subroutine's return value ( list of HTML::Element objects ) by mldvx4 (Friar) on Sep 27, 2023 at 10:21 UTC
Thanks. How can one tell when a 'new' variable is somehow connected to another?	[reply]
Re^3: Passing a list as a subroutine's return value ( list of HTML::Element objects ) by ikegami (Patriarch) on Sep 27, 2023 at 14:01 UTC
In general, you don't have to worry about that. As long as you hold a reference, the object will be kept alive. But you called a function which frees the tree including all of its nodes, which it does by hollowing out any existing references to its nodes.	[reply]