Remove bogus tree setting code from parentnode/tree.c

[nielsdos]

I don't know why this code was here in the first place, it is present since the initial implementation. It doesn't make sense because:

1. It would require updating the refcounts if the document wasn't actually already set.
2. We enforce that the document is the same as the target document by this point, so setting the tree is pointless.

[SakiTakamachi]

@nielsdos

It would require updating the refcounts if the document wasn't actually already set.

I'm not very familiar with DOM, but where is this processing?
(If someone else with more knowledge reviews it, you can forget about this question.)

[nielsdos]

@nielsdos

It would require updating the refcounts if the document wasn't actually already set.

I'm not very familiar with DOM, but where is this processing? (If someone else with more knowledge reviews it, you can forget about this question.)

You would need to call php_libxml_increment_doc_ref to increment the ref count of the document reference, see this code for example:

php-src/ext/dom/element.c

Lines 733 to 736 in ba54ceb

	if (attrp->doc == NULL && nodep->doc != NULL) {
	attrobj->document = intern->document;
	php_libxml_increment_doc_ref((php_libxml_node_object *)attrobj, NULL);
	}

[SakiTakamachi]

@nielsdos

Thanks!

In other words, does it mean that php_libxml_increment_doc_ref is always called before this function is called?

[nielsdos]

@nielsdos

Thanks!

In other words, does it mean that php_libxml_increment_doc_ref is always called before this function is called?

Not necessarily right before this function, but it must've been called at some point. Usually this is at the time that the node object was created, but it could also have happened during adoption. If this wasn't called then a WRONG_DOCUMENT_ERR type of DOMException is thrown.

[SakiTakamachi]

@nielsdos

Thank you for the explanation, I understand.

I was concerned that the execution path was a little complicated, but since use ZEND_ASSERT, I don't think it will be a problem.