Confirmed users
146
edits
DOM/ContentIterator Issues: Difference between revisions
Jump to navigation
Jump to search
m
→Potential Solutions
mNo edit summary |
|||
| Line 36: | Line 36: | ||
We know the node types which are likely to give us trouble. For instance, we could create a mCommonDocument value to hold a reference to a document, if that were the common ancestor. This may lead to memory bloat, however (mCommonDocument, mCommonAttribute, mCommonEntityRef, mCommonDocumentFragment, and who knows what else we'll need pointers for, most of which will be null). It's also not very scalable. | We know the node types which are likely to give us trouble. For instance, we could create a mCommonDocument value to hold a reference to a document, if that were the common ancestor. This may lead to memory bloat, however (mCommonDocument, mCommonAttribute, mCommonEntityRef, mCommonDocumentFragment, and who knows what else we'll need pointers for, most of which will be null). It's also not very scalable. | ||
=== Replace usage of nsIContent with nsINode === | |||
<pre> | |||
mrbkap nsINode | |||
WeirdAl mrbkap: dump nsIContent for nsINode in that case? | |||
mrbkap Yeah. | |||
WeirdAl mrbkap: I'll need to think about that. I like the idea -- DOM stuff sometimes relies a little too much on content, I've sensed | |||
Pike which is a perf decision | |||
</pre> | |||
=== Your Ideas Welcome Here === | === Your Ideas Welcome Here === | ||