For some reason, iframes seem to be omitted when I use XPatch on a page I've fetched. If I uncheck the "Emit items as a string" option, I can see the iframe text so it's definitely being fetched. Is there a way to avoid this problem?
We sanitize feeds due to potential malicious objects (re: activeX and others) and push those tags into an iframe.
This is done for safety of our users who view the output. Feeds that have enclosure nodes will have the video embedded in various readers.
So this seems to suggest iframes are fine so they shouldn't be stripped out...
I encountered the same kind of problem before, and I don't think it happens only for iframes. Basically, I think it's not filtered out, but just interpreted, as formatting or whatever, and thus disappear from the content.
In his first message, Antlane says he does succeed in fetching the iframe content while not checking the string output option: thus the html markup is stored as tree branches and not actual html markup which could be interpreted.
Hence my answer to your problem would be (if I'm not mistaken so far): work with your content as a tree of data instead of trying to work with it as a big fat string.