On the first test, its a bug in the activesupport code:
def to_hash(hash={})
if text?
raise LibXML::XML::Error if content.length >= LIB_XML_LIMIT
hash[CONTENT_ROOT] = content # !!!!! INCORRECT !!!!!
else
This is the xml:
<root>
good
<products>
hello everyone
</products>
morning
</root>
Look at the sequence of nodes as they are processed:
root
text (child)
element (products)
text (hello everyone)
text (morning)
When the text node (morning) is processed:
hash[CONTENT_ROOT] = content
Thus overwriting the text "good" with "morning"
You want something more like this:
hash[CONTENT_ROOT] = (hash[CONTENT_ROOT] || '') + content
Charlie
smime.p7s
Description: S/MIME Cryptographic Signature
_______________________________________________ libxml-devel mailing list libxml-devel@rubyforge.org http://rubyforge.org/mailman/listinfo/libxml-devel
