I found an odd misnomer/bug on PM. Take a node, view the xml of it, not formated in mozilla, but the actual XML loaded from the site. Values are stored between tags, but on PM, it seems to start with an "\n" for the values.

Here's an example of a random one. (Try it on this one too).

<?xml version="1.0" encoding="ISO-8859-1"?>
<node id="108933" title="Re: Re: Re: Convenient way to track function 
+calls?" created="2001-08-29 22:29:46" updated="">
<type id="11">
note</type>
<author id="12209">
clintp</author>
<data>
<field name="doctext">
That's exactly what I was suggesting.  I was in a bit of a rush and yo
+u noticed the figurative handwaving as I went out the door.  Nice job
+ of interpretation.  :)&lt;p&gt;

You might want a wantarray thrown in there somewhere to handle functio
+n calls in different contexts, if that's important.&lt;p&gt;

The other suggestion I have is use the Devel::Prof module and simply c
+ook the output from the dprofpp.  I think that assumes you have a Per
+l built for debugging though.

</field>
<field name="root_node">
108838</field>
<field name="parent_node">
108849</field>
</data>
</node>
[download]

I pasted this verbatim.

See how the values for <field>...</field> has \n's in them? It gets parsed in with the value... so the node type is actually \nnote. For the field with name of doctext, it makes sense, as someone hit entre somewhere. Otherwise, \n's need to be filtered out of the values every time. I'm sure it's not stored in the DB like that.

Just a heads up.