[prev in list] [next in list] [prev in thread] [next in thread] 

List:       mozilla-rdf
Subject:    Re: RDF Generic Builder change re: containment and ignore attributes
From:       Robert John Churchill <rjc () netscape ! com>
Date:       1999-12-03 2:10:57
[Download RAW message or body]

Yeah, I left the code in because I didn't want to prevent someone from
doing:

<tree containment="foo" datasources="rdf:blah">
  <treechildren>
      <treeitem id="container1" containment="containmentprop1"
ignore="ignoreprop1" container="true" />
      <treeitem id="container2" containment="containmentprop2"
ignore="ignoreprop2" container="true" />
  </treechildren>
</tree>

Granted, none of our current trees do this, so perhaps it isn't worth
keeping in.

I just figured that since we haven't identified this as a performance hit,
there was no real cost to leaving it in, especially since the "need
performance improvements" areas such as mail/news do set containment on the
root node, so they hit the optimal case.

Robert


Chris Waterson wrote:

> Robert, don't bother with the fallback case. Just look on the root, and
> change the addressbook code.
>
> If you leave the fallback case in, you'll get bizarre results when you
> do:
>
> <tree containment="foo" datasources="rdf:blah">
>   <treechildren containment="bar">
>     <!-- oops, in here, "foo" will be the containment property, not
> "bar" -->
>   </treechildren>
> </tree>

[prev in list] [next in list] [prev in thread] [next in thread]