It was a problem with the record initialisation code and the
introspect method not being handled correctly. Fixed in:
37a59389bf65d4c558ff906427d52fb6fae5d912
Thanks to Ross for the help :-)
Cheers, Tim
On Feb 17, 6:11 pm, Ross Mellgren <dri...@gmail.com> wrote:
> It is definitely the new. I thought that it was always this way because I
> observed it (I thought) before I made those changes. I just reproduced your
> exact case:
>
> val rec1 = MyMetaRecord.createRecord
> val res1 = rec1.foo("zippy").fields.map(_.name)
> val rec2 = new MyRecord()
> val res2 = rec2.foo("zippy").fields.map(_.name)
>
> <div>{ String.valueOf(res1) }</div> ++ <div>{ String.valueOf(res2)
> }</div>
>
> Result is:
>
> List(foo)
> List(null)
>
> So createRecord is doing the introspection. I'll start by reverting out my
> change to record and see if that makes the problem go away, to narrow it down.
>
> Do you have IM? We could cycle faster.
>
> -Ross
>
> On Feb 17, 2010, at 1:02 PM, Timothy Perrett wrote:
>
>
>
> > The field in question is a subtype of:
>
> > sealed abstract class TextADORField[OwnerType <: Record[OwnerType]]
> > (rec: OwnerType)
> > extends ADORField[OwnerType, String](rec){
> > def setFromString(s: String): Box[String] = {
> > dirty_?(true)
> > Full(set(s))
> > }
>
> > def setFromAny(in: Any): Box[String] = {
> > dirty_?(true)
> > in match {
> > case seq: Seq[_] if !seq.isEmpty => seq.map(setFromAny)(0)
> > case (s: String) :: _ => Full(set(s))
> > case null => Full(set(null))
> > case s: String => Full(set(s))
> > case Some(s: String) => Full(set(s))
> > case Full(s: String) => Full(set(s))
> > case None | Empty | Failure(_, _, _) => Full(set(null))
> > case o => Full(this.set(o.toString))
> > }
> > }
>
> > def defaultValue = ""
> > }
>
> > As you say, they should both go to setBox... Although, look at this:
>
> > scala> Visitor.createRecord.fields.map(_.name)
> > res25: List[String] = List(email, last_name, first_name)
>
> > scala> new Visitor().fields.map(_.name)
> > res27: List[String] = List(null, null, null)
>
> > That never used to be the case, you used to be able to call new and
> > still get the introspection.... any ideas?
>
> > Cheers, Tim
>
> > On Feb 17, 5:52 pm, Ross Mellgren <dri...@gmail.com> wrote:
> >> Name should be set by introspect during the record initialization, so I'm
> >> not sure why it would matter. Also, setFromString and set should both
> >> degenerate to setBox, which should do the actual work. Can you show your
> >> setFromString implementation? Are you overriding anything other than
> >> setFromAny and setFromString in the set department?
>
> >> -Ross
>
> >> On Feb 17, 2010, at 12:51 PM, Timothy Perrett wrote:
>
> >>> Not a generic one - to reproduce it you would need my custom record
> >>> implementation and access to internal work servers (neither of which I
> >>> can give out).
>
> >>> I have just realised however that the "name" field is present and
> >>> correct when the field value is set:
>
> >>> field.setFromString
>
> >>> So it appears that its just a problem with set(<param>) ?
>
> >>> Cheers, Tim
>
> >>> On Feb 17, 5:46 pm, Ross Mellgren <dri...@gmail.com> wrote:
> >>>> I'm starting to look at this. Do you have a repro case?
>
> >>>> -Ross
>
> >>>> On Feb 17, 2010, at 12:41 PM, Timothy Perrett wrote:
>
> >>>>> Guys,
>
> >>>>> Something strange is happening in Lift Record now - I have production
> >>>>> work that for some reason is now not behaving the way it should. The
> >>>>> custom record implementation that I have relies on knowing the field
> >>>>> name, and for some reason, it appears to not be being set:
>
> >>>>> (where Visitor is an example implementation of my Record):
>
> >>>>> scala> new Visitor().email("dfsfdsf").fields.map(_.name)
> >>>>> TRACE - Set the value of 'null' with 'dfsfdsf' and marked as dirty.
> >>>>> (true)
>
> >>>>> I added some tracing and I see that the "def name" is always null
> >>>>> now... this is a major problem for me. My field apply method looks
> >>>>> like:
>
> >>>>> override def apply(in: FieldValueType): OwnerType = if
> >>>>> (owner.meta.mutable_?){
> >>>>> set(in)
> >>>>> dirty_?(true)
> >>>>> Log.trace("Set the value of '"+name+"' with '"+in.toString+"' and
> >>>>> marked as dirty. ("+dirty_?.toString+")")
> >>>>> owner
> >>>>> } else {
> >>>>> owner.meta.createWithMutableField(owner, this, Box.!!(in))
> >>>>> }
>
> >>>>> Im really not sure what has broken this, but i've not changed my code
> >>>>> other than adding the legacy check to accommodate ross' latest
> >>>>> changes.
>
> >>>>> I URGENTLY need a fix for this - we have production code relying on
> >>>>> this.
>
> >>>>> Cheers, Tim
>
> >>>>> --
> >>>>> You received this message because you are subscribed to the Google
> >>>>> Groups "Lift" group.
> >>>>> To post to this group, send email to lift...@googlegroups.com.
> >>>>> To unsubscribe from this group, send email to
> >>>>> liftweb+unsubscr...@googlegroups.com.
> >>>>> For more options, visit this group
> >>>>> athttp://groups.google.com/group/liftweb?hl=en.
>
> >>> --
> >>> You received this message because you are subscribed to the Google Groups
> >>> "Lift" group.
> >>> To post to this group, send email to lift...@googlegroups.com.
> >>> To unsubscribe from this group, send email to
> >>> liftweb+unsubscr...@googlegroups.com.
> >>> For more options, visit this group
> >>> athttp://groups.google.com/group/liftweb?hl=en.
>
> > --
> > You received this message because you are subscribed to the Google Groups
> > "Lift" group.
> > To post to this group, send email to lift...@googlegroups.com.
> > To unsubscribe from this group, send email to
> > liftweb+unsubscr...@googlegroups.com.
> > For more options, visit this group
> > athttp://groups.google.com/group/liftweb?hl=en.
--
You received this message because you are subscribed to the Google Groups
"Lift" group.
To post to this group, send email to lift...@googlegroups.com.
To unsubscribe from this group, send email to
liftweb+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/liftweb?hl=en.
