Chris, Andreas, others LocalFS product (1.3-andreas) seems to be having a problem with the <dtml-tree> tag in Zope 2.9.2/Python 2.4.2. The <dtml-tree> behavior is very different for a folder with three nested sub-folders (a list of the three nested sub-folders) that what is seen when pointed to a LocalFS with three nested subdirectories, even though they should be similar or the same. The problem seems to be with the tpValues() program in the LocalFS product which returns an empty list. The controlling flag, o.isPrincipiaFolderish, seems to be set to zero when it ought to have a value of 1 when the object being referenced is a LocalDirectory instance. As a result, instead of returning a list of sub-directories it returns an empty list. There seems to be some trickery going on with the isPrincipiaFolderish and the tree_view attributes, but I don't quite get what's supposed to happen. SmileyChris? I think these were your mods. Incidentally, the ZMI panel which uses <dtml-tree> seems to have the same sort of problems I am seeing when there is a LocalFS. Has anyone else encountered this problem? I plan to move to the 1.7-rc1 version, but there does not appear to be a fix for a bug of this sort in this latest release.
Dennis Allison wrote:
LocalFS product (1.3-andreas) seems to be having a problem with the <dtml-tree> tag in Zope 2.9.2/Python 2.4.2.
Then don't use DTML, it does, after all suck (tm) ;-)
The problem seems to be with the tpValues() program in the LocalFS product which returns an empty list.
then tpValues() is likely wrong for localfs, unless localfs doesn't want to play ball with the tree tag. Given that the tree tag is used by the zmi, and you don't really want to display potentially millions of sub-objects in the zmi, it might be perfectly legitimate for local-fs to do what it's doing...
There seems to be some trickery going on with the isPrincipiaFolderish and the tree_view attributes, but I don't quite get what's supposed to happen. SmileyChris? I think these were your mods.
Not me, I don't smile, but you knew that already ;-)
Incidentally, the ZMI panel which uses <dtml-tree> seems to have the same sort of problems I am seeing when there is a LocalFS.
See my comments above. I'd roll your own code without using dtml and just iterate over the objectIds method... cheers, Chris -- Simplistix - Content Management, Zope & Python Consulting - http://www.simplistix.co.uk
Or just do a custom tpValues equivalent for the particular special case. I suspect that is what we will try next... ;-) On Mon, 8 May 2006, Chris Withers wrote:
Dennis Allison wrote:
LocalFS product (1.3-andreas) seems to be having a problem with the <dtml-tree> tag in Zope 2.9.2/Python 2.4.2.
Then don't use DTML, it does, after all suck (tm) ;-)
The problem seems to be with the tpValues() program in the LocalFS product which returns an empty list.
then tpValues() is likely wrong for localfs, unless localfs doesn't want to play ball with the tree tag. Given that the tree tag is used by the zmi, and you don't really want to display potentially millions of sub-objects in the zmi, it might be perfectly legitimate for local-fs to do what it's doing...
There seems to be some trickery going on with the isPrincipiaFolderish and the tree_view attributes, but I don't quite get what's supposed to happen. SmileyChris? I think these were your mods.
Not me, I don't smile, but you knew that already ;-)
Incidentally, the ZMI panel which uses <dtml-tree> seems to have the same sort of problems I am seeing when there is a LocalFS.
See my comments above.
I'd roll your own code without using dtml and just iterate over the objectIds method...
cheers,
Chris
--
Dennis Allison wrote at 2006-5-7 19:21 -0700:
... The problem seems to be with the tpValues() program in the LocalFS product which returns an empty list. The controlling flag, o.isPrincipiaFolderish, seems to be set to zero when it ought to have a value of 1 when the object being referenced is a LocalDirectory instance. As a result, instead of returning a list of sub-directories it returns an empty list.
A good analysis. Thus, you change what is necessary to let "tpValues" behave well ... -- Dieter
participants (3)
-
Chris Withers -
Dennis Allison -
Dieter Maurer