Re: [Zope] - Document modification time
At 03:36 PM 12/7/98, Kent wrote:
I did this, and it doesn't update if I change the index_html document at all. In fact, it actually seems to have the date/time of the update of standard_html_footer, which isn't terribly useful :-)
Actually, it depends... play with it a bit more and you will see even more interesting behavior. :^) Part of the 'problem' is that it is often hard to 'set' or determine what the object is that you are dealing with. Is it an object or is it a method? Depends... If you think this is interesting, just you wait. :^) :^) These sorts of issues will help determine how you build your site.
...
Am I asking the impossible? Also, it'd be nice to be able to figure out the last update time on a StructuredText document that was included with <!--#var blah fmt="structured-text"-->. I'd be happy with a solution that could scan through all the objects included and figure out which one was most recent.
I agree that this is a bit of a problem. When you have Documents calling Documents its sometimes hard to keep things straight. Acquisition is part of the confusion, but I don't really think its the issue here. Basically you have an object docA calling another docB. The docB is told to execute a method that both it and docA implement. So it calls its own method. Really you would like to specify that it should call the method on DocA. This is hard to do, if you just specify the method name. I propose that there should be something analogous to the PARENTS list for Documents. Maybe is should be called CALLERS or DOCUMENTS or something. It should be a list of Documents, so that when index_html calls standard_html_header which calls mydoc, mydoc should be able do something like: <!--#var "CALLERS[-1].bobobase_modification_time()"--> To call a method on index_html (the first CALLER). Thoughts? -Amos P.S. This facility would make the view dtml source widget on the Zope site work, as well :-)
Hi Amos (Amos Latteier), in <3.0.5.32.19981207142024.00adf3e0@mail.aracnet.com> on Dec 07 you wrote:
I propose that there should be something analogous to the PARENTS list for Documents. Maybe is should be called CALLERS or DOCUMENTS or something. It should be a list of Documents, so that when index_html calls standard_html_header which calls mydoc, mydoc should be able do something like:
<!--#var "CALLERS[-1].bobobase_modification_time()"--> To call a method on index_html (the first CALLER).
Thoughts?
Close... What about a parent document which lists child documents? I need to be able to query the list of child objects which are referenced in the parent document and find the latest date. And I'd like to be able to extend that behavior to things like like sql query results so they can contribute their 'modification date'. And no, I don't want to discuss issues about walking the tree right now. :^) In the past, I have tried to examine subobjects in this manner in order to produce an examination result and have met with varying degrees of success. For example, I was trying to examine the objectValues of a series of Conferas: <!--#in "objectValues('Confera Topic')" skip_unauthorized=1 sort=bobobase_modification_time --> <!--#in "_['sequence-item'].objectValues()" --> <!--#var uniqueId--> : <!--#var id--> : <!--#var meta_type--><BR> <!--#/in objectValues--> <!--#/in objectValues--> Though the documentation indicates that Conferas support objectValues, I get the following error when examining Conferas and don't when examining Folders. (this is with Principia): Error type: AttributeError Error value: objectValues Error traceback:Traceback (innermost last): File ./Document.py, line 80, in __call__ (Object: listConferas) File ./DT_String.py, line 471, in __call__ (Object: listConferas) File ./DT_In.py, line 634, in renderwob (Object: objectValues('Confera Topic')) File ./DT_In.py, line 583, in renderwob (Object: _['sequence-item'].objectValues()) File ./DT_Util.py, line 223, in eval (Object: _['sequence-item'].objectValues()) File <string>, line 0, in ? File ./DT_Util.py, line 97, in careful_getattr AttributeError: objectValues </pre> <!-- Traceback (innermost last): File /opt/home/htdb/Principia/lib/python/Components/cgi_module_publisher.py, line 1457, in publish_module File /opt/home/htdb/Principia/lib/python/Components/cgi_module_publisher.py, line 933, in publish File ./Document.py, line 80, in __call__ (Object: listConferas) File ./DT_String.py, line 471, in __call__ (Object: listConferas) File ./DT_In.py, line 634, in renderwob (Object: objectValues('Confera Topic')) File ./DT_In.py, line 583, in renderwob (Object: _['sequence-item'].objectValues()) File ./DT_Util.py, line 223, in eval (Object: _['sequence-item'].objectValues()) File <string>, line 0, in ? File ./DT_Util.py, line 97, in careful_getattr AttributeError: (see above) --> ===================================================================== Kent
Hi Kent, you dummy. You wrote:
Close...
What about a parent document which lists child documents? I need to be able to query the list of child objects which are referenced in the parent document and find the latest date. And I'd like to be able to extend that behavior to things like like sql query results so they can contribute their 'modification date'. And no, I don't want to discuss issues about walking the tree right now. :^)
In the past, I have tried to examine subobjects in this manner in order to produce an examination result and have met with varying degrees of success. For example, I was trying to examine the objectValues of a series of Conferas:
<!--#in "objectValues('Confera Topic')" skip_unauthorized=1 sort=bobobase_modification_time --> <!--#in "_['sequence-item'].objectValues()" --> <!--#var uniqueId--> : <!--#var id--> : <!--#var meta_type--><BR> <!--#/in objectValues--> <!--#/in objectValues-->
Ummm. I forgot where I was going... Got stuck in the exit lane and am now suffering from vapor lock, stuck here in the traffic... What I was thinking about was implementing it like one of those computed attribute thingies in Tabula <!--#var latest_bobobase_modification_time--> which I could do in an external method, as I discussed before... How would that fit in with your suggestion? (sitting here, playing with the clutch and flipping the turn signal switch) Kent
participants (2)
-
Amos Latteier -
kent＠eaenki.nde.swri.edu