- Zope-Dev - Zope lists

Re: [Zope-dev] dtml-in: next-batches
by Brett Carter 02 Mar '01

02 Mar '01

I've also discovered that setting overlap = 0 actually causes everything to overlap by 1. I didn't see anything in the collector about this either... -Brett >>>>> "Brett" == Brett Carter <brett(a)kavi.com> writes: Brett> I'm running Zope 2.2.2, and I'm trying to use the batching stuff for Brett> the dtml-in tag, however, I can't seem to access any attributes off Brett> the 'next-batches' or 'previous-batches' mapping object that's Brett> supposed to be available, with the attributes 'batch-start-index', Brett> 'batch-end-index', and 'batch-size'. However, my example code (see Brett> below) blows up with a key error when trying to access any of these. Brett> Is this a bug, or am I just misunderstanding how to use this thing? Brett> I wrote an external method to dir the 'next-batches' object, and it Brett> seems the only attributes available are: Brett> ['data', 'items', 'query_string', 'start_name_re'] Brett> -Brett Brett> Brett> <dtml-in "mystuff.retseq()" size=5 start=start> Brett> <dtml-var sequence-item> Brett> </dtml-in> Brett> Brett> <dtml-in "mystuff.retseq()" size=5 start=start previous> Brett> <a href="<dtml-var absolute_url>/batch_me?start=<dtml-var previous-sequence-start-index>">Previous</a> Brett> </dtml-in> Brett> <dtml-in "mystuff.retseq()" size=5 start=start next overlap=-1> Brett> <dtml-try> Brett> <dtml-in next-batches> Brett> <dtml-var batch-start-index> Brett> </dtml-in> Brett> <dtml-except> Brett> <pre> Brett> <dtml-var error_type> Brett> <dtml-var error_value> Brett> <dtml-var error_tb> Brett> </pre> Brett> </dtml-try> Brett> <a href="<dtml-var absolute_url>/batch_me?start=<dtml-var next-sequence-start-index>">Next</a> Brett> </dtml-in> Brett> Brett> _______________________________________________ Brett> Zope-Dev maillist - Zope-Dev(a)zope.org Brett> http://lists.zope.org/mailman/listinfo/zope-dev Brett> ** No cross posts or HTML encoding! ** Brett> (Related lists - Brett> http://lists.zope.org/mailman/listinfo/zope-announce Brett> http://lists.zope.org/mailman/listinfo/zope )

1 0

Re: [Zope-dev] SessionDataManager oddness
by Chris McDonough 02 Mar '01

02 Mar '01

Not really. In this case it seems appropriate, in others its not really necessary. I hope my suggestion works. ;-) ----- Original Message ----- From: "Matt Hamilton" <matth(a)netsight.co.uk> To: "Chris McDonough" <chrism(a)digicool.com> Sent: Friday, March 02, 2001 10:43 AM Subject: Re: [Zope-dev] SessionDataManager oddness > On Fri, 2 Mar 2001, Chris McDonough wrote: > > > Try putting the code which adds the session data manager from your class' > > __init__ into a manage_afterAdd method of your class, e.g.: > > > > def manage_afterAdd(self, item, container): > > ob = SessionDataManager( > > id='session_data_mgr', title='', path=None, timeout_mins=20) > > self._setObject('session_data_mgr', ob) > > > > manage_afterAdd is called automagically by the _setObject machinery at init > > time, but it gets the object in its acquisition-wrapped context. > > OK, thanks, I'll give that a try. For future reference, is there where I > should be putting in all code to add objects to a new instance of a > folderish class? Currently it is in __init__ (I add a couple of > BTreeFolders, and a ZCatalog). > > -Matt > > -- > Matt Hamilton matth(a)netsight.co.uk > Netsight Internet Solutions, Ltd. Business Vision on the Internet > http://www.netsight.co.uk +44 (0)117 9090901 > Web Hosting | Web Design | Domain Names | Co-location | DB Integration > > >

1 0

SessionDataManager oddness
by Matt Hamilton 02 Mar '01

02 Mar '01

Dear All, I am trying to add session tracking to a Product of mine. I am trying to add a SessionDataManager to a Folderish object in the __init__ method of the class, but having troubles. I think I am importing everything: from Globals import HTMLFile from Globals import Persistent from Globals import default__class_init__ import Acquisition import AccessControl.Role import OFS.Folder from Products.ZCatalog import ZCatalog from Products.BTreeFolder import BTreeFolder from Products.CoreSessionTracking.SessionDataManager import SessionDataManager from Products.CoreSessionTracking.SessionDataContainer import SessionDataContainerMount from Products.CoreSessionTracking.SessionData import SessionData try: from Products.CoreSessionTracking.SessioningInterfaces import SessionDataManagerInterface except: # pre-2.3 Zopes don't have an Interfaces module class SessionDataManagerInterface: pass then in my __init__ method: ob = SessionDataManager( id='session_data_mgr', title='', path=None, timeout_mins=20) self._setObject('session_data_mgr', ob) However I get an attribute error when trying to add an intance of the product. _sdc not found. I have traced it using pdb, and from what I can find out it appears as if it fails in _setObject. More specifically SessionDataContainerMount seems to not returning the correct thing. I have tried tracing the normal addition of a SessionDataManager in Zope and it works fine. SessionDataContainerMount doesn't have an __init__ method, but it looks like when I add a S.D.M via the normal web interface __of__ gets called, but when my product tries to add it __del__ gets called. Any ideas? I am at the end of my Zope internals knowledge.... if there is anything else I can provide to help diagnose, let me know. TIA, Matt -- Matt Hamilton matth(a)netsight.co.uk Netsight Internet Solutions, Ltd. Business Vision on the Internet http://www.netsight.co.uk +44 (0)117 9090901 Web Hosting | Web Design | Domain Names | Co-location | DB Integration

2 1

ZDebug conflicts with MailHost?
by Itai Tavor 02 Mar '01

02 Mar '01

Hi, On Zope 2.3.0 and 2.3.1b1, the following method produces an error when ZDebug is installed; If I remove ZDebug, it works fine. <dtml-sendmail mailhost=MailHost> To: itai(a)optusnet.com.au From: itai(a)optusnet.com.au Subject: Test This is a test </dtml-sendmail> With ZDebug the content of the sendmail tag is ignored; sendmail.send is called with all arguments empty. Traceback: Error type: TypeError Error value: len() of unsized object Traceback (innermost last): File /opt/Zope-2.3.0-linux2-x86/lib/python/ZPublisher/Publish.py, line 222, in publish_module File /opt/Zope-2.3.0-linux2-x86/lib/python/ZPublisher/Publish.py, line 187, in publish File /opt/Zope-2.3.0-linux2-x86/lib/python/Zope/__init__.py, line 221, in zpublisher_exception_hook (Object: ApplicationDefaultPermissions) File /opt/Zope-2.3.0-linux2-x86/lib/python/ZPublisher/Publish.py, line 171, in publish File /opt/Zope-2.3.0-linux2-x86/lib/python/ZPublisher/mapply.py, line 160, in mapply (Object: mailtest) File /opt/Zope-2.3.0-linux2-x86/lib/python/ZPublisher/Publish.py, line 112, in call_object (Object: mailtest) File /opt/Zope-2.3.0-linux2-x86/lib/python/OFS/DTMLMethod.py, line 189, in __call__ (Object: mailtest) File /opt/Zope/lib/python/Products/ZDebug/DTMLLocator.py, line 103, in __call__ (Object: mailtest) File /opt/Zope-2.3.0-linux2-x86/lib/python/DocumentTemplate/DT_String.py, line 538, in __call__ (Object: mailtest) File /opt/Zope/lib/python/Products/ZDebug/DTMLDebug.py, line 307, in debug_render_blocks File /opt/Zope/lib/python/Products/ZDebug/DTMLDebug.py, line 239, in debugException (Object: MailHost) File /opt/Zope/lib/python/Products/ZDebug/DTMLDebug.py, line 305, in debug_render_blocks File /opt/Zope-2.3.0-linux2-x86/lib/python/Products/MailHost/SendMailTag.py, line 188, in render (Object: MailHost) File /opt/Zope-2.3.0-linux2-x86/lib/python/Products/MailHost/MailHost.py, line 222, in send (Object: MailHost) File /opt/Zope-2.3.0-linux2-x86/lib/python1.5/smtplib.py, line 465, in sendmail File /opt/Zope-2.3.0-linux2-x86/lib/python1.5/smtplib.py, line 349, in mail File /opt/Zope-2.3.0-linux2-x86/lib/python1.5/smtplib.py, line 118, in quoteaddr File /opt/Zope-2.3.0-linux2-x86/lib/python1.5/rfc822.py, line 451, in parseaddr File /opt/Zope-2.3.0-linux2-x86/lib/python1.5/rfc822.py, line 496, in getaddrlist File /opt/Zope-2.3.0-linux2-x86/lib/python1.5/rfc822.py, line 504, in getaddress File /opt/Zope-2.3.0-linux2-x86/lib/python1.5/rfc822.py, line 484, in gotonext TypeError: (see above) -- -- Itai Tavor -- "Je sautille, donc je suis." -- itai(a)optusnet.com.au -- - Kermit the Frog -- -- -- -- "If you haven't got your health, you haven't got anything" --

1 0

dtml-in: next-batches
by Brett Carter 02 Mar '01

02 Mar '01

I'm running Zope 2.2.2, and I'm trying to use the batching stuff for the dtml-in tag, however, I can't seem to access any attributes off the 'next-batches' or 'previous-batches' mapping object that's supposed to be available, with the attributes 'batch-start-index', 'batch-end-index', and 'batch-size'. However, my example code (see below) blows up with a key error when trying to access any of these. Is this a bug, or am I just misunderstanding how to use this thing? I wrote an external method to dir the 'next-batches' object, and it seems the only attributes available are: ['data', 'items', 'query_string', 'start_name_re'] -Brett <dtml-in "mystuff.retseq()" size=5 start=start> <dtml-var sequence-item> </dtml-in> <dtml-in "mystuff.retseq()" size=5 start=start previous> <a href="<dtml-var absolute_url>/batch_me?start=<dtml-var previous-sequence-start-index>">Previous</a> </dtml-in> <dtml-in "mystuff.retseq()" size=5 start=start next overlap=-1> <dtml-try> <dtml-in next-batches> <dtml-var batch-start-index> </dtml-in> <dtml-except> <pre> <dtml-var error_type> <dtml-var error_value> <dtml-var error_tb> </pre> </dtml-try> <a href="<dtml-var absolute_url>/batch_me?start=<dtml-var next-sequence-start-index>">Next</a> </dtml-in>

1 0

Zope 2.3.0 AccessControl/SecurityInfo.py
by Randall F. Kern 01 Mar '01

01 Mar '01

Sorry if these have already been fixed in 2.3.1, I haven't had time to upgrade yet. I fixed a small bug in SecurityInfo that made SecurityInfo.setPermissionDefault() impossible to use: 191c191 > self.roles[permission_name] = rdict.keys() --- < self.roles[permission_name] = rdict def setPermissionDefault(self, permission_name, roles): """Declare default roles for a permission""" rdict = {} for role in roles: rdict[role] = 1 if self.roles.get(permission_name, rdict) != rdict: LOG('SecurityInfo', WARNING, 'Conflicting default role' 'declarations for permission "%s"' % permission_name) self._warnings = 1 self.roles[permission_name] = rdict.keys() My other problem was that I wanted to setup some default permissions that are used programatically in my product (i.e. they aren't used to protect any specific methods, but the methods themselves check for the permission.) A small change to ClassSecurityInfo.apply() allows you to call SecurityInfo.setPermissionDefault() to create new permissions, even if they aren't attached to any methods: 239a240,242 > for permission_name in self.roles.keys(): > if not ac_permissions.has_key(permission_name): > ac_permissions[permission_name] = () def apply(self, classobj): """Apply security information to the given class object.""" dict = classobj.__dict__ # Check the class for an existing __ac_permissions__ and # incorporate that if present to support older classes or # classes that haven't fully switched to using SecurityInfo. if dict.has_key('__ac_permissions__'): for item in dict['__ac_permissions__']: permission_name = item[0] self._setaccess(item[1], permission_name) if len(item) > 2: self.setDefaultRoles(permission_name, item[2]) # Set __roles__ for attributes declared public or private. # Collect protected attribute names in ac_permissions. ac_permissions = {} for name, access in self.names.items(): if access in (ACCESS_PRIVATE, ACCESS_PUBLIC, ACCESS_NONE): dict['%s__roles__' % name] = access else: if not ac_permissions.has_key(access): ac_permissions[access] = [] ac_permissions[access].append(name) # Now transform our nested dict structure into the nested tuple # structure expected of __ac_permissions__ attributes and set # it on the class object. getRoles = self.roles.get __ac_permissions__ = [] for permission_name in self.roles.keys(): if not ac_permissions.has_key(permission_name): ac_permissions[permission_name] = () permissions = ac_permissions.items() permissions.sort() for permission_name, names in permissions: roles = getRoles(permission_name, ()) if len(roles): entry = (permission_name, tuple(names), tuple(roles)) else: entry = (permission_name, tuple(names)) __ac_permissions__.append(entry) dict['__ac_permissions__'] = tuple(__ac_permissions__) # Take care of default attribute access policy access = getattr(self, 'access', _marker) if access is not _marker: dict['__allow_access_to_unprotected_subobjects__'] = access if getattr(self, '_warnings', None): LOG('SecurityInfo', WARNING, 'Class "%s" had conflicting ' 'security declarations' % classobj.__name__) -Randy

1 0

CopySupport issues
by richard＠bizarsoftware.com.au 01 Mar '01

01 Mar '01

These comments are against the CVS of about a week ago. I can't get to the CVS to update at the moment - the connection times out. Is there a reason why the manage_main call at the end of manage_copyObjects and manage_cutObjects in CopySupport gets called as self.manage_main(REQUEST) instead of self.manage_main(self, REQUEST)? Using the first form, we found that the manage_main call was dying because it couldn't acquire various attributes and methods it needed (eg, manage_page_header :). The second form just makes sense... In the Copy operation of manage_pasteObjects, the manage_afterClone is called before the object is added to its container. We'd much prefer the object to be both added to the container and re-got so it the manage_afterClone call may make use of acquired information. This is of use specifically for us so that the object may determine its new URL and catalog itself with our (Z)Catalog. Interestingly, a call to absolute_url() on the object breaks because self.REQUEST is '<Special Object Used to Force Acquisition>' (yes, a string). Buggered if I know where that came from (actually, I know it originates in Acquisition.c, but that's a scary, scary place :) Richard -- Richard Jones richard(a)bizarsoftware.com.au Senior Software Developer, Bizar Software (www.bizarsoftware.com.au)

1 1

Increased system CPU cycles under heavy load?
by Erik Stephens 01 Mar '01

01 Mar '01

Hello all, I apologize if this might be more appropriate for a Linux list. From our time-series graphs, we are noticing that the percentage of CPU cycles spent on 'system' tasks grows considerably under heavy load. Does anyone know what kinds of operations fall under 'system'? Task switching, paging, these kinds of things? When load is light, it appears that the same 1 or 2 Zope threads will handle the requests. As load increases, then more threads get into the mix. How likely is it that this increase in 'system' CPU cycles is stemming from having to switch between the Zope threads? Perhaps Zope goes into a near dead-lock state and starts to spend a bunch of time waiting for other threads to release a lock? The reason I mention this is that we don't notice the increased amount of 'system' CPU when the load increases on other servers (Apache, MySQL) on other machines. System info: Zope 2.1.6, Redhat 6.2, PC arch Thanks, Erik

1 0

REQUEST.steps
by Chris Withers 01 Mar '01

01 Mar '01

Is it alright to rely on this attribute of the REQUEST object? If so, is it affected by virtual hosting? cheers, Chris

4 4

Synchronize GUD or Worldpilot with IPAQ
by Valérie Aulnette 28 Feb '01

28 Feb '01

Hi ! Is it possible to synchronize GUD or Worldpilot with Pocket PC ? or with Outlook ? Did someone try ? And how ? Thanks. Valerie. ___________________________________________________________ Do You Yahoo!? -- Pour dialoguer en direct avec vos amis, Yahoo! Messenger : http://fr.messenger.yahoo.com

2 1