December 2002 - Zope-Dev

more bug day status
by Chris McDonough 16 Dec '02

16 Dec '02

26 issues closed. Only 222 more to go. ;-) Notable bugs fixed: - multiple selection unicode bug (#697) - make sure passwords aren't encrypted twice ;-) (#256) - don't fail in History tab if Historical object is stored in a non-history-supporting storage (#700) 10 people currently in the #zope-dev channel on irc.openprojects.net. - C

1 0

RE: [Zope-dev] ClientCache and Large Files
by sean.upton＠uniontrib.com 16 Dec '02

16 Dec '02

This sounds like the most appropriate solution, as long as you are careful to make sure that Squid or other forward cache is configured to cache files that large, it isn't likely going to be by default. Sean -----Original Message----- From: Casey Duncan [mailto:casey@zope.com] Sent: Monday, December 16, 2002 8:21 AM To: Guido van Rossum; Brian R Brinegar Cc: zope-dev(a)zope.org Subject: Re: [Zope-dev] ClientCache and Large Files On Monday 16 December 2002 10:53 am, Guido van Rossum wrote: > > We're running ZEO with Zope 2.5.1. We currently have a ZEO Cache > > (ClientCache) of 200 meg. Occasionally someone will download a 300 > > meg file that completely blows away the client cache. Is there a way > > to prevent this? You're best bet would be to put a proxy cache in front of Zope. Zope's not very efficient at serving big static files like this compared to Apache or squid. > I'm afraid not; this isn't a very typical use case. You could > implement a file size limit, or increase the cache size to be more > than twice the largest expected dowload (the cache is implemented as > two files, each of which is limited to half the nominal cache size). I think it would be nice to be able to control cacheability of certain objects. However it smells like a YAGNI, especially with good front end caches that can take care of it. -Casey _______________________________________________ Zope-Dev maillist - Zope-Dev(a)zope.org http://lists.zope.org/mailman/listinfo/zope-dev ** No cross posts or HTML encoding! ** (Related lists - http://lists.zope.org/mailman/listinfo/zope-announce http://lists.zope.org/mailman/listinfo/zope )

1 0

debugging Zope -> ImportError: ./ExtensionClass.so: undefined symbol: PyObject_Init
by Jens Wolk 16 Dec '02

16 Dec '02

Hi everybody, I tried to follow "Chapter 7: Testing and Debugging" of "The Zope Developer's Guide (Zope 2.4 edition)" http://www.zope.org/Documentation/Books/ZDG/current/TestingAndDebugging.stx However, on my Linux server I get the following error: (Zope is v2.5.1) [root@prefect Zope-2.5.1-linux2-x86-dev]# cd lib/python [root@prefect python]# python Python 1.5.2 (#1, Apr 3 2002, 18:16:26) [GCC 2.96 20000731 (Red Hat Linux 7.2 2 on linux-i386 Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam >>> import Zope, ZPublisher Traceback (innermost last): File "<stdin>", line 1, in ? File "Zope/__init__.py", line 21, in ? import ZODB, ZODB.ZApplication, imp File "ZODB/__init__.py", line 14, in ? import sys, ExtensionClass, TimeStamp, cPersistence, Persistence ImportError: ./ExtensionClass.so: undefined symbol: PyObject_Init >>> The system is a RedHat 7.2 system with the following python packages installed: [root@prefect python]# rpm -qa | grep python python-1.5.2-38 python-popt-0.8.8-7.x.2 python-xmlrpc-1.5.1-7.x.3 python2-docs-2.2-16 python2-2.2-16 python-clap-1.0.0-3 rpm-python-4.0.4-7x.18 python-docs-1.5.2-38 BTW, if I try to import ZPublisher only, the result is again an error: [root@prefect python]# python Python 1.5.2 (#1, Apr 3 2002, 18:16:26) [GCC 2.96 20000731 (Red Hat Linux 7.2 2 on linux-i386 Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam >>> import ZPublisher Traceback (innermost last): File "<stdin>", line 1, in ? File "ZPublisher/__init__.py", line 23, in ? from Publish import publish_module, Retry File "ZPublisher/Publish.py", line 19, in ? from Response import Response File "ZPublisher/Response.py", line 13, in ? import HTTPResponse File "ZPublisher/HTTPResponse.py", line 79, in ? status_codes[''.join(val.split(' ')).lower()]=key AttributeError: 'string' object has no attribute 'join' >>> It seems I missed something important. Can anybody give me hint? Yours. Jens

2 2

realtime Bug Day report
by Chris McDonough 16 Dec '02

16 Dec '02

So far the good folks participating in the bug day on irc.openprojects.net #zope-dev have resolved 13 issues in the collector. Present in the channel are 9 people. Feel free to join in. Thanks, - C

1 0

ClientCache and Large Files
by Brian R Brinegar 16 Dec '02

16 Dec '02

Hello, We're running ZEO with Zope 2.5.1. We currently have a ZEO Cache (ClientCache) of 200 meg. Occasionally someone will download a 300 meg file that completely blows away the client cache. Is there a way to prevent this? Thanks, -Brian Brinegar ECN Purdue University

3 2

Re: [ZDP] Suggested documentation update.
by Chris McDonough 16 Dec '02

16 Dec '02

Sounds good. Maybe we can just incorporate it into the Zope Book and ship the Book with Zope. On Mon, 2002-12-16 at 15:12, Lennart Regebro wrote: > doc/WEBSERVERS.txt only mentiones CGI. I'd suggest that we rename it to > CGI.txt and create a new APACHE.txt that takes up how to use Zope with > Apache, and make a new WEBSERVERS.txt that refers to these two documents, > with some description of when each should be used. > > I can make the APACHE.txt building on my Apache how-to. > > Thoughts? > > > _______________________________________________ > ZDP maillist - ZDP(a)zope.org > http://lists.zope.org/mailman/listinfo/zdp

2 2

Looking for a way to override the export/import of an persistent object
by Johan Carlsson [EasyPublisher] 14 Dec '02

14 Dec '02

Hi, I'm looking for someway the override the export/import for persistent Zope objects. For instance having an ExtensionClass based class I could override the the __setstate__/__getstate__ or __reduce__ methods to change some of the attributes. This would be useful for serializing objects that stores some parts in the file system (for instance "External Files"). But due to the persistent nature of ZODB these methods doesn't work as usual. Is this at all possible with the current Zope/ZODB? Regards, Johan Carlssons -- Torped Strategi och Kommunikation AB Johan Carlsson johanc(a)easypublisher.com Mail: Birkagatan 9 SE-113 36 Stockholm Sweden Visit: Västmannagatan 67, Stockholm, Sweden Phone +46-(0)8-32 31 23 Fax +46-(0)8-32 31 83 Mobil +46-(0)70-558 25 24 http://www.easypublisher.com http://www.torped.se

1 0

Proper Use of __init__ inside Zope Products?
by Jeff Rush 13 Dec '02

13 Dec '02

Can a Zope-internals guru provide some enlightenment regarding the mysteries of __init__? I'm writing some zproducts that subclass the existing Folder class, and then are themselves subclassed. What I've got works, but I'm not sure its _correct_. As I'm writing a "Zope Best Practices" document, I'd like to get it right. As I understand it, the use of __init__ should be avoided when possible, since it isn't invoked (necessarily) when persistent objects are reloaded from disk. Therefore Zope tends to do instance init within manage_addMYPRODUCT global-to-Zope functions, although this doesn't seem to be fully consistent throughout the Zope community contributions. The scenario is something like the following: def manage_addFolder(self, id, title='', ...): instance = Folder() instance.id = str(id) instance.title = title self._setObject(id, instance) class Folder(...): # no __init__ method def manage_addLargeFolder(self, id, title='', ...): instance = LargeFolder() instance.id = str(id) instance.title = title self._setObject(id, instance) class LargeFolder(Folder): def __init__(self): # parent class has no __init__ so don't call it! #Folder.__init__() self._tree = OOBTree() def manage_addSpecialLargeFolder(self, id, title='', ...): instance = SpecialLargeFolder() instance.id = str(id) instance.title = title self._setObject(id, instance) class SpecialLargeFolder(LargeFolder): def __init__(self): # parent class DOES have an __init__ so call it LargeFolder.__init__() self._other = OOBTree() (A) I reluctantly found it necessary to add an __init__ to my LargeFolder class because if I init'd self._tree in manage_addLargeFolder(), then it won't get init'd when manage_addSpecialLargeFolder() is used to create instances. manage_addLargeFolder doesn't get invoked in that case. (B) Currently the Folder class lacks an __init__ so I cannot call it from LargeFolder.__init__, but if someday ZC decides to add one, my code will fail to init Folder and break. Is this correct? I wish Python auto-provided a default no-nothing __init__ for this case so I could always call it. So the Guiding Rule seems to be "initialize those attributes that are common to all Zope objects, e.g. id and title, within the factory functions, but init any attributes you introduce using an __init__ method." Correct? -Jeff Rush

3 2

Calling in Context
by Brian R Brinegar 13 Dec '02

13 Dec '02

Howdy, Overview: I'm working on a helper product for Zope Presentation Templates. The goal of this product is to allow people to drop static content into a site developed with Zope Presentation Templates using WebDAV clients. (This is similar to the Kube Product but it's not folderish and supports WebDAV.) At first I suggested creating some ZPT with an Id like "view". Have users drop in static content as File objects. Then link to the files like "file_id/view". The "view" ZPT would reference the document_src method of the file object to pull the static content into the page. I like this approach, and I think this is how Presentation Templates are meant to be used. Problem: Our users will not change links created by WYSIWYG tools to include the "/view" so we need a product that removes the need for this. I am calling the product HTMLBlock. Version one of the product is a simple data holder with a document_src method which returns the content. It does not override index_html, instead acquiring the index_html from the context. I create an index_html page template with all of my presentation and some code like: <div tal:replace="here/document_src">Content Goes Here</div> When an HTMLBlock is viewed it shows up with all of the dynamic presentation of the index_html ZPT with the HTMLBlocks static content dropped in. Great! Almost there! :-) The problem is that people cannot create a default HTMLBlock with the Id index_html. It ends up trying to render itself much like a folder with the Id index_html. No Good. :-( So I temporarily fixed this by changing the template to redirect to an object with the Id "index" if document_src doesn't exist. When someone views a folder the index_html template is accessed, document_src is not defined so it redirects to "index" which is an HTMLBlock which renders correctly. I don't like this. :*( Still No Good. What I want is for the HTMLBlock product to have a property which specifies the Id of the template to use. Implement an index_html or __call__ method to locate the template and call it in the context of itself. I would like this to work generically so that the template could be DTML, Python, a ZPT, or anything else. Casey Duncan has some uber useful code on ZopeLabs.com to call an object generically using the ZPublisher. I borrowed it and came up with the code below, but it doesn't work. It calls the template, but the template cannot access methods of this object. Suggestions? Is there a better approach? (Note: template_id is the Id of the template to render) def __call__(self, client=None, REQUEST={}, RESONSE=None, **kw): """ Call the selected template in the context of myself. """A securityManager = getSecurityManager() securityManager.addContext(self) # Check for the template if not hasattr(self.aq_parent, self.template_id): return str(self) obj = getattr(self.aq_parent, self.template_id) if not securityManager.checkPermission('View', obj): return default return mapply( object=obj, positional=REQUEST.args, keyword=REQUEST, debug=call_object, maybe=1, missing_name=missing_name, handle_class=dont_publish_class, context=self, bind=1) The code below works, but requires a PageTemplate with the Id "index" to work def __call__(self, client=None, REQUEST={}, RESONSE=None, **kw): return self.index() Thanks, -Brian

1 1

post-publishing hook
by Carlo Giomini 13 Dec '02

13 Dec '02

Dear all, I need a sort of post-publishing hook (so to say). I need Zope to call a function of mine as the very last action of publishing a request, i.e. after having built the response (and sent it to the browser possibly), but before starting publishing a new request. My Zope server is running as a single-threaded app. Does Zope support this any way? Thanks for helping, cheers, Carlo.

6 5