January 2000 - Zope-Dev

Re: IMAP Server
by Butch Landingin 19 Jan '00

19 Jan '00

Hi all, I'm not doing anything IMAP-related but I found something which might be of use to you guys: http://horde.org/papers/Scalable_webmail_HOWTO.php3 Maybe you could incorporate some of his ideas into whatever you guys are building? HTH. Regards, Butch __________________________________________________ Do You Yahoo!? Talk to your friends online with Yahoo! Messenger. http://im.yahoo.com

1 0

Fix for ZPublisher.Client
by Hunter Kelly 19 Jan '00

19 Jan '00

Hi, I was having trouble using the load_site script with directories that had spaces in the names. Turns out that ZPublisher.Client.Function does not quote the url when it does the post. If there are problems with doing this, or there is a better way, please let me know. Please CC me on replies as I am not on the Zope-Dev mailing list. Hunter Kelly --- Client.py.orig Wed Jan 19 17:51:48 2000 +++ Client.py Wed Jan 19 17:50:52 2000 @@ -263,7 +263,7 @@ else: d[k]=v else: d[k]=v - rq=[('POST %s HTTP/1.0' % self.rurl),] + rq=[('POST %s HTTP/1.0' % quote(self.rurl)),] for n,v in self.headers.items(): rq.append('%s: %s' % (n,v)) if self.username and self.password:

1 0

RE: [Zope] AW: [Zope-dev] Re: [Zope] ZCatalog Find Items broken i n ZOPE 2.1? FIXED! n ZOPE 2.1? FIXED!
by Michel Pelletier 19 Jan '00

19 Jan '00

> -----Original Message----- > From: Markus Strickler [mailto:mstrickler@gmx.de] > Sent: Wednesday, January 19, 2000 4:10 AM > To: Stephen Pitts > Cc: Jason Spisak; Brian Lloyd; zope-dev(a)zope.org; zope(a)zope.org > Subject: [Zope] AW: [Zope-dev] Re: [Zope] ZCatalog Find Items > broken in > ZOPE 2.1? FIXED! > > > Did you follow the discussion about / and id in ZCatalog? > I haven't checked whether the apply_path bug has been fixed > in the cvs as > well. This was fixed along with an earlier bug in the code. Thanks for the patch. -Michel > > I just got hit by this as well. The culprit is in line 536 of > > lib/python/Products/ZCatalog/ZCatalog.py: > > > > apply_func(ob, (apply_path+'/'+p)) > > > > If apply_path is null, then an extra '/' is prepended to p > (the object's > > path) for no apparent reason. A quick fix is to replace this with: > > > > if apply_path: > > apply_func(ob, (apply_path+'/'+p)) > > else: > > apply_func(ob, p) > > > > A patch to that effect is attached > > -- > > Stephen Pitts > > smpitts(a)midsouth.rr.com > > webmaster - http://www.mschess.org > > > > > _______________________________________________ > Zope maillist - Zope(a)zope.org > http://lists.zope.org/mailman/listinfo/zope > ** No cross posts or HTML encoding! ** > (Related lists - > http://lists.zope.org/mailman/listinfo/zope-announce > http://lists.zope.org/mailman/listinfo/zope-dev ) >

1 0

RE: [Zope] Very important for CatalogWare zopists
by Brian Lloyd 19 Jan '00

19 Jan '00

> I have submitted this to the Collector and it is major for people who > have items in their ZODB that are Catalog aware. > > To create the bug follow these steps: > > 1.Create a Catalog Aware ZClass > 2.Create an instance of that class > 3.Make sure it is indexed by looking in the Catalog "Catolaged items" > tab > 4.Notice that there is *no* slash in front of the indexed url: > "MyFolder/myobject" > 5.Now clear the Catalog > 6.Go to the "Find Items to Catalog" > 7.Select just your New ZClass meta type so it is quick > 8.Click "Find" > 9.Make sure it is indexed by looking in the Catalog "Catolaged items" > tab > 10. Oh no! It has a different url. A leading slash! > "/MyFolder/myobject" > 11. Click on the link. It's broken! Hi Jason - these steps are equivalent to what is currently done on Zope.org (which is running on a checkout equivalent to 2.1.2), and this does not seem to be a problem there. While I won't totally rule out the possibility of something in Zope, the fact that we don't see this behavior seems to suggest that there might be something going awry in your particular ZClasses. In particular, I'd look for some difference in the way that it registers itself with the catalog at the time that it is created versus when it is "found" by the catalog. The difference could be a difference in either behavior, state or environment (i.e. is the acquisition context, etc. the same in both cases when the "absolute_url" method gets called to generate the url for the catalog? Is there any custom code around this logic?). The object constructor is a good place to start looking for this sort of thing - try to trace through the activity in the constructor versus what happens when the existing object is "found" by the catalog. If possible, try to see what the value of the url given to the catalog is what you expected it to be... > What this means is that if you have a lot of items in you > Catalog don't > ever ever ever use the "Find" machinery until Digital fixes this(It > worked in 2.0.0 Final). You will end up not being able to > edit/reindex > your objects, and the links in your applications may not > function. I have not heard any other reports of a problem like this, so until we can verify that it is a Zope problem I think that this is a little strong. If other Zopistas can reproduce this I would certainly like to hear about it though. Brian Lloyd brian(a)digicool.com Product Manager 540.371.6909 Digital Creations http://www.digicool.com

6 8

IMAP Server
by Michel Pelletier 19 Jan '00

19 Jan '00

Brad, Hi, I think I recall you being interested in writing an IMAP server in Python. I have begun writing an IMAP server, and I was curious if you'd like to join me in the development. I have an IMAP server based on asychat, my own protocol handler, and ZODB. Zope itegration is imminently planned. Below is the readme. I'll be setting up a site on starship tonight/tommorow to host the project. Are interested? -Michel Test: What the hell should I call this thing? This thing is an IMAP server written in Python. It consists of a simple asychat dispatcher, a generic protocol handler class, some mail objects, and an abstract backend storage mechanism. This thing requires the following software: asyncore/asynchat - Standard with Python. Also comes with ZServer (Zope) ZODB - The default namespace storage for this thing requrires ZODB. The backend is totaly plugable however, so I don't see any reason why some other mapping storage cannot be used (like gdbm, sleepycat berkeley, etc...) Currently, ZODB is not shipped seperatly from Zope, so this thing requires Zope to run. In fact, this thing hijacks quite a bit of Zope. To get it fired up, ungzip and untar the file into your Zope top level directory. edit lib/python/start_imap.py, change the path.append to reflect your medusa directory. I know, this is an ugly hack, bear with it for now. A Sample interaction (this is for real! not a simulation!): [michel@localhost python]$ telnet localhost 143 Trying 127.0.0.1... Connected to localhost. Escape character is '^]'. * OK IMAP4rev1 v11.241 Welcome 1 capability * CAPABILITY IMAP4 IMAP4rev1 NAMESPACE AUTH=LOGIN XUNDO 1 OK CAPABILITY completed 2 login mike pass 2 OK LOGIN complete 2 create aNewBox 2 OK create completed 3 select aNewBox * 0 EXISTS* 0 RECENT* [UNSEEN 1] Messages 1 is first unseen * [UIDVALIDITY 1] UIDs valid * FLAGS (\Flag1 Flag2)* [PERMANENTFLAGS %s] Limited 3 OK [READ-WRITE] SELECT completed 5 close aNewBox 5 OK close completed 6 namespace * (("~" "/")) (("Public/" "/")) (("" "/")) 6 OK namespace completed 7 logout * BYE BYE logging out 7 BYE LOGOUT completed Connection closed by foreign host. Because it is based on really cool peices of software, this thing has a number of benefits. High performance -- Ok that's not proven yet, but asyncore in general has been proven pretty speedy and the select() model is regarded as a high performance solution. Further, then entire enchelada, if Zope itegration is not used, runs in one process space. It can even run in the same ZServer thread space as a Zope server. ZODB database cache keeps commonly used objects in memory. Also, the entire IMAP mailbox namespace is used as keys to the ZODB dict, so the namespace is kept in memory at all times. This makes listing and searching through mailbox names very fast. Easy possbile Zope itegration -- The storage backend for this server can be realized by a simple layer over a mapping storage like object. So gbdm, berkley, or a bunch of Zope objects can be used to store mail objects. If Zope were used, it would be possible for mail and mailbox objects to have web interfaces, and either interface, IMAP or HTTP (or XML-RPC, WebDAV and FTP) can manipulate the SAME objects. It's the same old classic dream of multi-protocol publishing realized once again in IMAP. This has not been thought out much, specificly how the "security system" (tm) in this thing can interface with Zope's. In the end, it might require medling the entire thingy into Zope's framework, but I want to avoid that if possible. ZODB provides plugable storage For now, this thing depends on ZODB for storage. This has lots of benefits, for one, ZODB supports pluggable storage formats. At the moment there are several, like FileStorage, BerkeleyStorage, DemoStorage and ReadOnly. Digital Creations will likely in the future add other storage types, and this thing might as well let them do all the work. Fine grained multi-tasking Every client request and every response gives asyncore

3 2

tree tag, expand_all and branches.
by Anthony Baxter 19 Jan '00

19 Jan '00

Has anyone started/finished work on making the tree tag's 'expand_all' functionality work when you have 'branches_expr' but not 'branches'? There's 4 different reports in the collector for this bug, but all are 'under review', so I can't see if there's any patches for these in any of them. I can't find anything in my mail about this except a message from Martijn Pieters (in his pre-DC days) last October that he thought this was fixed in CVS - it sure doesn't seem to be fixed right now. Anthony

1 0

ZSql search interface!
by Victor.Zhai＠ogilvy.com 19 Jan '00

19 Jan '00

Hi,all It good for me to use PSql Input wizard to collect information from user. But.. I want to modify the search interface of ZSQL.Because I want to show user a good interface to use my project! Can anyone provide me some information! Thanks a lot! "WorldSecure Server <ogilvy.com>" made the following annotations on 01/18/00 23:27:49 ------------------------------------------------------------------------------ Privileged/Confidential Information may be contained in this message. If you are not the addressee indicated in this message (or responsible for delivery of the message to such person), you may not copy or deliver this message to anyone. In such case, you should destroy this message and kindly notify the sender by reply email. Please advise immediately if you or your employer does not consent to email for messages of this kind. Opinions, conclusions and other information in this message that do not relate to the official business of the Ogilvy Group shall be understood as neither given nor endorsed by it. ==============================================================================

1 0

ZopeClient
by Tres Seaver 18 Jan '00

18 Jan '00

I am playing with the wxPython client Amos posted as part of his XML-RPC example (very nice, BTW!) and would like to allow for browsing _all_ objects, rather than just folders and the two DTML objects. So, I tried this:: def OnItemExpanding(self, event): item=event.GetItem() if self.tree.GetChildrenCount(item, false): return server=self.getServer(item) # get the list of sub-folders and sub-documents try: children = server.objectValues( None ) for child in children: print child.id if child.isPrincipiaFolderish(): childNode =self.tree.AppendItem( item, child.id, 0 ) self.tree.SetItemHasChildren( childNode , true ) else: self.tree.AppendItem( item, child.id, 1 ) except Exception, e: print e return Unfortunately, the XML client library balks at passing None; so I tried '[]', but now Zope's special handling for None is not invoked, and I get nothing back. Can anyone suggest a remedy? BTW, I have added a Notebook to the right hand page, with pages for Document Source and Properties. I would like to change the page tabs on the fly, using the "manage_options" list, but I'll worry about that later. :) -- ========================================================= Tres Seaver tseaver(a)palladion.com 713-523-6582 Palladion Software http://www.palladion.com

2 2

A little more debugging than I wanted
by Ross Boylan 18 Jan '00

18 Jan '00

I have noticed that when I start debugging Zope it is often hard to turn debugging off. The "c" command to the debugger doesn't continue all the way--in fact, sometimes it seems to be single-stepping through the code. I wonder if anyone else has noticed a similar phenomenon, and if anyone has any suggestions for a cure. Background: Zope 2.x on NT 4.0, SP4, AMD K6-2. I start using the default start script, which uses -D. In my python product, or sometimes in the Zope code, I insert import pdb pdb.set_trace() I start a browser, and do something which makes the code with the breakpoint get hit. I debug for awhile--everything is normal. I hit "c" (continue) but I end up back in the debugger, typically somewhere in the object publishing code. Further "c"'s yield the same results (with pauses where I must send something new from the browser).

1 0

Confera error
by Rodrigo Leme de Mello 18 Jan '00

18 Jan '00

Hi all, I´ve installed the confera topic product from digital creations, i am using the Zserver and Zope 2.1.2. My zope directory installation is /zope The directory of the product is /zope/lib/python/Products/Confera I am able to post a message in a non moderated topic, but when i try to reply this message i got an error. Zope Error Zope has encountered an error while publishing this resource. Error Type: AttributeError Error Value: regs ---------------------------------------------------------------------------- ---- Troubleshooting Suggestions The URL may be incorrect. The parameters passed to this resource may be incorrect. A resource that this resource relies on may be encountering an error. For more detailed information about the error, please refer to the HTML source for this page. If the error persists please contact the site maintainer. Thank you for your patience. Traceback (innermost last): File /zope/lib/python/ZPublisher/Publish.py, line 214, in publish_module File /zope/lib/python/ZPublisher/Publish.py, line 179, in publish File /zope/lib/python/Zope/__init__.py, line 202, in zpublisher_exception_hook (Object: ElementWithAttributes) File /zope/lib/python/ZPublisher/Publish.py, line 165, in publish File /zope/lib/python/ZPublisher/mapply.py, line 160, in mapply (Object: addMessage) File /zope/lib/python/ZPublisher/Publish.py, line 102, in call_object (Object: addMessage) File lib/python/Products/Confera/Confera.py, line 287, in addMessage (Object: ElementWithAttributes) File /zope/lib/python1.5/regsub.py, line 38, in gsub AttributeError: (see above) Does anybody know what is going on ? Thanks

1 0