- Zope - Zope lists

question about string manipulation and tal:attributes
by Laura McCord 31 Aug '04

31 Aug '04

I need to concatenate strings is this a way I can do it? resString is a variable that contains a string. <DIV tal:attributes="id string:$resString + 'Parent'"> thanks.

2 1

__bobo_traverse__ (Cont.)
by Garito 31 Aug '04

31 Aug '04

Hi In my last mail I ask you (thanks for your responses) about an error publishing the object in __bobo_traverse__ The solution passes by returning an object instead a string (I would like to return a string :( ) Then my __bobo_traverse__ is something like: def __bobo_traverse__(self, REQUEST, name): obj = getattr(self, name, None) if obj is not None: return obj else: return getattr(self, 'Result') But I need to process some information in bobo_traverse before returns Result If you change the line return getattr(self, 'Result') for return getattr(self, 'Result')('someresults') doesn't work For that I need to do something like: _Results = dict() # I put these declaration in my class def __bobo_traverse__(self, REQUEST, name): obj = getattr(self, name, None) if obj is not None: return obj else: self._Results[str(REQUEST.SESSION.id)] = self.MyResult() return getattr(self, 'Result') def MyResult(self): """MyResult""" return 'This is my test result' def GetResults(self, Session): """ GetResults""" return self._Results[str(Session)] Then I create a Python Script called Result like: return context.GetResults(context.REQUEST.SESSION.id) But all these code raise: Traceback (innermost last): Module ZPublisher.Publish, line 91, in publish Module ZPublisher.BaseRequest, line 302, in traverse Module Products.MyProducts.MyProduct, line 116, in __bobo_traverse__ TypeError: object does not support item assignment I can't undersand why can't I assign a dictionary. I can understand if I try to assign a number like dictionary key (but I convert the session id to a string) Some ideas? THANKS AGAIN!

1 0

RE: [Zope] Re: Physical path of objects in portal_skins tool of a CMF site
by Pascal Peregrina 31 Aug '04

31 Aug '04

I agree with the "very strange way"... ...but not with the "bad application design" :) If I have some time, I will try to explain why I need this... Anyway, thanks again. Pascal -----Message d'origine----- De : zope-bounces(a)zope.org [mailto:zope-bounces@zope.org]De la part de yuppie Envoyé : mardi 31 août 2004 11:18 À : zope(a)zope.org Objet : [Zope] Re: Physical path of objects in portal_skins tool of a CMF site Pascal Peregrina wrote: > I found a workaround : setting a variable in the manage_afterAdd method of > my object that uses container.getPhysicalPath() to compute the path I need. > It looks like copy/paste/rename operations work fine, even inside > portal_skins, even when copying a whole folder. > > What do you think ? I think you're using the skins tool in a very strange way. Smells like bad application design. But your workaround should work. Cheers, Yuppie _______________________________________________ Zope maillist - Zope(a)zope.org http://mail.zope.org/mailman/listinfo/zope ** No cross posts or HTML encoding! ** (Related lists - http://mail.zope.org/mailman/listinfo/zope-announce http://mail.zope.org/mailman/listinfo/zope-dev ) ________________________________________________________________________ This e-mail has been scanned for all viruses by Star Internet. The service is powered by MessageLabs. For more information on a proactive anti-virus service working around the clock, around the globe, visit: http://www.star.net.uk ________________________________________________________________________ ________________________________________________________________________ This e-mail has been scanned for all viruses by Star Internet. The service is powered by MessageLabs. For more information on a proactive anti-virus service working around the clock, around the globe, visit: http://www.star.net.uk ________________________________________________________________________

1 0

RE: [Zope] Re: Physical path of objects in portal_skins tool of a CMF site
by Pascal Peregrina 31 Aug '04

31 Aug '04

You are right... (I just read the SkinnableObjectManager code for a while). I found a workaround : setting a variable in the manage_afterAdd method of my object that uses container.getPhysicalPath() to compute the path I need. It looks like copy/paste/rename operations work fine, even inside portal_skins, even when copying a whole folder. What do you think ? Anyway, thanks for the quick answers so far ! Pascal -----Message d'origine----- De : zope-bounces(a)zope.org [mailto:zope-bounces@zope.org]De la part de yuppie Envoyé : lundi 30 août 2004 18:40 À : zope(a)zope.org Objet : [Zope] Re: Physical path of objects in portal_skins tool of a CMF site Hi! Pascal Peregrina wrote: > Ok, object is /site_staging/portal_skins/DEFAULT/testpath.html So 'DEFAULT' is a skin layer and 'default' a skin that includes the 'DEFAULT' layer? > When invoking it using http://myserver/site_staging/testpath.html (so I > use CMF skins traversal with implicit skin=default) > aq_parent(self).id returns site_staging, which looks ok to me. > And getPhysicalPath() will return : '','site_staging','testpath.html' > > But I really need (for some complicated to explain reasons) to get the > full physical path of the object. > So something returning : > '','site_staging','portal_skins','DEFAULT','testpath.html' I guess you have no chance. SkinnableObjectManager.__getattr__ removes the original acquisition context from the object. So if you invoke it as attribute of 'site_staging', the object knows nothing about its original context. HTH, Yuppie BTW: There is a CMF list, see <http://mail.zope.org/mailman/listinfo/zope-cmf> _______________________________________________ Zope maillist - Zope(a)zope.org http://mail.zope.org/mailman/listinfo/zope ** No cross posts or HTML encoding! ** (Related lists - http://mail.zope.org/mailman/listinfo/zope-announce http://mail.zope.org/mailman/listinfo/zope-dev ) ________________________________________________________________________ This e-mail has been scanned for all viruses by Star Internet. The service is powered by MessageLabs. For more information on a proactive anti-virus service working around the clock, around the globe, visit: http://www.star.net.uk ________________________________________________________________________ ________________________________________________________________________ This e-mail has been scanned for all viruses by Star Internet. The service is powered by MessageLabs. For more information on a proactive anti-virus service working around the clock, around the globe, visit: http://www.star.net.uk ________________________________________________________________________

2 1

[ANN] MailBoxer 3.3 released
by Maik Jablonski 31 Aug '04

31 Aug '04

Hi, I've just released a new version of MailBoxer (3.3). # What is MailBoxer? MailBoxer is a mailinglist-/ newsletter-/ mailarchive-framework for Zope and is easy to use and customize. MailBoxer runs all mailinglists of the German Zope User Group (www.dzug.org) with several hundred subscriptions and mails per day. # Where can I see MailBoxer working? Have a look at the zope(a)zope.org-archives mirrored by MailBoxer: http://www.dzug.org/mailinglisten/zope-org-zope/ # Where can I download MailBoxer? http://mjablonski.zope.de/MailBoxer # What changed in 3.3? - added support for Jens Vagelpohls MaildropHost (http://www.dataflake.org/software/maildrophost) so MailBoxer works transactional (no more doubled mails due to read-/write-zodb-conflicts). Using MaildropHost is the RECOMMENDED setup for MailBoxer from now on!!! Thanks to Jens for a really great product, which should be the default MailHost-Product in Zope! - added a customizable view (mail_approve) with pending requests for moderation. Please use this template instead of the old plain-text messages. If no template is found, MailBoxer falls back to the old plain-texts. I'll drop this fall-back behaviour in one of the next releases. - added a unique id for queued messages, so a accidentially page-reload won't approve / discard a newly arrived mail. - added a unique message-id for each mail which is bulked to the list, so problems under some setups with cross-posts to several lists should be gone. Thanks to Christian Theune and his inital "5-minute-hacked-patch", which needed some improvements. Hope you like it this way, Christian...;) - fixed bug: if the url given to the smtp2zope.py is not found (404) (maybe some typo in the alias-file), the mail is bounced with "user-doesn't-exist" and not(!) queued by the MTA anymore for a next retry. Thanks to Christian Theune for patch! - fixed bug: the mail-bulk-code was protected with a "historical" try-except, which could result in loss of emails when something went wrong while talking to the smtp-server. - fixed bug: the smtp-connection is now closed / reopened after each batch, so MTAs which only accept a limited number of mails per connection won't defer your mails anymore. - fixed bug: under some wired circumstances the Mail in the REQUEST wasn't marshalled as string by Zope, so string-operations were broken on the mail-data. Added a getMailFromRequest()-method which does the string-conversion, so we always have a string. Thanks to Dirk Datzert for tracking this down! - fixed bug: searchMailBoxerArchive() used unrestrictedTraverse for traversing the Catalog. Changed to restrictedTraverse. - some minor refactorings & code-cleanups. Happy mailboxing, Cheers, Maik -- Maik Jablonski Universität Bielefeld - Zentrum für Lehrerbildung http://www.zfl.uni-bielefeld.de/personal/mjablonski/

1 0

Thierry Belgium : Zope backup trouble "Attempt to store an object from a foreign database connection"
by GOYVAERTS Thierry 31 Aug '04

31 Aug '04

Hello, In the previous versions of Zope (like : 2.4.2); I used two external methods to make entire backup and restore of all the Zope documents, see : the source script and procedure below. Now, I install Zope 2.7.2 and then add my two external methods as usual to restore all my applications from the previous version (2.4.2). It works very well but when I try to use the backup method I receive the following error : "Attempt to store an object from a foreign database connection" Is it a bug or some modified functionality of Zope ? Somebody could guide me ? Thierry Thierry.Goyvaerts(a)skynet.be Procedure to install and use the full backup/restore methods : Create a module, backup.py, with the contents of the source (given in the tied code file) within the Extensions directory of your e-Solution instance. Create two external methods, "restore" and "backup" in your root folder by pointing them at Extensions/backup.py -> restore and Extensions/backup.py -> backup. After creating the two external methods, protect each of them with a management-level permission. To back up your e-Solution, visit the "test" tab of the backup external method. A file named "e-Solution_backup.zexp" will be downloaded to your local machine via your browser. To restore your e-solution, place the "e-Solution_backup.zexp" file in the "import" directory of your e-Solution instance, and press the "test" tab of the restore external method. Be careful, this overwrites anything that exists with the same name. Note that the "zexp" file created by 'backup' may not successfully be imported using the standard import feature, it needs to be imported via the "restore" external method. Source file Extensions\backup.py : import Globals import os from Acquisition import aq_base from cStringIO import StringIO from Globals import MessageDialog from ZODB.PersistentMapping import PersistentMapping def backup(self, REQUEST=None, download=1, filename='e-Solution_backup.zexp'): """ saves backup file to var or allows download of backup file""" if download: f = StringIO() else: f = os.path.join(Globals.data_dir, filename) self.c = PersistentMapping() for k, v in self.objectItems(): self.c[k] = aq_base(v) get_transaction().commit() self.c._p_jar.exportFile(self.c._p_oid, f) del self.c if download: REQUEST.RESPONSE.setHeader('Content-type', 'application/data') REQUEST.RESPONSE.setHeader('Content-Disposition', 'inline;filename=%s' % filename) return f.getvalue() else: return MessageDialog( title="e-Solution backed up successfully", message="<EM>All items in root</EM> sucessfully\ exported to <pre>%s</pre>." % f, action="manage_main") def restore(self, REQUEST=None, filename='e-Solution_backup.zexp'): """ restores backup file from 'import' """ dirname, file=os.path.split(filename) if dirname: raise 'Bad Request', 'Invalid file name %s' % filename instance_home = INSTANCE_HOME software_home = os.path.join(SOFTWARE_HOME, '..%s..' % os.sep) software_home = os.path.normpath(software_home) for impath in (instance_home, software_home): filepath = os.path.join(impath, 'import', filename) if os.path.exists(filepath): break else: raise 'Bad Request', 'File does not exist: %s' % filename conn = self._p_jar ob = conn.importFile(filepath) for k, v in ob.items(): try: self._delObject(k) except AttributeError: pass self._setObject(k, v) return MessageDialog( title="e-Solution restored successfully", message="All items from <EM>%s</EM> sucessfully\ imported into <pre>root</pre>." % filepath, action="manage_main")

1 0

Zope returns a blank page
by Tim Edwards 30 Aug '04

30 Aug '04

Hi, We are running an application on Zope that connects back to an MS SQL Server 2000 database (Using mxODBC). The app is written using ZPTs, Python scripts and ZSQL methods. For performance reasons we have put all our queries in stored procedures, often one stored procedure will be called (through a ZSQL method) to retreive the data for a page and another to save data from that page. We have a problem where some of our users get a blank page while trying to load up one of the screens in our app. This screen contains a list of employees and their hours worked per day and in total for a week. The stored procedure for this screen is a little slow (the sum calculations are done in the SQL), it takes about 2 seconds to run through the Query Analyser. When doing a 'View Source' we get just the basic HTML tags - DOCTYPE,HTML, HEAD, BODY with nothing inside of them. So far this problem has only affected users at remote (low bandwidth) sites. Any ideas what could be causing Zope to return a blank page like this? Any suggestions? Thanks Tim Edwards

1 0

Zope VHM Rewrite Help
by Heather Perkins 30 Aug '04

30 Aug '04

Hello everyone. I am trying to rewrite the URL for my Plone Portal using the Zope Virtual Hosting Monster and Apache's RewriteEngine. However, I can't quite get the rewritten URL to work. My site is located at: http://pyrite.truman.edu:8201/miller My desired url is: http://pyrite.truman.edu/millerj My VirtualHost block in Apache's httpd.conf file looks like: <VirtualHost pyrite.truman.edu> ServerName pyrite.truman.edu RewriteEngine On # currently used RewriteRule (delete this when the Zope # VHM rewrite rule is correctly configured) RewriteRule ^/millerj(.*)$ http://pyrite.truman.edu:8201/miller$1 [L,P] # uncomment this RewriteRule when you figure out how to # anchor the site properly # RewriteRule ^/millerj(.*)$ http://pyrite.truman.edu:8201/miller/VirtualHostBase/http/pyrite.truman.edu… </VirtualHost> Any help would be greatly appreciated. Thanks, Heather Coons

2 1

Out of environment space
by sboggess 30 Aug '04

30 Aug '04

I'm trying to install Zope 2.7.2 on Windows 98se, and when click "run in consol" the command prompt says "out of environment space." The Zope Book lists many environment variables, but I'm not sure which if any to change. Can someone please help me out? Thanks

2 1

Re: [Zope] dict
by Johan Carlsson 30 Aug '04

30 Aug '04

Vahur Rebas wrote: > Hi, > > I was doing some dict modifing like testvar = {'test1':{'test2':'test2value'}} > so, everything was nice until I hit my browsers refresh button as fast > as I could and my little method which did some > if not testvar.has_key('test1') > wrote my default values to my testvar > tried setting self._p_changed = 1 > but after some time(usually sooner) my values disapear and after zope restart they even don't exist anymore :( > Tried PersistentMapping, that also didn't help much :( What probably happens is that when you stress the server the REQUEST end up in different threads, with different thread contexts. Try commiting the changes by doing a get_transaction().commit() to propagate cross threads and see if that works? Regards, Johan

1 0