August 2001 - Zope-Dev

Shallow copies/links/references
by Mario Valente 08 Mar '02

08 Mar '02

Hi: OK everyone, I've managed to get object links to work. As discussed previously by others (re: object references) and asked by myself: I had the need to be able to refer to objects from different points of the folder hierarchy without duplicating those objects. Yes I know that can be done *programatically*. That is not the idea. The idea is to have a folder with differente widgets/components and to be able to copy and paste references to those widgets throughout the folder hierarchy. Of course that the usual copy/paste is not OK, since that *duplicates* the objects and as such further changes to the original object wont be reflected in the copies. With something like this + Ordered Folders you are able to have an index_html that just renders the objects within a folder and you can then drop objects/widgets into the folder to produce a page. What I did: - edited CopySupport.py - copied the manage_pasteObjects method to a manage_pasteMonikers method - commented the #ob=ob._getCopy(self) line (the duplicate object part) - added a "Paste Ref" button to the lib/python/OFS/dtml/main.dtml file that calls the pasteMonikers method """ <input class="form-element" type="submit" name="manage_pasteMonikers:method" value="Paste Ref." /> """ It works. It does what I want/need. When you change a reference, the original also changes. When you change the original, the references change. You can delete either the reference or the original and all is OK. Issues/TODO - find a way to distinguish references from the original (tried to change the meta_type at paste time but no luck). At least the meta_type/icon should change to provide a visual cue. - provide tool(s) to find impact of changing a reference. Provide a link from a reference object to the original. Provide a link from an object to its several references. Comments and experiences are welcome. C U! -- Mario Valente

2 2

Merhabalar;
by Can Canbilek 13 Dec '01

13 Dec '01

Öncelikle rahatsýz ediyorsak özür dileriz ama bu maili okursanýz sanýrým bize teþekkür edeceksiniz. Sizlere, hiçbir masraf yapmadan yalnýzca internete girerek para kazanacaðýnýzý iddia ediyoruz. Hiçbir yalan dolan, hile olmadan.... Üstelik kredi kartý numaranýzý bile vermeniz gerekmiyor. Ufak bir formu dolduruyorsunuz, dünyaca ünlü bir þirket olan ve dünyanýn heryerine ( Türkiye dahil olmak üzere ) para yollayan bu þirketin küçük bir programýný download ediyorsunuz. Daha sonra her internete giriþinizde bu programý açýyorsunuz ve program nete ne kadar girdiðinizi hesaplýyor. Nete girdiðiniz her saat baþý 50 cent alýyorsunuz ve bunu karþýlýðýnda ekranýnýzýn küçük bi bölümünde reklamlar görünüyor... Ayda yalnýzca 60 saat internete girerek 30$ kazanýyorsunuz tabii daha fazla kazanmak isterseniz arkadaþlarýnýzý üye ediyorsunuz. Ýster saadet zinciri, isterseniz zeka kullanýmý deyin, bence bundan daha iyi bir yöntem yok.... Ayrýntýlý bilgi gelen çeklerimizin resimleri ve olasý sorularýnýzýn cevaplarý için aþagýdaki sitelerden herhangi birine giriniz(hepsi aynýdýr): http://www25.brinkster.com/firsatlar/index.htm http://www25.brinkster.com/hizmetlerimiz/index.htm http://www25.brinkster.com/spediainc/index.htm http://www25.brinkster.com/sendekazan/index.htm Saygýlarýmla... !!!!!!!SORULARINIZ ÝÇÝN LUTFEN cancanbilek2(a)hotmail.com ADRESÝNE MAÝL ATINIZ, BU MAÝLE REPLY YAPMAYINIZ!!!! ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- BU MAILI YANLISLIKLA ALDIGINIZA INANIYORSANIZ, YADA BIRDAHA BIZDEN MAIL ALMAK ISTEMIYORSANIZ http://www21.brinkster.com/dogangoko/index.htm ADRESÝNE GÝRÝN VE SAYFADAKÝ MAÝLLÝSTÝMÝZDEN ÇIKMAK ÝÇÝN BURAYA EMAÝLÝNIZÝ YAZINIZ BOLUMUNE EMAÝLÝNIZI YAZINIZ... ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- EGER BU MAILI 2CÝ KEZ ALIYORSANIZ LÜTFEN DÝKKATE ALMAYINIZ.... ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

4 3

Resize a image stored in a Oracle's BLOB Field
by Juan Javier Carrera Obrero 27 Nov '01

27 Nov '01

Hello, Thanks, I have displayed the images which them was stored in a Oracle's BLOB field, with: <dtml-call expr="RESPONSE.setHeader('content-type','image/jpeg')"> <dtml-in SQLMethod_with_a_Only_Row> <dtml-var expr="BLOB_FIELD.read()"> </dtml-in> However, I would like to resize this image. I can not do it with the "height" and "width" attributes, because the image is obtained from a Oracle's BLOB field, and no from image object (i.e.: I can not use <dtml-var "ImageObject.tag(border='5', align='left')">). How can I do it ? Thanks again !

4 3

manage_pasteObjects and REQUEST
by Danny William Adair 19 Sep '01

19 Sep '01

Hi! Does anyone know when manage_PasteObjects will stop requiring REQUEST to be passed? I want to mimick the Cut/Copy/Paste Support of the management interface in a UI of my own. my method "objectsCut" holds: <dtml-call "manage_cutObjects(REQUEST['ids'], REQUEST)"> while my "objectsPaste" method holds: <dtml-call "manage_pasteObjects(_.None, REQUEST)"> Pasting objects will sometimes (!?!) redirect me to the management interface, which I want to hide. I found these last lines in manage_pasteObjects (CopySupport.py): def manage_pasteObjects(self, cb_copy_data=None, REQUEST=None): """Paste previously copied objects into the current object. If calling manage_pasteObjects from python code, pass the result of a previous call to manage_cutObjects or manage_copyObjects as the first argument.""" ... ... ... if REQUEST is not None: REQUEST['RESPONSE'].setCookie('__cp', 'deleted', path='%s' % cookie_path(REQUEST), expires='Wed, 31-Dec-97 23:59:59 GMT') REQUEST['__cp'] = None return self.manage_main(self, REQUEST, update_menu=1, cb_dataValid=0) I don't want to hack this file. The only way to get the cookie set seems to be passing the REQUEST object. "...pass the result of a previous call to manage_cutObjects or manage_copyObjects as the first argument" This doesn't help me very much. It would help me if I wanted to Copy/Paste or Cut/Paste at the same time: <dtml-call "REQUEST.set('my_clipboard_data', manage_cutObjects(REQUEST['ids'], REQUEST))"> <dtml-call "manage_pasteObjects(my_clipboard_data)"> works fine, but if I want to do Copy and Paste at completely different times, I have to pass my_clipboard_data along the requests all the time. Is there a simple solution? Is it something I should put in the Bug Collector? Or did I miss something? Thank you very much for your help, Danny

3 2

Problems with CatalogAware?
by sean.upton＠uniontrib.com 10 Sep '01

10 Sep '01

I have a python product that has a class derived from CatalogAware and Folder. I programmatically add instances of these (1000s of them) on an automated basis every day in the early morning. My problem is that whenever I do this, the Catalog is updated with some problems: 1 - Metadata is skewed: DateTime objects are cataloged, and they end up with the wrong date ('2001/01/01') instead of today's date, even though the actual property contains today's date. The index in question is built from a property. 2 - Text indexes only get a portion of their text stored in the metadata, not all of it. The index in question is built from a class method in the product code. Using a bulk reindex solves this problem (subtransaction off, haven't tried it any other way). I'm using Zope 2.3.2 with ZEO 1.0b3 using a VHM. The automated process queries SQL for data, and uses those queries to create object instances of the previously mentioned catalog-aware class, and I create anywhere from 2000-7000 objects at a time. I'm not sure if this is due to the size of the transaction I am running. I haven't tried this in a smaller transaction yet; I was wondering if anyone else has seen this. Any thoughts? Sean ========================= Sean Upton Senior Programmer/Analyst SignOnSanDiego.com The San Diego Union-Tribune 619.718.5241 sean.upton(a)uniontrib.com =========================

2 1

cvs.zope.org
by Chris Withers 31 Aug '01

31 Aug '01

Just me or is cvs.zope.org dead? cheers, Chris

3 3

OOP Suggestion: allow local roles to restrict and not only extend user's roles.
by Philipp Robbel 31 Aug '01

31 Aug '01

Dear everyone, I'm badly stuck with my problem and while I think that I'm using valid OOP design I don't think that Zope currently supports this. Here's a description of my current design: In our company there are different research groups whom I gave their own subfolders. /databases /bioinformatics etc. Within each group there are pseudo-managers defined (i.e. they have most of the full manager role, i.e. can create/delete documents etc., so I'll use "manager" as role when describing my problem.) I want all groups to share a common set of both DTML methods and SQL functions, so I placed them in the root folder. In the root folder there is for example a new management function that - when called like /databases/show_management_screen - will show the management options for the group specified by the containing folder. It will also allow database inserts depending on the containing folder. I do that in my inherited "show_management_screen" dtml method by using statements like: <dtml-call "sqlAddProjects(ResearchGroupName = id, ...)"> which will do: INSERT INTO table (...) VALUES(...) WHERE ResearchGroupName=<dtml-sqlvar id type=nb> It's critical that a manager from /databases is *not* allowed to access the sql method in the /bioinformatics group, else he would be able to insert values under a fake GroupID by calling the method like /bioinformatics/sqlAddProjects directly. Another thing that's critical is that no researcher from any group nor the anonymous user has direct access to this Database method: that's why I not only placed them in the root folder but also restricted the "Access contents information", "Use Database Methods" and "View" permission of that sqlAddProjects Method to Managers. Problem which is solved by that: no anonymous user can access that method. Problem that comes up: no Researcher (i.e. subfolder manager) can access that method either, because root manager is, of course, != subfolder manager. Possible solution for inherited DTML methods: a) Disable that "Access contents information", "Use DB Methods" and "View" permissions protection in root folder. b) Check for permissions in the method itself. <dtml-if "AuthenticatedUser.has_role(this(), ['Manager'])"> as outlined in http://www.zope.org/Members/tseaver/user_settable_passwords. It's supposed to check, whether the accessing user has the Manager Role in the context the DTML method is called on. But this fails for SQL Methods and I'm pretty confident that there should be a simpler way in Zope itself. My idea to fully support this OOP design was the following: Enable local roles not only to extend the user's roles but also restrict them. I.e. in my case I would define all researchers from the various groups in the root acl_users folder as Managers. But - at the same time - change the local roles of the root folder to _not_ give the researchers the Manager role in the root folder (understandable?) In the /databases folder, however, I would give the database-researcher back his "Manager" role, via local roles. This would solve my whole problem: Step 1: restrict again all critical dtml and sql methods in the root folder to managers (as already outlined above: "...but also restricted the "Access contents information", "Use Database Methods" and "View" permission of that sqlAddProjects Method to Managers"). Step 2: If the database researcher calls the "show_management_screen" dtml method like /databases/show_management_screen it would work because a) he is defined as Manager in the root folder (but local roles restrict them there to be a "real" Manager) b) he has the proper local roles set in his /databases folder so that there he finally is a "real" Manager. He couldn't, however, access the function like /show_management_screen (i.e. in the root folder) or like /bioinformatics/show_management_screen: both would be restricted by the local rules. Shouldn't that be valid OOP design? Please let me know what you think or - if I oversaw another solution to my problem - please let me know how to solve it :-) My original problem was outlined in my previous post at http://groups.yahoo.com/group/zope/message/70099 where I unfortunately didn't get any replies. Thanks in advance, Philipp Robbel

1 0

How to run some external method as diffrent user
by Maciek 31 Aug '01

31 Aug '01

Hello, Is any elegant way to do that in ZOPE, without running zope as root? Zope is running on Linux, so I use sudo and commands.getstatusoutput (python function) for now. Best regards from Poland, Maciej Łotysz

1 0

view data in ZODB
by Sunil Kumar Gupta 31 Aug '01

31 Aug '01

Hi, If any one of you know how to view data stored in ZODB please reply. Regards, Sunil

2 1

python w/o py_malloc
by Dario Lopez-Kästen 31 Aug '01

31 Aug '01

Hello! I am having my Zope.2.4.0 repeatedly restarting (using z-monitor). I am using python 2.1.1 with Zope source-release. There was a discussion on the list recently about recompiling python w/o py_malloc, and that that would solve the problems. How do I recompile python w/o py_malloc? I've tried the python dist but couldn't find any docs... Please help... /dario - -------------------------------------------------------------------- Dario Lopez-Kästen Systems Developer Chalmers Univ. of Technology dario(a)ita.chalmers.se ICQ will yield no hits IT Systems & Services

2 2