February 2000 - Zope-Dev

Properties Introspection
by Josh Zeidner 28 Feb '00

28 Feb '00

Hello, I am trying to develop a way to dynamically introspect an objects properties( specifically, a ZClass ). How can I determine which sub-objects of a Zope object are properties? Can I do this in DTML or do I need to use python? -thanks Josh Zeidner ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com

1 0

Re: [Zope-dev] Compiling Zope with cygwin
by Tres Seaver 28 Feb '00

28 Feb '00

Michel Pelletier <michel(a)digicool.com> > To: zope-dev(a)zope.org > Subject: > > Has anyone gotten this to work? I can get Python to compile, and I can > get Zope to compile, but I can't get Python to compile w/ threads, and > thus Zope complains and bombs. Does anyone know if it's possible to > compile a python extension against one python but run it with another? > I'll keep playing if no one has a better clue. I think you want to be using the mingwin32 stuff to get threading right (the plain ole cygwin libc is flaky about threads -- mingwin32 uses the MS RTL). Robert Kern has a set of directions up about it on the starship, under: http://starship.python.net/crew/kernr/ -- ========================================================= Tres Seaver tseaver(a)palladion.com 713-523-6582 Palladion Software http://www.palladion.com

1 0

zClasses with zClasses - can they act in a folderish way?
by morton＠dennisinter.com 28 Feb '00

28 Feb '00

Ive been experimenting with zClasses nested within zClasses. I can get all kinds of behaviours happening and its extremely usefull. What I cant seem to achieve is to get a nested zClass to display in a folderish manner in the managment tree. I can get zClasses at the top level of a product to display in a folderish way (i.e. having a '+' control to expand its subtrees), but I cant get a zClass embedded inside a zClass to display this way. Ive tried subclassing from Folder, ObjectManager, nothing, etc etc, but nothing seems to work. Any suggestions anyone?

1 0

Compiling Zope with cygwin
by Michel Pelletier 28 Feb '00

28 Feb '00

Has anyone gotten this to work? I can get Python to compile, and I can get Zope to compile, but I can't get Python to compile w/ threads, and thus Zope complains and bombs. Does anyone know if it's possible to compile a python extension against one python but run it with another? I'll keep playing if no one has a better clue. -Michel

1 0

Re: [Zope-dev] Zopifying an FTP'd/POST'ed HTML document?
by Graham Fawcett 28 Feb '00

28 Feb '00

Chris Withers wrote: > Hi, > > Could one of you two please forward the response to the question to the zope > list, 'cos it didn't make it and I'd really like to know what it was :-) > > cheers, > > Chris Here it is, Chris. Itamar Shtull-Trauring wrote: > Graham Fawcett wrote: > > > > Has anywone worked on a mechanism for taking HTML documents that are FTP'd > > or HTTP-POST'ed by end-users and manipulating them into Zopish documents > > -- inserting and <..._footer> tags, for instance? > > Check out load_site.py in Zope's utilities folder, expecially the -D > option. Or my version of this code: > http://www.zope.org/Members/itamar/load_sitepossible only between equals > > Itamar S.T. itamars(a)ibm.net A good approach to the problem--it keeps all the hacked-off bits of the original message. When I get the time, I think I'll still hook this into a custom FTPd to Zopify/unZopify HTML documents published to the server. (The usefulness of an unZopify method might be questionable in some cases, but I'd like to be able to "give back what I got" if that's what the client asks for.) And BTW, Itamar, sorry for the (twice! oops..) misplaced thanks--whether you're the author or not, I'm equally grateful for your help. -- Graham

1 0

RE: Why missing connection._storage.load?
by Martijn Pieters 28 Feb '00

28 Feb '00

From: Loren Stafford [mailto:lstaffor@dynalogic.com] > > My product's initialize(context) procedure creates a ZCatalog if that > catalog doesn't already exist. I've defined an additional method > armed_event() for the Catalog object. Later, the same product > _init_.py > module calls armed_event() (OK -- it's called from a thread spawned by > initialize(context)). Do you create a new DB connection? > > When the catalog is created in the same product load process, > the call to > the catalog method armed_event() yields this output from pdb: > > > > C:/Zope214\lib\python\Products\ZScheduler\ZSchedule.py(47)arme > d_event() > > ...which is what I would expect. > > However, when the catalog already exists (from a previous > product load), > the call to the catalog method yields this output: > > > C:\Zope214\lib\python\ZODB\Connection.py(386)setstate() > > I didn't write any setstate methods, so I presume this is > extension class or > persistence voodoo. If so, the fact that pdb doesn't show how > it got to > setstate must be due to the fact that that logic is written > in C. Somewhere, > somehow, (perhaps via an inherited __setstate__ method) Zope > has figured out > that the catalog object has to be loaded from the ZOSB before it can > proceed, and calls on Connection.py\setstate to help do so. > Right, so far? > Indeed, via the Persistence class, the ZODB is asked to retrieve the object. How do you access the Catalog? You should obtain a new Root object for your new thread, and traverse to the Catalog via that reference. When you use the ZServer Monitor, you have to do the same thing: import Zope app=Zope.app() You then have your own connection to the database. When you are done with it (basically every time you go to sleep in the thread), you close it again: app._p_jar.close() del app This way you keep conflicts down, performance up. > OK, tracing along in setstate(), we get to this statement and > exception: > > > C:\Zope214\lib\python\ZODB\Connection.py(391)setstate() > -> p, serial = self._storage.load(oid, self._version) > (Pdb) s > AttributeError: "'None' objec...ribute 'load'" > > C:\Zope214\lib\python\ZODB\Connection.py(391)setstate() > -> p, serial = self._storage.load(oid, self._version) > (Pdb) > > OOps! This connection's storage object has no load method. > How can that > happen? Because pdb isn't showing how it got here, it's hard > to see what the > cause is. Can someone who's been here before help me out? > > Was the Catalog object not defined properly or not > initialized properly? If > not, how is the Catalog storage object supposed to be initialized? The storage is not defined. Probably because the Conection has been closed in the meantime, which is what happens as soon as initialisation of products is done. See lib/python/Zope/__init__.py HTH -- Martijn Pieters, Software Engineer | Digital Creations http://www.digicool.com | Creators of Zope http://www.zope.org | mailto:mj@digicool.com ICQ: 4532236 | PGP: http://wwwkeys.nl.pgp.net:11371/pks/lookup?op=get&search=0xA8A32149 -------------------------------------------

2 1

Re: [Zope-dev] Zopifying an FTP'd/POST'ed HTML document?
by Itamar Shtull-Trauring 28 Feb '00

28 Feb '00

Graham Fawcett wrote: > Thanks, Itamar! This is brilliant -- you've done a thorough job here. Not me - Digital Creations and Oleg Broytmann did all the work. -- -= This is NOT a pyramid scheme =- The SNAFU Principle: True communication is possible only between equals Itamar S.T. itamars(a)ibm.net

1 0

large data.fs with zope
by ethan mindlace fremen 28 Feb '00

28 Feb '00

zopistas, The 2.3.27+ version of the kernel supports large file sizes. I have 2.3.47 installed, and I've set the block size on ext2 to 4k, so with my kernel, i should be able to do files of 1TB. However, many of the userspace programs in linux don't recognize a file that's larger than 4GB: for example, ls reports that my 16GB test file is 64MB. What I'm wondering is whether Zope would have any problems with a large Data.fs on linux and/or any suggestions on how to test it. I expect my Data.fs to grow to 26GB in size... Thanks, Ethan Fremen http://imeme.net

4 3

File name
by Petr Knapek 28 Feb '00

28 Feb '00

Hi all, my question is very easy. Bug? When I uploaded an external file from Linux system into my zope powered site, the parser correctly distinguished the name and path. On the other hand in the case an external file uploaded from MS DOS like system the parser didn't distinguished it and put the name together with path into the variable 'filename' of uploaded 'fileObject'. I use Zope deb package 2.1.3 on Linux. Is it the problem of filename parser and thus bug? Thanks, Petr -- Petr Knápek PVT.NET, e-mail: petr.knapek(a)pvt.net, tel:+420-5-41 558 394 PVT a.s. - pvt.net, Veveří 102, 659 10 Brno, Czech Republic

1 0

Why missing connection._storage.load?
by Loren Stafford 28 Feb '00

28 Feb '00

My product's initialize(context) procedure creates a ZCatalog if that catalog doesn't already exist. I've defined an additional method armed_event() for the Catalog object. Later, the same product _init_.py module calls armed_event() (OK -- it's called from a thread spawned by initialize(context)). When the catalog is created in the same product load process, the call to the catalog method armed_event() yields this output from pdb: > C:/Zope214\lib\python\Products\ZScheduler\ZSchedule.py(47)armed_event() ...which is what I would expect. However, when the catalog already exists (from a previous product load), the call to the catalog method yields this output: > C:\Zope214\lib\python\ZODB\Connection.py(386)setstate() I didn't write any setstate methods, so I presume this is extension class or persistence voodoo. If so, the fact that pdb doesn't show how it got to setstate must be due to the fact that that logic is written in C. Somewhere, somehow, (perhaps via an inherited __setstate__ method) Zope has figured out that the catalog object has to be loaded from the ZOSB before it can proceed, and calls on Connection.py\setstate to help do so. Right, so far? OK, tracing along in setstate(), we get to this statement and exception: > C:\Zope214\lib\python\ZODB\Connection.py(391)setstate() -> p, serial = self._storage.load(oid, self._version) (Pdb) s AttributeError: "'None' objec...ribute 'load'" > C:\Zope214\lib\python\ZODB\Connection.py(391)setstate() -> p, serial = self._storage.load(oid, self._version) (Pdb) OOps! This connection's storage object has no load method. How can that happen? Because pdb isn't showing how it got here, it's hard to see what the cause is. Can someone who's been here before help me out? Was the Catalog object not defined properly or not initialized properly? If not, how is the Catalog storage object supposed to be initialized? -- Thanks -- Loren

1 0