September 2000 - Zope-Dev

Zope Error message
by Andy McKay 27 Sep '00

27 Sep '00

So now I need to change it so most normal errors (such as 404) are handled by standard_error_message. So I went to HTTPResponse.py and found this: def _error_html(self,title,body): # XXX could this try to use standard_error_message somehow? I gather someone smarter than me has tried this before. Before waste a day going down this path, does anyone have any advice? -- Andy McKay, Developer. ActiveState.

2 1

Proposal: Local ZClasses
by Jay, Dylan 27 Sep '00

27 Sep '00

Problem: Encapsulation for plugable brains and other purposes are too hard. ZClasses are a global kind of thing that requires going into a separate area of the site which you have to have permission to use. Also I find that I evolve my solutions. I will produce something in folders that works for what I want and then try and generalize it into ZClasses later. However if my solution in the folders itself wants to use classes then the only choice I have is external methods or ZClasses. Solution: Local ZClasses: A ZClass that can be created in any folder and is available with the acquisition path. It can then be used for pluggable brains with sql methods also where the Class can be reached in the acquisition path. e.g. I have a few fancy display methods that I use for display results of a sql query for getting phone contact info e.g. create a vCard for each entry. I create a ZClass with a property sheet matching those returned by the sql method. I create my display methods within this ZClass. I associate the sql method with the ZClass. I call the query with dtml-in. Within the loop I have access to the ZClasses methods. Second solution: Prototype objects: A folder is an object with methods etc. A ZClass is not much different except that it has a constructor and a list of the properties that are always there. With a bit of syntax we could make a local folder into a local ZClass. We just need a method that says, using this folder as a prototype, create me a new object. e.g. <dtml-call "REQUEST.set('myobj',_.new(users_class, name='Dylan'))">. This would effectively make a new non persistent object that inherits from the prototype folder I guess. This could be done with plugable brains as well by just specifying the folder as if it was a class and then new objects would created for each row. If you enable the ability to make these prototype objects persistent then you get the side effect that you can create symbolic links. You can access a whole other acquisition path using inheritance. In fact you could have a property of a folder which is its inheritance objects which you could change over its life time. In fact any object could have this property could affect an objects behavior without moving the object or modifying its underlying classes. Also it would mean overload any kind of object. Just create a folder that inherits from an object or class and add methods. Personally I think the second solution has a lot of merit. What do others think?

1 0

New Proposal: PersistentBlob
by John D. Heintz 26 Sep '00

26 Sep '00

Hi all, Yesterday I put up a new proposal on the http://dev.zope.org site that documents my vision of PersistentBlob. The main idea is the be able to mark blocks of strings (text or binary) content for storage in a file system as individual files. (Or blobs in a relation table.) The two motivations for this are keeping a ZODB FileStorage small and exposing blocks of string content to external processes (i.e. full-text indexing). Thanks for taking a look. John -- John D. Heintz DataChannel, Inc. Senior Engineer 512-633-1198 jheintz(a)isogen.com

1 0

ZServer and ftp port-1 data channel
by kent＠tiamat.goathill.org 26 Sep '00

26 Sep '00

It appears we still have a problem with ZServer's port 20 data channel response behavior with the latest released Zope (Zope 2.2.1 (source release, python 1.5.2, linux2) and Active Mode ftp clients. In particular, when you select the port-1 behavior, medusa can't seem to bind to port 20 when port 21 command channel is specified. "ftp_server.py" line 278: if self.bind_local_minus_one: cdc.bind (('', self.server.port - 1)) After verifying that self.server.port - 1 = 20, the current behavior is that medusa simply hangs at the bind statement. If there was a permissions problem my understanding (and experience) is that one would receive: socket.error: (13, 'Permission denied') and not a process hang. Any idea why medusa simply hangs on the attempted bind? Thanks

2 3

Testing Zope applications (was Re: [Zope-ZEO] Advice)
by Jim Fulton 26 Sep '00

26 Sep '00

Note that this conversation hasn't had anything to do with ZEO for some time, so I'm moving it over to zope-dev. Toby Dickenson wrote: > (snip) > >I think it is really much easier to use ZPublisher/Test > >(which is also available as Zope.debug: > > > > import Zope > > Zope.debug(url) > > > >This provides a much thinner and more easily > >controlled environment. > > ... but not a complete replication of the deployed environment. For > example, you cant test RESPONSE.write. This sounds like a bug that needs to be fixed, not a limitation in the approach. Would you mind submitting a collector item on it? > From a philosophical point of view, using Zope.debug is wrong because > the whole purpose is an integration test of the interactions of your > pre-tested components with the Zope environment (depending on the > complexity of your glue, you may also think of it as a unit test of > that glue). But Zope.debug uses the Zope environment. Unless you consider ZServer a critical part of the Zope environment. It is extremely rare for the difference between something like ZServer and ZPublisher.Test to have any noticable impact on the application behavior. (snip) > >> The key to this is (as always) ensuring that your design process > >> considers the testability of the glue. > > > >What do you mean by glue? > > The layer which allows Zope-ignorant code to be used as a Zope > product. ZCatalog.py is an example of glue for Catalog.py. I think that ZPublisher.Test provides a pretty good test harness for anything below the publisher level, such as ZCatalog. (snip) > >How does this make anything untestable? You certainly can decide > >correct and incorrect behavior. I think there is much more possibility > >of problems due to UI changes or dynamism that make analysis of > >test results difficult. > > I suspect our difference in opinion is the scope of what we are > testing. Is it a bug if a product can be broken by a Manager using the > management interface of an instance of a different product? > > Should this 'bug' be tested for? > > If so, then a test plan must verify that every dtml-namespace and > every acquisition name lookup can never be subverted by an > ObjectManager or PropertyManager instance, no matter how the instances > are arranged in containers. > > The alternative is that the Manager take responsibility for not > creating any objects with a conflicting name - but who decides which > names are disallowed. How do you test that this list is complete? First, as an aside, you can't prove that non-trivial software is correct via testing. :) Second, part of a test is deciding what the interface to the tested software is. One way to limit the scope of testing is to limit the promises made, for example to leave the behavior undefined under some circumstances. > I dont see a way to test this constraint, and it has proven impossible > to avoid the problems using design rules. I recently checked some of > our recent products using strategically placed debugging __getattr__ > hooks - with initially horrifying results. It sounds like this is a problem that needs to be addressed. It's not really a testing problem, but an architecture problem I'm not sure exactly what problem you are refering to. It sound's like an issue of depending on a specific acquired name and having the name overridden with something bogus. Is that it? There have been a couple of recent developments that I think could help this: 1. It's much easier than it used to be to name objects using physical paths and converting the physical paths to objects. This would allow you to refer to a particular object using an absolute reference. 2. There will be a new interface in Zope 2.3 that will allow you to prevent a name from being used lower in a containement hierarchy. This change is documented at: http://dev.zope.org/Wikis/DevSite/Proposals/ReplaceableProperty. The work has already been done and checked into CVS. I've asked Shane, the author, to update the interfaces wiki to capture this change. Jim -- Jim Fulton mailto:jim@digicool.com Python Powered! Technical Director (888) 344-4332 http://www.python.org Digital Creations http://www.digicool.com http://www.zope.org Under US Code Title 47, Sec.227(b)(1)(C), Sec.227(a)(2)(B) This email address may not be added to any commercial mail list with out my permission. Violation of my privacy with advertising or SPAM will result in a suit for a MINIMUM of $500 damages/incident, $1500 for repeats.

2 4

aq_inner: don't call it!
by Chris Withers 26 Sep '00

26 Sep '00

After more playing I've discovered that calling the resutl of aq_inner() appears to be a very bad idea ;-) The line: print self.aq_inner() in the __call__ method of a python product caused the python process to dump an 'unknown software exception'. Should that be so? Also, I thought aq_inner gave the contaiment context (presumably as a list?) but it just returns a single object. So, how can I get the containment context of an object as a list? cheers, Chris

5 7

ZCatalog : UTF-8 Chinese
by Sin Hang Kin 26 Sep '00

26 Sep '00

HI, I have a project which need search with Chinese. I think I can make Zcatalog to search Chinese in utf8. So I change Voodoo Kludge Splitter.py to convert the input string to unicode (assume it is utf8) and make a version of split of my own (see the attached splitter.py). I borrow (stolen) from Interscript the utf8 encoding conversion scheme. I separate the chinese with space by hand hoping the Zcatalog will work. After changing these, I have a catalog which looks good : I can see from the volucably the chinese were actually there (except with some which have html encode like < inside the utf8. I generate the search interface, and test it. However, the search of the index terms return nothings. I search most entries found in the vocubalury but none works, those work will return many unwanted results also. What is causing this failure? What I can do to go further? Rgs, Kent Sin --------------------------------- kentsin.weblogs.com kentsin.imeme.net

2 2

__call__ hell :-(
by Chris Withers 26 Sep '00

26 Sep '00

:-( Right, I have a Python product with a __call__ method that is (hopefully) going to be used to render a navigation tree. To do this (and have the tree expand in the right places) I need a list of objects that need to be expanded. I'm currently using the following expression to get this list: objects = copy(self.REQUEST.PARENTS) objects.reverse() Now, to this I need to append the PARENTS[0-1] object, if you see what I mean. In DTML, it would be this(), but I'm stumped as to what I should append in Python. self is just the navigator object. If I redefine __call__ as def __call__(self, client, REQUEST) I just get a TypeError: Error value: not enough arguments; expected 3, got 1 If I use a safer signature def __call__(self,client=None) then client is always None, and so not what I'm after at all. I had a look at DTMLMethod.py and PythonMethod.py to try and get a clue btu just got very scared. Can someone please help me out here? cheers, Chris PS: If there's a totally different way to get the containment list of objects that need to be expanded, I'd love to hear it :-)

1 0

Re: [Zope-dev] ZSQL Methods from Python?
by Jim Fulton 25 Sep '00

25 Sep '00

"Phillip J. Eby" wrote: > > At 08:43 AM 9/22/00 -0400, Jim Fulton wrote: > >> Also, does anyone know of any work done to extend ZSQL Methods to allow > >> stored-procedure calls? > > > >No, but I'd love to see someone tackle it. The semantics > >of stored procedures varies so widely accross databases, that > >I doubt that it would be easiliy generalizable. I think, at > >least for starters, a form of stored procedure support for > >Oracle would make alot of sense. > > > > Ty and I have put together a Stored Procedure method for Sybase; it > requires a minor patch to ZSybaseDA, however, to allow for the status code > return. I'm not sure how useful it would be to anyone else, though, since > all we ever do with stored procedures is process the return value, and > ignore any actual SELECT results. I think it would be useful. Can you get multiple return values? Jim -- Jim Fulton mailto:jim@digicool.com Python Powered! Technical Director (888) 344-4332 http://www.python.org Digital Creations http://www.digicool.com http://www.zope.org Under US Code Title 47, Sec.227(b)(1)(C), Sec.227(a)(2)(B) This email address may not be added to any commercial mail list with out my permission. Violation of my privacy with advertising or SPAM will result in a suit for a MINIMUM of $500 damages/incident, $1500 for repeats.

3 3

Suggestion for better Zope error response
by Skip Montanaro 25 Sep '00

25 Sep '00

Suppose you have a simple form like: <html> <form action="http://www.musi-cal.com/search" method="post"> City: <input name="performers" type="text"> Radius: <input name="radius:int" type="text"> <input type="submit" name="submit" value="Search"> </form> </html> and the user submits the form with an invalid input for the radius. Assuming ZServer was not run with -D, other than the standard boilerplate, all the user sees is exceptions.ValueError However, field2int raised ValueError with a string like ValueError: An integer was expected in the value 'wer' Making this available to the user, even in the non-debug case, seems like it would be useful feedback for the user that could help guide them to their mistake. -- Skip Montanaro (skip(a)mojam.com) http://www.mojam.com/ http://www.musi-cal.com/

1 0