- Zope-Dev - Zope lists

Re: Zope-Dev digest, Vol 1 #428 - 17 msgs
by Otto Hammersmith 10 Mar '00

10 Mar '00

Wow, three XML posts so close together. :) First, I'm sure DC wouldn't have any problems with outsiders working on core parts of Zope. Witness PythonMethods. Of course, it would have to be *good* code. That said, I'd be willing to do some contract work to get such a thing done. I wasn't going to post the URL to this for a while, but since it's germange to the discussion... http://www.zope.org/Members/otto/bluesky_xml. I started writing it about a week and a half ago, just cleaned it up a bit so it's worth putting out there. Not cataloged, or even requested to be, yet. The scope there is a little larger than just XSL.. personally, I don't think it's overkill, but someone may disagree with me. :) Comments and questions welcomed... though if you post to the list, please cc me. I subscribe to the digest and will likely not see replies for a long while unless they're cc'd. -Otto. -----Original Message----- From: zope-dev-admin(a)zope.org Sent: Fri, 10 Mar 2000 11:44:20 -0800 (PST) To: zope-dev(a)zope.org Subject: Zope-Dev digest, Vol 1 #428 - 17 msgs Message: 15 From: morton(a)dennisinter.com (Damian Morton) To: <zope-dev(a)zope.org>, <zope(a)zope.org> Subject: [Zope-dev] XSL for Zope - a bounty on this? Date: Fri, 10 Mar 2000 12:20:26 -0500 charset="iso-8859-1" A number of people have expressed a desire to have XSL support in Zope. The projects I am working on also seem to be needing the application of XSL, and whilst we can (and will) use some external XSL processor, it would seem that a 'cleaner' solution would be to have XSL support directly in Zope. I originally came to want to use Zope because I needed a tool that allowed the manipulation of a persistant tree of objects; my prototype was done in XML, but it needed persistance, an interface, and a programming language, hence the use of Zope. Even having only XPath support would be an extremely usefull way for packaging up simple 'queries' on the Zope database. Ive been looking around at the Python XSL implementations and the FourThought 4suite is looking pretty damned good, especially considering their goal of eventual ODMG 2.0 support. Zope desparately needs some standards compliance, even if only for the buzzwords. The question on my mind is: how much do other people want XSL in Zope? Im willing and able to put some $$$ up for a bounty on this. Is anyone else? I guess Id also like to heard what DC has to say about this; whether they are willing to have an external group contribute/merge a key technology into their codebase, and if they have any preferences about who this group might be. [snip] --__--__-- Message: 16 From: "Josh Zeidner" <jmz_phylogenic(a)hotmail.com> To: zope-dev(a)zope.org Subject: [Zope-dev] friday afternoon comments on XML and Zope Date: Fri, 10 Mar 2000 13:27:36 EST hello, I have been watching the activity on this newsgroup while developing a commercial system in Zope, and I feel I have made a few observations which might be useful to the Zope community at large. I am seeing a lot of similarities in the phylogeny of the XML spec and Zope. Currently, I have been involved in finding a solution for the the transformation of structured data in Zope. We ( myself and Damien Morton ) have been searching for a simple way of expressing selection and expansions of a tree-like data structure. It seems to me the best way to express these type of problems is with languages like XPath and XSLT. In addition, the development of a "zope data pointer" is remarkably similar to the problem that the drafters of the XLink specification are working out( in much larger numbers and levels of experience ). So... what I am implying is, as was touched upon before in a reply to one of my posts : is that if Zope were to be more tightly coupled to the XML specifiation, it would greatly increase Zope's user base and list of adaptable technologies. Zope is just not ready and willing to have XSL, XPath, and Xlink plugged right in. In order to use such non-proprietary technologies, custom logic must be applied to Zope: why should this be? Should Zope( being that it practically IS XML already ) be more *XML compliant*? In a nutshell- more XML friendly features built into the core of Zope is something I would like to see in future releases. Other than that, thanks for all your responses- viva la Zope. -josh zeidner ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com --__--__-- Message: 17 Date: Fri, 10 Mar 2000 10:49:21 -0800 From: Nitin Borwankar <nitin(a)borwankar.com> To: Damian Morton <morton(a)dennisinter.com> CC: zope-dev(a)zope.org, zope(a)zope.org Subject: [Zope-dev] Re: [Zope] XSL for Zope - a bounty on this? Damian Morton <morton(a)dennisinter.com> wrote > The question on my mind is: how much do other people want > XSL in Zope? Im willing and able to put some $$$ up for a > bounty on this. Is anyone else? I am willing to put up $$$ for this also assuming it is to go Open Source. Nitin Borwankar, nitin(a)borwankar.com ______________________________________________________________________ Get Visto.com! Private groups, event calendars, email, and much more. Visto.com. Life on the Dot. Check it out @ http://www.visto.com/info

1 0

friday afternoon comments on XML and Zope
by Josh Zeidner 10 Mar '00

10 Mar '00

hello, I have been watching the activity on this newsgroup while developing a commercial system in Zope, and I feel I have made a few observations which might be useful to the Zope community at large. I am seeing a lot of similarities in the phylogeny of the XML spec and Zope. Currently, I have been involved in finding a solution for the the transformation of structured data in Zope. We ( myself and Damien Morton ) have been searching for a simple way of expressing selection and expansions of a tree-like data structure. It seems to me the best way to express these type of problems is with languages like XPath and XSLT. In addition, the development of a "zope data pointer" is remarkably similar to the problem that the drafters of the XLink specification are working out( in much larger numbers and levels of experience ). So... what I am implying is, as was touched upon before in a reply to one of my posts : is that if Zope were to be more tightly coupled to the XML specifiation, it would greatly increase Zope's user base and list of adaptable technologies. Zope is just not ready and willing to have XSL, XPath, and Xlink plugged right in. In order to use such non-proprietary technologies, custom logic must be applied to Zope: why should this be? Should Zope( being that it practically IS XML already ) be more *XML compliant*? In a nutshell- more XML friendly features built into the core of Zope is something I would like to see in future releases. Other than that, thanks for all your responses- viva la Zope. -josh zeidner ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com

1 0

Re: [Zope-dev] ZCatalog: Forgot to mention
by Anthony Baxter 10 Mar '00

10 Mar '00

>>> Rik Hoekstra wrote > But you can with Python/External Methods? With or without parameters? > Where then lies the difference, and why? Is it because of DTML > evaluation? I do not understand this. If your Python Method has 'self' as it's first arg, then you can access the namespace through self.foo when being cataloged. Anthony, who's going to be out of circulation for a week and a half from tomorrow - have fun! -- Anthony Baxter <anthony(a)interlink.com.au> It's never too late to have a happy childhood.

1 0

Re: [Zope-dev] ZCatalog: Forgot to mention
by Rik Hoekstra 10 Mar '00

10 Mar '00

> Tom, > > Michel can correct me if I'm wrong here, but I believe only properties > get indexed in the Catalog. YOu can also only retuen properties as > meta_data. If you want the value of a given method that comes up fora > Catalog result, you do this: > > <dtml-in "Catalog.searchResults(title='foo')"> > <dtml-with "resolve_url(getpath(data_record_id_),REQUEST)"> > <dtml-var getParentEx> > </dtml-with> > </dtml-in> > > This will return the value of getParentEx method for the record > returned. You can catalog mthod results as well. But as catalog won't pass in any parameters, you cannot use DTML Methods or Documents. Use an External Method or Python Method instead. [RH>Hm, I didn't know about this feature either (is this documented somewhere, if not, I'll try and document it when i understand it ;-)), so I was experimenting a little. Using a DTML Document/method did work. However, I did not get to the point of getting it working for a method other than with a simple text content (like '1 2 3'). This the Catalog did return. I did not yet figure out how to query it. A standard search method did not return any results. However, the report from the search interface called on its own (without using the search form from the search interface) actually _did_ return all results. Could someone help me to understand this further? Michel? By the way, wasn't it also possible to index sql methods (if called as records through the web?) Rik

2 2

Advanced ZClass Creation Question
by James W. Howe 09 Mar '00

09 Mar '00

I'm building a product with a handful of ZClasses. I've built a couple of them and I find that I'm duplicating method code everytime I create a new ZClass definition. What I would like to do is abstract out some of the commonality into a base Python class that all my ZClasses would derive from. However, my brain starts to hurt when I think about the difference between sending messages to the ZClass object itself (say createInObjectManager), and sending messages to the instance which gets created by that creation method. What I want to be able to do is have a common object creation method as well as some common instance methods. Would these all be defined in one base class? For example, right now when I create an instance of my ZClass, I compute a unique ID and then pass that value into the createInObjectManager method. To do this, I'm thinking I would create a class like this: class MyCommonZClassBase: def createNewObject(self, REQUEST=None): newid = self.computeNewId() return self.createInObjectManager(newid, REQUEST) When I defined my ZClass, I would include "MyCommonZClassBase" as a base class of my ZClass (actually I would probably create a custom Python class for each ZClass that I wanted to create and this class would derive from MyCommonZClassBase). I assume that I could then create a new object by doing: MyZClass.createNewObject(REQUEST) Now that I've created my new object, I also have some things that I want to do whenever any of my ZClasses gets instantiated. For example, I want to set properties and update a Catalog (with an arbitrary name) To do this, I would might write a method like this: def initNewObject(self, REQUEST): REQUEST.set('title', REQUEST['name']) self.unindex_object() self.manage_editCataloger(self.catalogName, REQUEST) self.index_object() My question (and confusion) is this, would I define this method in MyCommonZClassBase, or would it have to be defined somewhere else? In Smalltalk, this would be handled by having a class method called "createNewObject" and an instance method called "initNewObject". Python doesn't seem to have that distinction, so I'm confused as to what I would need to do to make this work. Any tips or insights on this matter would be greatly appreciated. Thanks. James W. Howe mailto:jwh@allencreek.com Allen Creek Software, Inc. pgpkey: http://ic.net/~jwh/pgpkey.html Ann Arbor, MI 48103

1 0

Re: [Zope-dev] Globally unique id's in Zope in a new way
by Tres Seaver 09 Mar '00

09 Mar '00

Jason Spisak wrote: > > So this is only a Win32 module? > > Tres Seaver writes: > > > "James W. Howe" <jwh(a)allencreek.com> wrote: > > > > > > At 07:49 PM 3/8/00 -0500, Jason Spisak wrote: > > > >Zopsters, > > > > > > > >I use the Catalog for quite a bit of jumping around the ZODB, and have a > > > >30,000 plus objects in a system. I like the idea of not having to worry > > > >about id's conflicting with other numbers that might appear (invoice > > > >numbers, employee numbers, etc...) so I don't like to use a counter for > > > >ids. [...] > > > > > > I have the same problem. I want to automatically create objects and I'm > > > currently using the time method. When I bulk add objects, I would get a > > > collision on the time value, so I simply remembered the last id and added 1 > > > to that value. It works for my current needs, but I wish Zope provided a > > > standard mechanism for generating unique id's. > > > > > > For example, it would actually be nice if Zope provided an implementation > > > of UUID (called GUID by Microsoft). A UUID is a guaranteed unique 64-bit > > > value. It's used in COM, CORBA and other systems which need to have > > > automatically generated guaranteed unique values. The nice thing about > > > UUID's is that they are guaranteed to be unique network wide. In other > > > words, you (supposedly) can't generate a UUID which would ever conflict > > > with a UUID that I generated. It does this by an algorithm which takes > > > into account things like time-of-date, network card address, etc. I don't > > > have the details on the algorithm but I know it is publically > > > available. It would probably best be implemented at the C level. > > > > > > Just an idea. > > > > On Win32, this is trivial in Python:: > > > > import pywintypes > > guid = pywintypes.CreateGuid() > > return str( guid ) > > > > I *think* DCE RPC has a cognate function, uuid_create(), but I don't have > > documentation for it (Sun RPC probably has a similar call). Hmmm, a search of > > the Vaults of Parnassus (http://www.vex.net/parnassus) doesn't show any existing > > wrapper for either major flavor of RPC. pywintypes is one of the Mark Hammond's Win32 extensions -- mostly they are pythonified wrappers around various useful Win32 APIs. In at least one case (the popen() call), Mark has a working version, while the version in the standard Python "os" module is broken on Win32. Mark also provides tools for building COM clients and servers with Python. Mark's extensions are at: http://starship.python.net/crew/mhammond/win32/Downloads.html -- ========================================================= Tres Seaver tseaver(a)palladion.com 713-523-6582 Palladion Software http://www.palladion.com

1 0

Re: [Zope-dev] Globally unique id's in Zope in a new way
by Tres Seaver 09 Mar '00

09 Mar '00

"James W. Howe" <jwh(a)allencreek.com> wrote: > > At 07:49 PM 3/8/00 -0500, Jason Spisak wrote: > >Zopsters, > > > >I use the Catalog for quite a bit of jumping around the ZODB, and have a > >30,000 plus objects in a system. I like the idea of not having to worry > >about id's conflicting with other numbers that might appear (invoice > >numbers, employee numbers, etc...) so I don't like to use a counter for > >ids. [...] > > I have the same problem. I want to automatically create objects and I'm > currently using the time method. When I bulk add objects, I would get a > collision on the time value, so I simply remembered the last id and added 1 > to that value. It works for my current needs, but I wish Zope provided a > standard mechanism for generating unique id's. > > For example, it would actually be nice if Zope provided an implementation > of UUID (called GUID by Microsoft). A UUID is a guaranteed unique 64-bit > value. It's used in COM, CORBA and other systems which need to have > automatically generated guaranteed unique values. The nice thing about > UUID's is that they are guaranteed to be unique network wide. In other > words, you (supposedly) can't generate a UUID which would ever conflict > with a UUID that I generated. It does this by an algorithm which takes > into account things like time-of-date, network card address, etc. I don't > have the details on the algorithm but I know it is publically > available. It would probably best be implemented at the C level. > > Just an idea. On Win32, this is trivial in Python:: import pywintypes guid = pywintypes.CreateGuid() return str( guid ) I *think* DCE RPC has a cognate function, uuid_create(), but I don't have documentation for it (Sun RPC probably has a similar call). Hmmm, a search of the Vaults of Parnassus (http://www.vex.net/parnassus) doesn't show any existing wrapper for either major flavor of RPC. Tres. -- ========================================================= Tres Seaver tseaver(a)palladion.com 713-523-6582 Palladion Software http://www.palladion.com

2 1

Excuses
by Tom Deprez 09 Mar '00

09 Mar '00

I'm terribly sorry if I annoy somebody here on the list with all my stupid q'ns. Sorry. Tom.

3 3

objects that don't acquire from themselves?
by Michael Halle 09 Mar '00

09 Mar '00

I'm trying to write a zope object that has two characteristics: 1. it is an ObjectManager, and has a collection of methods as children, 2. it acts as a document, and uses methods with the same name as its children. However, no matter from where the object is referenced (either directly or from somewhere else in the object tree by acquisition), I don't want the object to use its own child method definitions: either is acquires the methods from somewhere else or the references should fail. Is there a canonical way of doing this? Thanks. Michael Halle mhalle(a)bwh.harvard.edu

1 0

Re: [Zope-dev] Globally unique id's in Zope in a new way
by Jason Spisak 09 Mar '00

09 Mar '00

Ross, On my dual SMP PII 350 I still only create 48 objects using this short code per second. <dtml-var standard_html_header> <dtml-in "_.range(0,1000)"> <dtml-with "ClientClass.createInObjectManager(_.getitem('unique_id',1), REQUEST)"> <dtml-var id> </dtml-with> </dtml-in> <dtml-var standard_html_footer> _.float works terrific. I suppose a blend of the two techniques would give a pretty solid id, even on very fast parallel machines, because they would have to get the same random number from a large amount of possiblities in under a millisecond. Thanks for both your help on this. <dtml-var "_.str(_.float(ZopeTime())) + _.str(_.whrandom.randint(1,100000))"> All my best, Ross Lazarus writes: > ZopeTime() to float to string conversion works even better since you get > milliseconds - there is always the risk that in an SMP processor > environment, you JUST might get bitten, but that's also true of any > hash algorithm - you're down to probabilities which are exceedingly > small. > > Of course, you could enclose your new id generating code (hash or float > time) in some loop which checks to see if an object with that id > exists, if it does, wait a small random time, and add a letter to the > end of the string, retest and loop until it IS unique, then create the > object. The time taken to do all that means you could be "safe" from > multiple processors doing precisely the same thing at precisely the > same time.. > > jason spisak <44-(a)hiretechs.com> wrote: > original article:http://www.egroups.com/group/zope/?start=25478 > > Zopsters, > > > > numbers, employee numbers, etc...) so I don't like to use a counter > for > > ids. The ZopeTime() to int to string conversion works unless you are > > programatically creating multiple items in the same folder, and have a > > descent processor. I have derived the following bad hack from the > code > > Jason Spisak CIO HireTechs.com 6151 West Century Boulevard Suite 900 Los Angeles, CA 90045 P. 310.665.3444 F. 310.665.3544 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.

1 0