April 2001 - Zope-Dev

Catalog in 2.3.1
by Adrian Hungate 19 Apr '01

19 Apr '01

Are there some problems with the new catalog? I have spotted two possible problems: 1) In CatalogAware, there is a function reindex_all, which appears to have a few problems, like calling index_object instead of reindex_object. Also, why "sub = self.objectValues()" when sub is never used? 2) If I search a text index for a value that it does not contain, I get nothing back (No surprises there), but if I search for a word that is a substring of a word in the index, I get a Type Error (expected integer key). Did I do something wrong, or have I hit a bug? Here is the traceback of the following search. A Title index that contains the word "wibble", search for the single word "wib": Traceback (innermost last): File C:\PROGRA~1\WebSite\lib\python\ZPublisher\Publish.py, line 223, in publish_module File C:\PROGRA~1\WebSite\lib\python\ZPublisher\Publish.py, line 187, in publish File C:\PROGRA~1\WebSite\lib\python\Zope\__init__.py, line 221, in zpublisher_exception_hook (Object: catalog_support) File C:\PROGRA~1\WebSite\lib\python\ZPublisher\Publish.py, line 171, in publish File C:\PROGRA~1\WebSite\lib\python\ZPublisher\mapply.py, line 160, in mapply (Object: results) File C:\PROGRA~1\WebSite\lib\python\ZPublisher\Publish.py, line 112, in call_object (Object: results) File C:\PROGRA~1\WebSite\lib\python\App\special_dtml.py, line 127, in __call__ (Object: results) File C:\PROGRA~1\WebSite\lib\python\DocumentTemplate\DT_String.py, line 538, in __call__ (Object: results) File C:\PROGRA~1\WebSite\lib\python\DocumentTemplate\DT_Let.py, line 146, in render (Object: rs="Catalog.searchResults(REQUEST)") File C:\PROGRA~1\WebSite\lib\python\DocumentTemplate\DT_Util.py, line 334, in eval (Object: Catalog.searchResults(REQUEST)) (Info: REQUEST) File <string>, line 0, in ? File C:\PROGRA~1\WebSite\lib\python\Products\ZCatalog\ZCatalog.py, line 535, in searchResults (Object: Traversable) File C:\PROGRA~1\WebSite\lib\python\Products\ZCatalog\Catalog.py, line 651, in searchResults File C:\PROGRA~1\WebSite\lib\python\Products\ZCatalog\Catalog.py, line 542, in _indexedSearch File C:\PROGRA~1\WebSite\lib\python\SearchIndex\UnTextIndex.py, line 513, in _apply_index File C:\PROGRA~1\WebSite\lib\python\SearchIndex\UnTextIndex.py, line 576, in query File C:\PROGRA~1\WebSite\lib\python\SearchIndex\UnTextIndex.py, line 616, in evaluate File C:\PROGRA~1\WebSite\lib\python\SearchIndex\UnTextIndex.py, line 448, in __getitem__ TypeError: (see above) -- Adrian Hungate Manager, European I.S. Acucorp UK Limited

3 2

Building a Zope Site from a Source Distribution
by Eric Roby 19 Apr '01

19 Apr '01

Could anyone point me to documentation or instructions on how to build Zope to a particular version of Python. My goal is to understand the process so that I can do this on the Win32, Linux RPM and Solaris distributions (make directives, etc...). My first attempt will be focused on the Win32 distribution. I have Borland's C++ compiler (the one that has been released to the public). I have looked at the Win32 source distribution and the build_extensions.py file makes reference to setup20 and setup15 files in the lib/python folder. They aren't there so I am a bit puzzled as to the purpose of this build script. Eric

1 0

html version of zope site
by Mike Blake 19 Apr '01

19 Apr '01

Hi! First of all: We have a client that (because of lame internal policy) will not be able to install zope on the final server. However, they _do_ want zope as a publishing solution for a web that will be practically nothing but dtml docs, images, and folders. They're only intested in accessing the web interface to modify text in custom zclass product interfaces we will make for them. Provided that there are no complicated methods integrated into the zope structure, is it possible to generate a navegable static html version of a zope hierarchy? Our idea is to let them access our server to manage and edit content and to give them a publish option (written in python or perl) that generates the html and ftps the site to a remote, non-zope webserver. The day that the client wants a more dynamic site with searching and such, we would discuss installing zope again on the remote server. Please give me your thoughts on this workaround solution. thanks, -- [morcilla(a)ketden.com chikoon]$ whoami Mike Blake

2 1

RE: [Zope-dev] Catalog in 2.3.1
by Adrian Hungate 19 Apr '01

19 Apr '01

Does anyone know if/when this will make it into a release? I am about to deploy a Helpdesk/Knowledge Base product which is very heavily Catalog based. -- Adrian Hungate Manager, European I.S. Acucorp UK Limited -----Original Message----- From: seb bacon [mailto:seb@jamkit.com] Sent: Thursday, 19 April 2001 10:13 To: Adrian Hungate Cc: Zope Developers List (E-mail) Subject: Re: [Zope-dev] Catalog in 2.3.1 * Adrian Hungate <ahungate(a)acucorp.com> [010419 10:09]: > Are there some problems with the new catalog? > > I have spotted two possible problems: > 1) In CatalogAware, there is a function reindex_all, which appears to have a > few problems, like calling index_object instead of reindex_object. Also, why > "sub = self.objectValues()" when sub is never used? > 2) If I search a text index for a value that it does not contain, I get > nothing back (No surprises there), but if I search for a word that is a > substring of a word in the index, I get a Type Error (expected integer key). > Did I do something wrong, or have I hit a bug? Don't know much about (1), but (2) is certainly a bug, and seems to have been fixed in CVS. seb

1 0

RE: [Zope-dev] How to choose: Or, Not and And when using search Results().
by Marco Nova 19 Apr '01

19 Apr '01

> > On Wed, 18 Apr 2001, Chris McDonough wrote: > > > > > You need to manually do unions or intersections on > results from multiple > > > calls to searchRequest currently. > > > > Is this a feature to be implemented? If not, why not? > > DC has no no concrete plans to implement operators or precedence in > catalog queries, although I think it's a really > super-worthwhile idea. > We're currently focusing on the things in the Zope 2.4 plan (see > http://dev.zope.org/Resources/zope_240_plan.html) If > someone wanted to > try to implement it, they could create a fishbowl project and I can > provide assistance. Why don't integrate already working search engine such as pls (www.pls.com) ? It shouldn't be so much difficult to integrate. - mn

3 3

Help! - I want to understand Acquisition...
by Bernie Rossi 19 Apr '01

19 Apr '01

Hi, First of all I'm converting an existing (and working) Zope product from version 2.1.4 to 2.3.1 and I'm having the following problem when trying to run the (new) converted version. In a dtml document which is run many times for different types of objects ie. something like <dtml-in " ... " > . . . <IMG ... src="<dtml-var expr="getIcon(REQUEST)">" ... > . . . </dtml-in> where it iterates through a list of objects, I get mixed results which I don't understand. When it works the getIcon function returns the path (using the REQUEST namespace) and icon for the particular object type and then this gets rendered as expected. But when it fails the icon for the type does not get rendered at all (understandably). After some debugging (ie. print statements) I found that when it failed the REQUEST namespace was not passing what I expected (it was empty), and when I tried to print the namespace it printed the following; <Special Object Used to Force Acquisition> So the python function (getIcon()) was expecting the REQUEST object but actually getting (it seems) is a derived (and unexpected) object, and failing when looking for particular attributes. I don't understand this behaviour (why it works for some and not others and why it worked in a previous version of Zope) and therefore how to fix it. Any help or pointers on the above or other relevant issues to upgrading older products would be appreciated. Thanks in advance Bernie Rossi.

1 0

ZCatalog catalogable types and nonproducts
by Karl Anderson 19 Apr '01

19 Apr '01

I have a Python Product with several subobjects which are not products in their own right. I'm trying to make these subobjects selectable in the Find Objects tab of a ZCatalog, so I can restrict the cataloged objects to these types. Whats the right way to do this? Since they're not addable products, they don't go in the "Find objects by type" selection by default. Rooting through the source, I see that this selection is populated by all_meta_types, and ObjectManager creates that list by appending a _product_meta_types attr onto an acquired value. I can't create a _product_meta_types property on the ZCatalog. I can add a _products_meta_type dict to the ZCatalog via Python: app.catalogtest.catalog._product_meta_types = ({'permission': 'Dummy', 'name': 'Manageable Element', 'action':'Dummy'},) I added the dummy permission and action becuase the contents view doesn't display without them. This seems to work, and these meta types aren't displayed in the add list for the container view (becuase I don't have the 'Dummy' perm), which is what I want. But should I be telling the users of my product to do this? -- Karl Anderson karl(a)digicool.com

1 0

Re: [Zope-dev] memory leak
by Karl Anderson 19 Apr '01

19 Apr '01

This turns out to be a memory leak in ParsedXML. DOM references created during a parse aren't released. We hope to have a new release very soon that fixes this. Restarting Zope is a temporary workaround - DOM refs retrieved from the DB aren't being leaked. -- Karl Anderson karl(a)digicool.com

1 0

Re: zope nautilus cabal
by Federico Di Gregorio 18 Apr '01

18 Apr '01

this is true. i tried to acces zope on my local machine at the address http://localhost:9673/ (note: *without* the manage!) and nautilus showed me everything. both mozilla and netscape show the welcome page. if i add manage nautilus simply refuse to authenticate. my impression is that even if http: is specified, nautilus uses some other kind of protocol (webdav?) to read the contents of the url. the bad thing is that you can read (not modify) every single document in the zope db. ciao, federico Scavenging the mail folder uncovered andrea(a)debian.org's letter: > Hi, here in mixad have found a "mysterious" bug with zope and > nautilus. We are investigating if is a bug or a feature. > > The problem is that nautilus can browse the internals of zope directory > without authentication. > > The method is pointing nautilus to http://www.foo.bar:9673 simply. > > Please can someone try to reproduce the bug ? The version of the > sw is: > > > ii libnautilus0 1.0-3 Shared libraries that part of Nautilus > ii libncurses5 5.2.20010318-1 Shared libraries for terminal handling > > ii nautilus 1.0-3 file manager and graphical shell > rc nautilus-trilo 1.0-2 Nautilus component framework for services > > ii zope 2.3.1-1 The Z Object Publishing Environment > > > > Regards > > Andrea Fanfani > -- > Andrea Fanfani > Era talmente intelligente che, datogli in mano un cubo di Rubik, > riusciva a mangiarlo in 15 secondi netti. (Anonimo) > -- Federico Di Gregorio MIXAD LIVE Chief of Research & Technology fog(a)mixadlive.com Debian GNU/Linux Developer & Italian Press Contact fog(a)debian.org Abandon the search for Truth; settle for a good fantasy. -- Anonymous

9 9

mxodbc no help
by Andy McKay 18 Apr '01

18 Apr '01

Well I tried running mxodbc and that seems even worse then zodbca when connecting to MS SQLServer... for example queries have to be in exactly the right order: This does not work: select Subtype_Cookbook.discussion, Subtype_Cookbook.resource_id from Subtype_Cookbook This works: select Subtype_Cookbook.resource_id, Subtype_Cookbook.discussion from Subtype_Cookbook Is there something im missing here? Many other people here have connected using zmxodbc to SQLServer with no problem... Cheers. -- Andy McKay.

1 0