- Zope - Zope lists

Re: [Zope] get non-acquired attributes of an instance
by thomas desvenain 26 Jun '06

26 Jun '06

2006/6/26, Andreas Jung <lists(a)zopyx.com>: > > > --On 26. Juni 2006 10:53:39 +0200 thomas desvenain > <thomas.desvenain(a)gmail.com> wrote: > > > aq_base(aq_inner(obj)).__dict__ ? > > > > hmm it doesn't works, i get metadata only, > > 'metadata'? This is an unkown term in the context of Zope/Python objects > except for catalog brain objects...what you you mean? sorry, it's because i work with at/plone objects : toto.__dict__ displays only some identifiers, metadata and 'content' of toto object {'_EtagSupport__etag': 'ts51320374.17', '_at_uid': 'ffa61909947ef9021cab5bcbccfb5f54', '_md': {'excludeFromNav': None, 'description': '', 'contributors': (), 'rights': '', 'language': '', 'effectiveDate': None, 'expirationDate': None, 'allowDiscussion': None, 'creators': (u'admin',), 'subject': ()}, '_objects': ({'id': 'connection', 'meta_type': 'Z Psycopg Database Connection'},), 'at_references': <Folder at at_references>, 'connection': <Connection at connection>, 'creation_date': DateTime('2006/06/26 13:10:40.161 GMT+2'), 'demarshall_hook': None, 'id': 'toto', 'marshall_hook': None, 'modification_date': DateTime('2006/06/26 13:12:54.164 GMT+2'), 'title': ''} and so does aq_base(aq_inner(toto)).__dict__ > > > > > and with > > > > dir(aq_base(aq_inner(test.toto))) > > > > i get acquired methods too... > > Of course yes...that's standard Python behavior. > > -aj > > -- Thomas Desvenain Junior Plone/Zope analyst-developer, Looking for a Job ! Now, trainee at Institut National de Recherche Agronomique Mathématiques, Informatique et Génôme laboratory 78350 Jouy en Josas +33 (0)1 34 65 29 43

2 1

No joy on 8080.
by Ignacio Valdes 26 Jun '06

26 Jun '06

Hello all, I installed Zope on Fedora Core 5 using yum install zope, I opened port 8080 with lokkit, then I started it up using zopectl start. This is all on a headless server by ssh so I can't check http://localhost:8080 but http://server ip address:8080 on my browser doesn't work. I do have xampp apache running as well and can get the xampp welcome screen to appear but with xampp apache started or stopped http://server ip address:8080 doesn't do anything. I'm baffled as to what to do next. What should I try? Thanks! -- IV

6 12

RE: [Zope] Searching for and displaying properties
by Sinang, Danny 26 Jun '06

26 Jun '06

> manage_addProperty does not trigger recataloguing of the object even if its a subclass of CatalogAwareness. You should ? > either call > obj.reindex_object() or at least obj.index_object(), if your class has the CatalogAwareness-Mixin or you use the catalog > API: > yourcatalog.catalog_object(obj,obj.absolute_url_path()) > After that, your metadata should show up nicely w/o indexing the whole object tree. Thanks Tino. I'd like to try your suggestion, but I don't know how to refer to the object that needs recataloguing. Can you help ? Again, here's my code that calls manage_addProperty. <code> wf = context.GeneralWorkflow instance_id = wf.addInstance( process_id=process, customer=jobname, comments=comment, title=subject, activation=1) wf.getInstance(instance_id).manage_addProperty('jobname', 'ABC12345', 'string') </code> - Danny

2 1

Re: [Zope] get non-acquired attributes of an instance
by thomas desvenain 26 Jun '06

26 Jun '06

aq_base(aq_inner(obj)).__dict__ ? hmm it doesn't works, i get metadata only, and with dir(aq_base(aq_inner(test.toto))) i get acquired methods too... 2006/6/26, Andreas Jung <lists(a)zopyx.com>: > > > --On 26. Juni 2006 10:30:50 +0200 thomas desvenain > <thomas.desvenain(a)gmail.com> wrote: > > > hello, > > > > is it possible to get the list of all an object's attributes, excepted > > the ones acquired from parents ? > > > > aq_base(aq_inner(obj)).__dict__ ? > > -aj > > -- Thomas Desvenain Junior Plone/Zope analyst-developer, Looking for a Job ! Now, trainee at Institut National de Recherche Agronomique Mathématiques, Informatique et Génôme laboratory 78350 Jouy en Josas +33 (0)1 34 65 29 43

2 1

[Ann] "AdvancedQuery 2.0" supports incremental ranking
by Dieter Maurer 25 Jun '06

25 Jun '06

"AdvancedQuery" is a search engine similar to Zope's "ZCatalog" that tries to remove many of "ZCatalog"s limitations and strives heavily for efficiency. "AdvancedQuery"s can be arbitrarily built from elementary index queries with "and", "or" and "not". This gives richer query expressiveness than "ZCatalog". "AdvancedQuery" supports incremental multi-level sorting. For highest efficiency, "AdvancedQuery" can (and probably should) be used with "ManagableIndex" and "IncrementalSearch2". New in version 2.0: support for incremental ranking. A rank determination rule can be specified by a sequence of weighted queries. The rank of a document is then either the sum or the maximum of the weights for queries that match the document. This kind of ranking is a sort specification and can be combined with other sort specifications in the usual way (leading to multi-level sorting). Ranking is performed incrementally. This way, ranking can be quite efficient when you only look at the first (few hundred) hits -- despite the fact that ranking is very expensive in general. More information and download: <http://www.dieter.handshake.de/pyprojects/zope> -- Dieter

1 0

Searching for and displaying properties
by Sinang, Danny 24 Jun '06

24 Jun '06

Dear All, I managed to add a property to an object by using the manage_addproperty () method. However, if I don't issue a manage.reindexIndex () for the new property, it won't come out in the catalog search results. So I issue reindexIndex (), but when I finally try to display the property, all I get is 'Missing Value'. But if I then issue the command refreshCatalog(clear=1), the property is now visible. Although refreshCatalog works, it would gobble up the server's CPU and memory resources everytime its run. My question is, is there another command I can use instead of refreshCatalog to make the property display ? Below is the code I'm talking about : wf = context.GeneralWorkflow instance_id = wf.addInstance( process_id=process, customer=jobname, comments=comment, title=subject, activation=1) wf.getInstance(instance_id).manage_addProperty('jobname', 'ABC12345', 'string') wf.Catalog.manage_reindexIndex(ids=['jobname'], REQUEST=None, RESPONSE=None, URL1=None) jobs = cat.searchResults({'jobname':'ABC12345'}) for j in jobs: print j.jobname Regards, Danny

3 2

RE: [Zope] Searching for and displaying properties
by Sinang, Danny 24 Jun '06

24 Jun '06

> This add a property to an object but it has nothing to do with adding > *metadata* to a ZCatalog instance. Please read the ZCatalog chapter of the Zope Book 2.7 edition. It carefully explains > what *metadata* in the ZCatalog context means. The short version: a Zcatalog search returns > *proxy* objects and the attrs of the proxy object contain only the values of the original object where you have > configured a metadata in the Zcatalog. Having read the Zope 2.7 ZCatalog chapter, I think I can say there's always been a metadata column called "jobname". So going back to my original question, is there a way for me not to have to mass-refresh my entire catalog just to be able to display my properties ? Regards, Danny

3 2

RE: [Zope] Searching for and displaying properties
by Sinang, Danny 24 Jun '06

24 Jun '06

> That's the official ZCatalog API. Otherwise find your objects through a catalog search and reindex them individually. I already did the reindexing just before the search and it didn't work. In any case, I may have found a workaround by using getObject(). According to http://www.plope.com/Books/2_7Edition/SearchingZCatalog.stx#2-10 : getObject() Returns the actual zope object from the result object. This is useful if you want to examine or show an attribute or method of the object that isn't in the metadata--once we have the actual object, we can get any normal attribute or method of it. However, be careful not to use this instead of defining metadata. Metadata, being stored in the catalog, is pre-calculated and quickly accessed; getting the same type of information by using getObject().attribute_name requires actually pulling your real object from the ZODB and may be a good deal slower. On the other hand, stuffing everything you might ever need into metadata will slow down all querying of your catalog, so you'll want to strike a balance. A good idea is to list in metadata those things that would normally appear on a tabular search results form; other things that might be needed less commonly (and for fewer result objects at a time) can be retried with getObject. I tried it and it works. My code now looks like this : <code> wf = context.GeneralWorkflow cat = wf.Catalog jobs = cat.searchResults({'jobname':'ABC12345'}) for j in jobs: print j.getObject().jobname return printed </code> My only question now is ... How much slower is this solution than using the jobname metadata (assuming it worked). - Danny

2 1

RE: [Zope] Searching for and displaying properties
by Sinang, Danny 24 Jun '06

24 Jun '06

> Click on "update catalog" in the Advanced tab or use the refreshCatalog() > method of the ZCatalog (described in its interfaces.py file). refreshCatalog() is exactly what I was forced to use in the first place. I was wondering if there exists an alternative function / method that won't require refreshing / reindexing the entire Catalog. Regards, Danny

2 1

RE: [Zope] Searching for and displaying properties
by Sinang, Danny 24 Jun '06

24 Jun '06

> 'jobname' is added as *metadata* to your catalog? Not sure if using manage_addProperty () to add jobname is equivalent to adding it as *metadata* . <code> manage_addProperty('jobname', 'ABC12345', 'string') </code>

2 1