September 2002 - Zope

<base href="http://mydomain.dk" />
by Preben Mikael Bohn 24 Sep '02

24 Sep '02

Hi all, I'm using Zope behind apache, using a virtual host monster. I have put a <dtml-var index.html> in my index_html. Everything works fine, but I just tried to validate my webpages using www.w3.org's validator, and it came out with an error... When I go into http://mydomain.dk/ Zope inserts a <base href="http://mydomain.dk/" /> in the <HEAD> section, which is wrong HTML (I think the "/" is illegal) If I go to http://mydomain.dk/index.html no base tag is inserted. Can anyone tell me what I am doing wrong? Best regards Preben

3 2

Structured text - inline images
by Chris Beaven 24 Sep '02

24 Sep '02

I'm trying to get inline images in structured text working and can't seem to figure it out... Quoted from http://www.zope.org/Members/jim/StructuredTextWiki/StructuredTextNGRules : Text encloded by double quotes followed by a colon, the string img, another colon, an URL, and concluded by punctuation plus white space, or just white space, is treated as an inline image. For example, "Zope icon":img:http://www.zope.org/p_/zopelogo_jpg is interpreted as (in HTML, <img alt="Zope icon" src="http://www.zope.org/p_/zopelogo_jpg">) But when I try it, all I get is a link to "img:http://www.zope.org/p_/zopelogo_jpg" titled "Zope icon". So obviously this rule doesn't seem to be working. Am I missing something obvious? Is there some other way I can get this working? Thanks Chris

3 2

Re: [Zope] manage_pasteObjects in manage_afterAdd raises AttributeError for getPhysicalRoot
by Grégoire Weber 24 Sep '02

24 Sep '02

Hi All, forgotten to post reply Tims and Roberts reply to the list. here ist is: sorry first, there was a typo in the snippet as Tim stated (container is correct instead of context in the last line): second, as robert supposed and I am casting suspicion on also is that the problem could be that obj has not yet anaquisition wrapper. Must this be done in addToCollection before container = ..., but how? third (Robert): Ok, I see in this special case the parent of object A is the same as the parent of object I. addToCollection is called in other contexts also, where this is not the case. So I can not change that. It seems not to be the containers problem. Greetings, Greg -------------------------------------------------------------------------- Snippet no 2: - now correct and slightly changed code snippet - The snippet now avoids generating obj out of the id and the container and then regenerating the id and the container out of obj - same error as before def manage_afterAdd(self, item, container): """ simplified code called after instanation of object A (self) in folder P (container) """ index_obj = getattr(container.aq_base, 'index_html', None) if (index_obj is not None) and (index_obj.meta_type == self.meta_type): # P has an object I (index_obj) of the same type as A is index_obj.addToCollection(container, self.getId()) # tried also to pass self wrapped by container: # index_obj.addToCollection(container, self.__of__(container).getId()) def addToCollection(self, container, id): """ simplified code in the same class called on index_html object """ # this copies object A (container/id, just instanated before # manage_afterAdd!!!) to object I (self, same as index_obj above) cb = container.manage_copyObjects(id) # OK self.manage_pasteObjects(cb) # raises ERROR (same as before) At 16:38 24.09.2002 +0100, you wrote: Grégoire Weber wrote: Hi, below an issue I don't have any idea where the problem lies. Are there some experts around that could give me some advice? Thanks! The code is for the redesign of the swiss community site http://www.zope.ch hopefully working in two weeks. So please, help! Are there any known issues copy&pasting a freshly added object A to another folderish object I in the A's manage_afterAdd hook? Doing Copy&Paste by hand (by GUI) after the object A was added works. Have a look at the simplified code snippet (Python Product): def manage_afterAdd(self, item, container): """ simplified code called after instanation of object A (self) in folder P (container) """ index_obj = getattr(container.aq_base, 'index_html', None) if (index_obj is not None) and (index_obj.meta_type == self.meta_type): # P has an object I (index_obj) of the same type as A is index_obj.addToCollection(container._getOb(self.getId())) def addToCollection(self, obj): """ simplified code in the same class called on index_html object """ container = obj.getParentNode() id = obj.getId() # this copies object A (obj, just instanated before manage_afterAdd!!!) # to object I (self, same as index_obj above) self.manage_pasteObjects(context.manage_copyObjects(id)) # raises ERROR below ^^^^^^^ <-- has to be container Not sure, but shouldn't 'context' be 'self'? I'm not sure because that doesn't seem to be the error that you are seeing. I would have thought that the use of 'context' would raise a NameError (or something like that). tim _____________________________________ Grégoire Weber Rigistr. 31 CH-8006 Zürich Switzerland phone: +41-(0)1-361 66 11 mobile: +41-(0)79-44 11 457 mailto:gregoire.weber@switzerland.org

1 0

Re: [Zope] Newbie: Case independent URLs?
by Fearless Froggie 24 Sep '02

24 Sep '02

The advantage with Oliver's approach is it gives you a chance to "educate" your users. Use the 404 page to show them the correct page, and explain to them what they did wrong. And then they'll explain it to two friends, and so on and so on . . . and we'll never have those kinds of user errors again! Rita Mikusch Fantasy Land **************************************************************************************** Robert Allyn wrote: > I agree it is a hassle. But... > > My most important "customer" is 60 - 90 years old, often having just started to use \ > a computer. For them typing, let alone email and the web are all new. They tend \ > to exchange web page URL's via 'snail' mail, or over the phone. The chance of them \ > using upper case when it is not needed is pretty high. I get a number of 404 \ > errors in my log from users typing in the wrong case in a URL. Because it is all \ > new to them, they figure they did it wrong and give up. I need to accommodate \ > these people for the site to be useable. > That is why I am willing to jump through some hoops to make this work. > > From a brief look around, it seems IIS and Apache on Windows are case insensitive, \ > as the OS is too dumb to know the difference. Maybe a dumb suggestion, but couldn't you just write a customized 404 error page which uses the zcatalog and some logic which offers one or more alternative, existing pages from your site? Like: "the address (http://yourserver/somedir/somedokument.PDF) which you entered does not exist, maybe you meant: http://yourserver/somedir/somedocument.pdf" Because from my expirience, confusing upper and lower case is not the only thing people confuse. cheers, oliver __________________________________________________ Do you Yahoo!? New DSL Internet Access from SBC & Yahoo! http://sbc.yahoo.com

1 0

Re: [Zope] Diacriticals in ZMI __S_O_L_V_E_D__
by Jaroslav Lukesh 24 Sep '02

24 Sep '02

Hi all that you have problems with diacriticals in ZMI. Here are set of files that I was change. All<head> sections was replaced by <head><meta http-equiv="Content-Type" content="text/html; charset=windows-1250"><meta http-equiv="Content-Language" content="cs"> Take look at http://www.zope.org/Members/lukesh/ (empty now, waiting for review, so be patient) here are all changed files. Regards JL.

1 0

Using a SQL database connection in Python scripts
by Andreas Tille 24 Sep '02

24 Sep '02

Hello, I just created a database connection and want to do some queries from a Python script and solving the problem without defining explicite ZSQL methods and just do some queries from the python script. In a former mail to this list I've got the following hint according to a similiar question: Mon, 13 May 2002, Gilles Lenfant <gilles(a)objectis.net> wrote: > Zope DAs have a "manage_test" method. The docstring says : > > results = someDA.manage_test(query, REQUEST=None) > > Executes the SQL in parameter query and returns results > > Note you must set the proxy to "manager" to a python script or DTML > that executes this method. Thus I tried results = context.myDBconnection.manage_test('SELECT * FROM table;', \ REQUEST=None) print results return printed And got <Shared.DC.ZRDB.Results.Results instance at 0x91327bc> Hmmm, it seems to work in any way - but how to parse the results object? I tired to find some solutions in the Zope Book but failed because it describes only DTML methods to query databases. Is there any reason to use only DTML with databases and my intention has real drawbacks or is it just a lack of documentation? Kind regards Andreas.

3 2

RE: [Zope] Newbie: Case independent URLs?
by Passin,Thomas B. (Tom) 24 Sep '02

24 Sep '02

[ Toby Dickenson] > > Like it or not, windows users are conditioned to expect case > insensitivity. > Our static site is full of files that mix up .pdf, .PDF, and > .Pdf, and we > need to use Apaches mod_speling to account for the > difference. Ive given up > on convincing the content authors that there is anything > wrong in this. > > I seem to remember IIS work this way by default? > According to the RFC (2396), the scheme part of the URI is to be all lowercase and also should be case insensitive in case uppercase is used anyway. Host names are also supposed to be case-insensitive. Otherwise the RFC does not constrain the case sensitivity-ness. So it would be up to the server. Cheers, Tom P

1 0

Zope Product question
by Bo M. Maryniuck 24 Sep '02

24 Sep '02

Hi, all. I would like to make my Product is called always. I mean, my product should be called all the time, even nobody ask him :-) It's like "acl_user". Where I should RTFM about or how to do this? -- Regards, Bogdan Win32-API is wonderful : Full of things to wonder about...

3 4

Re: OT: drop down menus
by Josef Meile 24 Sep '02

24 Sep '02

> Sorry for the off-topic subject, but I figure there are a fair few > people on this list who have experience/interest in this. It's quite a > common feature of websites. > > I'm trying to find the most cross-platform, and most gracefully > degrading drop-down/pop-out menu system. Hi, I have more websites that could interest you: Nice one Object oriented, good compatibility, good documentation, but a little bit complicated: http://www.twinhelix.com/dynamic.html Works with Netscape 4.x and 6.x, Opera 6, IE 6. I don't know about other browsers: http://www.milonic.co.uk/menu/ I know there is a Zope product based on the previous menus: http://www.zope.org/Members/haqa/ZFloatMenu Another Zope product: http://www.zope.org/Members/struktur/MenuItem More drop down menus: http://www.burmees.nl/menu/menus.htm http://developer.netscape.com/viewsource/smith_menu2/smith_menu2.html Nice alternatives to drop down menus: http://www.axpi.net/download.htm http://devedge.netscape.com/toolbox/examples/2001/xbCollapsibleLists/#examp… http://javascript.internet.com/navigation/slide-down-menu.html > Please share > experiences. Got any other ideas? If you're going to use drop down menus and you want search engines index your site, you have to do site maps to facilitate the cataloguing of your site. It seems like they don't index links inside JavaScripts. Hope it helps, Josef.

1 0

ZODBC and dictionaties
by Oliver Marx 24 Sep '02

24 Sep '02

Dear List I need to use Zope with an SQL server, but but but ZODBC which seems to be the standard DA for SQL server doesn't support dictionaries. Is there a proven solution to this problem? \Oliver

2 1