- Zope-Dev - Zope lists

Medusa and HTTP/1.1
by Seb Bacon 02 Apr '01

02 Apr '01

I've been yelling at my ISP (BTopenworld) for having a badly configured transparent proxy cache, which caches all content *unless* the source specifies otherwise. It's been really messing up my development, and of course it's a concern for my users too. Hopefully they'll sort it out. Anyway, while I was researching for ammunition, I read all the HTTP/1.1 specs (RFC2068). Interestingly although the RFC strongly recommends making unvalidated objects uncacheable, it stops short of forbidding it. I wondered if it might be a good idea to make the ZServer more HTTP/1.1 compliant, by always adding validators to all objects (at present validators are only added to image / file objects). This is the recommended behaviour for servers. In the case of the ZServer, I guess the correct behaviour would be to have every page object return a Last-modified header, which defaults to the modified date of the newest component of the page. This behaviour could then be modified using cache managers like the Accelerated HTTP Cache Manager. Just wondered what anyone else thought. seb

2 1

Re: [Zope-dev] Page Templates and nested templates
by Johan Carlsson 02 Apr '01

02 Apr '01

> There's a Tracker linked from the main ZPT Wiki page. You should be able to > add it there. On the other hand, I'd like to know how exactly you've tried > to nest macros, and how it failed for you. > > Cheers, > > Evan @ digicool & 4-am > Here is an examples: ID: here/master <html xmlns:tal="http://xml.zope.org/namespaces/tal" xmlns:metal="http://xml.zope.org/namespaces/metal" metal:define-macro="master"> <head metal:define-macro="head"> <title tal:content="here/title">sss title</title> <meta something=""> </head> <body bgcolor="#9999ff" metal:define-macro="body"> <div metal:define-macro="bodymacro"> <h1>This is a macro 21</h1> </div> <div metal:define-slot="bodySlot"> <h1>This is a H1 Header</h1> </div> </body> </html> Now notice this if I state use-macro for the METAL statement in the HTML-tag, all subsequent metal:define-macro statements gets replaced with the use-macro="here/master/macros/master>" ID: here/Page1 (version1) <html xmlns:tal="http://xml.zope.org/namespaces/tal" xmlns:metal="http://xml.zope.org/namespaces/metal"> <head metal:use-macro="here/master/macros/head"> <title tal:content="here/title">sss title</title> <meta something=""> </head> <body bgcolor="#9999ff" metal:use-macro="here/master/macros/body"> <div metal:use-macro="here/master/macros/body"> <h1>This is a macro 21</h1> </div> <div metal:define-slot="bodySlot"> <h1>This is a H1 Header</h1> </div> </body> </html> But if I instead apply the use-macro on the body and the head this are working perfectly, subsequent define-macros in the inner div gets turned in to use-macro="here/master/macros/body" though. ID: here/Page1 (version2) <html xmlns:tal="http://xml.zope.org/namespaces/tal" xmlns:metal="http://xml.zope.org/namespaces/metal"> <head metal:use-macro="here/master/macros/head"> <title tal:content="here/title">sss title</title> <meta something=""> </head> <body bgcolor="#9999ff" metal:use-macro="here/master/macros/body"> <div metal:use-macro="here/master/macros/body"> <h1>This is a macro 21</h1> </div> <div metal:define-slot="bodySlot"> <h1>This is a H1 Header</h1> </div> </body> </html> How I would want it to work is that if a higher level of macro is set to use-macro the subsequent level would remain unchanged (e.g. as define-macro) until I explicitly change it in to use-macro. I would then be able to set it to any macro definition, e.g. use-macro="here/master/macros/bodymacro" or use-macro="<some other path>/macros/<some other macro>" etc. I think what would make allot of sense and that's probably how it supposed to work, right? Best regards, Johan Carlsson

1 0

ANN: Localizer 0.1
by palomar＠sg.uji.es 02 Apr '01

02 Apr '01

Hi all, The product Localizer version 0.1 has been released, download from: http://sourceforge.net/projects/lleu It helps to internationalize and localize Python products and web sites. It's built around the GNU gettext utilities, to do that the gettext and locale modules have been back ported from Python 2.0 to Python 1.5.2 and are included (if you already use Python 2 remove both files). As an example the Localizer itself is localized to french and spanish, though both translations aren't finished it's enough to show how to internationalize/localize python products. It has a README.txt file explaining how to use it. Hope I'll get feedback soon :) jdavid

1 0

Re: [Zope-dev] Memory problems with ParsedXML ?
by Karl Anderson 02 Apr '01

02 Apr '01

Godefroid Chapelle <gotcha(a)swing.be> writes: > Hi, > > I have imported about 2500 xml files in ParsedXML objects through the > use of : > > manage_addProduct['ParsedXML'].manage_addParsedXML(id, '', xmlString) > > This works perfectly. > > But when trying to browse the folder containing the ParsedXML instances, > a whole lot of memory is used by Zope and it took my machine down. > > Have I done something wrong or does this info help you debug the product > ? Hm, it's been pointed out to me that the inefficiency of the get_size method of ParsedXML is compounded with many instances because the standard management interface uses it. It's possible that some stuff isn't getting garbage collected until the transaction is done, and calling get_size on many ParsedXML instances in one transaction is the problem. I'll look into it. I didn't think that anyone would want to see a management screen with 2500 ParsedXML instances - or 2500 instances of anything - at one time. Have you tried using a BTree folder? Another thing to keep in mind is that the ParsedXML *product* is a DOM tree and a management interface around it. The management interface only gives you a convenient UI and a place to store a few bits that the DOM doesn't know about (content type, namespace usage, etc.). It's always more efficient to just use the DOM tree and related utilities (parsing, printing), although it is less convenient - you have to know a little bit more about what you're doing. Do you need each of those 2500 product instances? See the createDOMDocument method to create a DOM document without the management interface. It's easy to wrap a ParsedXML product around an existing DOM tree - see the initFromDOMDocument method. Currently, you have to create the persistent ParsedXML product which wraps the DOM Document node, and install that product in a Zope ObjectManager (or subclass like Folder) somewhere, to avoid a few warts. Someday I'd like to be able to just instantiate a management wrapper when I want a management interface, and then just throw it away, without bothering to add it to a folder. -- Karl Anderson karl(a)digicool.com

2 2

Re: [Zope-dev] ANN: Proposal for ZCatalog "drop-in" indexes
by Dieter Maurer 02 Apr '01

02 Apr '01

Chris McDonough writes: > I'd like comments on this proposal if you've got the time: > > http://dev.zope.org/Wikis/DevSite/Proposals/DropInIndexes Looks good. When you are changing the indexing and catalog stuff: * make the interface "index" public, such that it can be used from outside the ZCatalog * standardize the result type of index searches such that partial search results can easily be combined * provide operators "union", "intersection" and "difference" on search results. What would you gain? We could leave the rigit search interface of the ZCatalog and provide flexible queries that and/or subqueries against different indexes. Especially CMFTopic would profit much from this. Dieter

4 3

ZCatalog
by Marco Nova 01 Apr '01

01 Apr '01

Hello Guys, I'm a newbie about python and zope (about a week and a half till now), so probably most of what I say now is probably wrong. I was trying to customize the ZDiscussion product with file upload, editing of previous posts and deletes of old posts and I guess I've find a bug in the ZCatalog.uncatalogObject method. In the code there is: for btree in (data, paths): try: del btree[rid] except KeyError: LOG('Catalog', ERROR, ('uncatalogObject unsuccessfully ' 'attempted to delete rid %s ' 'from paths or data btree.' % rid)) else: try: self.__len__.change(-1) except AttributeError: pass # No managed length that means for each successfull unindex item, the __len__ counter will be decreased by 2, that's wrong. The correct code should look like: for btree in (data, paths): try: del btree[rid] except KeyError: LOG('Catalog', ERROR, ('uncatalogObject unsuccessfully ' 'attempted to delete rid %s ' 'from paths or data btree.' % rid)) try: self.__len__.change(-1) except AttributeError: pass # No managed length Or if we want to handle situations where the unindexs are unsuccessfull, we have to rewrite better the method with the use of transactions. That's all. Regards, Marco Nova Tera S.p.A. P.S. Zope Version Zope 2.3.1b2 (binary release, python 1.5.2, win32-x86) Python Version 1.5.2 (#0, Jul 30 1999, 09:52:18) [MSC 32 bit (Intel)] System Platform win32

3 2

Re: [Zope-dev] Medusa and HTTP/1.1
by Dieter Maurer 31 Mar '01

31 Mar '01

Seb Bacon writes: > In the case of the ZServer, I guess the correct behaviour would be to > have every page object return a Last-modified header, which defaults to > the modified date of the newest component of the page. This behaviour > could then be modified using cache managers like the Accelerated HTTP > Cache Manager. This can be quite tricky, as it is not obvious at all what are the components of a page. Dieter

2 1

ANN: mxDateTime 0.1
by palomar＠sg.uji.es 31 Mar '01

31 Mar '01

Hi all, Download mxDateTime 0.1 from http://sourceforge.net/projects/lleu README.txt ---------- Introduction This product makes easier to use the mxDateTime python module from Zope. It allows to use it from Python Products and external methods using for example:: from Products.mxDateTime import DateTime It's also accessible from restricted code (DTML methods, Scripts, etc..) using the '_' variable, for example:: <dtml-var "_.mxDateTime.today()"> How it works (read this if youd don't use a Unix like OS!!) When the product is initialized for the first time a symlink is created to the DateTime directory. If symlinks are not available (for example in Windows) you have to copy the whole DateTime directory instead. To make the module accessible from restricted code it has been needed to wrap it so it plays nice with the Zope security machinery. Best regards, jdavid

1 0

[Mini-HowTo]: Multiple Selects in ZClass Edit Forms
by Michael R. Bernstein 31 Mar '01

31 Mar '01

Problem: ------- You want your ZClass instance edit form to correctly display the currently selected values for a multiple selection property. Assumptions: ----------- You have a ZClass that has a multiple select property named 'format' with a value set to 'format_list'. You have a 'format_list' lines property set where the ZClass *instances* can acquire it. You generated an 'Edit' type 'Property Sheet Interface' that contains the following code: <tr><th align=left valign=top>format</th> <td align=left valign=top></td> </tr> Solution: -------- Replace the above code with the following: <tr><th align=left valign=top>format</th> <td align=left valign=top> <select name="format:list" multiple size="4"> <dtml-in format_list> <option <dtml-if "_['sequence-item'] in _['format']"> SELECTED</dtml-if> > <dtml-var sequence-item></option> </dtml-in> </select> </td> </tr> This will correctly display the currently selected values for your ZClass instance in the edit form, and capture the changed values when the form is submitted, casting even single selections to a list. Note: If the 'format_list' lines property is set where the Zclass instances can acquire it, but not where the ZClass definition can acquire it, then the Property Sheet management screen will *not* display the values. If you want to provide a default set of values that the ZClass definition can acquire as well as the ZClass instances (of course you can also override this locally), set the format_list property on the root folder. HTH, Michael Bernstein.

1 0

Wrong documentation?
by Tim Ansell 31 Mar '01

31 Mar '01

In the help documentation for the 'Authenticated_User' object it says the following.. hasRole(object, roles): Return a value that is true if the user has the given roles on the given object and return false otherwise. Permission - Always available getRoles(object): Returns a list of the roles the user has on the given object (in the current context?) Permission - Always available Neither seem to work (when given an object) with the standard user folder or LoginManager. In Login Manager (and it appears in the normal User Folder, though i haven't tested it) you have to use getRolesInContext(object). As well it appears that hasRole returns 1 when it's not suppose to. The has_role(roles, object) appears to work the way it hasRole is suppose to. Has the documentation become out of date or is their something else going on that i have missed? Mithro

1 0