November 2000 - Zope

time.sleep() in dtml
by Martin Winkler 06 Nov '00

06 Nov '00

Hello all, some python functions work in dtml, mostly mathematical ones. But what I do need at the moment is the "time.sleep(x)" function to prevent an immediate reply to a request of the users. I tried <dtml-var "_.time.sleep(1)"> and <dtml-var "_.sleep(1)">, but Zope returned an "Attribute error". Any ideas? Or do I have to create a special python method for this? (I'm not used to python until now...) Martin

2 1

Mirroring Zope sites
by Christopher J. Kucera 06 Nov '00

06 Nov '00

Hello, I'd like to be able to take a "snapshot" of a Zope site (probably using the recursive "wget") so that it can be served from Apache alone. I'm running into problems coping with files named "x_html" (most notably "index_html") . . . Apache serves up the files as plaintext, so you just get the HTML in your browser window. I've been messing around with srm.conf in my Apache configuration to try and accept "_html" as a valid extention for "text/html" files, but to no avail. Has anyone managed to get Apache to serve "_html" as text/html? Thanks much! -CJ

1 0

Re: [Zope] Zope Book Beta
by dan＠sol.control.com 06 Nov '00

06 Nov '00

"Jason Cunliffe" <jasonic(a)nomadicsltd.com> writes: > ----- Original Message ----- > From: Chris Withers <chrisw(a)nipltd.com> > > > Kinda tough to check out/read/find mistakes in a book that's only > > available as a set of html pages that aren't full text searchable.... > > This is a really good point. > > Just as open source benefits from real-time peer review, so do manuscripts > get better when fresh eyes and minds and go over them. > > How about 'releasing' a PDF / HTML Doc with header/footers clearly marked: > 'Preview Review copy for Copy-Editing use only. Not for general public or > commercial distribution' I like this suggestion. There are three of us (ranging from novice to pretty experienced) here that were planning to print out copies of PDF and go over them at home, none of us has the time to sit at work and read/review the whole thing on a web browser.

5 8

recursive constructions
by Robin Becker 06 Nov '00

06 Nov '00

I would like to duplicate a folderish zclass path and copy some properties across from one tree /R/A/B/C to another /L/A/B/C. The request to create an object is http://localhost/R/manage_addProduct/myProd/myProd_factory/myProd_add?id=A,… I would like to then create /L/A/B/C etc using some kind of recursive scheme, but I don't know how to call a method to do this recursively and maintain the original permissions. Somehow I seem to lose something along the way. the actual sequence of calls would be http://localhost/L/manage_addProduct/myProd/myProd_factory/myProd_add?id=A,… http://localhost/L/A/manage_addProduct/myProd/myProd_factory/myProd_add?id=… http://localhost/L/A/B/manage_addProduct/myProd/myProd_factory/myProd_add?i… -- Robin Becker

1 0

Need win32 dll for DynPersist !
by m.kobald 06 Nov '00

06 Nov '00

Hi all, we need the DynPerisist.dll (Win2000) for presentation purpose. If someone has compilied this dll, would you please mail it to me. (ZPattern 0.4.3 beta 1) Normally we develop on Linux .................. ;) . Thanx and greetings from Heidelberg/Germany Micha

2 1

Properly intergating reportlab and Zope
by ed colmar 06 Nov '00

06 Nov '00

Hi all! I'm still trying to get this to work, so I'm going to try the lists for assistance again. TIA!!! I am working on a project using zope and reportlab. The goal is to have uploaded images end up in a formatted PDF file. I have both of these halves completed, I need to figure out how to connect them. I have been using SquishFile.py as the wrapper around uploaded images, as is done in squishdot. The few image types I have tried have displayed fine on screen, but PIL/Reportlab is not able to recognise the image. I know how to get an image into reportlab from the filesystem, but is there a way to use one that is in the Zodb? How would you go about wrapping the file in a PIL.Image() ? I've been trying things similar to: cardimage=PIL.Image.open(self.uploadedimage) c.drawInlineImage(cardimage, 1,1) This gives an error in reportlab.canvas.convert() I tried using PIL.image.fromstring(), but I can't quite get it to work out: filebytes=[str(self.uploadedimage.file_bytes()), 'bytes'] cardimage=PIL.Image.fromstring(self.uploadedimage.content_type(), filebytes, self.uploadedimage.file_data()) c.drawInlineImage(cardimage, 1,1) This gives a typeerror: PIL/Image.py, line 848, in new has anyone succesfully done this? can I just Leave the SquishFile wrapper off if it is ONLY going to be images that get uploaded? Thanks for any ideas! -ed- _______________________________________________ Zope maillist - Zope(a)zope.org http://lists.zope.org/mailman/listinfo/zope ** No cross posts or HTML encoding! ** (Related lists - http://lists.zope.org/mailman/listinfo/zope-announce http://lists.zope.org/mailman/listinfo/zope-dev )

1 0

Zope and Sybase Stored Procedures
by b.pouye＠paris.infonie.fr 06 Nov '00

06 Nov '00

Hi, I am working with zope and I want to securize my access to the data stored in a Sybase Server. So I am trying to do all my SQL requests by using stored procédures. The only thing that goes wrong is that I don't know how to make a call to a Sybase Stored Procedure within Zope. Can anyone help me ? Sincerly, Badara POUYE Infosources 16, rue Hoche Tour Kupka B 92906 Paris La Défense Tel.: (+33)141028000

6 6

Re: [Zope] Where is the code?
by Rik Hoekstra 06 Nov '00

06 Nov '00

Red Pineseed wrote: > > Thanks Rik, > > You answered my question perfectly. This was my guess and > found the Data.fs. This makes me wonder though how the > implementation of the three tier separation, ie > data/logic/presentation is achieved. Do you have any idea? Yes I do. The key point is that in Zope everything _is_ content, DC says. I fact I think it is more proper to say that in Zope everything is an object. But there are several kinds of objects, as they 'specialize' in something, so to say. This is to say that objects have behaviour and this behaviour is of a certain type. That's why you can tier up your application in three tiers, or even more if that's useful. An example may clarify this. Say we have a simple application that presents the users with data retrieved from a relational database. It has a navigation bar of a breadcrumbs style (like home > mypages > myapp) and there is a stylesheet attached to it. In schematic form a page from the app looks like this: * Header * Breadcrumbs * data from a database * footer in Zope you could make the following objects that represent different parts of the application 1. stylesheet object (containing stylesheet) 2. breadcrumbs object (depends on context) 3. header object (consists of HTTP header including stylesheet and breadcrumbs) 4. database adapter (managing the connection to your relational database) 5. ZSQL Method (a querying object) 6. Queryform (for inputting query values optional) 7. Report object (formatting retrieved data) 8. footer (to close the page. Might also the breadcrumbs) > What is the advantage of putting everything in a db? The db is the object repository. It would be hard to have persistent object without repository ;-) > Apparently if the db down, we will lose everything and > recovering from a db is pain as you may already know. Um, I'm no systems administrator, so there are probably many people who can say much more sensible things about this then I can. Anyway, a few remarks. First, the zope database is quite stable and it will be difficult to break it... Recovering is mostly just done by restarting the server. Usually SQL databases will give you much more trouble in this respect. Second, if you are concerned about failover of your server using the ZEO (Zope Enterprise Option) makes it easy to make your server distributed. This is also where Zope flexibility comes in: in default you data are stored in the zope database, but there are a number of other storage possibilities, including directory storage, bsddb storage, demo storage (storing everything in RAM so that you can run Zope from CD) and recently there Oracle storage has been added, though I'm not sure if that's public or stable as yet (I don't have Oracle myself anyway) If you insist on putting files on the filesystem, there is a Zope product called LocalFS that will include local directories and serve the contents from Zope. Moreover there are strategies for hybrid solutions in which you may serve static content or large files from Apache and do the dynamic stuff from Zope. > though it is great performance improvement to load the db in > RAM at run time, do you know of an other benefits? the db as a whole won't load in RAM, objects will only be 'woken up' when they are called. Benefits... Uh, it would be hard to imagine another way of making persistent object available especially doing it the way Zope does. Reversely it would be hard to make everything an object (and thus factor out common behaviour and make reusable code) without a persistent object system. Zope as a whole makes your applications easier to build and to maintain, because of the everything is an object approach. But like I wrote above, if you insist you can put part of your files on the filesystem. Or in relational databases for that matter. so much for now Rik

1 0

RE: [Zope] How to render a HTMLFile in a method in a product
by Max Møller Rasmussen 06 Nov '00

06 Nov '00

>From: Chris Withers [mailto:chrisw@nipltd.com] I got this answer to my question about how to render a dtml document from a method. >testisDocTemp = 1 >def test(self,ignored,md): > " A simple test method " > return HTMLFile('skins/' + self.skin, globals())(None,md) And I have been busy trying to read the source code to understand the advise: As far as I understand the story goes a little like this: HTMLFile returns an object (through inheritance) "String" in file "DT_String.py" that has "__call__()" implemented, which means that the class can be called as a function... Hmm ... I had actually not seen this method used in Python before. def __call__(self,client=None,mapping={},**kw): This method accepts an object in which it will look up values to insert in the dtml file, and a mapping object (dictionary) that also will look up values to insert. So i guess that is what (None,md) does. But I dont understand what "ignored" does in my method, as it is unused, unless it is used by md >def test(self,ignored,md): Furthermore I dont understand what md is. Some kind of mapping object naturally, but where does it come from? Also what is testisDocTemp supposed to do? If I grep the source a "isDocTemp" shows up but is isn't easy to see what this does either. Regards Max M

2 1

RE: [Zope] How to render a HTMLFile in a method in a product
by Max Møller Rasmussen 06 Nov '00

06 Nov '00

From: Chris Withers [mailto:chrisw@nipltd.com] Thanks for the reply >testisDocTemp = 1 >def test(self,ignored,md): > " A simple test method " > return HTMLFile('skins/' + self.skin, globals())(None,md) Is this documented anywhere? I have a little trouble using code I don't understand. Pretty Please. Regards Max M

2 1